From neilb at suse.de  Tue May  1 05:46:12 2012
From: neilb at suse.de (NeilBrown)
Date: Tue, 1 May 2012 13:46:12 +1000
Subject: [Gta04-owner] Discussion: what are your dreams for the Openmoko
 Community
In-Reply-To: <E371F348-7143-498E-90DA-E200791D2CF4@goldelico.com>
References: <6E626A96-2C27-4FC8-B024-FB1969C5BDD6@goldelico.com>
	<4F9DC7D6.9030605@swissmail.org>
	<E371F348-7143-498E-90DA-E200791D2CF4@goldelico.com>
Message-ID: <20120501134612.77b1d553@notabene.brown>

On Mon, 30 Apr 2012 07:38:13 +0200 "Dr. H. Nikolaus Schaller"
<hns at goldelico.com> wrote:

> Wow,
> this is a really impressive list with new ideas
> I have not yet heard of!
> 
> And the most interesting thing is that I think
> almost all can be done and don't have major
> technical hurdles to overcome. The main
> challenge is to make them user friendly
> and bug free.
> 
> It appears that we more have a lack of active
> developers doing it. And are missing some
> coordination to get a complete solution.
> 
> Maybe we should again think about a more
> formal organization of the Openmoko.org
> (software) project?
> 
> Some Foundation or Association?

What benefit would a Foundation or Association bring?
I think they bring value in co-ordination when you have lots of people who
are finding it difficult to work together.  But I don't think that is the
problem that we have.  As you said: "lack of active developers".

What can we do to encourage those developers that we do have, and to entice
some on-lookers who aren't developers yet but might be in the future?

Maybe instead of asking "what would you like to see" (which certainly does
have a place) we should be asking something more down to earth, immediate,
and practical.

 - what are you working on?
 - what have you recently achieved?
 - what one thing would you like help with?

For myself:
 I'm mostly working on aligning the kernel we use with the upstream kernel.
 I recently submitted a collection of patches upstream and am hoping
 some will stick.

 I've recently discovered that my approach to power-management for wifi
 and bluetooth is less than perfect.  Among other things it doesn't actually
 set the voltage properly, so we are running on 2.8V rather than 3.1V.  I
 think I know how to fix it though.

 I recently tried 3.4-rc4 on my GTA04 and it was less than a brilliant
 success. No sound cards appear and the X server keeps crashing. I would love
 it if someone else had a look ....  You can find the code in the
 'gta04/mainline' branch of   git at github.com:neilbrown/linux.git

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120501/a4ad99e3/attachment.bin>

From cvwillegen at gmail.com  Tue May  1 06:37:13 2012
From: cvwillegen at gmail.com (Christ van Willegen)
Date: Tue, 1 May 2012 06:37:13 +0200
Subject: [Gta04-owner] Discussion: what are your dreams for the Openmoko
	Community
In-Reply-To: <20120501134612.77b1d553@notabene.brown>
References: <6E626A96-2C27-4FC8-B024-FB1969C5BDD6@goldelico.com>
	<4F9DC7D6.9030605@swissmail.org>
	<E371F348-7143-498E-90DA-E200791D2CF4@goldelico.com>
	<20120501134612.77b1d553@notabene.brown>
Message-ID: <CA+Ot1OymaHM2NxY_z=dMA832062+jHu2xBAjkU_WUCxwebne-w@mail.gmail.com>

On Tue, May 1, 2012 at 05:46, NeilBrown <neilb at suse.de> wrote:
> Maybe instead of asking "what would you like to see" (which certainly does
> have a place) we should be asking something more down to earth, immediate,
> and practical.
>
> ?- what are you working on?
> ?- what have you recently achieved?
> ?- what one thing would you like help with?

At the moment, I'm working on nothing GTA-related, since I don't have one.

I made a few patches to the SHR dialer and SMS application to
implement a few features ('Suggest' button that shows a popup with
contacts that have the part of the number your dialled in their
contact info, and '...' buttont that lets you select a phone number
from your contacts list to insert into an SMS). I also looked into
Navit to see if I could speed it up a bit, but since it's hard (for
me) to get it to build, I couldn't test it on the device.

I would like to be able to contribute more, but setting up a working
build system always seems to be a major PITA! It's taken me many hours
over many attempts to get SHR to compile and build. And then, when I
make changes its hard to get them to the device to test. And if it
doesn't work, you're left with plain old 'dump to a file' debugging...

If building, installing and testing were easier, I'd try to get
involved in UI programming.

Christ van Willegen
-- 
09 F9 11 02 9D 74 E3 5B D8 41 56 C5 63 56 88 C0

From neilb at suse.de  Tue May  1 09:43:24 2012
From: neilb at suse.de (NeilBrown)
Date: Tue, 1 May 2012 17:43:24 +1000
Subject: [Gta04-owner] Discussion: what are your dreams for the Openmoko
 Community
In-Reply-To: <CA+Ot1OymaHM2NxY_z=dMA832062+jHu2xBAjkU_WUCxwebne-w@mail.gmail.com>
References: <6E626A96-2C27-4FC8-B024-FB1969C5BDD6@goldelico.com>
	<4F9DC7D6.9030605@swissmail.org>
	<E371F348-7143-498E-90DA-E200791D2CF4@goldelico.com>
	<20120501134612.77b1d553@notabene.brown>
	<CA+Ot1OymaHM2NxY_z=dMA832062+jHu2xBAjkU_WUCxwebne-w@mail.gmail.com>
Message-ID: <20120501174324.02b904fb@notabene.brown>

On Tue, 1 May 2012 06:37:13 +0200 Christ van Willegen <cvwillegen at gmail.com>
wrote:

> On Tue, May 1, 2012 at 05:46, NeilBrown <neilb at suse.de> wrote:
> > Maybe instead of asking "what would you like to see" (which certainly does
> > have a place) we should be asking something more down to earth, immediate,
> > and practical.
> >
> > ?- what are you working on?
> > ?- what have you recently achieved?
> > ?- what one thing would you like help with?
> 
> At the moment, I'm working on nothing GTA-related, since I don't have one.
> 
> I made a few patches to the SHR dialer and SMS application to
> implement a few features ('Suggest' button that shows a popup with
> contacts that have the part of the number your dialled in their
> contact info, and '...' buttont that lets you select a phone number
> from your contacts list to insert into an SMS). I also looked into
> Navit to see if I could speed it up a bit, but since it's hard (for
> me) to get it to build, I couldn't test it on the device.
> 
> I would like to be able to contribute more, but setting up a working
> build system always seems to be a major PITA! It's taken me many hours
> over many attempts to get SHR to compile and build. And then, when I
> make changes its hard to get them to the device to test. And if it
> doesn't work, you're left with plain old 'dump to a file' debugging...
> 
> If building, installing and testing were easier, I'd try to get
> involved in UI programming.

With the GTA04 (once you get one) it is quite practical to build on the GTA04
itself so hopefully testing will be easier.
When I work this way I use emacs' 'tramp' mode to edit files on the GTA04 in
my laptop's emacs - it accesses the files with ssh/scp.  Then I make/test in
an 'ssh' window.  The usb2.0 networking is noticably faster than the usb1.1 on
the GTA02.

So hopefully when yo do get a GTA04 it will all be a lot easier!!

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120501/ab199a75/attachment.bin>

From hns at goldelico.com  Tue May  1 11:14:32 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Tue, 1 May 2012 11:14:32 +0200
Subject: [Gta04-owner] Discussion: what are your dreams for the Openmoko
	Community
In-Reply-To: <20120501134612.77b1d553@notabene.brown>
References: <6E626A96-2C27-4FC8-B024-FB1969C5BDD6@goldelico.com>
	<4F9DC7D6.9030605@swissmail.org>
	<E371F348-7143-498E-90DA-E200791D2CF4@goldelico.com>
	<20120501134612.77b1d553@notabene.brown>
Message-ID: <5819F955-E7C0-4A4A-88E7-527967FE8209@goldelico.com>


Am 01.05.2012 um 05:46 schrieb NeilBrown:

> On Mon, 30 Apr 2012 07:38:13 +0200 "Dr. H. Nikolaus Schaller"
> <hns at goldelico.com> wrote:
> 
>> Wow,
>> this is a really impressive list with new ideas
>> I have not yet heard of!
>> 
>> And the most interesting thing is that I think
>> almost all can be done and don't have major
>> technical hurdles to overcome. The main
>> challenge is to make them user friendly
>> and bug free.
>> 
>> It appears that we more have a lack of active
>> developers doing it. And are missing some
>> coordination to get a complete solution.
>> 
>> Maybe we should again think about a more
>> formal organization of the Openmoko.org
>> (software) project?
>> 
>> Some Foundation or Association?
> 
> What benefit would a Foundation or Association bring?
> I think they bring value in co-ordination when you have lots of people who
> are finding it difficult to work together.  But I don't think that is the
> problem that we have.  As you said: "lack of active developers".

Hm. I think we *have* a problem finding people to work together.
Is there a roadmap? Are there big visible and coordinated activities
to recruit new developers on big events like FOSDEM, LinuxTag etc.?

I think a foundation/association can help here because it has the ability to
collect some money and distribute to specific efforts. And can elect
persons to do something...

Currently, all developers (and marketeers) are pure volunteers driven
by their own agenda. Although this allows impressive progress in some
areas just because it is not coordinated (like your or Radek's work) it
has limits if we want to maintain and improve the overall quality of the
software stacks. And get the visibility of e.g. Debian or Suse.

I regularily hear complaints that the (user space) software isn't complete,
is broken in areas that had been working before, important features are
missing etc.

Well, this mailing lists are one means of collecting such complaints and
making it known to (potential) developers. But I can imagine another level
of development process, i.e. hackathons (hosted by some Openmoko
association or foundation and sponsored by donations). But usually good
developers aren't that good in organizing such events. And vice versa.

It has also been suggested to me that Golden Delicious Comp. should take
such a role (like Openmoko. Inc. did have) - but I am hesitating.

I see the role of GDC to provide future open hardware but remain software
agnostic, i.e. completely open. And some of our community software is
hardware agnostic (well, to some degree). E.g. SHR runs on several
different platforms. This IMHO excludes such a leadership role.

> 
> What can we do to encourage those developers that we do have, and to entice
> some on-lookers who aren't developers yet but might be in the future?

Such an organization could give more visibility. It can give some "empowerment"
if someone becomes elected as project leader. Maybe even provide some financial
compensation (e.g. for travel expenses, free participation in Hackathons).

Finally, who is the person for press contacts? If someone wants to write
an article about Openmoko software - whom to contact first place?

But overall: we (and new users and developers) must feel that it is important
for mankind that we are member of this community (being formal or informal)
and contribute.

> 
> Maybe instead of asking "what would you like to see" (which certainly does
> have a place) we should be asking something more down to earth, immediate,
> and practical.
> 
> - what are you working on?
> - what have you recently achieved?
> - what one thing would you like help with?
> 
> For myself:
> I'm mostly working on aligning the kernel we use with the upstream kernel.
> I recently submitted a collection of patches upstream and am hoping
> some will stick.

++

> I've recently discovered that my approach to power-management for wifi
> and bluetooth is less than perfect.  Among other things it doesn't actually
> set the voltage properly, so we are running on 2.8V rather than 3.1V.  I
> think I know how to fix it though.

2.8V is a little low which reduces transmission power.
But on the GTA04A4 you can't set the voltage any more. It has a fixed 3.3V
LDO.

> I recently tried 3.4-rc4 on my GTA04 and it was less than a brilliant
> success. No sound cards appear and the X server keeps crashing. I would love
> it if someone else had a look ....  You can find the code in the
> 'gta04/mainline' branch of   git at github.com:neilbrown/linux.git
> 
> NeilBrown

Nikolaus



From hns at goldelico.com  Tue May  1 18:06:23 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Tue, 1 May 2012 18:06:23 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
Message-ID: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>

Hi,

we have developed a prototype for a 80 button QWERTY keyboard PCB
that could eventually be connected/integrated into the GTA04. It should fit
into a specially designed battery cover so that you can easily stow it away
if not needed. Such battery covers could be produced individually through
3D printing solving the issue of manageing and stocking 20 different key
layouts.

But watch yourself how we think it can look like:

	http://youtu.be/WM94%5fR5eKcc

There is also a new video showing a comparison with some other keyboards:

	http://youtu.be/wGASnE1zGh4

Pleas tell us if you like this idea and what you would like to pay for such an
extension unit through the Wishlist function of our shop:

	http://www.handheld-linux.com/wiki.php?page=GTA04%3AKeyboard

Two issues are still to be developed:

a) how to reliably connect it to the GTA04 PCB (soldering copper wires or a
    FFCs is a little difficult so it should have a tiny, flexible but robust B2B cable).

    Maybe, we can use a micro-USB socket or similar (we need to connect 6 wires).
    This may also need a redesign of the GTA04 board (for a nice plug)

b) design a 3D printable case with key-caps that is robust enough

If you want to support us for developing this idea, please give us a kickstart
donation.


Nikolaus

PS: the keyboard driver for the TCA8418 is already part of Linux 3.3 - and has
been backported to the 2.6.32-kernel. This has been tested to work on a
BeagleBoard XM.

http://git.goldelico.com/?p=gta04-kernel.git;a=commit;h=f19d5c430458bbce8955bc9e04dd161f6a80347d

It just needs platform data in the board file:

http://git.goldelico.com/?p=gta04-kernel.git;a=blobdiff;f=arch/arm/mach-omap2/board-omap3gta04.c;h=8a7e4b0803920f635e7101bfbd5a60b6b84b1107;hp=3e49efef2de0b42cd419a46a9cd45448fd04a44c;hb=4b2de3db742abce9212c1af2cc576e2a3a64b0d9;hpb=1d7c6b5f043661621ec374d96c3c4a4454f9bb7b

From hns at goldelico.com  Tue May  1 18:14:58 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Tue, 1 May 2012 18:14:58 +0200
Subject: [Gta04-owner] Detecting headset plug: twl4030-madc IRQ issue
In-Reply-To: <20120429083223.509aac7c@notabene.brown>
References: <1335633337.1729.11.camel@aldrin>
	<20120429083223.509aac7c@notabene.brown>
Message-ID: <89D760AE-9013-4214-BA7A-BCFB132497AB@goldelico.com>


Am 29.04.2012 um 00:32 schrieb NeilBrown:

> On Sat, 28 Apr 2012 19:15:37 +0200 PaulK <paulk at paulk.fr> wrote:
> 
>> I am now working on adding sound to Replicant. Dealing with the controls
>> seems very clear to me now, thanks to the docs on:
>> http://projects.goldelico.com/p/gta04-kernel/doc/
>> 
>> Currently, I am trying to get the twl4030-madc to report headset
>> plug/unplug. If I read correctly the docs, ADC7 is MICSENSE. 
>> Now the driver reports data through channels, so I guess what I'm
>> looking for is chan 7. 
>> 
>> A threaded IRQ is requested at probe and, if I got it correctly, is
>> supposed to call the twl4030_madc_threaded_irq_handler function when the
>> value of one ADC changes. Except that I didn't see that function getting
>> called when the headset is plugged or unplugged. I even tried removing
>> the battery (as it should report battery-related infos), it it didn't
>> get called either. 
>> 
>> Now there are a few confusing elements: there is a
>> twl4030_madc_enable_irq function that could lead me to think that it is
>> somehow disabled by default and the ADCs chip must somehow be told to
>> make use of the IRQ line. 
>> 
>> I have never dealt with IRQ interrupts before so I may be missing
>> something obvious here. Though, I would appreciate help discovering what
>> is actually going on and why this handler function never gets called.
>> 
> 
> 
> I might be missing something, but I don't think you can get an interrupt when
> the headset is plugged in.

Yes, there is no hardware provision for generating an interrupt.

> 
> All that ADC7 allows if for polling - you can ask it to measure the voltage
> level and then it sends an interrupt when the voltage has been measured.
> Then you can read it and interpret it.
> 
> The voltage is proportional to the current through R706.  To get current
> through this resistor you need to set HSMICBIAS_EN so that VHSMIC.OUT is
> driven.
> 
> When nothing is plugged in this will drive 2.2V through a resistor network of 
>  R707 - 330
>  R706 - 33
>  R705 - 2K4
>  R703 - 330
> 
> I assume they are all Kohms, so that is about 700Kohm so we get about 3.1
> microamps so ADC7 sees about 0.3volts (10uA/V).
> 
> When 3-connector headset is plugged in, R703 is shorted, so the resistance
> becomes 365K and we get 6uA, or 0.6 Volts.

> 
> When a 3-connector headset with a mic is plugged in ... I suspect we don't
> see any DC current, just an AC current from the mic, so maybe ADC7 will see
> 0V???
> 
> In any case, it is possible to ask and find out what it connected, but I
> cannot see any way to get an interrupt when that changes.
> 
> Is that right Nikolaus??

Yes,
this is the basic idea. It is intended that one can connect a remote control
unit with several buttons and a resistor ladder. Similar to what the Sharp Zaurus
did have.

	http://www.penguin.cz/~utx/zaurus/hardware
	http://www.penguin.cz/~utx/zaurus/photos#ce-rh2

And, it is planned that it can detect headsets with and without microphones.

But this circuit has not yet been tested if it does anything useful in practice.
So if some volunteer wants to play with it... And make some proposals how
to modify this circuit for future GTA04 devices.

So how important is it for Replicant to work useful to detect the headset
being plugged in? Can there be a too to manually switch off the speakers?
This could be a workaround until we have tested and made the MADC working.

> 
> 
> NeilBrown

BR,
Nikolaus



From rah at settrans.net  Wed May  2 11:21:35 2012
From: rah at settrans.net (Bob Ham)
Date: Wed, 02 May 2012 10:21:35 +0100
Subject: [Gta04-owner] Home printed cases
Message-ID: <1335950495.32304.134.camel@myrtle.6gnip.net>

Hi all,

When people started talking about 3D-printing cases I was surprised to
see them referring to commercial services, rather than printing at home
using a Reprap? or similar printer.  This, to me, home printing is in
perfect alignment with the ethic behind the GTA04.

Recently, I went to a maker night? at my local maker/hack/co-working
space? where they have a Makerbot Cupcake CNC?.  I managed to print a
back plate using Slyon's back.stl model?.  Unfortunately, the Cupcake
has quite a small print area (about the size of a cupcake :-)  Because
of this, I was only able to print the part at 80% size.  I've
photographed the print and put the photographs here:

  http://settrans.net/~rah/gta04-0.8-print/

Ideally I would have split the back.stl 3D mesh in half and printed both
sides but I didn't have time on the night to get to grips with the 3D
editors I had available (you'd think it would be a simple task! :-)  I
will do this at home when I get some time and then print the two halves
at another maker night in future.


What I've learned from this so far, is that the print does seem quite
sturdy.  However, the definition of the 3D printer is nowhere near the
quality needed to produce a usable print of the existing case design.
Notably the tangs on the side of the back plate are completely missing
from this 80% print as they are too thin.

So, the future direction of my endeavour is to design a new case that
(1) is printable by a home 3D printer, (2) can house all the parts for a
GTA04 and (3) is usable as a daily phone.  This is no small endeavour,
not least because it involves building my own Reprap printer to make use
of the larger print area.

My intention is to start with a gross, simple design without concern for
aesthetics; just a rectangular block big enough to hold the phone
components and be strong enough to survive a pocket.  I expect to use
screws to allow access to a battery compartment.  Once this design is
usable, I will refine it to be more aesthetic.

I will also, of course, keep you all updated on my progress.

Bob


? http://reprap.org/wiki/Main_Page
? http://makernight.co.uk/
? http://doesliverpool.com/
? https://en.wikipedia.org/wiki/MakerBot_Industries#Cupcake_CNC
? http://www.slyon.de/CAD/

-- 
Bob Ham <rah at settrans.net>
Diaspora: rah at pod.settrans.net

for (;;) { ++pancakes; }
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120502/7986cc81/attachment.bin>

From lukasmaerdian at googlemail.com  Wed May  2 12:56:29 2012
From: lukasmaerdian at googlemail.com (=?ISO-8859-1?Q?Lukas_M=E4rdian?=)
Date: Wed, 02 May 2012 12:56:29 +0200
Subject: [Gta04-owner] 3D printed Shapeways case
In-Reply-To: <1335947444.32304.97.camel@myrtle.6gnip.net>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<1335947444.32304.97.camel@myrtle.6gnip.net>
Message-ID: <4FA112DD.8020001@googlemail.com>

On 02.05.2012 10:30, Bob Ham wrote:
> On Tue, 2012-05-01 at 18:06 +0200, Dr.H.NikolausS wrote:
> At present, we do not have this.  As I understand it, Slyon's cases do
> not fit together properly and so we cannot build a complete GTA04 case
> from them yet (has this changed?)  There are also no case component kits
> available.
> 
> I think it would be more advantageous to spend time dealing with these
> basic problems.  That way, we will have a solid base from which to build
> more advanced designs, including designs that incorporate a keyboard.
> 
> Also, I believe we should not limit ourselves to the existing case
> shape.  I will say more about that in a separate email to the GTA04
> list.
> 

Actually my latest prototype case fits together pretty well and is
suitable for daily use (if you use the CaseKit replacement parts like
earpiece, vibracall motor, ...).

I've done some improvements to it, which I didn't have the possibility
to test, yet. Therefore I didn't announce anything new until now. But
this very latest version was presented at the Openmobility conference.

But still, the very latest (and probably pretty usable) version is
available [0] and ready to use. There are a few people out there, who
already have one.

I've also had access to a RepRap at FOSDEM and did experiment with it
(image attached). But my conclusion is, that RepRaps just aren't precise
enough (yet) to print the very fine structures needed for a GTA04 case.
It's probably possible do design a very basic case, which is printable
by a RepRap, but the question is, if this is good/comfortable enough for
daily usage..

Some RepRaps seem to evolve in a pretty good direction already. My top 3 is:

http://reprappro.com/Huxley
http://www.wix.com/3dprinterfaq/veloso3d
http://store.solidoodle.com/index.php?route=product/product&product_id=56

But still there is the problem with non-printable overhangs and
availability of those RepRaps.

Cheers,
  Lukas 'Slyon' M?rdian

[0] http://blog.slyon.de/3d-printed-gta04-case/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: MakerBot.jpg
Type: image/jpeg
Size: 158962 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120502/c27708b8/attachment-0001.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 490 bytes
Desc: OpenPGP digital signature
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120502/c27708b8/attachment-0001.bin>

From farny.guillaume at yahoo.fr  Wed May  2 16:22:15 2012
From: farny.guillaume at yahoo.fr (Farny Guillaume)
Date: Wed, 2 May 2012 15:22:15 +0100 (BST)
Subject: [Gta04-owner] Re : sir
In-Reply-To: <CA+JMvaX6NknfTGC_1zx4URa2rHM2sYQek6A0ZP1wMsMjNi4tPg@mail.gmail.com>
References: <1335895781.11832.YahooMailNeo@web132304.mail.ird.yahoo.com>
	<CA+JMvaX6NknfTGC_1zx4URa2rHM2sYQek6A0ZP1wMsMjNi4tPg@mail.gmail.com>
Message-ID: <1335968535.18176.YahooMailNeo@web132306.mail.ird.yahoo.com>

Ackowledged.
I transmit to the gta04-owner at goldelico.com mailing list.


I renew my offer : if need be, I can rule a domain with git or other services, and be responsive and responsible.

Regards,


--


Guillaume Farny


CCAS de Bagnolet
Place Salvador Allende
93170 Bagnolet
FRANCE


06.16.52.42.32
farny.guillaume at yahoo.fr



________________________________
 De?: Sean Moss-Pultz <sean at openmoko.com>
??: Farny Guillaume <farny.guillaume at yahoo.fr> 
Envoy? le : Mercredi 2 mai 2012 14h37
Objet?: Re: sir
 

Dear?Guillaume

I'm in the process of contacting the admins. Please give me a few days as they are in different timezones.?

Sean



On Wed, May 2, 2012 at 2:09 AM, Farny Guillaume <farny.guillaume at yahoo.fr> wrote:

Sir,
>
>
>I am an openmoko project contributer.
>
>
>
>Obviously some openmoko.org subdomains don't respond.
>Can you tell what is going wrong ?
>
>?
>Do you face a problem ?
>Do you need assistance or someone to take responsibility of the servers ?
>
>Regards,
>
>
>
>
>--
>
>
>
>Guillaume Farny
>
>
>
>CCAS de Bagnolet
>Place Salvador Allende
>93170 Bagnolet
>FRANCE
>
>
>
>06.16.52.42.32
>farny.guillaume at yahoo.fr
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120502/fddd7fdb/attachment.html>

From christoph.mair at gmail.com  Wed May  2 19:41:44 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Wed, 2 May 2012 19:41:44 +0200
Subject: [Gta04-owner] Home printed cases
In-Reply-To: <1335950495.32304.134.camel@myrtle.6gnip.net>
References: <1335950495.32304.134.camel@myrtle.6gnip.net>
Message-ID: <CAHvJMeGXvfxSjj9w6ruUpMzZTJReZ+P0GAF3rtqvASPxRju92w@mail.gmail.com>

On Wed, May 2, 2012 at 11:21 AM, Bob Ham <rah at settrans.net> wrote:
> Hi all,
> Recently, I went to a maker night? at my local maker/hack/co-working
> space? where they have a Makerbot Cupcake CNC?. ?I managed to print a
> back plate using Slyon's back.stl model?. ?Unfortunately, the Cupcake
> has quite a small print area (about the size of a cupcake :-) ?Because
> of this, I was only able to print the part at 80% size. ?I've
> photographed the print and put the photographs here:

I've seen a RepRap during the OpenMobility converence in Prague and it
seems to be able to produce better quality parts than the Makerbot
(seen at FOSDEM). The group which brought the RepRap developed a
nozzle with 0.35mm diameter and an example part I saw was very
promising. They tried to "slice" the front cover for this new nozzle
and it worked without problems. Unfortunately there was not enough
time to actually print it..

Christoph

From christoph.mair at gmail.com  Wed May  2 19:58:33 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Wed, 2 May 2012 19:58:33 +0200
Subject: [Gta04-owner] 3D printed Shapeways case
In-Reply-To: <4FA112DD.8020001@googlemail.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<1335947444.32304.97.camel@myrtle.6gnip.net>
	<4FA112DD.8020001@googlemail.com>
Message-ID: <CAHvJMeG0BH_OT0WZmdDGL4PfSOGqpzKMQj0=E_E-awgCdUvTgA@mail.gmail.com>

On Wed, May 2, 2012 at 12:56 PM, Lukas M?rdian
<lukasmaerdian at googlemail.com> wrote:
> On 02.05.2012 10:30, Bob Ham wrote:
>> On Tue, 2012-05-01 at 18:06 +0200, Dr.H.NikolausS wrote:
>> At present, we do not have this. ?As I understand it, Slyon's cases do
>> not fit together properly and so we cannot build a complete GTA04 case
>> from them yet (has this changed?) ?There are also no case component kits
>> available.
>>
>> I think it would be more advantageous to spend time dealing with these
>> basic problems. ?That way, we will have a solid base from which to build
>> more advanced designs, including designs that incorporate a keyboard.
>>
>> Also, I believe we should not limit ourselves to the existing case
>> shape. ?I will say more about that in a separate email to the GTA04
>> list.
>>
>
> Actually my latest prototype case fits together pretty well and is
> suitable for daily use (if you use the CaseKit replacement parts like
> earpiece, vibracall motor, ...).
>
> I've done some improvements to it, which I didn't have the possibility
> to test, yet. Therefore I didn't announce anything new until now. But
> this very latest version was presented at the Openmobility conference.
>
> But still, the very latest (and probably pretty usable) version is
> available [0] and ready to use. There are a few people out there, who
> already have one.

I ordered a complete case and it arrived just in time to show it at
the Openmobility conference. The outer parts fit quite well, just the
speaker box seems to have some small problems. I could no debug them
yet but I'll try to get it done during the next weekend. I will attend
the Linuxwochen in Vienna on friday and saturday. If you want to see
my case parts (or anything related) come to my talk on friday at 16:00
[1] or try to catch me during the two days.

Best regards,
  Christoph

[1] http://linuxwochen.at/index.php?option=com_content&view=article&id=307&Itemid=83

From lukasmaerdian at googlemail.com  Wed May  2 20:08:29 2012
From: lukasmaerdian at googlemail.com (=?UTF-8?B?THVrYXMgTcOkcmRpYW4=?=)
Date: Wed, 02 May 2012 20:08:29 +0200
Subject: [Gta04-owner] 3D printed Shapeways case
In-Reply-To: <CAHvJMeG0BH_OT0WZmdDGL4PfSOGqpzKMQj0=E_E-awgCdUvTgA@mail.gmail.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<1335947444.32304.97.camel@myrtle.6gnip.net>
	<4FA112DD.8020001@googlemail.com>
	<CAHvJMeG0BH_OT0WZmdDGL4PfSOGqpzKMQj0=E_E-awgCdUvTgA@mail.gmail.com>
Message-ID: <4FA1781D.2080602@googlemail.com>

On 02.05.2012 19:58, Christoph Mair wrote:
> On Wed, May 2, 2012 at 12:56 PM, Lukas M?rdian
> <lukasmaerdian at googlemail.com> wrote:
>> On 02.05.2012 10:30, Bob Ham wrote:
>>> On Tue, 2012-05-01 at 18:06 +0200, Dr.H.NikolausS wrote:
>>> At present, we do not have this.  As I understand it, Slyon's cases do
>>> not fit together properly and so we cannot build a complete GTA04 case
>>> from them yet (has this changed?)  There are also no case component kits
>>> available.
>>>
>>> I think it would be more advantageous to spend time dealing with these
>>> basic problems.  That way, we will have a solid base from which to build
>>> more advanced designs, including designs that incorporate a keyboard.
>>>
>>> Also, I believe we should not limit ourselves to the existing case
>>> shape.  I will say more about that in a separate email to the GTA04
>>> list.
>>>
>>
>> Actually my latest prototype case fits together pretty well and is
>> suitable for daily use (if you use the CaseKit replacement parts like
>> earpiece, vibracall motor, ...).
>>
>> I've done some improvements to it, which I didn't have the possibility
>> to test, yet. Therefore I didn't announce anything new until now. But
>> this very latest version was presented at the Openmobility conference.
>>
>> But still, the very latest (and probably pretty usable) version is
>> available [0] and ready to use. There are a few people out there, who
>> already have one.
> 
> I ordered a complete case and it arrived just in time to show it at
> the Openmobility conference. The outer parts fit quite well, just the
> speaker box seems to have some small problems. I could no debug them
> yet but I'll try to get it done during the next weekend. I will attend
> the Linuxwochen in Vienna on friday and saturday. If you want to see
> my case parts (or anything related) come to my talk on friday at 16:00
> [1] or try to catch me during the two days.

Unfortunately I won't make it to Vienna...


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 490 bytes
Desc: OpenPGP digital signature
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120502/5bdd115f/attachment.bin>

From butrus.butrus at gmail.com  Thu May  3 10:03:12 2012
From: butrus.butrus at gmail.com (Butrus Damaskus)
Date: Thu, 3 May 2012 10:03:12 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
Message-ID: <CALM6Ch_hKeeB4WJw=jHCpEFq67Xg9VtM_YZiYG6K3iw7+JYpcg@mail.gmail.com>

Very interresting, but I'm a little bit puzzled as how it should work
in the real life...

On Tue, May 1, 2012 at 6:06 PM, Dr. H. Nikolaus Schaller
<hns at goldelico.com> wrote:
> Hi,
>
> we have developed a prototype for a 80 button QWERTY keyboard PCB
> that could eventually be connected/integrated into the GTA04. It should fit
> into a specially designed battery cover so that you can easily stow it away
> if not needed. Such battery covers could be produced individually through
> 3D printing solving the issue of manageing and stocking 20 different key
> layouts.
>
> But watch yourself how we think it can look like:
>
> ? ? ? ?http://youtu.be/WM94%5fR5eKcc
>
> There is also a new video showing a comparison with some other keyboards:
>
> ? ? ? ?http://youtu.be/wGASnE1zGh4
>
> Pleas tell us if you like this idea and what you would like to pay for such an
> extension unit through the Wishlist function of our shop:
>
> ? ? ? ?http://www.handheld-linux.com/wiki.php?page=GTA04%3AKeyboard
>
> Two issues are still to be developed:
>
> a) how to reliably connect it to the GTA04 PCB (soldering copper wires or a
> ? ?FFCs is a little difficult so it should have a tiny, flexible but robust B2B cable).
>
> ? ?Maybe, we can use a micro-USB socket or similar (we need to connect 6 wires).
> ? ?This may also need a redesign of the GTA04 board (for a nice plug)
>
> b) design a 3D printable case with key-caps that is robust enough
>
> If you want to support us for developing this idea, please give us a kickstart
> donation.
>
>
> Nikolaus
>
> PS: the keyboard driver for the TCA8418 is already part of Linux 3.3 - and has
> been backported to the 2.6.32-kernel. This has been tested to work on a
> BeagleBoard XM.
>
> http://git.goldelico.com/?p=gta04-kernel.git;a=commit;h=f19d5c430458bbce8955bc9e04dd161f6a80347d
>
> It just needs platform data in the board file:
>
> http://git.goldelico.com/?p=gta04-kernel.git;a=blobdiff;f=arch/arm/mach-omap2/board-omap3gta04.c;h=8a7e4b0803920f635e7101bfbd5a60b6b84b1107;hp=3e49efef2de0b42cd419a46a9cd45448fd04a44c;hb=4b2de3db742abce9212c1af2cc576e2a3a64b0d9;hpb=1d7c6b5f043661621ec374d96c3c4a4454f9bb7b
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

From hns at goldelico.com  Thu May  3 10:26:47 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Thu, 3 May 2012 10:26:47 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <CALM6Ch_hKeeB4WJw=jHCpEFq67Xg9VtM_YZiYG6K3iw7+JYpcg@mail.gmail.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<CALM6Ch_hKeeB4WJw=jHCpEFq67Xg9VtM_YZiYG6K3iw7+JYpcg@mail.gmail.com>
Message-ID: <B508BBBD-F3F3-48D7-9B5B-0C4F5E018E16@goldelico.com>


Am 03.05.2012 um 10:03 schrieb Butrus Damaskus:

> Very interresting, but I'm a little bit puzzled as how it should work
> in the real life...

That is a task for further study and optimization. We have just built one unit
and we need some 3D printed cases to do a better evaluation.

Some more general thought: I am not sure how we should handle such ideas and
how this community sees its role.

Should we announce/publish immediately as soon as we have a prototype? And ask
the community for comments and participation in improving it? Or should we silently
finish some design and work with beta testers not on this list...

I think this is a fundamental difference to open software development. There we can
have a stable, a testing and an experimental branch and everyone can download
the one he/she likes. With hardware we have to build physical protoypes and make
experiments (which are not for free) to find out what a good solution is.

Nikolaus



> 
> On Tue, May 1, 2012 at 6:06 PM, Dr. H. Nikolaus Schaller
> <hns at goldelico.com> wrote:
>> Hi,
>> 
>> we have developed a prototype for a 80 button QWERTY keyboard PCB
>> that could eventually be connected/integrated into the GTA04. It should fit
>> into a specially designed battery cover so that you can easily stow it away
>> if not needed. Such battery covers could be produced individually through
>> 3D printing solving the issue of manageing and stocking 20 different key
>> layouts.
>> 
>> But watch yourself how we think it can look like:
>> 
>>        http://youtu.be/WM94%5fR5eKcc
>> 
>> There is also a new video showing a comparison with some other keyboards:
>> 
>>        http://youtu.be/wGASnE1zGh4
>> 
>> Pleas tell us if you like this idea and what you would like to pay for such an
>> extension unit through the Wishlist function of our shop:
>> 
>>        http://www.handheld-linux.com/wiki.php?page=GTA04%3AKeyboard
>> 
>> Two issues are still to be developed:
>> 
>> a) how to reliably connect it to the GTA04 PCB (soldering copper wires or a
>>    FFCs is a little difficult so it should have a tiny, flexible but robust B2B cable).
>> 
>>    Maybe, we can use a micro-USB socket or similar (we need to connect 6 wires).
>>    This may also need a redesign of the GTA04 board (for a nice plug)
>> 
>> b) design a 3D printable case with key-caps that is robust enough
>> 
>> If you want to support us for developing this idea, please give us a kickstart
>> donation.
>> 
>> 
>> Nikolaus
>> 
>> PS: the keyboard driver for the TCA8418 is already part of Linux 3.3 - and has
>> been backported to the 2.6.32-kernel. This has been tested to work on a
>> BeagleBoard XM.
>> 
>> http://git.goldelico.com/?p=gta04-kernel.git;a=commit;h=f19d5c430458bbce8955bc9e04dd161f6a80347d
>> 
>> It just needs platform data in the board file:
>> 
>> http://git.goldelico.com/?p=gta04-kernel.git;a=blobdiff;f=arch/arm/mach-omap2/board-omap3gta04.c;h=8a7e4b0803920f635e7101bfbd5a60b6b84b1107;hp=3e49efef2de0b42cd419a46a9cd45448fd04a44c;hb=4b2de3db742abce9212c1af2cc576e2a3a64b0d9;hpb=1d7c6b5f043661621ec374d96c3c4a4454f9bb7b
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner


From morphis at gravedo.de  Thu May  3 11:07:55 2012
From: morphis at gravedo.de (Simon Busch)
Date: Thu, 03 May 2012 11:07:55 +0200
Subject: [Gta04-owner] Modem closes serial line after active calls
	(sometimes)
In-Reply-To: <201204291356.06717.psonek2@seznam.cz>
References: <4F9ABBE7.2010905@gravedo.de>
	<87r4v8yxto.fsf@neil-laptop.ossau.uklinux.net>
	<4F9BE61A.90807@gravedo.de> <201204291356.06717.psonek2@seznam.cz>
Message-ID: <4FA24AEB.8090708@gravedo.de>

On 29.04.2012 15:56, Radek Polak wrote:
> On Saturday 28 April 2012 12:44:10 Simon Busch wrote:
> 
>> As far as I got the details from that thread there is no clean solution
>> for the disconnecting modem just a workaround as Radek implemented it
>> for qtmoko, right?
> 
> Yup, i found no clean solution. Currently i switch to 2G on modem 
> initialization - this seems to help + after 30s of unsuccessful reads i 
> restart whole QtMoko + some other hacks:

Yeah, 2G seems to solve the problem. I tried with limiting the radio
access to 2G only and never got a disconnect from the modem.

Disabling 3G generally for the GTA04 is nothing we really want. So I am
thinking about just limiting the radio access to 2G when dealing with
calls. When a call is incoming or outgoing we have to switch the radio
access to 2G and switch it back to ANY when we're done with call
handling. Let's see if this works not just in theory ...

regards,
Simon

-- 
Simon Busch - http://mm.gravedo.de/blog/

From rah at settrans.net  Thu May  3 11:18:16 2012
From: rah at settrans.net (Bob Ham)
Date: Thu, 03 May 2012 10:18:16 +0100
Subject: [Gta04-owner] Modem closes serial line after active calls
	(sometimes)
In-Reply-To: <4FA24AEB.8090708@gravedo.de>
References: <4F9ABBE7.2010905@gravedo.de>
	<87r4v8yxto.fsf@neil-laptop.ossau.uklinux.net>
	<4F9BE61A.90807@gravedo.de> <201204291356.06717.psonek2@seznam.cz>
	<4FA24AEB.8090708@gravedo.de>
Message-ID: <4FA24D58.9070104@settrans.net>

On 03/05/2012 10:07, Simon Busch wrote:
> On 29.04.2012 15:56, Radek Polak wrote:

>> Currently i switch to 2G on modem 
>> initialization - this seems to help + after 30s of unsuccessful reads i 
>> restart whole QtMoko + some other hacks:
> 
> Yeah, 2G seems to solve the problem. I tried with limiting the radio
> access to 2G only and never got a disconnect from the modem.
> 
> Disabling 3G generally for the GTA04 is nothing we really want. So I am
> thinking about just limiting the radio access to 2G when dealing with
> calls.

I feel I should point out that my phone company does not allow 2G usage;
they are 3G only.  The connection will roam to 2G networks through third
parties if there is no 3G signal but if you do this regularly they will
complain.  If you continue to do this after they've complained they will
cut off your connection.

-- 
Bob Ham <rah at settrans.net>
Diaspora: rah at pod.settrans.net

for (;;) { ++pancakes; }

From morphis at gravedo.de  Thu May  3 11:46:40 2012
From: morphis at gravedo.de (Simon Busch)
Date: Thu, 03 May 2012 11:46:40 +0200
Subject: [Gta04-owner] Modem closes serial line after active calls
	(sometimes)
In-Reply-To: <4FA24D58.9070104@settrans.net>
References: <4F9ABBE7.2010905@gravedo.de>
	<87r4v8yxto.fsf@neil-laptop.ossau.uklinux.net>
	<4F9BE61A.90807@gravedo.de> <201204291356.06717.psonek2@seznam.cz>
	<4FA24AEB.8090708@gravedo.de> <4FA24D58.9070104@settrans.net>
Message-ID: <4FA25400.3050109@gravedo.de>

On 03.05.2012 11:18, Bob Ham wrote:
> On 03/05/2012 10:07, Simon Busch wrote:
>> On 29.04.2012 15:56, Radek Polak wrote:
> 
>>> Currently i switch to 2G on modem 
>>> initialization - this seems to help + after 30s of unsuccessful reads i 
>>> restart whole QtMoko + some other hacks:
>>
>> Yeah, 2G seems to solve the problem. I tried with limiting the radio
>> access to 2G only and never got a disconnect from the modem.
>>
>> Disabling 3G generally for the GTA04 is nothing we really want. So I am
>> thinking about just limiting the radio access to 2G when dealing with
>> calls.
> 
> I feel I should point out that my phone company does not allow 2G usage;
> they are 3G only.  The connection will roam to 2G networks through third
> parties if there is no 3G signal but if you do this regularly they will
> complain.  If you continue to do this after they've complained they will
> cut off your connection.

Supporting use case like this and providing a common solution is
somewhat hard. I am not planing to do this for FSO as the one and only
solution but trying to see if it helps. There will be a reinitialization
of the transport if it HUP's like Radek did for qtmoko too.

regards,
Simon

-- 
Simon Busch - http://mm.gravedo.de/blog/

From psonek2 at seznam.cz  Thu May  3 13:10:42 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Thu, 03 May 2012 13:10:42 +0200
Subject: [Gta04-owner] Modem closes serial line after active calls
	(sometimes)
In-Reply-To: <4FA25400.3050109@gravedo.de>
References: <4F9ABBE7.2010905@gravedo.de> <4FA24D58.9070104@settrans.net>
	<4FA25400.3050109@gravedo.de>
Message-ID: <1986775.jfjNFk0U4Y@rp-core>

On Thursday, May 03, 2012 11:46:40 AM Simon Busch wrote:

> Supporting use case like this and providing a common solution is
> somewhat hard. I am not planing to do this for FSO as the one and only
> solution but trying to see if it helps. There will be a reinitialization
> of the transport if it HUP's like Radek did for qtmoko too.

It would be better if we could handle this on kernel level somehow. Right now 
we need these ugly workarounds for QtMoko, FSO and oFono. I am not happy at 
all having code like in QSerialPort:

	if(not read anything for 10s && port=/dev/ttyHSO_Applications)
		restart_qtmoko();

IIRC the problem with USB reenumerating happens even when on 2G - but i am not 
100% sure - i havent seen it for long time.

Maybe one well placed delay in kernel could solve this? Nikolaus reported that 
if you enable debug/verbose for the hso driver, it's not happending. Maybe 
crazy idea - replace all printk's with small delay and see what happens?

Regards

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120503/5e2e1076/attachment.html>

From neil at ossau.homelinux.net  Thu May  3 13:16:12 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Thu, 03 May 2012 13:16:12 +0200
Subject: [Gta04-owner] Modem closes serial line after active calls
	(sometimes)
In-Reply-To: <4FA24AEB.8090708@gravedo.de> (Simon Busch's message of "Thu, 03
	May 2012 11:07:55 +0200")
References: <4F9ABBE7.2010905@gravedo.de>
	<87r4v8yxto.fsf@neil-laptop.ossau.uklinux.net>
	<4F9BE61A.90807@gravedo.de> <201204291356.06717.psonek2@seznam.cz>
	<4FA24AEB.8090708@gravedo.de>
Message-ID: <87lil9wm4j.fsf@neil-laptop.ossau.uklinux.net>

Simon Busch <morphis at gravedo.de> writes:

> On 29.04.2012 15:56, Radek Polak wrote:
>> On Saturday 28 April 2012 12:44:10 Simon Busch wrote:
>> 
>>> As far as I got the details from that thread there is no clean solution
>>> for the disconnecting modem just a workaround as Radek implemented it
>>> for qtmoko, right?
>> 
>> Yup, i found no clean solution. Currently i switch to 2G on modem 
>> initialization - this seems to help + after 30s of unsuccessful reads i 
>> restart whole QtMoko + some other hacks:
>
> Yeah, 2G seems to solve the problem. I tried with limiting the radio
> access to 2G only and never got a disconnect from the modem.
>
> Disabling 3G generally for the GTA04 is nothing we really want. So I am
> thinking about just limiting the radio access to 2G when dealing with
> calls. When a call is incoming or outgoing we have to switch the radio
> access to 2G and switch it back to ANY when we're done with call
> handling. Let's see if this works not just in theory ...

Nice idea!

In parallel, though, I believe Nikolaus was going to look into the
possibilities for modem firmware upgrade.  Nikolaus, is there any news
on that?

      Neil

From andreas at kemnade.info  Thu May  3 14:23:49 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Thu, 3 May 2012 14:23:49 +0200
Subject: [Gta04-owner] Home printed cases
In-Reply-To: <1335950495.32304.134.camel@myrtle.6gnip.net>
References: <1335950495.32304.134.camel@myrtle.6gnip.net>
Message-ID: <20120503142349.4c7d4b96@kemnade.info>

Hi,

you have seen my experiments using my reprap, a 0.5mm nozzle with the back part?
http://misc.andi.de1.cc/f1.jpg
http://misc.andi.de1.cc/f2.jpg


They look a bit better I think. 
I still think that redesigning the thing would be a good idea, but then I think
it should be parameterized, so that you can easily change thickness of walls
and so on. The retirement option for my GTA02 board is to steer my reprap, and
should get a simple case for that. I plan to do some simple thing with openscad.

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120503/3b56480e/attachment.bin>

From morphis at gravedo.de  Thu May  3 14:41:08 2012
From: morphis at gravedo.de (Simon Busch)
Date: Thu, 03 May 2012 14:41:08 +0200
Subject: [Gta04-owner] Modem closes serial line after active calls
	(sometimes)
In-Reply-To: <87lil9wm4j.fsf@neil-laptop.ossau.uklinux.net>
References: <4F9ABBE7.2010905@gravedo.de>
	<87r4v8yxto.fsf@neil-laptop.ossau.uklinux.net>
	<4F9BE61A.90807@gravedo.de> <201204291356.06717.psonek2@seznam.cz>
	<4FA24AEB.8090708@gravedo.de>
	<87lil9wm4j.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <4FA27CE4.5070006@gravedo.de>

On 03.05.2012 13:16, Neil Jerram wrote:
> Simon Busch <morphis at gravedo.de> writes:
> 
>> On 29.04.2012 15:56, Radek Polak wrote:
>>> On Saturday 28 April 2012 12:44:10 Simon Busch wrote:
>>>
>>>> As far as I got the details from that thread there is no clean solution
>>>> for the disconnecting modem just a workaround as Radek implemented it
>>>> for qtmoko, right?
>>>
>>> Yup, i found no clean solution. Currently i switch to 2G on modem 
>>> initialization - this seems to help + after 30s of unsuccessful reads i 
>>> restart whole QtMoko + some other hacks:
>>
>> Yeah, 2G seems to solve the problem. I tried with limiting the radio
>> access to 2G only and never got a disconnect from the modem.
>>
>> Disabling 3G generally for the GTA04 is nothing we really want. So I am
>> thinking about just limiting the radio access to 2G when dealing with
>> calls. When a call is incoming or outgoing we have to switch the radio
>> access to 2G and switch it back to ANY when we're done with call
>> handling. Let's see if this works not just in theory ...
> 
> Nice idea!
> 
> In parallel, though, I believe Nikolaus was going to look into the
> possibilities for modem firmware upgrade.  Nikolaus, is there any news
> on that?

Both sounds a lot bettern than doing such ugly workarounds in userspace.
Can someone investigate the kernel approach Radek mentioned?

Upgrading the firmware would another nice way but should only be done
when it's safe and doesn't break the modem.

regards,
Simon

-- 
Simon Busch - http://mm.gravedo.de/blog/

From andreas at kemnade.info  Sat May  5 16:07:27 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Sat, 5 May 2012 16:07:27 +0200
Subject: [Gta04-owner] bad wifi performance in kernel 3.3 but not with 2.6.32
Message-ID: <20120505160727.15782d2f@kemnade.info>

Hi,

I found out that I get
around 1700KByte/s with kernel 2.6.32 and 2.8V for wifi at one place.
With Kernel 3.3 I get only 70KByte/s at the same place. iwconfig
shows 5.5 Mb/s then.
Some days ago I tried 1m away from another accesspoint with kernel 3.2,
also only 70KByte/s.

Anyone knows what is going wrong there? What is the way to debug that?

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120505/c894c34d/attachment.bin>

From christoph.mair at gmail.com  Sat May  5 20:09:13 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Sat, 5 May 2012 20:09:13 +0200
Subject: [Gta04-owner] bad wifi performance in kernel 3.3 but not with
	2.6.32
In-Reply-To: <20120505160727.15782d2f@kemnade.info>
References: <20120505160727.15782d2f@kemnade.info>
Message-ID: <CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>

On Sat, May 5, 2012 at 4:07 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
> Hi,
>
> I found out that I get
> around 1700KByte/s with kernel 2.6.32 and 2.8V for wifi at one place.
> With Kernel 3.3 I get only 70KByte/s at the same place. iwconfig
> shows 5.5 Mb/s then.
> Some days ago I tried 1m away from another accesspoint with kernel 3.2,
> also only 70KByte/s.
>
> Anyone knows what is going wrong there? What is the way to debug that?

Maybe just the firmware changed? This should be rather easily
detectable. Other differences require more work to find them.

Christoph

From guru at unixarea.de  Sat May  5 20:34:53 2012
From: guru at unixarea.de (Matthias Apitz)
Date: Sat, 5 May 2012 20:34:53 +0200
Subject: [Gta04-owner] bad wifi performance in kernel 3.3 but not with
 2.6.32
In-Reply-To: <CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>
References: <20120505160727.15782d2f@kemnade.info>
	<CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>
Message-ID: <20120505183452.GA1106@tiny>

El d?a Saturday, May 05, 2012 a las 08:09:13PM +0200, Christoph Mair escribi?:

> On Sat, May 5, 2012 at 4:07 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
> > Hi,
> >
> > I found out that I get
> > around 1700KByte/s with kernel 2.6.32 and 2.8V for wifi at one place.
> > With Kernel 3.3 I get only 70KByte/s at the same place. iwconfig
> > shows 5.5 Mb/s then.
> > Some days ago I tried 1m away from another accesspoint with kernel 3.2,
> > also only 70KByte/s.
> >
> > Anyone knows what is going wrong there? What is the way to debug that?

How do you measure the 1700KByte/s and 70KByte/s exactly? Such diff
reminds me more being a routing or DNS issue...

	matthias
-- 
Matthias Apitz
e <guru at unixarea.de> - w http://www.unixarea.de/
UNIX since V7 on PDP-11, UNIX on mainframe since ESER 1055 (IBM /370)
UNIX on x86 since SVR4.2 UnixWare 2.1.2, FreeBSD since 2.2.5

From andreas at kemnade.info  Sun May  6 08:14:22 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Sun, 06 May 2012 08:14:22 +0200
Subject: [Gta04-owner] bad wifi performance in kernel 3.3 but not with
 2.6.32
In-Reply-To: <20120505183452.GA1106@tiny>
References: <20120505160727.15782d2f@kemnade.info>
	<CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>
	<20120505183452.GA1106@tiny>
Message-ID: <1336284862.2565.2435.camel@localhost>

Hi,

On Sat, 2012-05-05 at 20:34 +0200, Matthias Apitz wrote:
> El d?a Saturday, May 05, 2012 a las 08:09:13PM +0200, Christoph Mair escribi?:
> 
> > On Sat, May 5, 2012 at 4:07 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
> > > Hi,
> > >
> > > I found out that I get
> > > around 1700KByte/s with kernel 2.6.32 and 2.8V for wifi at one place.
> > > With Kernel 3.3 I get only 70KByte/s at the same place. iwconfig
> > > shows 5.5 Mb/s then.
> > > Some days ago I tried 1m away from another accesspoint with kernel 3.2,
> > > also only 70KByte/s.
> > >
> > > Anyone knows what is going wrong there? What is the way to debug that?
> 
> How do you measure the 1700KByte/s and 70KByte/s exactly? Such diff
> reminds me more being a routing or DNS issue...
> 
ssh ip_of_some_host_in_the_lan "cat /dev/zero" | buffer -z 128k
>/dev/null
and then I waited for the values to stabilize.
apt-get shows similar low rates with 3.2 and 3.3 even
when big files are transferred.
I always look also at the number of bytes transferred. 

Greetings
Andreas Kemnade



From andreas at kemnade.info  Sun May  6 10:12:46 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Sun, 06 May 2012 10:12:46 +0200
Subject: [Gta04-owner] bad wifi performance in kernel 3.3 but not with
 2.6.32
In-Reply-To: <CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>
References: <20120505160727.15782d2f@kemnade.info>
	<CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>
Message-ID: <1336291966.2565.2805.camel@localhost>

Hi,

On Sat, 2012-05-05 at 20:09 +0200, Christoph Mair wrote:
> On Sat, May 5, 2012 at 4:07 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
> > Hi,
> >
> > I found out that I get
> > around 1700KByte/s with kernel 2.6.32 and 2.8V for wifi at one place.
> > With Kernel 3.3 I get only 70KByte/s at the same place. iwconfig
> > shows 5.5 Mb/s then.
> > Some days ago I tried 1m away from another accesspoint with kernel 3.2,
> > also only 70KByte/s.
> >
> > Anyone knows what is going wrong there? What is the way to debug that?
> 
> Maybe just the firmware changed? This should be rather easily
> detectable. Other differences require more work to find them.
> 
Userspace (so also /lib/firmware) was same (mostly the basic LXDE rootfs
with some additions).
The LXDE rootfs and also qtmoko ships with
the debian libertas-firmware package.
I tried also the firmware from the firmware-libertas package in debian
unstable/backports after that. It does not change things much.

Greetings
Andreas Kemnade



From guru at unixarea.de  Sun May  6 10:23:53 2012
From: guru at unixarea.de (Matthias Apitz)
Date: Sun, 6 May 2012 10:23:53 +0200
Subject: [Gta04-owner] bad wifi performance in kernel 3.3 but not with
 2.6.32
In-Reply-To: <1336291966.2565.2805.camel@localhost>
References: <20120505160727.15782d2f@kemnade.info>
	<CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>
	<1336291966.2565.2805.camel@localhost>
Message-ID: <20120506082352.GA2763@tinyCurrent>

El d?a Sunday, May 06, 2012 a las 10:12:46AM +0200, Andreas Kemnade escribi?:

> > > Anyone knows what is going wrong there? What is the way to debug that?
> > 
> > Maybe just the firmware changed? This should be rather easily
> > detectable. Other differences require more work to find them.
> > 
> Userspace (so also /lib/firmware) was same (mostly the basic LXDE rootfs
> with some additions).
> The LXDE rootfs and also qtmoko ships with
> the debian libertas-firmware package.
> I tried also the firmware from the firmware-libertas package in debian
> unstable/backports after that. It does not change things much.

Hi,

Pls check this on all involved points:
-- netmasks and MTU on all interfaces
-- routings
-- with tcpdump if there are misrouted pkgs and redirects

What gives PING from the nearest IP to the FR's Wifi IP?
Is this with WEP or WPA? Try changing this to the other mode.

HIH

	matthias
-- 
Matthias Apitz
t +49-89-61308 351 - f +49-89-61308 399 - m +49-170-4527211
e <guru at unixarea.de> - w http://www.unixarea.de/
UNIX since V7 on PDP-11 | UNIX on mainframe since ESER 1055 (IBM /370)
UNIX on x86 since SVR4.2 UnixWare 2.1.2 | FreeBSD since 2.2.5

From andreas at kemnade.info  Sun May  6 16:08:27 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Sun, 06 May 2012 16:08:27 +0200
Subject: [Gta04-owner] bad wifi performance in kernel 3.3 but not with
 2.6.32
In-Reply-To: <20120506082352.GA2763@tinyCurrent>
References: <20120505160727.15782d2f@kemnade.info>
	<CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>
	<1336291966.2565.2805.camel@localhost>
	<20120506082352.GA2763@tinyCurrent>
Message-ID: <1336313307.3586.421.camel@localhost>

Hi,

On Sun, 2012-05-06 at 10:23 +0200, Matthias Apitz wrote:
> El d?a Sunday, May 06, 2012 a las 10:12:46AM +0200, Andreas Kemnade escribi?:
> 
> > > > Anyone knows what is going wrong there? What is the way to debug that?
> > > 
> > > Maybe just the firmware changed? This should be rather easily
> > > detectable. Other differences require more work to find them.
> > > 
> > Userspace (so also /lib/firmware) was same (mostly the basic LXDE rootfs
> > with some additions).
> > The LXDE rootfs and also qtmoko ships with
> > the debian libertas-firmware package.
> > I tried also the firmware from the firmware-libertas package in debian
> > unstable/backports after that. It does not change things much.
> 
> Hi,
> 
> Pls check this on all involved points:
> -- netmasks and MTU on all interfaces
> -- routings
> -- with tcpdump if there are misrouted pkgs and redirects
> 

Why should this also affect the bitrate iwconfig shows (with 3.3 it
stays at 5.5Mb/s)? 
But I doublechecked that points anyways again to be sure. 
The unusual low rate was in different networks. And other devices work
there. So this points can be ruled out.

> What gives PING from the nearest IP to the FR's Wifi IP?

--- 192.168.9.119 ping statistics ---
33 packets transmitted, 33 packets received, 0% packet loss
round-trip min/avg/max = 4.7/12.4/20.5 ms


> Is this with WEP or WPA? Try changing this to the other mode.
> 
That was with WPA. But I'll set up another accesspoint for testing soon.

Greetings
Andreas Kemnade


From openmoko at ginguppin.de  Sun May  6 19:14:45 2012
From: openmoko at ginguppin.de (arne anka)
Date: Sun, 06 May 2012 19:14:45 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
Message-ID: <op.wdwcuv2i41evsb@ginguppin>

> But watch yourself how we think it can look like:
>
> 	http://youtu.be/WM94%5fR5eKcc
>
> There is also a new video showing a comparison with some other keyboards:
>
> 	http://youtu.be/wGASnE1zGh4

both videos are "currently unavailable". does anyone know, what's wrong?

From gta04 at metalstrolche.de  Sun May  6 19:22:50 2012
From: gta04 at metalstrolche.de (Stefan Wildemann)
Date: Sun, 06 May 2012 19:22:50 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <op.wdwcuv2i41evsb@ginguppin>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<op.wdwcuv2i41evsb@ginguppin>
Message-ID: <4FA6B36A.60603@metalstrolche.de>

Am 06.05.2012 19:14, schrieb arne anka:
>> But watch yourself how we think it can look like:
>>
>>     http://youtu.be/WM94%5fR5eKcc
>>
>> There is also a new video showing a comparison with some other
>> keyboards:
>>
>>     http://youtu.be/wGASnE1zGh4
>
> both videos are "currently unavailable". does anyone know, what's wrong?
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

NACK - Both are available for me (from Germany).
Greetings,
Stefan

From butrus.butrus at gmail.com  Mon May  7 08:47:17 2012
From: butrus.butrus at gmail.com (Butrus Damaskus)
Date: Mon, 7 May 2012 08:47:17 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <B508BBBD-F3F3-48D7-9B5B-0C4F5E018E16@goldelico.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<CALM6Ch_hKeeB4WJw=jHCpEFq67Xg9VtM_YZiYG6K3iw7+JYpcg@mail.gmail.com>
	<B508BBBD-F3F3-48D7-9B5B-0C4F5E018E16@goldelico.com>
Message-ID: <CALM6Ch8Jq=JMfxOjpNT3en7n0YPeu+Gbe+e3rXmJv1TvEYOK9w@mail.gmail.com>

On Thu, May 3, 2012 at 10:26 AM, Dr. H. Nikolaus Schaller
<hns at goldelico.com> wrote:
>
> Am 03.05.2012 um 10:03 schrieb Butrus Damaskus:
>
>> Very interresting, but I'm a little bit puzzled as how it should work
>> in the real life...
>
> That is a task for further study and optimization. We have just built one unit
> and we need some 3D printed cases to do a better evaluation.

Ok, so let's start discuss that ;-).

My question was if you foresee the keyboard to be detachable somehow
(than placing it on the rear side
of OpenMoko would make sense) or whether there would be some mechanism
to "open" the keyboard.
Otherwise I cannot understand how the keboard on the backside would be used...

> Some more general thought: I am not sure how we should handle such ideas and
> how this community sees its role.
>
> Should we announce/publish immediately as soon as we have a prototype?

Yes, please ;-)

> And ask the community for comments and participation in improving it?

I would say yes :)

> Or should we silently
> finish some design and work with beta testers not on this list...
>
> I think this is a fundamental difference to open software development. There we can
> have a stable, a testing and an experimental branch and everyone can download
> the one he/she likes. With hardware we have to build physical protoypes and make
> experiments (which are not for free) to find out what a good solution is.
>
> Nikolaus



>>
>> On Tue, May 1, 2012 at 6:06 PM, Dr. H. Nikolaus Schaller
>> <hns at goldelico.com> wrote:
>>> Hi,
>>>
>>> we have developed a prototype for a 80 button QWERTY keyboard PCB
>>> that could eventually be connected/integrated into the GTA04. It should fit
>>> into a specially designed battery cover so that you can easily stow it away
>>> if not needed. Such battery covers could be produced individually through
>>> 3D printing solving the issue of manageing and stocking 20 different key
>>> layouts.
>>>
>>> But watch yourself how we think it can look like:
>>>
>>> ? ? ? ?http://youtu.be/WM94%5fR5eKcc
>>>
>>> There is also a new video showing a comparison with some other keyboards:
>>>
>>> ? ? ? ?http://youtu.be/wGASnE1zGh4
>>>
>>> Pleas tell us if you like this idea and what you would like to pay for such an
>>> extension unit through the Wishlist function of our shop:
>>>
>>> ? ? ? ?http://www.handheld-linux.com/wiki.php?page=GTA04%3AKeyboard
>>>
>>> Two issues are still to be developed:
>>>
>>> a) how to reliably connect it to the GTA04 PCB (soldering copper wires or a
>>> ? ?FFCs is a little difficult so it should have a tiny, flexible but robust B2B cable).
>>>
>>> ? ?Maybe, we can use a micro-USB socket or similar (we need to connect 6 wires).
>>> ? ?This may also need a redesign of the GTA04 board (for a nice plug)
>>>
>>> b) design a 3D printable case with key-caps that is robust enough
>>>
>>> If you want to support us for developing this idea, please give us a kickstart
>>> donation.
>>>
>>>
>>> Nikolaus
>>>
>>> PS: the keyboard driver for the TCA8418 is already part of Linux 3.3 - and has
>>> been backported to the 2.6.32-kernel. This has been tested to work on a
>>> BeagleBoard XM.
>>>
>>> http://git.goldelico.com/?p=gta04-kernel.git;a=commit;h=f19d5c430458bbce8955bc9e04dd161f6a80347d
>>>
>>> It just needs platform data in the board file:
>>>
>>> http://git.goldelico.com/?p=gta04-kernel.git;a=blobdiff;f=arch/arm/mach-omap2/board-omap3gta04.c;h=8a7e4b0803920f635e7101bfbd5a60b6b84b1107;hp=3e49efef2de0b42cd419a46a9cd45448fd04a44c;hb=4b2de3db742abce9212c1af2cc576e2a3a64b0d9;hpb=1d7c6b5f043661621ec374d96c3c4a4454f9bb7b
>>> _______________________________________________
>>> Gta04-owner mailing list
>>> Gta04-owner at goldelico.com
>>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

From rah at settrans.net  Mon May  7 13:13:45 2012
From: rah at settrans.net (Bob Ham)
Date: Mon, 07 May 2012 12:13:45 +0100
Subject: [Gta04-owner] Home printed cases
In-Reply-To: <20120503142349.4c7d4b96@kemnade.info>
References: <1335950495.32304.134.camel@myrtle.6gnip.net>
	<20120503142349.4c7d4b96@kemnade.info>
Message-ID: <1336389225.7305.23.camel@myrtle.6gnip.net>

On Thu, 2012-05-03 at 14:23 +0200, Andreas Kemnade wrote:

> you have seen my experiments using my reprap, a 0.5mm nozzle with the back part?

Yes, I thought I had seen a reprap'd case in the past, thanks.


> They look a bit better I think. 

They do, yes.  I think part of the problem with my print is that the
Cupcake is not the best printer.  It has quite an old, large nozzle,
it's pretty rickety, etc.  As I mentioned, I'll be building my own
reprap so perhaps if the prints are of high enough quality it won't be
necessary to start a new design from scratch.

> I still think that redesigning the thing would be a good idea, but then I think
> it should be parameterized, so that you can easily change thickness of walls
> and so on.

I'll be happy to test any designs you come up with ;-)

-- 
Bob Ham <rah at settrans.net>

for (;;) { ++pancakes; }
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120507/489aa503/attachment.bin>

From hns at goldelico.com  Mon May  7 15:31:11 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Mon, 7 May 2012 15:31:11 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <CALM6Ch8Jq=JMfxOjpNT3en7n0YPeu+Gbe+e3rXmJv1TvEYOK9w@mail.gmail.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<CALM6Ch_hKeeB4WJw=jHCpEFq67Xg9VtM_YZiYG6K3iw7+JYpcg@mail.gmail.com>
	<B508BBBD-F3F3-48D7-9B5B-0C4F5E018E16@goldelico.com>
	<CALM6Ch8Jq=JMfxOjpNT3en7n0YPeu+Gbe+e3rXmJv1TvEYOK9w@mail.gmail.com>
Message-ID: <4ACBDEE9-7D94-4A69-829C-C74DDC37B06A@goldelico.com>


Am 07.05.2012 um 08:47 schrieb Butrus Damaskus:

> On Thu, May 3, 2012 at 10:26 AM, Dr. H. Nikolaus Schaller
> <hns at goldelico.com> wrote:
>> 
>> Am 03.05.2012 um 10:03 schrieb Butrus Damaskus:
>> 
>>> Very interresting, but I'm a little bit puzzled as how it should work
>>> in the real life...
>> 
>> That is a task for further study and optimization. We have just built one unit
>> and we need some 3D printed cases to do a better evaluation.
> 
> Ok, so let's start discuss that ;-).
> 
> My question was if you foresee the keyboard to be detachable somehow
> (than placing it on the rear side
> of OpenMoko would make sense) or whether there would be some mechanism
> to "open" the keyboard.
> Otherwise I cannot understand how the keboard on the backside would be used...

The idea is that the keyboard is integrated in the back battery cover. And if you remove
it, both parts (main body and battery cover/keyboard) are connected through a small
ribbon cable.

I.e. you remove the keyboard, fold it by 180 degrees so that the keys show in the same
direction as the display and you can us the display in landscape format.

If that works out with the snap mechanism of the battery cover and/or if it needs
some physical connection between both parts has to be worked out.

A complete alternative could be a battery cover with a hinge. Like this sketch:

	http://download.goldelico.com/gta04/images/Scan-120507143506.jpeg

Your phantasy and creativity is asked for...

> 
>> Some more general thought: I am not sure how we should handle such ideas and
>> how this community sees its role.
>> 
>> Should we announce/publish immediately as soon as we have a prototype?
> 
> Yes, please ;-)
> 
>> And ask the community for comments and participation in improving it?
> 
> I would say yes :)
> 
>> Or should we silently
>> finish some design and work with beta testers not on this list...
>> 
>> I think this is a fundamental difference to open software development. There we can
>> have a stable, a testing and an experimental branch and everyone can download
>> the one he/she likes. With hardware we have to build physical protoypes and make
>> experiments (which are not for free) to find out what a good solution is.
>> 
>> Nikolaus
> 
> 
> 
>>> 
>>> On Tue, May 1, 2012 at 6:06 PM, Dr. H. Nikolaus Schaller
>>> <hns at goldelico.com> wrote:
>>>> Hi,
>>>> 
>>>> we have developed a prototype for a 80 button QWERTY keyboard PCB
>>>> that could eventually be connected/integrated into the GTA04. It should fit
>>>> into a specially designed battery cover so that you can easily stow it away
>>>> if not needed. Such battery covers could be produced individually through
>>>> 3D printing solving the issue of manageing and stocking 20 different key
>>>> layouts.
>>>> 
>>>> But watch yourself how we think it can look like:
>>>> 
>>>>        http://youtu.be/WM94%5fR5eKcc
>>>> 
>>>> There is also a new video showing a comparison with some other keyboards:
>>>> 
>>>>        http://youtu.be/wGASnE1zGh4
>>>> 
>>>> Pleas tell us if you like this idea and what you would like to pay for such an
>>>> extension unit through the Wishlist function of our shop:
>>>> 
>>>>        http://www.handheld-linux.com/wiki.php?page=GTA04%3AKeyboard
>>>> 
>>>> Two issues are still to be developed:
>>>> 
>>>> a) how to reliably connect it to the GTA04 PCB (soldering copper wires or a
>>>>    FFCs is a little difficult so it should have a tiny, flexible but robust B2B cable).
>>>> 
>>>>    Maybe, we can use a micro-USB socket or similar (we need to connect 6 wires).
>>>>    This may also need a redesign of the GTA04 board (for a nice plug)
>>>> 
>>>> b) design a 3D printable case with key-caps that is robust enough
>>>> 
>>>> If you want to support us for developing this idea, please give us a kickstart
>>>> donation.
>>>> 
>>>> 
>>>> Nikolaus
>>>> 
>>>> PS: the keyboard driver for the TCA8418 is already part of Linux 3.3 - and has
>>>> been backported to the 2.6.32-kernel. This has been tested to work on a
>>>> BeagleBoard XM.
>>>> 
>>>> http://git.goldelico.com/?p=gta04-kernel.git;a=commit;h=f19d5c430458bbce8955bc9e04dd161f6a80347d
>>>> 
>>>> It just needs platform data in the board file:
>>>> 
>>>> http://git.goldelico.com/?p=gta04-kernel.git;a=blobdiff;f=arch/arm/mach-omap2/board-omap3gta04.c;h=8a7e4b0803920f635e7101bfbd5a60b6b84b1107;hp=3e49efef2de0b42cd419a46a9cd45448fd04a44c;hb=4b2de3db742abce9212c1af2cc576e2a3a64b0d9;hpb=1d7c6b5f043661621ec374d96c3c4a4454f9bb7b
>>>> _______________________________________________
>>>> Gta04-owner mailing list
>>>> Gta04-owner at goldelico.com
>>>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>>> _______________________________________________
>>> Gta04-owner mailing list
>>> Gta04-owner at goldelico.com
>>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>> 
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner


From guru at unixarea.de  Mon May  7 15:44:00 2012
From: guru at unixarea.de (Matthias Apitz)
Date: Mon, 7 May 2012 15:44:00 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <4ACBDEE9-7D94-4A69-829C-C74DDC37B06A@goldelico.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<CALM6Ch_hKeeB4WJw=jHCpEFq67Xg9VtM_YZiYG6K3iw7+JYpcg@mail.gmail.com>
	<B508BBBD-F3F3-48D7-9B5B-0C4F5E018E16@goldelico.com>
	<CALM6Ch8Jq=JMfxOjpNT3en7n0YPeu+Gbe+e3rXmJv1TvEYOK9w@mail.gmail.com>
	<4ACBDEE9-7D94-4A69-829C-C74DDC37B06A@goldelico.com>
Message-ID: <20120507134359.GA2556@tiny>

El d?a Monday, May 07, 2012 a las 03:31:11PM +0200, Dr. H. Nikolaus Schaller escribi?:

> The idea is that the keyboard is integrated in the back battery cover. And if you remove
> it, both parts (main body and battery cover/keyboard) are connected through a small
> ribbon cable.
> 
> I.e. you remove the keyboard, fold it by 180 degrees so that the keys show in the same
> direction as the display and you can us the display in landscape format.
> 
> If that works out with the snap mechanism of the battery cover and/or if it needs
> some physical connection between both parts has to be worked out.
> 

I don't want to be negative (just reaistic), the list of disadvantages
of such a solution (keyboard in battery cover, or with hinge) is:

1) works only on desktop or table surface, i.e. not while walking or
standing;

2) open the battery cover many more times (as for hard reset with battery
lift) will break the case soon;

3) you are forced to use landscape applications, while most of the apps
today are in portait;

4) perhaps your battery will fall out (if not secured with something) in
a lot of cases, esp. if it is the most important situation :-)

to be continued...

	matthias

-- 
Matthias Apitz
e <guru at unixarea.de> - w http://www.unixarea.de/
UNIX since V7 on PDP-11, UNIX on mainframe since ESER 1055 (IBM /370)
UNIX on x86 since SVR4.2 UnixWare 2.1.2, FreeBSD since 2.2.5

From ed.kapitein at tno.nl  Mon May  7 16:07:41 2012
From: ed.kapitein at tno.nl (ed.kapitein at tno.nl)
Date: Mon, 07 May 2012 16:07:41 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <20120507134359.GA2556@tiny>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<4ACBDEE9-7D94-4A69-829C-C74DDC37B06A@goldelico.com>
	<20120507134359.GA2556@tiny>
Message-ID: <1753127.WRdF5SGY4M@lp-00079.dt01.tno.nl>

On Monday 07 May 2012 15:44:00 Matthias Apitz wrote:
> El d?a Monday, May 07, 2012 a las 03:31:11PM +0200, Dr. H. Nikolaus Schaller 
escribi?:
> > The idea is that the keyboard is integrated in the back battery cover. And
> > if you remove it, both parts (main body and battery cover/keyboard) are
> > connected through a small ribbon cable.
> > 
> > I.e. you remove the keyboard, fold it by 180 degrees so that the keys show
> > in the same direction as the display and you can us the display in
> > landscape format.
> > 
> > If that works out with the snap mechanism of the battery cover and/or if
> > it needs some physical connection between both parts has to be worked
> > out.
> 
> I don't want to be negative (just reaistic), the list of disadvantages
> of such a solution (keyboard in battery cover, or with hinge) is:
> 
> 1) works only on desktop or table surface, i.e. not while walking or
> standing;
> 
> 2) open the battery cover many more times (as for hard reset with battery
> lift) will break the case soon;
> 
> 3) you are forced to use landscape applications, while most of the apps
> today are in portait;
> 
> 4) perhaps your battery will fall out (if not secured with something) in
> a lot of cases, esp. if it is the most important situation :-)
> 
> to be continued...
> 
> 	matthias

If thinking out loud is allowed...
If you move the battery  *inside* the keyboard part and make the hinge a bit 
bigger, you could use it as a screen protector if you fold the keyboard in 
front of the display.
And if you fold it the otherway, it could fit in what currently is the battery 
area.
Just my 2 cents..

Kind regards,
Ed

This e-mail and its contents are subject to the DISCLAIMER at http://www.tno.nl/emaildisclaimer


From martin.jansa at gmail.com  Mon May  7 16:35:38 2012
From: martin.jansa at gmail.com (Martin Jansa)
Date: Mon, 7 May 2012 16:35:38 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <20120507134359.GA2556@tiny>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<CALM6Ch_hKeeB4WJw=jHCpEFq67Xg9VtM_YZiYG6K3iw7+JYpcg@mail.gmail.com>
	<B508BBBD-F3F3-48D7-9B5B-0C4F5E018E16@goldelico.com>
	<CALM6Ch8Jq=JMfxOjpNT3en7n0YPeu+Gbe+e3rXmJv1TvEYOK9w@mail.gmail.com>
	<4ACBDEE9-7D94-4A69-829C-C74DDC37B06A@goldelico.com>
	<20120507134359.GA2556@tiny>
Message-ID: <CA+chaQcYjeJ_W+Bk4jGw-KXPKkso7VhOFVtbAhw0P9HYcTN+DQ@mail.gmail.com>

On Mon, May 7, 2012 at 3:44 PM, Matthias Apitz <guru at unixarea.de> wrote:
> El d?a Monday, May 07, 2012 a las 03:31:11PM +0200, Dr. H. Nikolaus Schaller escribi?:
>
>> The idea is that the keyboard is integrated in the back battery cover. And if you remove
>> it, both parts (main body and battery cover/keyboard) are connected through a small
>> ribbon cable.
>>
>> I.e. you remove the keyboard, fold it by 180 degrees so that the keys show in the same
>> direction as the display and you can us the display in landscape format.
>>
>> If that works out with the snap mechanism of the battery cover and/or if it needs
>> some physical connection between both parts has to be worked out.
>>
>
> I don't want to be negative (just reaistic), the list of disadvantages
> of such a solution (keyboard in battery cover, or with hinge) is:
>
> 1) works only on desktop or table surface, i.e. not while walking or
> standing;
>
> 2) open the battery cover many more times (as for hard reset with battery
> lift) will break the case soon;
>
> 3) you are forced to use landscape applications, while most of the apps
> today are in portait;
>
> 4) perhaps your battery will fall out (if not secured with something) in
> a lot of cases, esp. if it is the most important situation :-)
>
> to be continued...
>
> ? ? ? ?matthias

I agree that physical connection between both parts is the most
important part which makes
such solution to work out of the box (and not like any other small BT
or usb keyboard).

Most of the time when I'm using hw keyboard on Sharp spitz or n900 I
hold device between
and type with just two thumbs and if needed I can hold it safely with
just one hand and open door
or something.

I don't know about other people, but I've never used keyboard with one
hand while
holding device in other hand (like sometimes shown in that comparison video).

I would even consider to give up overall oval shape in order to make
some sliding mechanism
easier to implement (I cannot imagine how to slide keyboard from
current shape if we want to use
whole lenght of device for keyboard part which I think we want/need).

just my 2c..

Cheers,

From ed at kapitein.org  Mon May  7 16:44:28 2012
From: ed at kapitein.org (ed at kapitein.org)
Date: Mon, 07 May 2012 16:44:28 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <20120507134359.GA2556@tiny>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<4ACBDEE9-7D94-4A69-829C-C74DDC37B06A@goldelico.com>
	<20120507134359.GA2556@tiny>
Message-ID: <2232380.TrNeQQoHO2@lp-00079.dt01.tno.nl>

On Monday 07 May 2012 15:44:00 Matthias Apitz wrote:
> El d?a Monday, May 07, 2012 a las 03:31:11PM +0200, Dr. H. Nikolaus Schaller 
escribi?:
> > The idea is that the keyboard is integrated in the back battery cover. And
> > if you remove it, both parts (main body and battery cover/keyboard) are
> > connected through a small ribbon cable.
> > 
> > I.e. you remove the keyboard, fold it by 180 degrees so that the keys show
> > in the same direction as the display and you can us the display in
> > landscape format.
> > 
> > If that works out with the snap mechanism of the battery cover and/or if
> > it needs some physical connection between both parts has to be worked
> > out.
> 
> I don't want to be negative (just reaistic), the list of disadvantages
> of such a solution (keyboard in battery cover, or with hinge) is:
> 
> 1) works only on desktop or table surface, i.e. not while walking or
> standing;
> 
> 2) open the battery cover many more times (as for hard reset with battery
> lift) will break the case soon;
> 
> 3) you are forced to use landscape applications, while most of the apps
> today are in portait;
> 
> 4) perhaps your battery will fall out (if not secured with something) in
> a lot of cases, esp. if it is the most important situation :-)
> 
> to be continued...
> 
> 	matthias


If thinking out loud is allowed...
If you move the battery  inside the keyboard part and make the hinge a bit 
bigger, you could use it as a screen protector if you fold the keyboard in 
front of the display.
And if you fold it the otherway, it could fit in what currently is the battery 
area.
Just my 2 cents..

Kind regards,
Ed
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120507/71124513/attachment-0001.html>

From psonek2 at seznam.cz  Mon May  7 20:29:06 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Mon, 7 May 2012 18:29:06 +0000
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <4ACBDEE9-7D94-4A69-829C-C74DDC37B06A@goldelico.com>
References: <5C4A6ED2-1F62-4FAA-8168-D39F59742F71@goldelico.com>
	<CALM6Ch8Jq=JMfxOjpNT3en7n0YPeu+Gbe+e3rXmJv1TvEYOK9w@mail.gmail.com>
	<4ACBDEE9-7D94-4A69-829C-C74DDC37B06A@goldelico.com>
Message-ID: <201205071829.06615.psonek2@seznam.cz>

On Monday 07 May 2012 13:31:11 Dr. H. Nikolaus Schaller wrote:

> The idea is that the keyboard is integrated in the back battery cover. And
> if you remove it, both parts (main body and battery cover/keyboard) are
> connected through a small ribbon cable.
> 
> I.e. you remove the keyboard, fold it by 180 degrees so that the keys show
> in the same direction as the display and you can us the display in
> landscape format.
> 
> If that works out with the snap mechanism of the battery cover and/or if it
> needs some physical connection between both parts has to be worked out.
> 
> A complete alternative could be a battery cover with a hinge. Like this
> sketch:
> 
> 	http://download.goldelico.com/gta04/images/Scan-120507143506.jpeg
> 
> Your phantasy and creativity is asked for...

Maybe we can just copy motorola backflip - check videos on youtube, it has big 
keyboard on the backside - it looks very comfortable and mechanically one of 
the easier solutions.

Btw i have now fullscreen sw keyboard in QtMoko. It has big keys and is very 
comfortable to use. I will post video later...

Regards

Radek

From eric.andre at rican.ch  Mon May  7 21:35:13 2012
From: eric.andre at rican.ch (Eric =?ISO-8859-1?Q?Andr=E9?=)
Date: Mon, 07 May 2012 21:35:13 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
References: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
Message-ID: <1336419313.2490.12.camel@Nokia-N900>


I'd love a GTA with a keybord.

If I dare, before it's too late for such an idea and if I am not the only one who would like that:

Keeping the keyboard behind the device, using, let's say six fingers, three of each hand (=six keys) for the whole alphabet?

Inspired from binary codes, it'd make 63 possible combinations.

Longer to learn but very easy then. And we could type while looking at the screen instead of searching for tiny keys to tap on.

Tell me if it's totally stupid or just a bit...
 Eric
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120507/c9dde2b2/attachment.html>

From rah at settrans.net  Mon May  7 21:42:22 2012
From: rah at settrans.net (Bob Ham)
Date: Mon, 07 May 2012 20:42:22 +0100
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <1336419313.2490.12.camel@Nokia-N900>
References: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
	<1336419313.2490.12.camel@Nokia-N900>
Message-ID: <1336419742.7305.167.camel@myrtle.6gnip.net>

On Mon, 2012-05-07 at 21:35 +0200, Eric Andr? wrote:

> Keeping the keyboard behind the device, using, let's say six fingers, three of each hand (=six keys) for the whole alphabet?

> Tell me if it's totally stupid or just a bit...

Neither:

https://en.wikipedia.org/wiki/Chorded_keyboard

A chorded keyboard could be useful on a phone.  In fact, chorded and
qwerty keyboards wouldn't be mutually exclusive :-)

-- 
Bob Ham <rah at settrans.net>

for (;;) { ++pancakes; }
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: This is a digitally signed message part
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120507/d687697f/attachment.bin>

From george.refseth at arxi.no  Mon May  7 21:48:18 2012
From: george.refseth at arxi.no (George Refseth)
Date: Mon, 07 May 2012 21:48:18 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <1336419313.2490.12.camel@Nokia-N900>
References: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
	<1336419313.2490.12.camel@Nokia-N900>
Message-ID: <4FA82702.5050906@arxi.no>

On 07/05/12 21:35, Eric Andr? wrote:
>
>
> I'd love a GTA with a keybord.
>
> If I dare, before it's too late for such an idea and if I am not the 
> only one who would like that:
>
> Keeping the keyboard behind the device, using, let's say six fingers, 
> three of each hand (=six keys) for the whole alphabet?
>
> Inspired from binary codes, it'd make 63 possible combinations.
>
> Longer to learn but very easy then. And we could type while looking at 
> the screen instead of searching for tiny keys to tap on.
>
> Tell me if it's totally stupid or just a bit...
> Eric
>
>
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner
Why not four sensors for each hand sensitive to directional movement as 
well as pressure, that way you multiply the possibilities.
Like one eight-touch pane.

regards,
George
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120507/04f2c2cd/attachment.html>

From andreas at kemnade.info  Tue May  8 17:23:26 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Tue, 8 May 2012 17:23:26 +0200
Subject: [Gta04-owner] USB-OTG hacking guide / how to be usb host
Message-ID: <20120508172326.067df72c@kemnade.info>

Hi,

I have some experience with the usb stuff now. Perhaps it is interesting for
other people, too, especially since the usb stuff is mostly excluded from
the SoC datasheet. I can also convert parts of the things below to clean patches.

One main difference is that Vbus handling is deeply integrated into the
hardware and many things are not software controlled. Basically the whole
usb otg state machine lives in the SoC in hardware, so you can not set host
role that easily.
The USB part in the SoC steers directly the configuration registers in the
twl4030. But the registers also can be accessed using I2C from the kernel.
Vbus power output can be controled using DRVVBUS in the OTG_CTRL register.
Charger settings override DRVVBUS.

Host with power from usb devices/y-cables (b_host):
Variant 1.
  1. connect the GTA04 to a usb host (id must not be grounded here)
  2. set MUSB_DEVCTL_HR in MUSB_DEVCTL
       (this would normally be done as a reaction of an USB request
       to the control endpoint to start HNP)
  3. Disconnect data lines from host, keep Vbus on
     GTA04 will switch on its own host pulldowns
  4. Connect the data lines to the peripheral
 
Variant 2:
  This mostly emulates some steps from above.
  1. provide power (again, the id pin must not be grounded)
  2. emulate a connection to an usb host by setting
     OTG_CTRL_DPPULLDOWN via i2c
  3. set MUSB_DEVCTL_HR in MUSB_DEVCTL
  4. emulate disconnection from host by clearing
     OTG_CTRL_DPPULLDOWN via i2c
  5. enable pulldown again by setting
     OTG_CTRL_DPPULLDOWN via i2c
     so that the phy can recognize connected devices.
  6. Connect data lines to the peripheral

Host gives out power (a_host):
standard variant:
  - connect usb device by using a cable with the ID pin grounded

If you do not have an usb mini plug with id grounded:
Variant 1
1. connect the openmoko charger to the GTA04 (not to
   the wall socket!)
2. disconnect it. (with 2.6.32 the usb system notices that,
   but not 3.x)
3. connect the device (tried that with usb-a to mini-b cables, and
   adapter usb a socket to usb a socket)

variant 2
   - fiddle with the MUSB_TEST_FORCE_HOST bit in the MUSB_TESTMODE register
     and connect the device. I do not understand that bit fully, so I do not
     give more precise instructions. In kernel 3.x, that bit can be accessed
     through debugfs.

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120508/48e84f27/attachment.bin>

From andreas at kemnade.info  Tue May  8 07:17:11 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Tue, 8 May 2012 07:17:11 +0200
Subject: [Gta04-owner] Home printed cases
In-Reply-To: <1336389225.7305.23.camel@myrtle.6gnip.net>
References: <1335950495.32304.134.camel@myrtle.6gnip.net>
	<20120503142349.4c7d4b96@kemnade.info>
	<1336389225.7305.23.camel@myrtle.6gnip.net>
Message-ID: <20120508071711.5003c35d@kemnade.info>

Hi,

On Mon, 07 May 2012 12:13:45 +0100
Bob Ham <rah at settrans.net> wrote:

> On Thu, 2012-05-03 at 14:23 +0200, Andreas Kemnade wrote:
> 
> > you have seen my experiments using my reprap, a 0.5mm nozzle with the back part?
> 
> Yes, I thought I had seen a reprap'd case in the past, thanks.
> 
> 
> > They look a bit better I think. 
> 
> They do, yes.  I think part of the problem with my print is that the
> Cupcake is not the best printer.  It has quite an old, large nozzle,
> it's pretty rickety, etc.  As I mentioned, I'll be building my own
> reprap so perhaps if the prints are of high enough quality it won't be
> necessary to start a new design from scratch.
> 
AFAIK, the design was adapted to meet the specifications of slyparts, but 
perhaps some walls can be made thicker where size is not that critital so
it can be printed with more 3d printers.
What is the easiest way to change the wall size? using the stl files in blender?
Using the non-stl files in freecad?

> > I still think that redesigning the thing would be a good idea, but then I think
> > it should be parameterized, so that you can easily change thickness of walls
> > and so on.
> 
> I'll be happy to test any designs you come up with ;-)

Well first, I try something very simple, just to tie it to my reprap where
requirements are low. 

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120508/e4db0ade/attachment.bin>

From psonek2 at seznam.cz  Tue May  8 23:45:22 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Tue, 8 May 2012 21:45:22 +0000
Subject: [Gta04-owner] QtMoko's new keyboard
Message-ID: <201205082145.22636.psonek2@seznam.cz>

Hi,
while there was discussion about hw keyboard, i was working on software 
onscreen keyboard with simple goal: implement the best onscreen keyboard in 
the world.

I think i am now very close ;-) The key is to have as big buttons as possible. 
Here is picture and video of how it looks now:

http://activationrecord.net/radekp/pub/keyboard.png
http://www.youtube.com/watch?v=fyN7wS66y_I

It still needs some more work, but it's currently very usable and i am really 
happy how it works. E.g. the video was taken with N900 in my left hand and 
still with the SMS layout i made no mistake.

My plan for now is to finish it up and remove all those 5!! QtMoko input 
methods in favour of this one. I can make installable packages of the old 
methods if anyone is interested. This should be part of v45. For v46 i can try 
to implement customizable layouts and unicode characters.

Regards

Radek

From hermann.schwaerzler at chello.at  Tue May  8 22:38:03 2012
From: hermann.schwaerzler at chello.at (=?ISO-8859-1?Q?Hermann_Schw=E4rzler?=)
Date: Tue, 08 May 2012 22:38:03 +0200
Subject: [Gta04-owner] QtMoko's new keyboard
In-Reply-To: <201205082145.22636.psonek2@seznam.cz>
References: <201205082145.22636.psonek2@seznam.cz>
Message-ID: <4FA9842B.20507@chello.at>

hello Radek!

Radek Polak schrieb:

> while there was discussion about hw keyboard, i was working on software 
> onscreen keyboard with simple goal: implement the best onscreen keyboard in 
> the world.
> 
> I think i am now very close ;-) The key is to have as big buttons as possible. 
> Here is picture and video of how it looks now:

looks cool!
I am not an expert but for me it looks like very close to the best 
onscreen keyboard as well! :-)

you are using both versions in portrait mode. wouldn't it be better to 
use landscape (at least for the qwert version)?

greetings
hermann

From psonek2 at seznam.cz  Wed May  9 05:53:46 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Wed, 09 May 2012 05:53:46 +0200
Subject: [Gta04-owner] QtMoko's new keyboard
In-Reply-To: <4FA9842B.20507@chello.at>
References: <201205082145.22636.psonek2@seznam.cz> <4FA9842B.20507@chello.at>
Message-ID: <9758054.ndlP8m2u8X@rp-core>

On Tuesday, May 08, 2012 10:38:03 PM Hermann Schw?rzler wrote:

> looks cool!
> I am not an expert but for me it looks like very close to the best
> onscreen keyboard as well! :-)
> 
> you are using both versions in portrait mode. wouldn't it be better to
> use landscape (at least for the qwert version)?

You can use both versions also in landscape:

http://scap.linuxtogo.org/files/56fc64d5c8809d6820cb7060536c6c1b.png
http://scap.linuxtogo.org/files/dcb6453bd094bac8b05741b5eb767598.png

Regards

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120509/981afe7e/attachment.html>

From marek.belisko at gmail.com  Thu May 10 21:53:48 2012
From: marek.belisko at gmail.com (Belisko Marek)
Date: Thu, 10 May 2012 21:53:48 +0200
Subject: [Gta04-owner] hmc5883 iio driver find way to mainline
Message-ID: <CAAfyv36=0iuKVnpc3WmnO6VCoh3u=Vet+=eexEtdVAXZ2hYbdQ@mail.gmail.com>

Hi,

last week I was working on adding hmc5883 to actual kernel but one
guy was faster and seems patches are mostly accepted by iio maintainer
(http://thread.gmane.org/gmane.linux.kernel.iio/4098)

Do we have some support in current userspace to test (qtmoko, shr?) such a
feature?

Regards,

marek


-- 
as simple and primitive as possible
-------------------------------------------------
Marek Belisko - OPEN-NANDRA
Freelance Developer

Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
Tel: +421 915 052 184
skype: marekwhite
twitter: #opennandra
web: http://open-nandra.com

From psonek2 at seznam.cz  Fri May 11 08:32:04 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Fri, 11 May 2012 08:32:04 +0200
Subject: [Gta04-owner] hmc5883 iio driver find way to mainline
In-Reply-To: <CAAfyv36=0iuKVnpc3WmnO6VCoh3u=Vet+=eexEtdVAXZ2hYbdQ@mail.gmail.com>
References: <CAAfyv36=0iuKVnpc3WmnO6VCoh3u=Vet+=eexEtdVAXZ2hYbdQ@mail.gmail.com>
Message-ID: <6894720.2ZWGsxzlEX@rp-core>

On Thursday, May 10, 2012 09:53:48 PM Belisko Marek wrote:

> last week I was working on adding hmc5883 to actual kernel but one
> guy was faster and seems patches are mostly accepted by iio maintainer
> (http://thread.gmane.org/gmane.linux.kernel.iio/4098)
> 
> Do we have some support in current userspace to test (qtmoko, shr?) such a
> feature?

Not yet in QtMoko. I have just finished support for GTA04 vibra motor. 
Accelerometer, compass and barometer are not yet done. I doubt there are even 
framework classes [1] for these, since such hw is quite new. Maybe for the 
compass and barometer it would be enough to have application that reads and 
displays values directly from kernel?

Anyway - any patches are welcome.

Regards

Radek


[1] http://radekp.github.com/qtmoko/api/classes.html
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120511/2e9cd6a4/attachment.html>

From hns at goldelico.com  Fri May 11 08:58:47 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Fri, 11 May 2012 08:58:47 +0200
Subject: [Gta04-owner] hmc5883 iio driver find way to mainline
In-Reply-To: <CAAfyv36=0iuKVnpc3WmnO6VCoh3u=Vet+=eexEtdVAXZ2hYbdQ@mail.gmail.com>
References: <CAAfyv36=0iuKVnpc3WmnO6VCoh3u=Vet+=eexEtdVAXZ2hYbdQ@mail.gmail.com>
Message-ID: <D1C1536C-ED33-4EF0-A819-FCCD01A1246B@goldelico.com>

Hi Marek,

Am 10.05.2012 um 21:53 schrieb Belisko Marek:

> Hi,
> 
> last week I was working on adding hmc5883 to actual kernel but one
> guy was faster and seems patches are mostly accepted by iio maintainer

What a bummer.

> (http://thread.gmane.org/gmane.linux.kernel.iio/4098)

Is that driver acceptable by us (GTA04 needs) or do we need something different?

> Do we have some support in current userspace to test (qtmoko, shr?) such a
> feature?

How does it report values? How can it be controlled?

> 
> Regards,
> 
> marek

rgds, Nikolaus

> 
> 
> -- 
> as simple and primitive as possible
> -------------------------------------------------
> Marek Belisko - OPEN-NANDRA
> Freelance Developer
> 
> Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
> Tel: +421 915 052 184
> skype: marekwhite
> twitter: #opennandra
> web: http://open-nandra.com
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner


From marek.belisko at gmail.com  Fri May 11 09:36:09 2012
From: marek.belisko at gmail.com (Belisko Marek)
Date: Fri, 11 May 2012 09:36:09 +0200
Subject: [Gta04-owner] hmc5883 iio driver find way to mainline
In-Reply-To: <D1C1536C-ED33-4EF0-A819-FCCD01A1246B@goldelico.com>
References: <CAAfyv36=0iuKVnpc3WmnO6VCoh3u=Vet+=eexEtdVAXZ2hYbdQ@mail.gmail.com>
	<D1C1536C-ED33-4EF0-A819-FCCD01A1246B@goldelico.com>
Message-ID: <CAAfyv36tZin2mqCdFavG3ABpPKJ_DLZrdynhMDqs1ABY4m5Znw@mail.gmail.com>

Hi,

On Fri, May 11, 2012 at 8:58 AM, Dr. H. Nikolaus Schaller
<hns at goldelico.com> wrote:
> Hi Marek,
>
> Am 10.05.2012 um 21:53 schrieb Belisko Marek:
>
>> Hi,
>>
>> last week I was working on adding hmc5883 to actual kernel but one
>> guy was faster and seems patches are mostly accepted by iio maintainer
>
> What a bummer.
Yes I have exactly the same patches as him (fortunately work take just
4 hours) so not that bad ;)
>
>> (http://thread.gmane.org/gmane.linux.kernel.iio/4098)
>
> Is that driver acceptable by us (GTA04 needs) or do we need something different?
GTA04 have hmc5883 which is working fine with that driver. We just
need to add some lines
to board file.
>
>> Do we have some support in current userspace to test (qtmoko, shr?) such a
>> feature?
>
> How does it report values? How can it be controlled?
Everything is done via sysfs (settings). For getting values there is
/dev file automatically created
which could be used for getting data.
>
>>
>> Regards,
>>
>> marek
>
> rgds, Nikolaus
>
>>
>>
>> --
>> as simple and primitive as possible
>> -------------------------------------------------
>> Marek Belisko - OPEN-NANDRA
>> Freelance Developer
>>
>> Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
>> Tel: +421 915 052 184
>> skype: marekwhite
>> twitter: #opennandra
>> web: http://open-nandra.com
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

regards,

marek

-- 
as simple and primitive as possible
-------------------------------------------------
Marek Belisko - OPEN-NANDRA
Freelance Developer

Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
Tel: +421 915 052 184
skype: marekwhite
twitter: #opennandra
web: http://open-nandra.com

From turricum at yahoo.de  Fri May 11 10:46:31 2012
From: turricum at yahoo.de (Manuel Christen)
Date: Fri, 11 May 2012 09:46:31 +0100 (BST)
Subject: [Gta04-owner] Sell GTA02 complete with Debug Board and GPS
	Extension antenna
In-Reply-To: <CAAfyv36tZin2mqCdFavG3ABpPKJ_DLZrdynhMDqs1ABY4m5Znw@mail.gmail.com>
References: <CAAfyv36=0iuKVnpc3WmnO6VCoh3u=Vet+=eexEtdVAXZ2hYbdQ@mail.gmail.com>
	<D1C1536C-ED33-4EF0-A819-FCCD01A1246B@goldelico.com>
	<CAAfyv36tZin2mqCdFavG3ABpPKJ_DLZrdynhMDqs1ABY4m5Znw@mail.gmail.com>
Message-ID: <1336725991.86757.YahooMailNeo@web171104.mail.ukl.yahoo.com>

Hello Mailinglist

I sell my GTA02A5 with QtMoko, Debug Board and GPS Extension antenna.

If someone would like to buy it pls let me know!

kind regards
Manuel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120511/2ec222a1/attachment.html>

From psonek2 at seznam.cz  Fri May 11 13:07:29 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Fri, 11 May 2012 13:07:29 +0200
Subject: [Gta04-owner] bq27000-battery not calibrated?
Message-ID: <16510293.FgAQ2JUM9r@rp-core>

Hi,
i started getting this message:

bq27000-battery bq27000-battery: battery is not calibrated! ignoring capacity 
values

Any idea what is going on? It also seems that capacity sysfs node is not 
available:

cat /sys/class/power_supply/bq27000-battery/capacity
cat: /sys/class/power_supply/bq27000-battery/capacity: No data available

and because of this charge indicator stopped working in QtMoko.

This is with both 3.3 and 3.2 kernels.

Regards

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120511/e04a6110/attachment.html>

From shulyaka at gmail.com  Mon May 14 10:55:49 2012
From: shulyaka at gmail.com (Denis Shulyaka)
Date: Mon, 14 May 2012 12:55:49 +0400
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <1336419742.7305.167.camel@myrtle.6gnip.net>
References: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
	<1336419313.2490.12.camel@Nokia-N900>
	<1336419742.7305.167.camel@myrtle.6gnip.net>
Message-ID: <CAM_Y7Lj4+Qq1o03GgNoU_P+eXGR3oWf5pCuHzga76hiLjbbAoQ@mail.gmail.com>

2012/5/7 Bob Ham <rah at settrans.net>:
> On Mon, 2012-05-07 at 21:35 +0200, Eric Andr? wrote:
>
>> Keeping the keyboard behind the device, using, let's say six fingers, three of each hand (=six keys) for the whole alphabet?
>
>> Tell me if it's totally stupid or just a bit...
>
> Neither:
>
> https://en.wikipedia.org/wiki/Chorded_keyboard
>
> A chorded keyboard could be useful on a phone. ?In fact, chorded and
> qwerty keyboards wouldn't be mutually exclusive :-)
>
> --
> Bob Ham <rah at settrans.net>

Speaking of blind-typing, I would suggest something like this:

http://www.sparkfun.com/products/10250

Having a 3x4 array of capacitive "buttons" on the back side is
sufficient for typing (just like we do on a feature-phone - we don't
even need to learn the layout) and can even be performed using one
hand. The PCB can be more thin and we don't need any sliding
mechanism. MPR121 uses I2C 3.3v interface which I guess we've got.

From christoph.mair at gmail.com  Mon May 14 11:01:18 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Mon, 14 May 2012 11:01:18 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <CAM_Y7Lj4+Qq1o03GgNoU_P+eXGR3oWf5pCuHzga76hiLjbbAoQ@mail.gmail.com>
References: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
	<1336419313.2490.12.camel@Nokia-N900>
	<1336419742.7305.167.camel@myrtle.6gnip.net>
	<CAM_Y7Lj4+Qq1o03GgNoU_P+eXGR3oWf5pCuHzga76hiLjbbAoQ@mail.gmail.com>
Message-ID: <CAHvJMeGu4VRyAHk0=d3y0mCH7ajUDMu_FXeGoUaDTJ8kcrFJ-g@mail.gmail.com>

On Mon, May 14, 2012 at 10:55 AM, Denis Shulyaka <shulyaka at gmail.com> wrote:
> MPR121 uses I2C 3.3v interface which I guess we've got.

This chip can also operate from 1.8V which is what we have on the
GTA04 (and the I2C bus also uses these levels)

Christoph

From hns at goldelico.com  Mon May 14 11:11:16 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Mon, 14 May 2012 11:11:16 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <CAHvJMeGu4VRyAHk0=d3y0mCH7ajUDMu_FXeGoUaDTJ8kcrFJ-g@mail.gmail.com>
References: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
	<1336419313.2490.12.camel@Nokia-N900>
	<1336419742.7305.167.camel@myrtle.6gnip.net>
	<CAM_Y7Lj4+Qq1o03GgNoU_P+eXGR3oWf5pCuHzga76hiLjbbAoQ@mail.gmail.com>
	<CAHvJMeGu4VRyAHk0=d3y0mCH7ajUDMu_FXeGoUaDTJ8kcrFJ-g@mail.gmail.com>
Message-ID: <353F0B10-69CF-4A96-924A-5FCA1A469613@goldelico.com>


Am 14.05.2012 um 11:01 schrieb Christoph Mair:

> On Mon, May 14, 2012 at 10:55 AM, Denis Shulyaka <shulyaka at gmail.com> wrote:
>> MPR121 uses I2C 3.3v interface which I guess we've got.
> 
> This chip can also operate from 1.8V which is what we have on the
> GTA04 (and the I2C bus also uses these levels)
> 
> Christoph

Christoph didn't mention that there is already a driver for the MPR121

	http://gitorious.org/freerunner-navigation-board/mpr121/trees/master/kernel

and hardware for experimentation was available through the Freerunner Navigation
board:

	http://www.mail-archive.com/community at lists.openmoko.org/msg62529.html
	http://wiki.openmoko.org/wiki/Freerunner_Navigation_Board_v3#Bottom_side

Unfortunately it is no longer in production now (lack of demand).

If someone wants to develop such a touch keyboard, we
can help to get it produced and distributed for the masses.

Nikolaus


From eric.andre at rican.ch  Mon May 14 20:39:24 2012
From: eric.andre at rican.ch (Eric =?iso-8859-1?Q?Andr=E9?=)
Date: Mon, 14 May 2012 20:39:24 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototyp
In-Reply-To: <mailman.3.1336989606.98357.gta04-owner@goldelico.com>
Message-ID: <1337020764-da8470729705ffc4a442b961f0f1cc11@rican.ch>


----- Message d'origine -----
De: gta04-owner-request at goldelico.com
Date: Mon, 14 May 2012 12:00:06 +0200
Sujet: Gta04-owner Digest, Vol 17, Issue 14
?: gta04-owner at goldelico.com

>Send Gta04-owner mailing list submissions to
>	gta04-owner at goldelico.com
>
>To subscribe or unsubscribe via the World Wide Web, visit
>	http://lists.goldelico.com/mailman/listinfo/gta04-owner
>or, via email, send a message with subject or body 'help' to
>	gta04-owner-request at goldelico.com
>
>You can reach the person managing the list at
>	gta04-owner-owner at goldelico.com
>
>When replying, please edit your Subject line so it is more specific
>than "Re: Contents of Gta04-owner digest..."
>
>
>Today's Topics:
>
>   1. Re: [ANN] GTA04 Keyboard prototype (Denis Shulyaka)
>   2. Re: [ANN] GTA04 Keyboard prototype (Christoph Mair)
>   3. Re: [ANN] GTA04 Keyboard prototype (Dr. H. Nikolaus Schaller)
>
>
>----------------------------------------------------------------------
>
>Having a 3x4 array of capacitive "buttons" on the back side is
>sufficient for typing (just like we do on a feature-phone - we don't
>even need to learn the layout) and can even be performed using one
>hand. The PCB can be more thin and we don't need any sliding
>mechanism. MPR121 uses I2C 3.3v interface which I guess we've got.

I Would'nt use a capacitive as this surface is allways touched. 
The machine will allways confuse between manipulation and information (typing).
Inmy humble opinion.
Best regards
Eric

From frederi1 at gmail.com  Mon May 14 21:45:50 2012
From: frederi1 at gmail.com (Fred .)
Date: Mon, 14 May 2012 21:45:50 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <353F0B10-69CF-4A96-924A-5FCA1A469613@goldelico.com>
References: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
	<1336419313.2490.12.camel@Nokia-N900>
	<1336419742.7305.167.camel@myrtle.6gnip.net>
	<CAM_Y7Lj4+Qq1o03GgNoU_P+eXGR3oWf5pCuHzga76hiLjbbAoQ@mail.gmail.com>
	<CAHvJMeGu4VRyAHk0=d3y0mCH7ajUDMu_FXeGoUaDTJ8kcrFJ-g@mail.gmail.com>
	<353F0B10-69CF-4A96-924A-5FCA1A469613@goldelico.com>
Message-ID: <CAP7RQn6OgSdtYp_5wFpamfLDgfWXUF9JeeX3Rh2o2aAa-e2L1w@mail.gmail.com>

Hello,
I have two question:
  Connecting a second "touchscreen panel" isn't simpler?
Maybe capacitive, sliding from a slot between the battery and the battery
cover.Or really flexible then you just roll it on the side!


  Is there a cad file available for the GTA04 board itself?
Probably useful for those designers...not so familiar with eagle, or
datasheets :)

Maybe someone already produced such drawing, in 3d?
I can try by myself beggining with the picture on the manual for example
(ch 6.19.4), i just need the dimensions of all the chips/connectors, maybe
goldelico can provide electrical and mechanicals datas, files, measures of
the room between two metal shields...

Best regards, Fred.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120514/b531c5e4/attachment.html>

From hns at goldelico.com  Fri May 18 09:03:22 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Fri, 18 May 2012 09:03:22 +0200
Subject: [Gta04-owner] Delivery of the GTA04
In-Reply-To: <1337281281.7508.4.camel@myrtle.6gnip.net>
References: <20120507194632.33ac4631@epigpbn01.ad-hugtip.local>
	<660E9FA3-8EC6-4881-A489-108C6A0AF343@goldelico.com>
	<1337281281.7508.4.camel@myrtle.6gnip.net>
Message-ID: <CF02C929-030A-4D63-840E-E6D3AE3E5A3E@goldelico.com>


Am 17.05.2012 um 21:01 schrieb Bob Ham:

> On Tue, 2012-05-08 at 09:26 +0200, Dr. H. Nikolaus Schaller wrote:
>> * I think it will need 1-2 weeks until we finally know and they have production ramped up
> 
> Do we know whether the GTA04 group tour units have production ramped up?

They told me this morning that they *think* they have solved the soldering problems
to a level they (and we) can live with. But they were not willing to make a prognosis
when all GTA04 group tour devices are produced and fault free. The first ones are
done and have passed the first automatic test (which does not cover all functions).

Sorry for that snail speed,
Nikolaus


PS: next time we must aim for higher number of units so that we get higher
priority (and lower price). Let's beat the iPhone 5 sales numbers :)

PPS: I know that there are many of you interested in a GTA04 out there but
did not participate in the Group Tour.

Please indicate your non-binding interest through the "Wishlist" function
so that we get some statistics for future demand:

http://www.handheld-linux.com/wiki.php?page=GTA04


From calypso at riseup.net  Sat May 19 09:40:08 2012
From: calypso at riseup.net (calypso at riseup.net)
Date: Sat, 19 May 2012 09:40:08 +0200
Subject: [Gta04-owner] GTA04 on Replicant - Website update
Message-ID: <1337413208.3388.15.camel@niklasdesktop>

Hallo,
news from Replicant with an updated Status page:
http://redmine.replicant.us/projects/replicant/wiki/GTA04Status


And my question from the replicant-mailinglist attached:

Le jeudi 17 mai 2012 ? 18:38 +0200, calypso at riseup.net a ?crit :
> I just wondered if Replicant is looking for Betatesters since there
were
> announcements at the gta04-owner-list and on your website.

Well, we are not particularly looking for beta-testers yet as most
components aren't working already (there is not a lot to test
currently).

> Do you already have a roadmap or timeline for gta04 implementation ?

Things are going slowly as I'm the only one involved, and it is hard for
me to find time to improve things significantly on a daily basis. 
To keep you posted: I'm currently working on the audio part. I have
nearly achieved the output part (the android music player works, etc)
but I still need a way to detect headphones plug and input is more
problematic as I need audio downsampling. That will be done though, and
certainly before the end of this week.

Next steps will be: power management, battery/backlight support and
graphics speed improvement. I think we will release a set of images at
this point. Then, we'll add support for the "hardest" parts, Modem/3G
and Wifi/Bluetooth, along with GPS and sensors. This is going to take
much longer and we'll probably release images when significant progress
is made. 

I have made a page to hold GTA04 status at:
http://redmine.replicant.us/projects/replicant/wiki/GTA04Status
and it'll probably give more details as development goes on.

-- 
Paul Kocialkowski, Replicant developer

Replicant is a fully free Android distribution

Website:        http://www.replicant.us
Wiki/Tracker:   http://redmine.replicant.us



From hns at goldelico.com  Sat May 19 10:22:38 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Sat, 19 May 2012 10:22:38 +0200
Subject: [Gta04-owner] GTA04 on Replicant - Website update
In-Reply-To: <1337413208.3388.15.camel@niklasdesktop>
References: <1337413208.3388.15.camel@niklasdesktop>
Message-ID: <74C89330-D9B2-4B50-8FF8-145F6F0EAFC7@goldelico.com>

Hi Paul,

Am 19.05.2012 um 09:40 schrieb calypso at riseup.net:

> Hallo,
> news from Replicant with an updated Status page:
> http://redmine.replicant.us/projects/replicant/wiki/GTA04Status

That is good!

> 
> 
> And my question from the replicant-mailinglist attached:
> 
> Le jeudi 17 mai 2012 ? 18:38 +0200, calypso at riseup.net a ?crit :
>> I just wondered if Replicant is looking for Betatesters since there
> were
>> announcements at the gta04-owner-list and on your website.
> 
> Well, we are not particularly looking for beta-testers yet as most
> components aren't working already (there is not a lot to test
> currently).
> 
>> Do you already have a roadmap or timeline for gta04 implementation ?
> 
> Things are going slowly as I'm the only one involved, and it is hard for

Hm. I think there might be some AoF people around
who might be interested in getting newer hardware for the future.

And, AFAIK, there is still the offer by some community member
to donate a free GTA04 board (from the group tour) to another
Replicant developer.

> me to find time to improve things significantly on a daily basis. 
> To keep you posted: I'm currently working on the audio part. I have
> nearly achieved the output part (the android music player works, etc)

++ I think getting a music player work is an important milestone.

> but I still need a way to detect headphones plug and input is more
> problematic as I need audio downsampling. That will be done though, and
> certainly before the end of this week.
> 
> Next steps will be: power management, battery/backlight support and
> graphics speed improvement. I think we will release a set of images at
> this point. Then, we'll add support for the "hardest" parts, Modem/3G
> and Wifi/Bluetooth, along with GPS and sensors. This is going to take
> much longer and we'll probably release images when significant progress
> is made. 

Yes, please do some "release early and often" strategy.

> 
> I have made a page to hold GTA04 status at:
> http://redmine.replicant.us/projects/replicant/wiki/GTA04Status
> and it'll probably give more details as development goes on.
> 
> -- 
> Paul Kocialkowski, Replicant developer
> 
> Replicant is a fully free Android distribution
> 
> Website:        http://www.replicant.us
> Wiki/Tracker:   http://redmine.replicant.us
> 
> 
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner


From tomek-k at wp.eu  Sat May 19 12:28:12 2012
From: tomek-k at wp.eu (Tomasz =?utf-8?q?Ka=C5=BAmierczak?=)
Date: Sat, 19 May 2012 12:28:12 +0200
Subject: [Gta04-owner] GTA02 for sale
Message-ID: <201205191228.12139.tomek-k@wp.eu>

Hello,

I have a GTA02 for sale if anyone is interested. My price would be ?80/?100/400PLN.
Please contact me directly so that we can discus payment and shipment details (my location is Pozna?, Poland).

The phone is working, battery is in a (subjectively) quite good state.

What I'm selling:
- GTA02, revision A5 (first mass production revision), not modified, european version (900/1800/1900 Mhz), qi bootloader, QtMoko OS
- battery (FIC 1200mAh, Li-ion)
- charger with adapters
- USB cable
- headphones
- 2GB ?SD card
- stylus
- cover

I can also send photos of the phone and the accesories.

Cheers,
Tomek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120519/1db94cd1/attachment.html>

From hns at goldelico.com  Sun May 20 12:45:15 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Sun, 20 May 2012 12:45:15 +0200
Subject: [Gta04-owner] Rainy Day Summer Activities: improvements for the
	GTA04...
Message-ID: <81CFC8E4-ED3A-4877-A930-0BCCB91346A8@goldelico.com>

For all those, who already have their GTA04 in their hands:

just in case you are looking to get ideas for rainy days during this summer,
you can help to improve the GTA04 to become the best open smartphone
we ever had.

I have listed some topics that are waiting for experiments and solutions (list
is not exhaustive and we all are open to new ideas):

* make 3D graphics acceleration work like it does on the BeagleBoard, OpenPandora
* camera driver (there is a partially working driver in the hw-validation kernel, but shows
   only noise and no image)
* experiment with the DSP built into the CPU chip
* add driver to enable/switch UART3 to IrDA mode
* further improvements on power management
* migrate more drivers/patches from hw-validation to kernel 3.x
* make the FM radio system work
* experiment with the RFID EEPROM chip
* add Kalman filters to GPS/Accelerometer/Gyroscope/Barometer for precision location tracking
* help on Replicant (100% free Android)
* help on SHR (adding, stabilizing, fixing packages)
* help on QtMoko
* start to port http://www.merproject.org/, http://plasma-active.org/, ...
...

Please also look at:

	http://projects.goldelico.com/p/gta04-kernel/issues/

This already collects some background information.

Happy summer of improvements,
Nikolaus



From marek.belisko at gmail.com  Sun May 20 14:24:47 2012
From: marek.belisko at gmail.com (Belisko Marek)
Date: Sun, 20 May 2012 14:24:47 +0200
Subject: [Gta04-owner] Rainy Day Summer Activities: improvements for the
	GTA04...
In-Reply-To: <81CFC8E4-ED3A-4877-A930-0BCCB91346A8@goldelico.com>
References: <81CFC8E4-ED3A-4877-A930-0BCCB91346A8@goldelico.com>
Message-ID: <CAAfyv35gC09NBpza6BB-HQ6QibQbdTFgc-rYi3oHfYL54Un5mg@mail.gmail.com>

On Sun, May 20, 2012 at 12:45 PM, Dr. H. Nikolaus Schaller
<hns at goldelico.com> wrote:
> For all those, who already have their GTA04 in their hands:
>
> just in case you are looking to get ideas for rainy days during this summer,
> you can help to improve the GTA04 to become the best open smartphone
> we ever had.
>
> I have listed some topics that are waiting for experiments and solutions (list
> is not exhaustive and we all are open to new ideas):
>
> * make 3D graphics acceleration work like it does on the BeagleBoard, OpenPandora
> * camera driver (there is a partially working driver in the hw-validation kernel, but shows
> ? only noise and no image)
> * experiment with the DSP built into the CPU chip
> * add driver to enable/switch UART3 to IrDA mode
> * further improvements on power management
> * migrate more drivers/patches from hw-validation to kernel 3.x
- hmc5883 driver seems to be merged soon (will try to add support to QTMoko)
> * make the FM radio system work
- I started to work on FM radio driver for (3.3 kernel)
> * experiment with the RFID EEPROM chip
> * add Kalman filters to GPS/Accelerometer/Gyroscope/Barometer for precision location tracking
> * help on Replicant (100% free Android)
> * help on SHR (adding, stabilizing, fixing packages)
> * help on QtMoko
> * start to port http://www.merproject.org/, http://plasma-active.org/, ...
> ...
>
> Please also look at:
>
> ? ? ? ?http://projects.goldelico.com/p/gta04-kernel/issues/
>
> This already collects some background information.
>
> Happy summer of improvements,
> Nikolaus
>
>
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

Regards,

marek

-- 
as simple and primitive as possible
-------------------------------------------------
Marek Belisko - OPEN-NANDRA
Freelance Developer

Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
Tel: +421 915 052 184
skype: marekwhite
twitter: #opennandra
web: http://open-nandra.com

From GNUtoo at no-log.org  Sun May 20 18:24:37 2012
From: GNUtoo at no-log.org (Denis 'GNUtoo' Carikli)
Date: Sun, 20 May 2012 18:24:37 +0200
Subject: [Gta04-owner] alsaloop and poll issues with the modem ALSA driver.
Message-ID: <1354263.ex7NO0GF0D@gnutoo-desktop>

Hi,
I am trying to get a good forwarder to work on the gta04 A3 for SHR
(The forwarder is the main blocking thing for me, since I've no gta04 A4).

Here are the result from debugging alsaloop:
If the modem is in call, the poll system call behave normally:
poll took 246094us //from alsaloop
but if the modem is not in a call:
the poll will take a lot less:
poll took 31us //from alsaloop
which will result in 100% or near 100% CPU usage.

I think that the right thing is to fix the audio driver or the ASOC framework.

Denis.


From hns at goldelico.com  Sun May 20 18:30:17 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Sun, 20 May 2012 18:30:17 +0200
Subject: [Gta04-owner] OpenPhoenux / GTA04 @ LinuxTag Berlin
In-Reply-To: <44D0EA2E-C6DB-4FCB-A0D6-C205604135EF@goldelico.com>
References: <44D0EA2E-C6DB-4FCB-A0D6-C205604135EF@goldelico.com>
Message-ID: <27CE5C49-CAC7-4135-9326-9607B3AC8ED0@goldelico.com>


Am 16.05.2012 um 08:19 schrieb Dr. H. Nikolaus Schaller:

> Lukas and myself will be in Berlin at Friday (25 may) afternoon and
> Saturday (26 may).
> 
> We are looking forward to meet as many of you as possible for face
> to face discussions! I think there will be many interesting things to discuss:
> 
> * GTA04 and future
> * 3D case printing
> * Keyboard prototype
> * Distro projects (SHR, QtMoko, Replicant/AoF, ...)
> * Openmoko.org organization
> * whatever you think is important
> 
> There will be a shared Project Meeting Point (Booth 7.2a 181c) as scheduled here:
> 
> 	http://wiki.linuxtag.org/w/fp:ProjectMeetingPoints
> 
> Please look up regularily since it may change on short notice.
> 
> And, we have made a late entry submission for a presentation, but it is
> not yet confirmed.

It has just been confirmed (thanks to the organizers for accepting
a very late submission!) for Saturday 26th, 11:00 - 11:30:

http://www.linuxtag.org/2012/de/program/program/vortragsdetails.html?no_cache=1&talkid=403

> 
> To make easier the decision to come, the LinuxTag organizers have
> given us some free entrance tickets (regular price is AFAIK 20 EUR).
> If you are interested, please send me a private mail and I can send you
> an electronic ticket that you can print yourself.

We still have such tickets to share.

> 
> CU at Berlin,
> 
> Nikolaus


From hns at goldelico.com  Sun May 20 19:09:14 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Sun, 20 May 2012 19:09:14 +0200
Subject: [Gta04-owner] alsaloop and poll issues with the modem ALSA
	driver.
In-Reply-To: <1354263.ex7NO0GF0D@gnutoo-desktop>
References: <1354263.ex7NO0GF0D@gnutoo-desktop>
Message-ID: <F9F396B3-4EE7-4C15-9414-5ACBC85BD061@goldelico.com>


Am 20.05.2012 um 18:24 schrieb Denis 'GNUtoo' Carikli:

> Hi,
> I am trying to get a good forwarder to work on the gta04 A3 for SHR
> (The forwarder is the main blocking thing for me, since I've no gta04 A4).
> 
> Here are the result from debugging alsaloop:
> If the modem is in call, the poll system call behave normally:
> poll took 246094us //from alsaloop
> but if the modem is not in a call:
> the poll will take a lot less:
> poll took 31us //from alsaloop
> which will result in 100% or near 100% CPU usage.
> 
> I think that the right thing is to fix the audio driver or the ASOC framework.

This may have to do with the PCM clock/frame sync that is coming
from the modem. And it starts only at the moment the connection is
established.

Maybe, a poll() before the clock arrives fails and returns some error.
Could be some error that isn't reported to the user space.

Radek & Neil may know a little more about this.

To reduce CPU usage, you could try to find some indication about
being in call/not in call and sleep the forwarder process a little.

Nikolaus


From morphis at gravedo.de  Sun May 20 20:42:24 2012
From: morphis at gravedo.de (Simon Busch)
Date: Sun, 20 May 2012 20:42:24 +0200
Subject: [Gta04-owner] alsaloop and poll issues with the modem ALSA
	driver.
In-Reply-To: <F9F396B3-4EE7-4C15-9414-5ACBC85BD061@goldelico.com>
References: <1354263.ex7NO0GF0D@gnutoo-desktop>
	<F9F396B3-4EE7-4C15-9414-5ACBC85BD061@goldelico.com>
Message-ID: <4FB93B10.4000704@gravedo.de>

On 20.05.2012 19:09, Dr. H. Nikolaus Schaller wrote:
> 
> Am 20.05.2012 um 18:24 schrieb Denis 'GNUtoo' Carikli:
> 
>> Hi,
>> I am trying to get a good forwarder to work on the gta04 A3 for SHR
>> (The forwarder is the main blocking thing for me, since I've no gta04 A4).
>>
>> Here are the result from debugging alsaloop:
>> If the modem is in call, the poll system call behave normally:
>> poll took 246094us //from alsaloop
>> but if the modem is not in a call:
>> the poll will take a lot less:
>> poll took 31us //from alsaloop
>> which will result in 100% or near 100% CPU usage.

What does "in call" really mean? An active or outgoing call?

>> I think that the right thing is to fix the audio driver or the ASOC framework.
> 
> This may have to do with the PCM clock/frame sync that is coming
> from the modem. And it starts only at the moment the connection is
> established.
> 
> Maybe, a poll() before the clock arrives fails and returns some error.
> Could be some error that isn't reported to the user space.
> 
> Radek & Neil may know a little more about this.
> 
> To reduce CPU usage, you could try to find some indication about
> being in call/not in call and sleep the forwarder process a little.

Can't we power up the alsaloop thing only when a call is active and
disable it just when a call ends?

Hm, if I look into the gsmvoice_alsa_forwarder plugin of fsoaudiod it's
already done this way or do I miss some tiny detail?

regards,
Simon

-- 
Simon Busch - http://mm.gravedo.de/blog/

From hns at goldelico.com  Sun May 20 21:01:26 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Sun, 20 May 2012 21:01:26 +0200
Subject: [Gta04-owner] alsaloop and poll issues with the modem ALSA
	driver.
In-Reply-To: <4FB93B10.4000704@gravedo.de>
References: <1354263.ex7NO0GF0D@gnutoo-desktop>
	<F9F396B3-4EE7-4C15-9414-5ACBC85BD061@goldelico.com>
	<4FB93B10.4000704@gravedo.de>
Message-ID: <0A15FAD2-B6D1-4E96-A925-3681EF7B9A57@goldelico.com>


Am 20.05.2012 um 20:42 schrieb Simon Busch:

> On 20.05.2012 19:09, Dr. H. Nikolaus Schaller wrote:
>> 
>> Am 20.05.2012 um 18:24 schrieb Denis 'GNUtoo' Carikli:
>> 
>>> Hi,
>>> I am trying to get a good forwarder to work on the gta04 A3 for SHR
>>> (The forwarder is the main blocking thing for me, since I've no gta04 A4).
>>> 
>>> Here are the result from debugging alsaloop:
>>> If the modem is in call, the poll system call behave normally:
>>> poll took 246094us //from alsaloop
>>> but if the modem is not in a call:
>>> the poll will take a lot less:
>>> poll took 31us //from alsaloop
>>> which will result in 100% or near 100% CPU usage.
> 
> What does "in call" really mean? An active or outgoing call?

The moment where the voice data connection is established.

For an incoming call with "RING" messages, it is some milliseconds
after accepting the call by some AT command.

For an outgoing call it is after doing ATD when the other side
picks up the received and this event has travelled back to the
mobile through the GSM signalling network.

> 
>>> I think that the right thing is to fix the audio driver or the ASOC framework.
>> 
>> This may have to do with the PCM clock/frame sync that is coming
>> from the modem. And it starts only at the moment the connection is
>> established.
>> 
>> Maybe, a poll() before the clock arrives fails and returns some error.
>> Could be some error that isn't reported to the user space.
>> 
>> Radek & Neil may know a little more about this.
>> 
>> To reduce CPU usage, you could try to find some indication about
>> being in call/not in call and sleep the forwarder process a little.
> 
> Can't we power up the alsaloop thing only when a call is active and
> disable it just when a call ends?

Yes, you can start it when ATD begins. But in my experience it may
need 5 seconds until the other side starts to ring and several ringing
tones until the other side picks up. So it may be for 10 seconds in this
100% CPU state - as far as my understanding goes.

> 
> Hm, if I look into the gsmvoice_alsa_forwarder plugin of fsoaudiod it's
> already done this way or do I miss some tiny detail?
> 
> regards,
> Simon

BR,
Nikolaus



From neilb at suse.de  Mon May 21 10:30:55 2012
From: neilb at suse.de (NeilBrown)
Date: Mon, 21 May 2012 18:30:55 +1000
Subject: [Gta04-owner] 3.4 kernel for GTA04.
Message-ID: <20120521183055.43437285@notabene.brown>


Hot on the heels for the release of 3.4 from Linus, I'm happy to
announce
   git://github.com/neilbrown/linux.git gta04/3.4.y
   git://neil.brown.name/gta04 3.4-gta04

There is very little new here apart from changes needs to make it work on 3.4
which included:
 - fixes to the pwm driver (drives the backlight) to match changes in dm_timer
    (OMAP dual-mode timer)
 - declare more consumers for the display sub system regulator so the driver
   can find its powersource
 - Use the new driver probe deferral code (instead of my own) to
   initialise the wifi chip only after the reset line is available.
   This required some changes and bug fixes to GPIO and MMC.
 - change the names of some DAI names so that audio works again.

Most of my time has been going to pushing various other things upstream -
hopefully some of the fixes I'm carrying will appear in 3.5-rc.

Reports of success and failure are always welcome.

NeilBrown

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120521/07778c65/attachment-0001.bin>

From psonek2 at seznam.cz  Mon May 21 15:21:54 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Mon, 21 May 2012 15:21:54 +0200
Subject: [Gta04-owner] 3.4 kernel for GTA04.
In-Reply-To: <20120521183055.43437285@notabene.brown>
References: <20120521183055.43437285@notabene.brown>
Message-ID: <2337169.UG2WUa0Mgu@rp-core>

On Monday, May 21, 2012 10:30:55 AM NeilBrown wrote:

> Hot on the heels for the release of 3.4 from Linus, I'm happy to
> announce
>    git://github.com/neilbrown/linux.git gta04/3.4.y
>    git://neil.brown.name/gta04 3.4-gta04
> Reports of success and failure are always welcome.

Nice, i am now running 3.3 kernel on the phone for 2 weeks and it looks good. 
I will start with 3.4 testing and if all goes right it will be the kernel for 
next qtmoko version.

Btw i have noticed one thing with 3.3 kernel. I have some 640x480 videos and 
they play nicely with 3.2 but with 3.3 they are quite slow. But i havent 
looked at it further. I will try look into it later when i have some time.

Regards

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120521/2d77ede6/attachment.html>

From neilb at suse.de  Tue May 22 00:40:33 2012
From: neilb at suse.de (NeilBrown)
Date: Tue, 22 May 2012 08:40:33 +1000
Subject: [Gta04-owner] 3.4 kernel for GTA04.
In-Reply-To: <2337169.UG2WUa0Mgu@rp-core>
References: <20120521183055.43437285@notabene.brown>
	<2337169.UG2WUa0Mgu@rp-core>
Message-ID: <20120522084033.3fb24f21@notabene.brown>

On Mon, 21 May 2012 15:21:54 +0200 Radek Polak <psonek2 at seznam.cz> wrote:

> On Monday, May 21, 2012 10:30:55 AM NeilBrown wrote:
> 
> > Hot on the heels for the release of 3.4 from Linus, I'm happy to
> > announce
> >    git://github.com/neilbrown/linux.git gta04/3.4.y
> >    git://neil.brown.name/gta04 3.4-gta04
> > Reports of success and failure are always welcome.
> 
> Nice, i am now running 3.3 kernel on the phone for 2 weeks and it looks good. 
> I will start with 3.4 testing and if all goes right it will be the kernel for 
> next qtmoko version.
> 
> Btw i have noticed one thing with 3.3 kernel. I have some 640x480 videos and 
> they play nicely with 3.2 but with 3.3 they are quite slow. But i havent 
> looked at it further. I will try look into it later when i have some time.

No idea what would cause that - the video subsystem is still mostly a
block-box to me.

How hard would it be for me to add qtmoko bits to my current Debian install
so I could test things out without committing to just qtmoko?  Is there a
debian package I can install?  Or an apt-source I can add?
Or can I just copy "/opt" from the tar ball and go from there?
And am I correct that QTmoko can run directly on the frame buffer or over
X11??

Thanks,
NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120522/088be5d5/attachment.bin>

From psonek2 at seznam.cz  Tue May 22 09:10:58 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Tue, 22 May 2012 09:10:58 +0200
Subject: [Gta04-owner] 3.4 kernel for GTA04.
In-Reply-To: <20120522084033.3fb24f21@notabene.brown>
References: <20120521183055.43437285@notabene.brown>
	<2337169.UG2WUa0Mgu@rp-core>
	<20120522084033.3fb24f21@notabene.brown>
Message-ID: <1506821.ugZtAhk3oY@rp-core>

On Tuesday, May 22, 2012 12:40:33 AM NeilBrown wrote:

> No idea what would cause that - the video subsystem is still mostly a
> block-box to me.

No problem, this has time.

> How hard would it be for me to add qtmoko bits to my current Debian install
> so I could test things out without committing to just qtmoko?  Is there a
> debian package I can install?  Or an apt-source I can add?
> Or can I just copy "/opt" from the tar ball and go from there?

There is apt repo:

	echo "deb http://qtmoko.sourceforge.net/debian/gta04 /" >> 
/etc/apt/sources.list

or you can just download and install the .deb from here:

	http://qtmoko.sourceforge.net/debian/gta04/qtmoko-gta04_44-1_armel.deb

Currently qtmoko adds itself to startup scripts so if you want to remove it 
and launch it manually you can do:

	update-rc.d qtmoko-gta04 remove

and use this for launching:

	/etc/init.d/qtmoko-gta04 start

or this:

	. /opt/qtmoko/qpe.env
	qpe

I think it should be running on clean debian without problems. It expects your 
udev rules for modem and vibra motor so this should be fine. You might find 
useful this howto [1] in case of problems.

> And am I correct that QTmoko can run directly on the frame buffer or over
> X11??

It currently runs on framebuffer.

It can run on emulated framebuffer in X11 (qvfb) - this is used for "emulator" 
on PC.

It was running on X11 using standard qt-X11 in some old openmoko ASU distro. 
It required some patches and it was very slow.

Regards

Radek


[1] 
https://github.com/radekp/qtmoko/blob/master/doc/txt/debian_rootfs_howto_gta04.txt
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120522/577cde99/attachment.html>

From paulk at paulk.fr  Tue May 22 20:54:15 2012
From: paulk at paulk.fr (PaulK)
Date: Tue, 22 May 2012 20:54:15 +0200
Subject: [Gta04-owner] Detecting headset plug: twl4030-madc IRQ issue
In-Reply-To: <89D760AE-9013-4214-BA7A-BCFB132497AB@goldelico.com>
References: <1335633337.1729.11.camel@aldrin>
	<20120429083223.509aac7c@notabene.brown>
	<89D760AE-9013-4214-BA7A-BCFB132497AB@goldelico.com>
Message-ID: <1337712855.1709.5.camel@aldrin>

After bringing audio output to Replicant on GTA04, I'm back to work on
headset plug detect. For now, I'm going with hw-validation kernel just
to make sure the ADCs are handled correctly. 

So apparently, ADC7 is MICSENSE (no headset / short / headset) [1],
exactly what I need. After trying some user-space madc code [2], it
seems that ADC7 reports pretty-much the same value (around raw=10)
regardless of headset plug state. 

What am I missing here? Any help would be greatly appreciated!

[1]: http://projects.goldelico.com/p/gta04-kernel/page/Devices-Overview/
[2]: https://github.com/scottellis/madc

-- 
Paul Kocialkowski, Replicant developer

Replicant is a fully free Android distribution

Website:        http://www.replicant.us
Wiki/Tracker:   http://redmine.replicant.us


From marek.belisko at gmail.com  Tue May 22 22:53:05 2012
From: marek.belisko at gmail.com (Belisko Marek)
Date: Tue, 22 May 2012 22:53:05 +0200
Subject: [Gta04-owner] hmc5883 iio driver find way to mainline
In-Reply-To: <CAAfyv36tZin2mqCdFavG3ABpPKJ_DLZrdynhMDqs1ABY4m5Znw@mail.gmail.com>
References: <CAAfyv36=0iuKVnpc3WmnO6VCoh3u=Vet+=eexEtdVAXZ2hYbdQ@mail.gmail.com>
	<D1C1536C-ED33-4EF0-A819-FCCD01A1246B@goldelico.com>
	<CAAfyv36tZin2mqCdFavG3ABpPKJ_DLZrdynhMDqs1ABY4m5Znw@mail.gmail.com>
Message-ID: <CAAfyv34Z-C5OoxkJ8Pr3Wa5Bx+pAhjJQSVFTSHz7uZXTS4eVpg@mail.gmail.com>

Hi,

On Fri, May 11, 2012 at 9:36 AM, Belisko Marek <marek.belisko at gmail.com> wrote:
> Hi,
>
> On Fri, May 11, 2012 at 8:58 AM, Dr. H. Nikolaus Schaller
> <hns at goldelico.com> wrote:
>> Hi Marek,
>>
>> Am 10.05.2012 um 21:53 schrieb Belisko Marek:
>>
>>> Hi,
>>>
>>> last week I was working on adding hmc5883 to actual kernel but one
>>> guy was faster and seems patches are mostly accepted by iio maintainer
>>
>> What a bummer.
> Yes I have exactly the same patches as him (fortunately work take just
> 4 hours) so not that bad ;)
>>
>>> (http://thread.gmane.org/gmane.linux.kernel.iio/4098)
>>
>> Is that driver acceptable by us (GTA04 needs) or do we need something different?
> GTA04 have hmc5883 which is working fine with that driver. We just
> need to add some lines
> to board file.
patches for hmc5883 are already in linux-next. Any ideas how to import
them to Neils kernel?
Try to do that manually but there's a lot of dependencies on other
changes which aren't in 3.4 :(
>>
>>> Do we have some support in current userspace to test (qtmoko, shr?) such a
>>> feature?
>>
>> How does it report values? How can it be controlled?
> Everything is done via sysfs (settings). For getting values there is
> /dev file automatically created
> which could be used for getting data.
>>
>>>
>>> Regards,
>>>
>>> marek
>>
>> rgds, Nikolaus
>>
>>>
>>>
>>> --
>>> as simple and primitive as possible
>>> -------------------------------------------------
>>> Marek Belisko - OPEN-NANDRA
>>> Freelance Developer
>>>
>>> Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
>>> Tel: +421 915 052 184
>>> skype: marekwhite
>>> twitter: #opennandra
>>> web: http://open-nandra.com
>>> _______________________________________________
>>> Gta04-owner mailing list
>>> Gta04-owner at goldelico.com
>>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>>
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>
> regards,
>
> marek
>
> --
> as simple and primitive as possible
> -------------------------------------------------
> Marek Belisko - OPEN-NANDRA
> Freelance Developer
>
> Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
> Tel: +421 915 052 184
> skype: marekwhite
> twitter: #opennandra
> web: http://open-nandra.com

thanks,

marek

-- 
as simple and primitive as possible
-------------------------------------------------
Marek Belisko - OPEN-NANDRA
Freelance Developer

Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
Tel: +421 915 052 184
skype: marekwhite
twitter: #opennandra
web: http://open-nandra.com

From neilb at suse.de  Tue May 22 22:57:34 2012
From: neilb at suse.de (NeilBrown)
Date: Wed, 23 May 2012 06:57:34 +1000
Subject: [Gta04-owner] Detecting headset plug: twl4030-madc IRQ issue
In-Reply-To: <1337712855.1709.5.camel@aldrin>
References: <1335633337.1729.11.camel@aldrin>
	<20120429083223.509aac7c@notabene.brown>
	<89D760AE-9013-4214-BA7A-BCFB132497AB@goldelico.com>
	<1337712855.1709.5.camel@aldrin>
Message-ID: <20120523065734.23b8f37f@notabene.brown>

On Tue, 22 May 2012 20:54:15 +0200 PaulK <paulk at paulk.fr> wrote:

> After bringing audio output to Replicant on GTA04, I'm back to work on
> headset plug detect. For now, I'm going with hw-validation kernel just
> to make sure the ADCs are handled correctly. 
> 
> So apparently, ADC7 is MICSENSE (no headset / short / headset) [1],
> exactly what I need. After trying some user-space madc code [2], it
> seems that ADC7 reports pretty-much the same value (around raw=10)
> regardless of headset plug state. 
> 
> What am I missing here? Any help would be greatly appreciated!

Maybe you need to make sure VHSMIC.OUT is generating current.
i.e. Set the HSMICBIAS_EN bit in MICBIAS_CTL

You should be able to do this with alsamixer by finding the control called

   Headset Mic Bias

and enabling that.
But maybe you already have and it is some other problem...

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120523/bd3ff21c/attachment.bin>

From andreas at kemnade.info  Tue May 22 23:07:55 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Tue, 22 May 2012 23:07:55 +0200
Subject: [Gta04-owner] bad wifi performance in kernel 3.3 but not with
 2.6.32
In-Reply-To: <1336313307.3586.421.camel@localhost>
References: <20120505160727.15782d2f@kemnade.info>
	<CAHvJMeHNsbCx57-+S0x4L-4WAH88F6FyUSE=rMJoGXKFbxBUig@mail.gmail.com>
	<1336291966.2565.2805.camel@localhost>
	<20120506082352.GA2763@tinyCurrent>
	<1336313307.3586.421.camel@localhost>
Message-ID: <20120522230755.3262390b@kemnade.info>

Hi,


> > Is this with WEP or WPA? Try changing this to the other mode.
> > 
> That was with WPA. But I'll set up another accesspoint for testing soon.
> 
I have tested with a Dlink DI-524 again, just with one computer connected to it
via lan and the gta04 via wifi.
WPA2/WEP/no encription.

Still the same picture. High speed with 2.6.32, low speed with 3.3/3.2
with 1m distance to the accesspoint.

I have tested with an older fritzbox (my ap at home) using WPA2-PSK,
some more high-end cisco accesspoints using WPA-PEAP-EAP/MSCHAPv2
and also a linksys ap using WPA2-PSK

I do not believe that I live in a world of misconfigured networks.

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120522/0545e6bd/attachment.bin>

From andreas at kemnade.info  Wed May 23 07:51:30 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Wed, 23 May 2012 07:51:30 +0200
Subject: [Gta04-owner] How to not charge your GTA04
Message-ID: <20120523075130.2e873872@kemnade.info>

Hi,

I found out some methods to get an empty battery. Well, just not plug it to a computer
or to the charger is a bit boring ;-) So lets be a bit more creative.

1. Using the openmoko charger:
- connect the charger first to the gta04 and then to the wall socket
- disconnecct the charger from the wall socket and connect it again
- have a short power failure

2. Connecting the gta04 to the computer
- use bad cables with a high resistance
  I have found several ones with a resistance between 1 and 2 ohms.
  Depending on battery state, the charger will detect an usb unplug
  at 4.3V or 4.5V usb voltage, and stops charging despite the voltage
  goes up again. If the battery is empty, you'll get the precharge done
  but then the main charge is not enabled. Not every device is so
  picky about the voltage, so you might have considered these cables
  ok. Using these cables for GTA02 is no problem for example.


about 1.
  checking for id resistor and not setting MUSB_SESSION in MUSB_DEVCTL
  if resistor is 102K should help. In all these scenarios,
  the gta04 starts providing vbus by itself and does not notice
  vbus from external.

about 2:
  setting kernel parameters allow_usb=N and the default current to 500mA
  should already help in several cases. And of course: check your
  cables! I'm not talking about extremely long cables here and I have
  also cables which look same like the bad ones but have half the resistance.

So I wish you happy battery discharging ;-)

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120523/29ee862e/attachment.bin>

From hns at goldelico.com  Thu May 24 10:54:55 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Thu, 24 May 2012 10:54:55 +0200
Subject: [Gta04-owner] Fwd: [ANNOUNCE] Cornucopia 0.11.0 - The Vala/C
	reference implementation of the freesmartphone.org APIs
References: <4FBDF559.4010003@gravedo.de>
Message-ID: <24F35135-2BD9-4DA6-B038-D0DBF61371DA@goldelico.com>

FYI

Anfang der weitergeleiteten E-Mail:

> Von: Simon Busch <morphis at gravedo.de>
> Datum: 24. Mai 2012 10:46:17 MESZ
> An: smartphones-userland at linuxtogo.org
> Betreff: [ANNOUNCE] Cornucopia 0.11.0 - The Vala/C reference implementation of the freesmartphone.org APIs
> 
> We'd like to announce the next release of all Freesmartphone.org
> middleware stack. This release includes several bug fixes and
> enhancements and a improved support for the GTA04 smartphone.

thanks to the FSO developers!
Nikolaus

> 
> Please note: starting with this release we will update the ABI version
> of all public libraries when required by API changes.
> 
> More information about the Freesmartphone.org middleware is available at
> our website: http://www.freesmartphone.org/
> 
> The release tarballs are available here:
>   http://downloads.freesmartphone.org/sources/cornucopia/0.11
> 
> The following components are included in this release:
> 
> * libfsosystem (0.11)
> * libfsobasics (0.11)
> * libfsoframework (0.11)
> * libfsotransport (0.11)
> * libfsoresource (0.11)
> * fsousaged (0.11)
> * fsodatad (0.11)
> * fsonetworkd (0.11)
> * fsotdld (0.11)
> * fsodeviced (0.11)
> * fsogsmd (0.11)
> 
> Followed by the release of the core stack there is a snapshot release of
> the API specification and it's glib-binding library too:
> 
> * specs (2012.05.24.1)
> * libfso-glib (2012.05.24.1)
> 
> The release tarballs are available here:
>   http://downloads.freesmartphone.org/sources/specs/
>   http://downloads.freesmartphone.org/sources/libfso-glib/
> 
> The FSO Team
> 
> _______________________________________________
> Smartphones-userland mailing list
> Smartphones-userland at linuxtogo.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/smartphones-userland


From morphis at gravedo.de  Thu May 24 10:58:02 2012
From: morphis at gravedo.de (Simon Busch)
Date: Thu, 24 May 2012 10:58:02 +0200
Subject: [Gta04-owner] Fwd: [ANNOUNCE] Cornucopia 0.11.0 - The Vala/C
 reference implementation of the freesmartphone.org APIs
In-Reply-To: <24F35135-2BD9-4DA6-B038-D0DBF61371DA@goldelico.com>
References: <4FBDF559.4010003@gravedo.de>
	<24F35135-2BD9-4DA6-B038-D0DBF61371DA@goldelico.com>
Message-ID: <4FBDF81A.5040606@gravedo.de>

On 24.05.2012 10:54, Dr. H. Nikolaus Schaller wrote:
> FYI
> 
> Anfang der weitergeleiteten E-Mail:
> 
>> Von: Simon Busch <morphis at gravedo.de>
>> Datum: 24. Mai 2012 10:46:17 MESZ
>> An: smartphones-userland at linuxtogo.org
>> Betreff: [ANNOUNCE] Cornucopia 0.11.0 - The Vala/C reference implementation of the freesmartphone.org APIs
>>
>> We'd like to announce the next release of all Freesmartphone.org
>> middleware stack. This release includes several bug fixes and
>> enhancements and a improved support for the GTA04 smartphone.
> 
> thanks to the FSO developers!
> Nikolaus

Thanks a lot :)

regards,
Simon

-- 
Simon Busch - http://mm.gravedo.de/blog/

From hns at goldelico.com  Thu May 24 14:24:03 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Thu, 24 May 2012 14:24:03 +0200
Subject: [Gta04-owner] Detecting headset plug: twl4030-madc IRQ issue
In-Reply-To: <1337712855.1709.5.camel@aldrin>
References: <1335633337.1729.11.camel@aldrin>
	<20120429083223.509aac7c@notabene.brown>
	<89D760AE-9013-4214-BA7A-BCFB132497AB@goldelico.com>
	<1337712855.1709.5.camel@aldrin>
Message-ID: <E9A0D857-484C-4840-8AF2-2EC5CE7B2372@goldelico.com>

Hi Paul,

Am 22.05.2012 um 20:54 schrieb PaulK:

> After bringing audio output to Replicant on GTA04, I'm back to work on
> headset plug detect. For now, I'm going with hw-validation kernel just
> to make sure the ADCs are handled correctly. 
> 
> So apparently, ADC7 is MICSENSE (no headset / short / headset) [1],
> exactly what I need. After trying some user-space madc code [2], it
> seems that ADC7 reports pretty-much the same value (around raw=10)
> regardless of headset plug state. 
> 
> What am I missing here? Any help would be greatly appreciated!

The circuit at MICSENSE tries to measure the resistance between ring 1
(the extra ring for a 4 pin headset) and 2. This should be different if there
is no headset (no current), a headset with microphone (ca. 2 kOhm), or
a 3 pin headset (short circuit) or some remote control unit with a switchable
resistor ladder.

So far the theory.

In practice I have checked the voltage VHSMIC.OUT the the current sensor
chip U710 and by default it is switched off. Therefore, the voltage at MICENSE
is 0 and raw=10 may just show some offset of the ADC.

So step 1 is to enable VHSMIC.OUT. I think by default it is enabled only
if you select the microphone start some arecord. This is to save some power.

According to

http://projects.goldelico.com/p/gta04-kernel/page/Sound/

it should be

echo 0 >/sys/devices/virtual/gpio/gpio23/value  # switch off video out
amixer set 'Analog Left Headset Mic' cap
arecord

This switches VHSMIC.OUT to 2.2V (with no headset).

I have tried to measure some MICSENSE voltages (not very precisely
and completely) and I did see 0.13V (no headset) and 0.85V (some headset).7

So you should be able to see different voltages on ADC7 (I have not tried that).

Hope this helps to get you to a reliable solution.

Please share any further observations.

Nikolaus

> 
> [1]: http://projects.goldelico.com/p/gta04-kernel/page/Devices-Overview/
> [2]: https://github.com/scottellis/madc
> 
> -- 
> Paul Kocialkowski, Replicant developer
> 
> Replicant is a fully free Android distribution
> 
> Website:        http://www.replicant.us
> Wiki/Tracker:   http://redmine.replicant.us
> 
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner


From hns at goldelico.com  Thu May 24 21:47:37 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Thu, 24 May 2012 21:47:37 +0200
Subject: [Gta04-owner] Update for Manual
Message-ID: <26894E2E-F9BF-4FB6-8D1C-7A50FF78D261@goldelico.com>

I have updated the manual to fix some glitches and added some more warnings during installation.
All changes are now summarized in a chapter 13.

http://projects.goldelico.com/p/gta04-main/downloads/46/

The main warning is about opening the display connector. This must be done carefully
not to break off the black lever. If it is broken, there is no way to repair it except unsoldering
the connector and soldering a new one. This needs professional SMD rework tools and
can't be DIY.

Here are the official instructions by Hirose (page 10&11):

http://www.hirose.co.jp/cataloge_hp/e58613007.pdf

Nikolaus

From neil at ossau.homelinux.net  Fri May 25 22:39:31 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Fri, 25 May 2012 22:39:31 +0200
Subject: [Gta04-owner] Using BMP085
Message-ID: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>

Hi all,

http://projects.goldelico.com/p/gta04-kernel/page/Mainline-Status/#wikititle_27
says that the BMP085 exists and works well, but I can't see the files
under /sys that are supposed to allow reading the current temperature
and pressure, which according to
http://www.mjmwired.net/kernel/Documentation/ABI/testing/sysfs-i2c-bmp085
are /sys/bus/i2c/devices/<busnum>-<devaddr>/pressure0_input and
/sys/bus/i2c/devices/<busnum>-<devaddr>/temp0_input.

I discovered that BMP085 is built (in Neil's kernels) as a module and so
have done "modprobe bmp085".  That caused the following BMP085-related
paths to appear under /sys:

gta04:~# find /sys -name *bmp*
/sys/bus/i2c/drivers/bmp085
/sys/module/bmp085
/sys/module/bmp085/drivers/i2c:bmp085

But I still can't find *pressure* anywhere.  Is there something else I
need to do to enable the full function of the BMP085?

Thanks,
        Neil


From christoph.mair at gmail.com  Fri May 25 23:21:20 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Fri, 25 May 2012 23:21:20 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>

Hi Neil,

The board file is not configured correctly to tell the kernel which
bus the chip is connected to.

Try: echo bmp085 0x77 > /sys/bus/i2c/devices/i2c-2/new_device

I'm not sure if 0x77 was the correct I2C address, but in general this
should make the chip available.

HTH,
  Christoph


On Fri, May 25, 2012 at 10:39 PM, Neil Jerram <neil at ossau.homelinux.net> wrote:
> Hi all,
>
> http://projects.goldelico.com/p/gta04-kernel/page/Mainline-Status/#wikititle_27
> says that the BMP085 exists and works well, but I can't see the files
> under /sys that are supposed to allow reading the current temperature
> and pressure, which according to
> http://www.mjmwired.net/kernel/Documentation/ABI/testing/sysfs-i2c-bmp085
> are /sys/bus/i2c/devices/<busnum>-<devaddr>/pressure0_input and
> /sys/bus/i2c/devices/<busnum>-<devaddr>/temp0_input.
>
> I discovered that BMP085 is built (in Neil's kernels) as a module and so
> have done "modprobe bmp085". ?That caused the following BMP085-related
> paths to appear under /sys:
>
> gta04:~# find /sys -name *bmp*
> /sys/bus/i2c/drivers/bmp085
> /sys/module/bmp085
> /sys/module/bmp085/drivers/i2c:bmp085
>
> But I still can't find *pressure* anywhere. ?Is there something else I
> need to do to enable the full function of the BMP085?
>
> Thanks,
> ? ? ? ?Neil
>
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

From neil at ossau.homelinux.net  Fri May 25 23:34:47 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Fri, 25 May 2012 23:34:47 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	(Christoph Mair's message of "Fri, 25 May 2012 23:21:20 +0200")
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
Message-ID: <87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>

Christoph Mair <christoph.mair at gmail.com> writes:

> Hi Neil,
>
> The board file is not configured correctly to tell the kernel which
> bus the chip is connected to.
>
> Try: echo bmp085 0x77 > /sys/bus/i2c/devices/i2c-2/new_device

Hooray, it works, thank you:

gta04:~# echo bmp085 0x77 > /sys/bus/i2c/devices/i2c-2/new_device
gta04:~# find /sys -name "*pressure*"
/sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
gta04:~# cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
92192

> I'm not sure if 0x77 was the correct I2C address, but in general this
> should make the chip available.

So is there a more correct fix that should go into the board file?  (Or
possibly already has, since I'm still only running 3.2.0-gta04+.)

      Neil

From christoph.mair at gmail.com  Fri May 25 23:47:57 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Fri, 25 May 2012 23:47:57 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <CAHvJMeHxXdEfX6k3Hf2x6yrqn6zbpkmq4x0QW8-TbhA84Padvw@mail.gmail.com>

On Fri, May 25, 2012 at 11:34 PM, Neil Jerram <neil at ossau.homelinux.net> wrote:
> So is there a more correct fix that should go into the board file? ?(Or
> possibly already has, since I'm still only running 3.2.0-gta04+.)

IIRC there is already something in the board file but either it is not
compiled in due to a wrong #ifdef or the entry is somehow wrong.
You have to specify that there is a chip on I2C bus "2" at address
"0x77" which can be controlled with the "bmp085" driver.. but I can't
remember the correct code ;-)

Best Regards,
  Christoph

From neil at ossau.homelinux.net  Sat May 26 09:19:02 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Sat, 26 May 2012 09:19:02 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <CAHvJMeHxXdEfX6k3Hf2x6yrqn6zbpkmq4x0QW8-TbhA84Padvw@mail.gmail.com>
	(Christoph Mair's message of "Fri, 25 May 2012 23:47:57 +0200")
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeHxXdEfX6k3Hf2x6yrqn6zbpkmq4x0QW8-TbhA84Padvw@mail.gmail.com>
Message-ID: <87wr3zqujd.fsf@neil-laptop.ossau.uklinux.net>

Christoph Mair <christoph.mair at gmail.com> writes:

> On Fri, May 25, 2012 at 11:34 PM, Neil Jerram <neil at ossau.homelinux.net> wrote:
>> So is there a more correct fix that should go into the board file? ?(Or
>> possibly already has, since I'm still only running 3.2.0-gta04+.)
>
> IIRC there is already something in the board file but either it is not
> compiled in due to a wrong #ifdef or the entry is somehow wrong.
> You have to specify that there is a chip on I2C bus "2" at address
> "0x77" which can be controlled with the "bmp085" driver.. but I can't
> remember the correct code ;-)

Thanks, I'll take a look at that.

     Neil

From neil at ossau.homelinux.net  Sat May 26 12:02:21 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Sat, 26 May 2012 12:02:21 +0200
Subject: [Gta04-owner] Debian testing touchscreen trouble
Message-ID: <87sjenqmz6.fsf@neil-laptop.ossau.uklinux.net>

Hi all,

I've just done a general Debian testing upgrade and have hit a
touchscreen problem: X restarts whenever I touch the screen.

It looks like the 1.12 release of the X server is still in early days,
so perhaps that's the source of the problem.

Anyway, others using Debian testing may want to refrain from upgrading
their X components right now...

     Neil

From christoph.mair at gmail.com  Sat May 26 14:21:14 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Sat, 26 May 2012 14:21:14 +0200
Subject: [Gta04-owner] Debian testing touchscreen trouble
In-Reply-To: <87sjenqmz6.fsf@neil-laptop.ossau.uklinux.net>
References: <87sjenqmz6.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <CAHvJMeF8b2sOXvbFxBNPcxpMTWfGRM5FxZCOj8157o0jxQ+DGQ@mail.gmail.com>

On Sat, May 26, 2012 at 12:02 PM, Neil Jerram <neil at ossau.homelinux.net> wrote:
> Hi all,
>
> I've just done a general Debian testing upgrade and have hit a
> touchscreen problem: X restarts whenever I touch the screen.
>
> It looks like the 1.12 release of the X server is still in early days,
> so perhaps that's the source of the problem.

Did you update goldelicos rootfs? IIRC there is exists a problem with
the tslib library and you have to stick to a older (or newer?)
version. There should be something in the ML archive about this topic.

Best regards,
  Christoph

From paulk at paulk.fr  Sat May 26 22:06:28 2012
From: paulk at paulk.fr (PaulK)
Date: Sat, 26 May 2012 22:06:28 +0200
Subject: [Gta04-owner] Detecting headset plug: twl4030-madc IRQ issue
In-Reply-To: <E9A0D857-484C-4840-8AF2-2EC5CE7B2372@goldelico.com>
References: <1335633337.1729.11.camel@aldrin>
	<20120429083223.509aac7c@notabene.brown>
	<89D760AE-9013-4214-BA7A-BCFB132497AB@goldelico.com>
	<1337712855.1709.5.camel@aldrin>
	<E9A0D857-484C-4840-8AF2-2EC5CE7B2372@goldelico.com>
Message-ID: <1338062788.1812.14.camel@aldrin>

Le jeudi 24 mai 2012 ? 14:24 +0200, Dr. H. Nikolaus Schaller a ?crit :
> So step 1 is to enable VHSMIC.OUT. I think by default it is enabled only
> if you select the microphone start some arecord. This is to save some power.
> 
> According to
> 
> http://projects.goldelico.com/p/gta04-kernel/page/Sound/
> 
> it should be
> 
> echo 0 >/sys/devices/virtual/gpio/gpio23/value  # switch off video out
> amixer set 'Analog Left Headset Mic' cap
> arecord

Indeed, it works (while recording): I get raw values around 887 when
there is a headset and around 21 when there is no headset. 
I have looked at other android devices code regarding jack detect and it
turns out Galaxy Nexus is using twl6030 ADC. Its driver is similar to
the twl4030 ADC one so it gives some hints. Though, micbias is set via
GPIO on Galaxy Nexus. 

Thanks a lot for your help. I think I have all the infos to get it to
work very soon!

> This switches VHSMIC.OUT to 2.2V (with no headset).
> 
> I have tried to measure some MICSENSE voltages (not very precisely
> and completely) and I did see 0.13V (no headset) and 0.85V (some headset).7
> 
> So you should be able to see different voltages on ADC7 (I have not tried that).
> 
> Hope this helps to get you to a reliable solution.
> 
> Please share any further observations.
> 
> Nikolaus



From neil at ossau.homelinux.net  Sun May 27 11:45:56 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Sun, 27 May 2012 11:45:56 +0200
Subject: [Gta04-owner] QtMoko v44
In-Reply-To: <CAGNKUN85pmWScjiTjgyuxvONasdRjTR69=dO1W+z5boDCUSfPg@mail.gmail.com>
	(Michele Brocco's message of "Fri, 13 Apr 2012 17:17:07 +0200")
References: <201204121540.45667.psonek2@seznam.cz>
	<6A9DAB94-6BE6-480C-8911-0E9AFBCBA5F9@gmx.net>
	<CAGNKUN85pmWScjiTjgyuxvONasdRjTR69=dO1W+z5boDCUSfPg@mail.gmail.com>
Message-ID: <87wr3yyn1n.fsf@neil-laptop.ossau.uklinux.net>

Michele Brocco <ssj2micvm at gmail.com> writes:

> hi Rene,
>
> this is exactly the problem I mentioned above. As stated by  Radek,
> change the file /etc/bluetooth/uart to  -s 115200 /dev/ttyO0 any
> 115200 flow and eventually increase the bluetooth speed afterwards
>
> On 4/13/12, Rene Leitner <rene.leitner at gmx.net> wrote:
>>
>> I have a problem with v44. When qtmoko is starting then the home screen is
>> changing all the time between an empty home screen and an home screen with
>> icons. And the clock is visible the hole in the middle of the screen.

I've just started using QtMoko v44 on my GTA04.  Just for the record, I
had the same problem as Michele and Rene, and the same fix (that Michele
describes above) worked for me.

       Neil

From neil at ossau.homelinux.net  Sun May 27 19:11:14 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Sun, 27 May 2012 19:11:14 +0200
Subject: [Gta04-owner] Debian testing touchscreen trouble
In-Reply-To: <CAHvJMeF8b2sOXvbFxBNPcxpMTWfGRM5FxZCOj8157o0jxQ+DGQ@mail.gmail.com>
	(Christoph Mair's message of "Sat, 26 May 2012 14:21:14 +0200")
References: <87sjenqmz6.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeF8b2sOXvbFxBNPcxpMTWfGRM5FxZCOj8157o0jxQ+DGQ@mail.gmail.com>
Message-ID: <87sjelzgzw.fsf@neil-laptop.ossau.uklinux.net>

Christoph Mair <christoph.mair at gmail.com> writes:

> On Sat, May 26, 2012 at 12:02 PM, Neil Jerram <neil at ossau.homelinux.net> wrote:
>> Hi all,
>>
>> I've just done a general Debian testing upgrade and have hit a
>> touchscreen problem: X restarts whenever I touch the screen.
>>
>> It looks like the 1.12 release of the X server is still in early days,
>> so perhaps that's the source of the problem.
>
> Did you update goldelicos rootfs?

Yes, I did; I switched it from stable to testing ages ago and have been
incrementally updating ever since.  (Because I wanted some things that
are not yet in stable.)

But now I see the problem with doing that!

> IIRC there is exists a problem with
> the tslib library and you have to stick to a older (or newer?)
> version. There should be something in the ML archive about this topic.

Yes, that rings a bell, thanks.

Well, I do have some Debian rootfs backups that I could try to revert
to, but I've decided instead to try working with QtMoko for a while - so
that's the solution to my problem for now.

Regards,
      Neil

From neilb at suse.de  Sun May 27 23:27:21 2012
From: neilb at suse.de (NeilBrown)
Date: Mon, 28 May 2012 07:27:21 +1000
Subject: [Gta04-owner] Debian testing touchscreen trouble
In-Reply-To: <87sjelzgzw.fsf@neil-laptop.ossau.uklinux.net>
References: <87sjenqmz6.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeF8b2sOXvbFxBNPcxpMTWfGRM5FxZCOj8157o0jxQ+DGQ@mail.gmail.com>
	<87sjelzgzw.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <20120528072721.36f10e4e@notabene.brown>

On Sun, 27 May 2012 19:11:14 +0200 Neil Jerram <neil at ossau.homelinux.net>
wrote:

> Christoph Mair <christoph.mair at gmail.com> writes:
> 
> > On Sat, May 26, 2012 at 12:02 PM, Neil Jerram <neil at ossau.homelinux.net> wrote:
> >> Hi all,
> >>
> >> I've just done a general Debian testing upgrade and have hit a
> >> touchscreen problem: X restarts whenever I touch the screen.
> >>
> >> It looks like the 1.12 release of the X server is still in early days,
> >> so perhaps that's the source of the problem.
> >
> > Did you update goldelicos rootfs?
> 
> Yes, I did; I switched it from stable to testing ages ago and have been
> incrementally updating ever since.  (Because I wanted some things that
> are not yet in stable.)
> 
> But now I see the problem with doing that!
> 
> > IIRC there is exists a problem with
> > the tslib library and you have to stick to a older (or newer?)
> > version. There should be something in the ML archive about this topic.
> 
> Yes, that rings a bell, thanks.
> 
> Well, I do have some Debian rootfs backups that I could try to revert
> to, but I've decided instead to try working with QtMoko for a while - so
> that's the solution to my problem for now.

Not that I want to discourage you from playing with QtMoko, but if you
do want to get X working again, download

xserver-xorg-core_1.11.4-1_armel.deb
xserver-xorg-input-tslib_0.0.6-7+b1_armel.deb
xserver-xorg-video-fbdev_0.4.2-4+b2_armel.deb
xserver-xorg-video-omap3_0.1.1.1-1+b2_armel.deb

from your local Debian mirror and

 dpkg -i xserver-xorg-*

Subsequent attempts to use "apt-get" to install things will fail until you use
   apt-get install -f
to fix it up.  But that will break X again.
I guess I really should submit a bug report.  In my spare time :-)

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120528/74999b6a/attachment.bin>

From neilb at suse.de  Mon May 28 02:24:25 2012
From: neilb at suse.de (NeilBrown)
Date: Mon, 28 May 2012 10:24:25 +1000
Subject: [Gta04-owner] Debian testing touchscreen trouble
In-Reply-To: <20120528072721.36f10e4e@notabene.brown>
References: <87sjenqmz6.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeF8b2sOXvbFxBNPcxpMTWfGRM5FxZCOj8157o0jxQ+DGQ@mail.gmail.com>
	<87sjelzgzw.fsf@neil-laptop.ossau.uklinux.net>
	<20120528072721.36f10e4e@notabene.brown>
Message-ID: <20120528102425.7b8f0c8c@notabene.brown>

On Mon, 28 May 2012 07:27:21 +1000 NeilBrown <neilb at suse.de> wrote:


> I guess I really should submit a bug report.  In my spare time :-)


http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=674821

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120528/9cf943c6/attachment.bin>

From marek.belisko at gmail.com  Mon May 28 23:53:47 2012
From: marek.belisko at gmail.com (Belisko Marek)
Date: Mon, 28 May 2012 23:53:47 +0200
Subject: [Gta04-owner] FM Transceiver - split frequency?
In-Reply-To: <CAAfyv36sJULYucODqgwpUw4ZSX+-wqA_ZDAhH4Q--xNBM4ec6w@mail.gmail.com>
References: <CA+Ot1Ow8r4kZtOw1SyiD_Ad7R6w-Aroy+5ujt+4frZwVGPC=xw@mail.gmail.com>
	<1333995641.2553.54.camel@localhost>
	<CAAfyv355=Q77Xq0uktgbtUFn76+0YTFBFVVjCbKuWWS+GsER+Q@mail.gmail.com>
	<20120412095745.26288cf3@kemnade.info>
	<1334265397.2742.46.camel@localhost>
	<CAAfyv36sJULYucODqgwpUw4ZSX+-wqA_ZDAhH4Q--xNBM4ec6w@mail.gmail.com>
Message-ID: <CAAfyv37RBMafLG=f=bTa0WzavyXCg+=EpfGnDP8JA=+XUFQ-dw@mail.gmail.com>

Hi Andreas,

any update in FM area? In previous email you mention about McBSP1 patch
but in your diff I cannot see any McBSP changes. Thanks

On Thu, Apr 19, 2012 at 11:21 PM, Belisko Marek <marek.belisko at gmail.com> wrote:
> On Thu, Apr 12, 2012 at 11:16 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
>> Hi,
>>
>> On Thu, 2012-04-12 at 09:57 +0200, Andreas Kemnade wrote:
>>> Hi,
>>>
>>> On Wed, 11 Apr 2012 23:53:47 +0200
>>> Belisko Marek <marek.belisko at gmail.com> wrote:
>>>
>>> > > About the driver. the status is:
>>> > > I can play audio data into the si4721. It arrives at the right pin.
>>> > > Clock comes from the SoC and looks sane.
>>> > Is there any driver available for that? I plan to work on FM receiver.
>>> >
>>> I have a userspace program to initialize the si4721 using /dev/i2c-2.
>>> Thhat seems to work according to the RSSI values I get. I have an unfinished
>>> patch for the 3.2 kernel for initializing McBSP1 accordingly. I'll
>>> send the diff this evening, so you can help debugging.
>>>
>> the userspace program is at:
>> http://misc.andi.de1.cc/si4721.c
>>
>> The kernel diff is attached (of course not cleaned up yet).
> I can't compile attached diff with Neils 3.3 kernel. I got following
> errors (fixes are in attachment):
> sound/soc/codecs/si47xx.c:54:29: error: conflicting types for
> ?soc_codec_dev_si47xx?
> sound/soc/codecs/si47xx.h:29:36: note: previous declaration of
> ?soc_codec_dev_si47xx? was here
>
> sound/soc/omap/gta04-fm.c:113:15: error: variable ?gta04_fm_devdata?
> has initializer but incomplete type
> sound/soc/omap/gta04-fm.c:114:2: error: unknown field ?card? specified
> in initializer
> sound/soc/omap/gta04-fm.c:114:2: warning: excess elements in struct
> initializer [enabled by default]
> sound/soc/omap/gta04-fm.c:114:2: warning: (near initialization for
> ?gta04_fm_devdata?) [enabled by default]
> sound/soc/omap/gta04-fm.c:115:2: error: unknown field ?codec_dev?
> specified in initializer
> sound/soc/omap/gta04-fm.c:115:2: warning: excess elements in struct
> initializer [enabled by default]
> sound/soc/omap/gta04-fm.c:115:2: warning: (near initialization for
> ?gta04_fm_devdata?) [enabled by default]
> sound/soc/omap/gta04-fm.c:116:2: error: unknown field ?codec_data?
> specified in initializer
> sound/soc/omap/gta04-fm.c:116:2: warning: excess elements in struct
> initializer [enabled by default]
> sound/soc/omap/gta04-fm.c:116:2: warning: (near initialization for
> ?gta04_fm_devdata?) [enabled by default]
> sound/soc/omap/gta04-fm.c:113:30: warning: ?gta04_fm_devdata? defined
> but not used [-Wunused-variable]
>
>> arecord -D default:CARD=gta04fm
>> should as far as I understand (clock has to come from the SoC) it should
>> spit out data even if the si4721 is not correctly configured.
>> But it does not. We should check the registers if there
>> waits data to be collected and dma ist just not working.
>>
>> Just one hint if you read datasheets and drivers: The chip family seems
>> to have different firmware versions, so that for older chips there
>> are chips which have a totally different interface and newer ones. For
>> the si4721 itself the problem seems not to exist. But a generic si47xx
>> driver made for older ones will not work.
>> Of course the userspace program should be merged with some driver in the
>> kernel.
>>
>> Greetings
>> Andreas Kemnade
>>
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>>
>
> regards,
>
> marek
>
> --
> as simple and primitive as possible
> -------------------------------------------------
> Marek Belisko - OPEN-NANDRA
> Freelance Developer
>
> Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
> Tel: +421 915 052 184
> skype: marekwhite
> twitter: #opennandra
> web: http://open-nandra.com

marek

-- 
as simple and primitive as possible
-------------------------------------------------
Marek Belisko - OPEN-NANDRA
Freelance Developer

Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
Tel: +421 915 052 184
skype: marekwhite
twitter: #opennandra
web: http://open-nandra.com

From neil at ossau.homelinux.net  Tue May 29 10:08:18 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Tue, 29 May 2012 10:08:18 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
Message-ID: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>

Hi Radek,

Attached is a patch for showing current pressure (from BMP085) on the
QtMoko home screen.  Please could you consider including it?

I believe it's correct and complete except for two points.

1. I probably haven't done the device-dependence properly - i.e. on a
device other than the GTA04, the home screen might show something wrong,
and there might be a pointless "Show Pressure" option in the settings.

2. I haven't added the UI to any themes apart from finxi (the default).

If you can advise about both those points, I'll be happy to do what's
needed to finish this off.

Regards,
        Neil


-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-gta04-Show-current-pressure-from-BMP085-on-home-scre.patch
Type: text/x-diff
Size: 14109 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120529/cd1c7de1/attachment.bin>

From george.refseth at arxi.no  Tue May 29 10:51:20 2012
From: george.refseth at arxi.no (George Refseth)
Date: Tue, 29 May 2012 10:51:20 +0200
Subject: [Gta04-owner] [ANN] GTA04 Keyboard prototype
In-Reply-To: <4FA82702.5050906@arxi.no>
References: <mailman.51.1336401925.152.gta04-owner@goldelico.com>
	<1336419313.2490.12.camel@Nokia-N900> <4FA82702.5050906@arxi.no>
Message-ID: <4FC48E08.8000303@arxi.no>

On 05/07/2012 09:48 PM, George Refseth wrote:
> On 07/05/12 21:35, Eric Andr? wrote:
>>
>>
>> I'd love a GTA with a keybord.
>>
>> If I dare, before it's too late for such an idea and if I am not the 
>> only one who would like that:
>>
>> Keeping the keyboard behind the device, using, let's say six fingers, 
>> three of each hand (=six keys) for the whole alphabet?
>>
>> Inspired from binary codes, it'd make 63 possible combinations.
>>
>> Longer to learn but very easy then. And we could type while looking 
>> at the screen instead of searching for tiny keys to tap on.
>>
>> Tell me if it's totally stupid or just a bit...
>> Eric
>>
>>
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
> Why not four sensors for each hand sensitive to directional movement 
> as well as pressure, that way you multiply the possibilities.
> Like one eight-touch pane.
Here is one way of accomplishing: Multi-input without using screen estate..
http://www.ocularlcd.com/library/specs_multi/F043-00224-100.pdf
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120529/cb48ab2e/attachment-0001.html>

From psonek2 at seznam.cz  Tue May 29 10:51:37 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Tue, 29 May 2012 10:51:37 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <1959700.PdKRYJD7rX@rp-core>

On Tuesday, May 29, 2012 10:08:18 AM Neil Jerram wrote:

> Hi Radek,
> 
> Attached is a patch for showing current pressure (from BMP085) on the
> QtMoko home screen.  Please could you consider including it?

Hi Neil!
nice works, it's now commited and ready for next release. I have got it 
running and it works great.

> I believe it's correct and complete except for two points.
> 
> 1. I probably haven't done the device-dependence properly - i.e. on a
> device other than the GTA04, the home screen might show something wrong,
> and there might be a pointless "Show Pressure" option in the settings.

No problem.

> 2. I haven't added the UI to any themes apart from finxi (the default).

No problem too.

Thanks a lot

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120529/c610846e/attachment.html>

From neil at ossau.homelinux.net  Tue May 29 11:18:52 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Tue, 29 May 2012 11:18:52 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net> (Neil Jerram's
	message of "Fri, 25 May 2012 23:34:47 +0200")
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <87hauzwdj7.fsf@neil-laptop.ossau.uklinux.net>

Neil Jerram <neil at ossau.homelinux.net> writes:

> Hooray, it works, thank you:
>
> gta04:~# echo bmp085 0x77 > /sys/bus/i2c/devices/i2c-2/new_device
> gta04:~# find /sys -name "*pressure*"
> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
> gta04:~# cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
> 92192

I've been looking now at temperature too
(/sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input).

I'm getting 33 degrees C [1], although the actual temperature in my
house is only 16 degrees C.  Is that because the temperature reported is
intentionally that *inside* the GTA04, and hence reasonably higher than
ambient?  If so, is 33 degrees a reasonable temperature to be measuring
inside the GTA04?

[1] To be precise, I mean that ((VALUE + 5) / 10) = 33, where VALUE is
`cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input`.

Regards,
        Neil

From hns at goldelico.com  Tue May 29 11:35:21 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Tue, 29 May 2012 11:35:21 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <87hauzwdj7.fsf@neil-laptop.ossau.uklinux.net>
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
	<87hauzwdj7.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <AE29E69D-F31C-4B8D-AA81-630E37106CC9@goldelico.com>

Hi Neil,

Am 29.05.2012 um 11:18 schrieb Neil Jerram:

> Neil Jerram <neil at ossau.homelinux.net> writes:
> 
>> Hooray, it works, thank you:
>> 
>> gta04:~# echo bmp085 0x77 > /sys/bus/i2c/devices/i2c-2/new_device
>> gta04:~# find /sys -name "*pressure*"
>> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>> gta04:~# cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>> 92192
> 
> I've been looking now at temperature too
> (/sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input).
> 
> I'm getting 33 degrees C [1], although the actual temperature in my
> house is only 16 degrees C.  Is that because the temperature reported is
> intentionally that *inside* the GTA04, and hence reasonably higher than
> ambient?  If so, is 33 degrees a reasonable temperature to be measuring

Yes.

> inside the GTA04?

Yes. It can go up to 45-50 degrees if the CPU is running at full power.
The BMP085 is mounted directly near the CPU and the PCB also
distributes the heat of other high-power components (WLAN). 

The temp measurements of the BMP085 are done for compensating the
pressure readings and are not intended to indicate the correct outside temp.

Maybe, we can try to use a thermopile sensor in some future device.

> 
> [1] To be precise, I mean that ((VALUE + 5) / 10) = 33, where VALUE is
> `cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input`.
> 
> Regards,
>        Neil
> 

BR,
Nikolaus


From shulyaka at gmail.com  Tue May 29 11:55:49 2012
From: shulyaka at gmail.com (Denis Shulyaka)
Date: Tue, 29 May 2012 13:55:49 +0400
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <AE29E69D-F31C-4B8D-AA81-630E37106CC9@goldelico.com>
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
	<87hauzwdj7.fsf@neil-laptop.ossau.uklinux.net>
	<AE29E69D-F31C-4B8D-AA81-630E37106CC9@goldelico.com>
Message-ID: <CAM_Y7LgbsXFEp7W7bk2Tyt_47btT6mig1y5jZWus1Y=g1YvMag@mail.gmail.com>

Hi,

2012/5/29 Dr. H. Nikolaus Schaller <hns at goldelico.com>:
> Hi Neil,
>
> Am 29.05.2012 um 11:18 schrieb Neil Jerram:
>
>> Neil Jerram <neil at ossau.homelinux.net> writes:
>>
>>> Hooray, it works, thank you:
>>>
>>> gta04:~# echo bmp085 0x77 > /sys/bus/i2c/devices/i2c-2/new_device
>>> gta04:~# find /sys -name "*pressure*"
>>> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>>> gta04:~# cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>>> 92192
>>
>> I've been looking now at temperature too
>> (/sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input).
>>
>> I'm getting 33 degrees C [1], although the actual temperature in my
>> house is only 16 degrees C. ?Is that because the temperature reported is
>> intentionally that *inside* the GTA04, and hence reasonably higher than
>> ambient? ?If so, is 33 degrees a reasonable temperature to be measuring
>
> Yes.
>
>> inside the GTA04?
>
> Yes. It can go up to 45-50 degrees if the CPU is running at full power.
> The BMP085 is mounted directly near the CPU and the PCB also
> distributes the heat of other high-power components (WLAN).

What about temperature sensors inside gyro and accelerometer? Are the
chips mounted farther from the CPU?

> The temp measurements of the BMP085 are done for compensating the
> pressure readings and are not intended to indicate the correct outside temp.
>
> Maybe, we can try to use a thermopile sensor in some future device.
>
>>
>> [1] To be precise, I mean that ((VALUE + 5) / 10) = 33, where VALUE is
>> `cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input`.
>>
>> Regards,
>> ? ? ? ?Neil
>>
>
> BR,
> Nikolaus
>
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

From hns at goldelico.com  Tue May 29 12:32:34 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Tue, 29 May 2012 12:32:34 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <CAM_Y7LgbsXFEp7W7bk2Tyt_47btT6mig1y5jZWus1Y=g1YvMag@mail.gmail.com>
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
	<87hauzwdj7.fsf@neil-laptop.ossau.uklinux.net>
	<AE29E69D-F31C-4B8D-AA81-630E37106CC9@goldelico.com>
	<CAM_Y7LgbsXFEp7W7bk2Tyt_47btT6mig1y5jZWus1Y=g1YvMag@mail.gmail.com>
Message-ID: <010EB839-4A80-4331-A3D4-12E749AFCFA0@goldelico.com>


Am 29.05.2012 um 11:55 schrieb Denis Shulyaka:

> Hi,
> 
> 2012/5/29 Dr. H. Nikolaus Schaller <hns at goldelico.com>:
>> Hi Neil,
>> 
>> Am 29.05.2012 um 11:18 schrieb Neil Jerram:
>> 
>>> Neil Jerram <neil at ossau.homelinux.net> writes:
>>> 
>>>> Hooray, it works, thank you:
>>>> 
>>>> gta04:~# echo bmp085 0x77 > /sys/bus/i2c/devices/i2c-2/new_device
>>>> gta04:~# find /sys -name "*pressure*"
>>>> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>>>> gta04:~# cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>>>> 92192
>>> 
>>> I've been looking now at temperature too
>>> (/sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input).
>>> 
>>> I'm getting 33 degrees C [1], although the actual temperature in my
>>> house is only 16 degrees C.  Is that because the temperature reported is
>>> intentionally that *inside* the GTA04, and hence reasonably higher than
>>> ambient?  If so, is 33 degrees a reasonable temperature to be measuring
>> 
>> Yes.
>> 
>>> inside the GTA04?
>> 
>> Yes. It can go up to 45-50 degrees if the CPU is running at full power.
>> The BMP085 is mounted directly near the CPU and the PCB also
>> distributes the heat of other high-power components (WLAN).
> 
> What about temperature sensors inside gyro and accelerometer? Are the
> chips mounted farther from the CPU?

Yes, but the PCB distributes the energy to all corners. So they show a little
lower temperature but also not any outside temp.

BTW: I think the CPU itself has another temp sensor inside. And the UMTS module.
As well as the TSC2007. So there are approx. 4-6 temp measurement positions. But
with different offsets and precision.

So, someone could try to work out a differential equation for energy diffusion through
the board and the case and use this to estimate the outside temp from the sensor
data :)

A graphical representation of the temperature profile would be similar to a
tent being raised by the temperature sources (CPU, WLAN, UMTS). The tent pegs
would be connected to the outside air temp. And the sensors are measuring the
value at different positions in between.

A simplified model could extrapolate the temp difference between two sensors
at some distance to the outside.

> 
>> The temp measurements of the BMP085 are done for compensating the
>> pressure readings and are not intended to indicate the correct outside temp.
>> 
>> Maybe, we can try to use a thermopile sensor in some future device.
>> 
>>> 
>>> [1] To be precise, I mean that ((VALUE + 5) / 10) = 33, where VALUE is
>>> `cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input`.
>>> 
>>> Regards,
>>>        Neil
>>> 
>> 
>> BR,
>> Nikolaus
>> 
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner


From ben_deering at swissmail.org  Tue May 29 13:16:54 2012
From: ben_deering at swissmail.org (Benjamin Deering)
Date: Tue, 29 May 2012 07:16:54 -0400
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <CAM_Y7LgbsXFEp7W7bk2Tyt_47btT6mig1y5jZWus1Y=g1YvMag@mail.gmail.com>
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
	<87hauzwdj7.fsf@neil-laptop.ossau.uklinux.net>
	<AE29E69D-F31C-4B8D-AA81-630E37106CC9@goldelico.com>
	<CAM_Y7LgbsXFEp7W7bk2Tyt_47btT6mig1y5jZWus1Y=g1YvMag@mail.gmail.com>
Message-ID: <4FC4B026.6020205@swissmail.org>

I did some work with temperature sensing on GTA02 for my ski wax chooser 
program.  I tried BMP085 mounted outside of the middle part of the case 
and still did not get good readings.  A later experiment showed that a 
lot of heat could transfer through the wires to affect the readings.  I 
ended up installing an MLX90614 infrared temperature sensor.  That works 
well.

Pictures of BMP085 install, BMP085 test, and MLX90614 install:
http://www.jeepingben.net/index.php?level=album&id=24
http://www.jeepingben.net/index.php?level=album&id=26

Ben

On 05/29/2012 05:55 AM, Denis Shulyaka wrote:
> Hi,
>
> 2012/5/29 Dr. H. Nikolaus Schaller<hns at goldelico.com>:
>> Hi Neil,
>>
>> Am 29.05.2012 um 11:18 schrieb Neil Jerram:
>>
>>> Neil Jerram<neil at ossau.homelinux.net>  writes:
>>>
>>>> Hooray, it works, thank you:
>>>>
>>>> gta04:~# echo bmp085 0x77>  /sys/bus/i2c/devices/i2c-2/new_device
>>>> gta04:~# find /sys -name "*pressure*"
>>>> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>>>> gta04:~# cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>>>> 92192
>>> I've been looking now at temperature too
>>> (/sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input).
>>>
>>> I'm getting 33 degrees C [1], although the actual temperature in my
>>> house is only 16 degrees C.  Is that because the temperature reported is
>>> intentionally that *inside* the GTA04, and hence reasonably higher than
>>> ambient?  If so, is 33 degrees a reasonable temperature to be measuring
>> Yes.
>>
>>> inside the GTA04?
>> Yes. It can go up to 45-50 degrees if the CPU is running at full power.
>> The BMP085 is mounted directly near the CPU and the PCB also
>> distributes the heat of other high-power components (WLAN).
> What about temperature sensors inside gyro and accelerometer? Are the
> chips mounted farther from the CPU?
>
>> The temp measurements of the BMP085 are done for compensating the
>> pressure readings and are not intended to indicate the correct outside temp.
>>
>> Maybe, we can try to use a thermopile sensor in some future device.
>>
>>> [1] To be precise, I mean that ((VALUE + 5) / 10) = 33, where VALUE is
>>> `cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input`.
>>>
>>> Regards,
>>>         Neil
>>>
>> BR,
>> Nikolaus
>>
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>


From hns at goldelico.com  Tue May 29 13:25:32 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Tue, 29 May 2012 13:25:32 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <4FC4B026.6020205@swissmail.org>
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
	<87hauzwdj7.fsf@neil-laptop.ossau.uklinux.net>
	<AE29E69D-F31C-4B8D-AA81-630E37106CC9@goldelico.com>
	<CAM_Y7LgbsXFEp7W7bk2Tyt_47btT6mig1y5jZWus1Y=g1YvMag@mail.gmail.com>
	<4FC4B026.6020205@swissmail.org>
Message-ID: <56CD9D51-D74C-43AD-83DC-38ECE5EC33BE@goldelico.com>


Am 29.05.2012 um 13:16 schrieb Benjamin Deering:

> I did some work with temperature sensing on GTA02 for my ski wax chooser program.  I tried BMP085 mounted outside of the middle part of the case and still did not get good readings.  A later experiment showed that a lot of heat could transfer through the wires to affect the readings.  I ended up installing an MLX90614 infrared temperature sensor.  That works well.

Yes, the MLX90614 is a chip like what I think we should add to a future GTA04
release for doing reliable outside temperature measurements.

> 
> Pictures of BMP085 install, BMP085 test, and MLX90614 install:
> http://www.jeepingben.net/index.php?level=album&id=24
> http://www.jeepingben.net/index.php?level=album&id=26
> 
> Ben

BR,
Nikolaus

> 
> On 05/29/2012 05:55 AM, Denis Shulyaka wrote:
>> Hi,
>> 
>> 2012/5/29 Dr. H. Nikolaus Schaller<hns at goldelico.com>:
>>> Hi Neil,
>>> 
>>> Am 29.05.2012 um 11:18 schrieb Neil Jerram:
>>> 
>>>> Neil Jerram<neil at ossau.homelinux.net>  writes:
>>>> 
>>>>> Hooray, it works, thank you:
>>>>> 
>>>>> gta04:~# echo bmp085 0x77>  /sys/bus/i2c/devices/i2c-2/new_device
>>>>> gta04:~# find /sys -name "*pressure*"
>>>>> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>>>>> gta04:~# cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>>>>> 92192
>>>> I've been looking now at temperature too
>>>> (/sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input).
>>>> 
>>>> I'm getting 33 degrees C [1], although the actual temperature in my
>>>> house is only 16 degrees C.  Is that because the temperature reported is
>>>> intentionally that *inside* the GTA04, and hence reasonably higher than
>>>> ambient?  If so, is 33 degrees a reasonable temperature to be measuring
>>> Yes.
>>> 
>>>> inside the GTA04?
>>> Yes. It can go up to 45-50 degrees if the CPU is running at full power.
>>> The BMP085 is mounted directly near the CPU and the PCB also
>>> distributes the heat of other high-power components (WLAN).
>> What about temperature sensors inside gyro and accelerometer? Are the
>> chips mounted farther from the CPU?
>> 
>>> The temp measurements of the BMP085 are done for compensating the
>>> pressure readings and are not intended to indicate the correct outside temp.
>>> 
>>> Maybe, we can try to use a thermopile sensor in some future device.
>>> 
>>>> [1] To be precise, I mean that ((VALUE + 5) / 10) = 33, where VALUE is
>>>> `cat /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/temp0_input`.
>>>> 
>>>> Regards,
>>>>        Neil
>>>> 
>>> BR,
>>> Nikolaus
>>> 
>>> _______________________________________________
>>> Gta04-owner mailing list
>>> Gta04-owner at goldelico.com
>>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>> _______________________________________________
>> Gta04-owner mailing list
>> Gta04-owner at goldelico.com
>> http://lists.goldelico.com/mailman/listinfo/gta04-owner
>> 
> 
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner


From neil at ossau.homelinux.net  Tue May 29 13:46:29 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Tue, 29 May 2012 13:46:29 +0200
Subject: [Gta04-owner] Using BMP085
In-Reply-To: <AE29E69D-F31C-4B8D-AA81-630E37106CC9@goldelico.com> (H. Nikolaus
	Schaller's message of "Tue, 29 May 2012 11:35:21 +0200")
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
	<87hauzwdj7.fsf@neil-laptop.ossau.uklinux.net>
	<AE29E69D-F31C-4B8D-AA81-630E37106CC9@goldelico.com>
Message-ID: <87d35nw6p6.fsf@neil-laptop.ossau.uklinux.net>

"Dr. H. Nikolaus Schaller" <hns at goldelico.com> writes:

>> I'm getting 33 degrees C [1], although the actual temperature in my
>> house is only 16 degrees C.  Is that because the temperature reported is
>> intentionally that *inside* the GTA04, and hence reasonably higher than
>> ambient?  If so, is 33 degrees a reasonable temperature to be measuring
>
> Yes.
>
>> inside the GTA04?
>
> Yes. It can go up to 45-50 degrees if the CPU is running at full power.
> The BMP085 is mounted directly near the CPU and the PCB also
> distributes the heat of other high-power components (WLAN). 
>
> The temp measurements of the BMP085 are done for compensating the
> pressure readings and are not intended to indicate the correct outside temp.

Thanks for the quick and clear replies!

     Neil

From neil at ossau.homelinux.net  Tue May 29 13:53:39 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Tue, 29 May 2012 13:53:39 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <1959700.PdKRYJD7rX@rp-core> (Radek Polak's message of "Tue, 29
	May 2012 10:51:37 +0200")
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<1959700.PdKRYJD7rX@rp-core>
Message-ID: <878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>

Radek Polak <psonek2 at seznam.cz> writes:

> Hi Neil!

Hi!

> nice works, it's now commited and ready for next release. I have got it running
> and it works great.

That was quick - but did your commit include the new files?
(gta04pressure.cpp and .h)

Also here's a further change to export the BMP085-reported temperature;
but doesn't add it to the UI by default, as it won't be meaningful to
most users.

It's not that useful, but maybe worth having in the codebase for
completeness?

Regards,
        Neil

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-gta04-also-export-internal-temperature-from-pressure.patch
Type: text/x-diff
Size: 1894 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120529/cb9603b8/attachment.bin>

From psonek2 at seznam.cz  Tue May 29 16:04:40 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Tue, 29 May 2012 16:04:40 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<1959700.PdKRYJD7rX@rp-core>
	<878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <2974371.6dpNV8Oylk@rp-core>

On Tuesday, May 29, 2012 01:53:39 PM Neil Jerram wrote:

> That was quick - but did your commit include the new files?
> (gta04pressure.cpp and .h)

ahh, sorry, it's fixed now

> Also here's a further change to export the BMP085-reported temperature;
> but doesn't add it to the UI by default, as it won't be meaningful to
> most users.
> 
> It's not that useful, but maybe worth having in the codebase for
> completeness?

yup, it's included now too.

Btw the sysfs path is a bit different in 3.4 kernel so i might change it if v45 
will have 3.4 kernel which is quite likely.

Regards

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120529/65cc8ce3/attachment-0001.html>

From neil at ossau.homelinux.net  Tue May 29 18:49:54 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Tue, 29 May 2012 18:49:54 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <2974371.6dpNV8Oylk@rp-core> (Radek Polak's message of "Tue, 29
	May 2012 16:04:40 +0200")
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<1959700.PdKRYJD7rX@rp-core>
	<878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>
	<2974371.6dpNV8Oylk@rp-core>
Message-ID: <874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>

Radek Polak <psonek2 at seznam.cz> writes:

> On Tuesday, May 29, 2012 01:53:39 PM Neil Jerram wrote:
>
>  
>
>> That was quick - but did your commit include the new files?
>
>> (gta04pressure.cpp and .h)
>
>  
>
> ahh, sorry, it's fixed now

Great, thanks.

>> Also here's a further change to export the BMP085-reported temperature;
>
>> but doesn't add it to the UI by default, as it won't be meaningful to
>
>> most users.
>
>>
>
>> It's not that useful, but maybe worth having in the codebase for
>
>> completeness?
>
>  
>
> yup, it's included now too.

And for that too.

> Btw the sysfs path is a bit different in 3.4 kernel so i might change it if v45
> will have 3.4 kernel which is quite likely.

Sure, makes sense.  (Do you have a trick for checking all sysfs paths when
moving to a new kernel?  Maybe just "grep -r /sys *"?)

I'm building myself and using the update-qtmoko script now.  If v45 has
a new kernel, I guess that means that I'll only have to download and
unpack the qtmoko-fat-v45.tar.gz file.  Does that sound right, or are
there other bits in the rootfs that I might miss?

     Neil

From andreas at kemnade.info  Tue May 29 22:45:06 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Tue, 29 May 2012 22:45:06 +0200
Subject: [Gta04-owner] Host AP mode
Message-ID: <1338324306.2543.655.camel@localhost>

Hi,

I found out that
http://corysohrakoff.wordpress.com/2011/09/13/enabling-wifi-ap-mode-on-a-gumstix-overo/
is a good howto about host ap mode. 
It is about using an alternate driver for the wifi chipset (libertas_tf)
which uses a thinner firmware, so the host can do ap mode. I could
configure hostapd so that I have an unencrypted accesspoint.
The libertras_tf driver needs some 
#include <linux/module.h> to work with 3.2 kernel. I have used hostapd
version 1.0.
Normal infrastructure mode does also work.

But my earlier mentioned wifi performance problem with 3.x, which does
not exist in 2.6.32, is not better with the libertas_tf driver.
Could the SD/MMC host driver be the bottleneck?

Greetings
Andreas Kemnade




From christoph.mair at gmail.com  Tue May 29 23:11:01 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Tue, 29 May 2012 23:11:01 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <1338324306.2543.655.camel@localhost>
References: <1338324306.2543.655.camel@localhost>
Message-ID: <CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>

On Tue, May 29, 2012 at 10:45 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
> Hi,
>
> I found out that
> http://corysohrakoff.wordpress.com/2011/09/13/enabling-wifi-ap-mode-on-a-gumstix-overo/
> is a good howto about host ap mode.
> It is about using an alternate driver for the wifi chipset (libertas_tf)
> which uses a thinner firmware, so the host can do ap mode. I could
> configure hostapd so that I have an unencrypted accesspoint.
> The libertras_tf driver needs some
> #include <linux/module.h> to work with 3.2 kernel. I have used hostapd
> version 1.0.
> Normal infrastructure mode does also work.

Great! The last time I gave it a try I could not connect to my
network; scanning worked fine.

> But my earlier mentioned wifi performance problem with 3.x, which does
> not exist in 2.6.32, is not better with the libertas_tf driver.
> Could the SD/MMC host driver be the bottleneck?

I tried to debug the MMC interface and found out that the chip
interface runs at 25MHz. I've read somewhere that the module should
support up to 50MHz but I could not convince the module to use this
speed. Anyways, 25MHz with a 4 bits wide interface should be enough
for everyone [TM].

Unfortunately I could not find out why the chip does not use higher
data rates than 5MBit/s. I've seen very short periods where iwconfig
displayed higher data rates (up to 48MBits/s) but the value returned
immediately to 5MBits/s when I called iwconfig again (a few times).

Best regards,
  Christoph

From GNUtoo at no-log.org  Wed May 30 00:08:24 2012
From: GNUtoo at no-log.org (Denis 'GNUtoo' Carikli)
Date: Wed, 30 May 2012 00:08:24 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <1338324306.2543.655.camel@localhost>
References: <1338324306.2543.655.camel@localhost>
Message-ID: <1338329304.23885.1.camel@gnutoo-desktop>

hi,
I made a hostap enabled kernel in the past for the gta04 but I had no
time to test it at runtime:
http://git.freesmartphone.org/?p=linux-2.6.git;a=shortlog;h=refs/heads/3.2-gta04%2Blibertas-tf-sdio

I think I even wrote to the gta04 mailing list(I don't remember if I
started a thread or responded to a mail).

Denis.



From butrus.butrus at gmail.com  Wed May 30 00:13:26 2012
From: butrus.butrus at gmail.com (Butrus Damaskus)
Date: Wed, 30 May 2012 00:13:26 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
Message-ID: <CALM6Ch9YobkS9_smw8TJhYg+T=LawqH5rBoEN3BZs0FQnAHm=w@mail.gmail.com>

On Tue, May 29, 2012 at 11:11 PM, Christoph Mair
<christoph.mair at gmail.com> wrote:
> On Tue, May 29, 2012 at 10:45 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
>> Hi,
>>
>> I found out that
>> http://corysohrakoff.wordpress.com/2011/09/13/enabling-wifi-ap-mode-on-a-gumstix-overo/
>> is a good howto about host ap mode.
>> It is about using an alternate driver for the wifi chipset (libertas_tf)
>> which uses a thinner firmware, so the host can do ap mode. I could
>> configure hostapd so that I have an unencrypted accesspoint.
>> The libertras_tf driver needs some
>> #include <linux/module.h> to work with 3.2 kernel. I have used hostapd
>> version 1.0.
>> Normal infrastructure mode does also work.
>
> Great! The last time I gave it a try I could not connect to my
> network; scanning worked fine.
>
>> But my earlier mentioned wifi performance problem with 3.x, which does
>> not exist in 2.6.32, is not better with the libertas_tf driver.
>> Could the SD/MMC host driver be the bottleneck?
>
> I tried to debug the MMC interface and found out that the chip
> interface runs at 25MHz. I've read somewhere that the module should
> support up to 50MHz but I could not convince the module to use this
> speed. Anyways, 25MHz with a 4 bits wide interface should be enough
> for everyone [TM].
>
> Unfortunately I could not find out why the chip does not use higher
> data rates than 5MBit/s. I've seen very short periods where iwconfig
> displayed higher data rates (up to 48MBits/s) but the value returned
> immediately to 5MBits/s when I called iwconfig again (a few times).


What is the signal strength/quality?

> Best regards,
> ?Christoph
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

From neilb at suse.de  Wed May 30 02:22:32 2012
From: neilb at suse.de (NeilBrown)
Date: Wed, 30 May 2012 10:22:32 +1000
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<1959700.PdKRYJD7rX@rp-core>
	<878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>
	<2974371.6dpNV8Oylk@rp-core>
	<874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <20120530102232.0ed6e55b@notabene.brown>

On Tue, 29 May 2012 18:49:54 +0200 Neil Jerram <neil at ossau.homelinux.net>
wrote:

> Sure, makes sense.  (Do you have a trick for checking all sysfs paths when
> moving to a new kernel?  Maybe just "grep -r /sys *"?)
> 

Try to avoid using paths that start /sys/devices.

Look for a link from /sys/class/ instead.  Or maybe /sys/bus/.  It is more
likely to be stable. And shorter.

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/c3fbbe3e/attachment.bin>

From christoph.mair at gmail.com  Wed May 30 08:13:17 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Wed, 30 May 2012 08:13:17 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <CALM6Ch9YobkS9_smw8TJhYg+T=LawqH5rBoEN3BZs0FQnAHm=w@mail.gmail.com>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<CALM6Ch9YobkS9_smw8TJhYg+T=LawqH5rBoEN3BZs0FQnAHm=w@mail.gmail.com>
Message-ID: <CAHvJMeF9EwQCRc-nfJey6nt3=A4+Djd3WpCmpqJwBL+p0BzL5w@mail.gmail.com>

On Wed, May 30, 2012 at 12:13 AM, Butrus Damaskus
<butrus.butrus at gmail.com> wrote:
> On Tue, May 29, 2012 at 11:11 PM, Christoph Mair
> <christoph.mair at gmail.com> wrote:
>> On Tue, May 29, 2012 at 10:45 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
>>> Hi,
>>>
>>> I found out that
>>> http://corysohrakoff.wordpress.com/2011/09/13/enabling-wifi-ap-mode-on-a-gumstix-overo/
>>> is a good howto about host ap mode.
>>> It is about using an alternate driver for the wifi chipset (libertas_tf)
>>> which uses a thinner firmware, so the host can do ap mode. I could
>>> configure hostapd so that I have an unencrypted accesspoint.
>>> The libertras_tf driver needs some
>>> #include <linux/module.h> to work with 3.2 kernel. I have used hostapd
>>> version 1.0.
>>> Normal infrastructure mode does also work.
>>
>> Great! The last time I gave it a try I could not connect to my
>> network; scanning worked fine.
>>
>>> But my earlier mentioned wifi performance problem with 3.x, which does
>>> not exist in 2.6.32, is not better with the libertas_tf driver.
>>> Could the SD/MMC host driver be the bottleneck?
>>
>> I tried to debug the MMC interface and found out that the chip
>> interface runs at 25MHz. I've read somewhere that the module should
>> support up to 50MHz but I could not convince the module to use this
>> speed. Anyways, 25MHz with a 4 bits wide interface should be enough
>> for everyone [TM].
>>
>> Unfortunately I could not find out why the chip does not use higher
>> data rates than 5MBit/s. I've seen very short periods where iwconfig
>> displayed higher data rates (up to 48MBits/s) but the value returned
>> immediately to 5MBits/s when I called iwconfig again (a few times).
>
>
> What is the signal strength/quality?

I can't remember exactly, but I also tested it standing next to the AP (2m).

Christoph

From hns at goldelico.com  Wed May 30 09:51:05 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Wed, 30 May 2012 09:51:05 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <CAHvJMeF9EwQCRc-nfJey6nt3=A4+Djd3WpCmpqJwBL+p0BzL5w@mail.gmail.com>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<CALM6Ch9YobkS9_smw8TJhYg+T=LawqH5rBoEN3BZs0FQnAHm=w@mail.gmail.com>
	<CAHvJMeF9EwQCRc-nfJey6nt3=A4+Djd3WpCmpqJwBL+p0BzL5w@mail.gmail.com>
Message-ID: <A8C6975B-0131-4046-9B5F-4EFAA1FC005F@goldelico.com>


Am 30.05.2012 um 08:13 schrieb Christoph Mair:

> On Wed, May 30, 2012 at 12:13 AM, Butrus Damaskus
> <butrus.butrus at gmail.com> wrote:
>> On Tue, May 29, 2012 at 11:11 PM, Christoph Mair
>> <christoph.mair at gmail.com> wrote:
>>> On Tue, May 29, 2012 at 10:45 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
>>>> Hi,
>>>> 
>>>> I found out that
>>>> http://corysohrakoff.wordpress.com/2011/09/13/enabling-wifi-ap-mode-on-a-gumstix-overo/
>>>> is a good howto about host ap mode.
>>>> It is about using an alternate driver for the wifi chipset (libertas_tf)
>>>> which uses a thinner firmware, so the host can do ap mode. I could
>>>> configure hostapd so that I have an unencrypted accesspoint.
>>>> The libertras_tf driver needs some
>>>> #include <linux/module.h> to work with 3.2 kernel. I have used hostapd
>>>> version 1.0.
>>>> Normal infrastructure mode does also work.
>>> 
>>> Great! The last time I gave it a try I could not connect to my
>>> network; scanning worked fine.
>>> 
>>>> But my earlier mentioned wifi performance problem with 3.x, which does
>>>> not exist in 2.6.32, is not better with the libertas_tf driver.
>>>> Could the SD/MMC host driver be the bottleneck?
>>> 
>>> I tried to debug the MMC interface and found out that the chip
>>> interface runs at 25MHz. I've read somewhere that the module should
>>> support up to 50MHz but I could not convince the module to use this
>>> speed. Anyways, 25MHz with a 4 bits wide interface should be enough
>>> for everyone [TM].
>>> 
>>> Unfortunately I could not find out why the chip does not use higher
>>> data rates than 5MBit/s. I've seen very short periods where iwconfig
>>> displayed higher data rates (up to 48MBits/s) but the value returned
>>> immediately to 5MBits/s when I called iwconfig again (a few times).
>> 
>> 
>> What is the signal strength/quality?
> 
> I can't remember exactly, but I also tested it standing next to the AP (2m).

I wonder if this is a single observation with Christoph's board (it was
one of the first that were produced) or a general issue. As far as I
followed this discussion, Andreas has high speed with the 2.6.32
kernel but not with 3.x.

Anyone else seeing this speed up/down effect?

BR,
Nikolaus


From andreas at kemnade.info  Wed May 30 08:23:49 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Wed, 30 May 2012 08:23:49 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
Message-ID: <20120530082349.7f88e606@kemnade.info>

Hi,

On Tue, 29 May 2012 23:11:01 +0200
Christoph Mair <christoph.mair at gmail.com> wrote:

> 
> > But my earlier mentioned wifi performance problem with 3.x, which does
> > not exist in 2.6.32, is not better with the libertas_tf driver.
> > Could the SD/MMC host driver be the bottleneck?
> 
> I tried to debug the MMC interface and found out that the chip
> interface runs at 25MHz. I've read somewhere that the module should
> support up to 50MHz but I could not convince the module to use this
> speed. Anyways, 25MHz with a 4 bits wide interface should be enough
> for everyone [TM].
>
Hmm, that transfers the 640KByte, which are enough for everyone in
approx 50ms. Maybe people can live with that.
Does the data get collected from there in time? Are interrupts
correctly handled?

> Unfortunately I could not find out why the chip does not use higher
> data rates than 5MBit/s. I've seen very short periods where iwconfig
> displayed higher data rates (up to 48MBits/s) but the value returned
> immediately to 5MBits/s when I called iwconfig again (a few times).

The libertastf driver adjusts the bit rates to higher values but
that does not lead to higher data rates.

But why does the standard wifi driver in kernel 2.6.32 does not
have the problem? Does the improved power management in 3.x play a
role here?
The most interesting question is probably when the problem did start.
We know:
2.6.32: no problem
3.2, 3.3, 3.4: problem

Is that problem only with GTA04A4+Kernel3.x or also with GTA04A3+Kernel3.x?
What about the first 3.x kernels?  
How to bisect that problem? Do some minimal gta04 porting for the kernels,
so that only the most important devices work? 

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/11252074/attachment.bin>

From neil at ossau.homelinux.net  Wed May 30 10:13:55 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Wed, 30 May 2012 10:13:55 +0200
Subject: [Gta04-owner] Using BMP085
References: <871um8dmgs.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeEQjL5HM3mbqM3G18hWSdT-DmRK3H3rnRzNiq+c1n5fwg@mail.gmail.com>
	<87wr40c5c8.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeHxXdEfX6k3Hf2x6yrqn6zbpkmq4x0QW8-TbhA84Padvw@mail.gmail.com>
Message-ID: <87r4u2qe64.fsf@neil-laptop.ossau.uklinux.net>

Christoph Mair <christoph.mair at gmail.com> writes:

> On Fri, May 25, 2012 at 11:34 PM, Neil Jerram <neil at ossau.homelinux.net> wrote:
>> So is there a more correct fix that should go into the board file? ?(Or
>> possibly already has, since I'm still only running 3.2.0-gta04+.)
>
> IIRC there is already something in the board file but either it is not
> compiled in due to a wrong #ifdef or the entry is somehow wrong.
> You have to specify that there is a chip on I2C bus "2" at address
> "0x77" which can be controlled with the "bmp085" driver.. but I can't
> remember the correct code ;-)

The code in the board file looks plausible to me, and appears to contain
those details that you mentioned:

static struct i2c_board_info __initdata gta04_i2c2_boardinfo[] = {
...
#ifdef CONFIG_BMP085
{
	I2C_BOARD_INFO("bmp085", 0x77),
	.type		= "bmp085",
	.platform_data	= &bmp085_info,
	.irq		=  OMAP_GPIO_IRQ(BMP085_EOC_IRQ_GPIO),
},
#endif
...
}

However in my (which means NeilBrown's) kernel build, the BMP085 driver
is built as a module, and a search for CONFIG_BMP085 gives

./include/generated/autoconf.h:3596:#define CONFIG_BMP085_MODULE 1

but nothing that defines CONFIG_BMP085 without the trailing "_MODULE".

Is is the case that that board file code would only work if the BMP085
driver was built in?

Regards,
        Neil

From psonek2 at seznam.cz  Wed May 30 10:13:13 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Wed, 30 May 2012 10:13:13 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<2974371.6dpNV8Oylk@rp-core>
	<874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <2196834.iaGdoPYacI@rp-core>

On Tuesday, May 29, 2012 06:49:54 PM Neil Jerram wrote:

> > Btw the sysfs path is a bit different in 3.4 kernel so i might change it
> > if v45 will have 3.4 kernel which is quite likely.
> 
> Sure, makes sense.  (Do you have a trick for checking all sysfs paths when
> moving to a new kernel?  Maybe just "grep -r /sys *"?)

I just check what get's broken and fix it ;-)

> I'm building myself and using the update-qtmoko script now.  If v45 has
> a new kernel, I guess that means that I'll only have to download and
> unpack the qtmoko-fat-v45.tar.gz file.  Does that sound right, or are
> there other bits in the rootfs that I might miss?

You should just install the new kernel (or do apt-get update && apt-get 
upgrade). The uImage bin is then automaticaly copied to fat by gta04-init.

I think there are no more bits needed (except that i added bmp085 module to 
/etc/modules).

Regards

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/faecf568/attachment.html>

From neil at ossau.homelinux.net  Wed May 30 10:20:14 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Wed, 30 May 2012 10:20:14 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <20120530082349.7f88e606@kemnade.info> (Andreas Kemnade's message
	of "Wed, 30 May 2012 08:23:49 +0200")
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
Message-ID: <87mx4qqdvl.fsf@neil-laptop.ossau.uklinux.net>

Andreas Kemnade <andreas at kemnade.info> writes:

> Is that problem only with GTA04A4+Kernel3.x or also with GTA04A3+Kernel3.x?

I have GTA04A3 and am running a 3.x kernel.  Subjectively I haven't
really noticed how fast (or slow) my Wifi is going, but if you would
like to specify a test for me to run, I'd be happy to run it and report
the result.

      Neil

From neil at ossau.homelinux.net  Wed May 30 10:21:56 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Wed, 30 May 2012 10:21:56 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <2196834.iaGdoPYacI@rp-core> (Radek Polak's message of "Wed, 30
	May 2012 10:13:13 +0200")
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<2974371.6dpNV8Oylk@rp-core>
	<874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
	<2196834.iaGdoPYacI@rp-core>
Message-ID: <87ipfeqdsr.fsf@neil-laptop.ossau.uklinux.net>

Radek Polak <psonek2 at seznam.cz> writes:

> I just check what get's broken and fix it ;-)

:-)

>> I'm building myself and using the update-qtmoko script now. If v45 has
>
>> a new kernel, I guess that means that I'll only have to download and
>
>> unpack the qtmoko-fat-v45.tar.gz file. Does that sound right, or are
>
>> there other bits in the rootfs that I might miss?
>
>  
>
> You should just install the new kernel (or do apt-get update && apt-get
> upgrade). The uImage bin is then automaticaly copied to fat by gta04-init.
>
> I think there are no more bits needed (except that i added bmp085 module to /
> etc/modules).

That's great, thanks.

       Neil

From neil at ossau.homelinux.net  Wed May 30 10:25:46 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Wed, 30 May 2012 10:25:46 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <20120530102232.0ed6e55b@notabene.brown> (NeilBrown's message of
	"Wed, 30 May 2012 10:22:32 +1000")
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<1959700.PdKRYJD7rX@rp-core>
	<878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>
	<2974371.6dpNV8Oylk@rp-core>
	<874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
	<20120530102232.0ed6e55b@notabene.brown>
Message-ID: <87ehq2qdmd.fsf@neil-laptop.ossau.uklinux.net>

NeilBrown <neilb at suse.de> writes:

> On Tue, 29 May 2012 18:49:54 +0200 Neil Jerram <neil at ossau.homelinux.net>
> wrote:
>
>> Sure, makes sense.  (Do you have a trick for checking all sysfs paths when
>> moving to a new kernel?  Maybe just "grep -r /sys *"?)
>> 
>
> Try to avoid using paths that start /sys/devices.
>
> Look for a link from /sys/class/ instead.  Or maybe /sys/bus/.  It is more
> likely to be stable. And shorter.

Thanks, I hadn't heard that before.

In the particular case of BMP085, though, there's unfortunately only
a /sys/devices path:

root at neo:~# find /sys -name "*pressure*"
/sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
root at neo:~# 

      Neil

From christoph.mair at gmail.com  Wed May 30 10:28:49 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Wed, 30 May 2012 10:28:49 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <87ehq2qdmd.fsf@neil-laptop.ossau.uklinux.net>
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<1959700.PdKRYJD7rX@rp-core>
	<878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>
	<2974371.6dpNV8Oylk@rp-core>
	<874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
	<20120530102232.0ed6e55b@notabene.brown>
	<87ehq2qdmd.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <CAHvJMeFgK1kq1GvAhcnTFHyT0T8sR02teeXN6eoAXDxqRy9+HA@mail.gmail.com>

On Wed, May 30, 2012 at 10:25 AM, Neil Jerram <neil at ossau.homelinux.net> wrote:
> NeilBrown <neilb at suse.de> writes:
>
>> On Tue, 29 May 2012 18:49:54 +0200 Neil Jerram <neil at ossau.homelinux.net>
>> wrote:
>>
>>> Sure, makes sense. ?(Do you have a trick for checking all sysfs paths when
>>> moving to a new kernel? ?Maybe just "grep -r /sys *"?)
>>>
>>
>> Try to avoid using paths that start /sys/devices.
>>
>> Look for a link from /sys/class/ instead. ?Or maybe /sys/bus/. ?It is more
>> likely to be stable. And shorter.
>
> Thanks, I hadn't heard that before.
>
> In the particular case of BMP085, though, there's unfortunately only
> a /sys/devices path:
>
> root at neo:~# find /sys -name "*pressure*"
> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
> root at neo:~#

There should be at least be a /sys/bus/i2c/devices/2-0077/.. (or
similar) path too.

Christoph

From psonek2 at seznam.cz  Wed May 30 10:30:28 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Wed, 30 May 2012 10:30:28 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <20120530102232.0ed6e55b@notabene.brown>
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
	<20120530102232.0ed6e55b@notabene.brown>
Message-ID: <1546418.nYmUQuBDnt@rp-core>

On Wednesday, May 30, 2012 02:22:32 AM NeilBrown wrote:

> Try to avoid using paths that start /sys/devices.
> 
> Look for a link from /sys/class/ instead.  Or maybe /sys/bus/.  It is more
> likely to be stable. And shorter.

Thanks Neil, i have found path same for 3.2 and 3.4 kernel:

https://github.com/radekp/qtmoko/commit/5621e51b490848d906fe1762cba66ca6fca9a264

Regards

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/606246dc/attachment.html>

From andreas at kemnade.info  Wed May 30 18:10:21 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Wed, 30 May 2012 18:10:21 +0200
Subject: [Gta04-owner] FM Transceiver - split frequency?
In-Reply-To: <CAAfyv37RBMafLG=f=bTa0WzavyXCg+=EpfGnDP8JA=+XUFQ-dw@mail.gmail.com>
References: <CA+Ot1Ow8r4kZtOw1SyiD_Ad7R6w-Aroy+5ujt+4frZwVGPC=xw@mail.gmail.com>
	<1333995641.2553.54.camel@localhost>
	<CAAfyv355=Q77Xq0uktgbtUFn76+0YTFBFVVjCbKuWWS+GsER+Q@mail.gmail.com>
	<20120412095745.26288cf3@kemnade.info>
	<1334265397.2742.46.camel@localhost>
	<CAAfyv36sJULYucODqgwpUw4ZSX+-wqA_ZDAhH4Q--xNBM4ec6w@mail.gmail.com>
	<CAAfyv37RBMafLG=f=bTa0WzavyXCg+=EpfGnDP8JA=+XUFQ-dw@mail.gmail.com>
Message-ID: <20120530181021.55749c30@kemnade.info>

Hi,

oh sorry, that was sending a patch without brain usage in a stupid way...
I'm sending it again in the next hours.

Greetings
Andreas Kemnade

On Mon, 28 May 2012 23:53:47 +0200
Belisko Marek <marek.belisko at gmail.com> wrote:

> Hi Andreas,
> 
> any update in FM area? In previous email you mention about McBSP1 patch
> but in your diff I cannot see any McBSP changes. Thanks
> 
> On Thu, Apr 19, 2012 at 11:21 PM, Belisko Marek <marek.belisko at gmail.com> wrote:
> > On Thu, Apr 12, 2012 at 11:16 PM, Andreas Kemnade <andreas at kemnade.info> wrote:
> >> Hi,
> >>
> >> On Thu, 2012-04-12 at 09:57 +0200, Andreas Kemnade wrote:
> >>> Hi,
> >>>
> >>> On Wed, 11 Apr 2012 23:53:47 +0200
> >>> Belisko Marek <marek.belisko at gmail.com> wrote:
> >>>
> >>> > > About the driver. the status is:
> >>> > > I can play audio data into the si4721. It arrives at the right pin.
> >>> > > Clock comes from the SoC and looks sane.
> >>> > Is there any driver available for that? I plan to work on FM receiver.
> >>> >
> >>> I have a userspace program to initialize the si4721 using /dev/i2c-2.
> >>> Thhat seems to work according to the RSSI values I get. I have an unfinished
> >>> patch for the 3.2 kernel for initializing McBSP1 accordingly. I'll
> >>> send the diff this evening, so you can help debugging.
> >>>
> >> the userspace program is at:
> >> http://misc.andi.de1.cc/si4721.c
> >>
> >> The kernel diff is attached (of course not cleaned up yet).
> > I can't compile attached diff with Neils 3.3 kernel. I got following
> > errors (fixes are in attachment):
> > sound/soc/codecs/si47xx.c:54:29: error: conflicting types for
> > ?soc_codec_dev_si47xx?
> > sound/soc/codecs/si47xx.h:29:36: note: previous declaration of
> > ?soc_codec_dev_si47xx? was here
> >
> > sound/soc/omap/gta04-fm.c:113:15: error: variable ?gta04_fm_devdata?
> > has initializer but incomplete type
> > sound/soc/omap/gta04-fm.c:114:2: error: unknown field ?card? specified
> > in initializer
> > sound/soc/omap/gta04-fm.c:114:2: warning: excess elements in struct
> > initializer [enabled by default]
> > sound/soc/omap/gta04-fm.c:114:2: warning: (near initialization for
> > ?gta04_fm_devdata?) [enabled by default]
> > sound/soc/omap/gta04-fm.c:115:2: error: unknown field ?codec_dev?
> > specified in initializer
> > sound/soc/omap/gta04-fm.c:115:2: warning: excess elements in struct
> > initializer [enabled by default]
> > sound/soc/omap/gta04-fm.c:115:2: warning: (near initialization for
> > ?gta04_fm_devdata?) [enabled by default]
> > sound/soc/omap/gta04-fm.c:116:2: error: unknown field ?codec_data?
> > specified in initializer
> > sound/soc/omap/gta04-fm.c:116:2: warning: excess elements in struct
> > initializer [enabled by default]
> > sound/soc/omap/gta04-fm.c:116:2: warning: (near initialization for
> > ?gta04_fm_devdata?) [enabled by default]
> > sound/soc/omap/gta04-fm.c:113:30: warning: ?gta04_fm_devdata? defined
> > but not used [-Wunused-variable]
> >
> >> arecord -D default:CARD=gta04fm
> >> should as far as I understand (clock has to come from the SoC) it should
> >> spit out data even if the si4721 is not correctly configured.
> >> But it does not. We should check the registers if there
> >> waits data to be collected and dma ist just not working.
> >>
> >> Just one hint if you read datasheets and drivers: The chip family seems
> >> to have different firmware versions, so that for older chips there
> >> are chips which have a totally different interface and newer ones. For
> >> the si4721 itself the problem seems not to exist. But a generic si47xx
> >> driver made for older ones will not work.
> >> Of course the userspace program should be merged with some driver in the
> >> kernel.
> >>
> >> Greetings
> >> Andreas Kemnade
> >>
> >> _______________________________________________
> >> Gta04-owner mailing list
> >> Gta04-owner at goldelico.com
> >> http://lists.goldelico.com/mailman/listinfo/gta04-owner
> >>
> >
> > regards,
> >
> > marek
> >
> > --
> > as simple and primitive as possible
> > -------------------------------------------------
> > Marek Belisko - OPEN-NANDRA
> > Freelance Developer
> >
> > Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
> > Tel: +421 915 052 184
> > skype: marekwhite
> > twitter: #opennandra
> > web: http://open-nandra.com
> 
> marek
> 
> -- 
> as simple and primitive as possible
> -------------------------------------------------
> Marek Belisko - OPEN-NANDRA
> Freelance Developer
> 
> Ruska Nova Ves 219 | Presov, 08005 Slovak Republic
> Tel: +421 915 052 184
> skype: marekwhite
> twitter: #opennandra
> web: http://open-nandra.com
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/00ab0b3a/attachment.bin>

From andreas at kemnade.info  Wed May 30 17:57:52 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Wed, 30 May 2012 17:57:52 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <87mx4qqdvl.fsf@neil-laptop.ossau.uklinux.net>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
	<87mx4qqdvl.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <20120530175752.084d26a1@kemnade.info>

Hi Neil,

On Wed, 30 May 2012 10:20:14 +0200
Neil Jerram <neil at ossau.homelinux.net> wrote:

> Andreas Kemnade <andreas at kemnade.info> writes:
> 
> > Is that problem only with GTA04A4+Kernel3.x or also with GTA04A3+Kernel3.x?
> 
> I have GTA04A3 and am running a 3.x kernel.  Subjectively I haven't
> really noticed how fast (or slow) my Wifi is going, but if you would
> like to specify a test for me to run, I'd be happy to run it and report
> the result.

Basically send something bigger over the network, exclude as much possible
bottlenecks as possible.
I did just a simple (besides of noticing that apt-get shows
low rates even with big files)
ssh somewhere_near_you "cat /dev/zero" | buffer -z 128k >/dev/null
on the freerunner. Let that run long enough so that values can stabilize.
In the beginning, values will be too low because the time needed for
possible dns resolving, logging in is also taken into account.
But the transfer is, that influence goes to zero. 
somewhere_near_you should be in your LAN.

Best is to have the same userspace with both kernels.

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/059edcdc/attachment.bin>

From psonek2 at seznam.cz  Wed May 30 11:00:24 2012
From: psonek2 at seznam.cz (Radek Polak)
Date: Wed, 30 May 2012 11:00:24 +0200
Subject: [Gta04-owner] graphics speed drop in 3.3 and 3.4 kernels
Message-ID: <3450316.xdxilV0xF8@rp-core>

Hi,
after some time on 3.3 and 3.4 kernel i installed now 3.2 kernel and i was 
quite surprised about the graphics or maybe overall speed difference.

3.2 kernel is much faster. You can easily check this on debian squeeze:

	apt-get install mplayer
	wget http://activationrecord.net/radekp/pub/impossible_is_nothing2.mp4

	cat > ~/.mplayer/config << __END__
vo=fbdev2
ao=alsa
[default]
afm=ffmpeg
vfm=ffmpeg
vf=scale=480:640
sws=0
__END__

	mplayer impossible_is_nothing2.mp4

With 3.2 kernel it plays nice, with 3.3 and 3.4 i get like 1 fps.

So i wonder if i did some mistake or everyone has this problem. Testing 
results and ideas are welcome.

Regards

Radek
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/7f190412/attachment.html>

From neil at ossau.homelinux.net  Wed May 30 10:54:46 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Wed, 30 May 2012 10:54:46 +0200
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <CAHvJMeFgK1kq1GvAhcnTFHyT0T8sR02teeXN6eoAXDxqRy9+HA@mail.gmail.com>
	(Christoph Mair's message of "Wed, 30 May 2012 10:28:49 +0200")
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<1959700.PdKRYJD7rX@rp-core>
	<878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>
	<2974371.6dpNV8Oylk@rp-core>
	<874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
	<20120530102232.0ed6e55b@notabene.brown>
	<87ehq2qdmd.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeFgK1kq1GvAhcnTFHyT0T8sR02teeXN6eoAXDxqRy9+HA@mail.gmail.com>
Message-ID: <87aa0qqca1.fsf@neil-laptop.ossau.uklinux.net>

Christoph Mair <christoph.mair at gmail.com> writes:

>> root at neo:~# find /sys -name "*pressure*"
>> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
>> root at neo:~#
>
> There should be at least be a /sys/bus/i2c/devices/2-0077/.. (or
> similar) path too.

You're right (of course, and as Radek has also found):

root at neo:~# cat /sys/bus/i2c/drivers/bmp085/2-0077/pressure0_input 
92895

So I've learned something else new: "find" apparently isn't reliable for
finding everything under /sys.

      Neil

From andreas at kemnade.info  Wed May 30 19:24:18 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Wed, 30 May 2012 19:24:18 +0200
Subject: [Gta04-owner] FM Transceiver - split frequency?
In-Reply-To: <CAAfyv37RBMafLG=f=bTa0WzavyXCg+=EpfGnDP8JA=+XUFQ-dw@mail.gmail.com>
References: <CA+Ot1Ow8r4kZtOw1SyiD_Ad7R6w-Aroy+5ujt+4frZwVGPC=xw@mail.gmail.com>
	<1333995641.2553.54.camel@localhost>
	<CAAfyv355=Q77Xq0uktgbtUFn76+0YTFBFVVjCbKuWWS+GsER+Q@mail.gmail.com>
	<20120412095745.26288cf3@kemnade.info>
	<1334265397.2742.46.camel@localhost>
	<CAAfyv36sJULYucODqgwpUw4ZSX+-wqA_ZDAhH4Q--xNBM4ec6w@mail.gmail.com>
	<CAAfyv37RBMafLG=f=bTa0WzavyXCg+=EpfGnDP8JA=+XUFQ-dw@mail.gmail.com>
Message-ID: <1338398658.3981.84.camel@localhost>

Hi,


On Mon, 2012-05-28 at 23:53 +0200, Belisko Marek wrote:
> Hi Andreas,
> 
> any update in FM area? In previous email you mention about McBSP1 patch
> but in your diff I cannot see any McBSP changes. Thanks
> 
My priority list was:
1. have reliable charging in all relevant situations
2. make my usb device at my bike work also with gta04.
   (provides some sensor input)
3. and then the audio stuff (including the fm stuff)

1. and 2. had some more problems then I thought and I have written a lot
about my experiences there. But they are mostly solved now. Only
producing some clean patches out of it might be interesting.

and of course everything which annoys me.

I have not many news about FM.
I did a register comparison between McBCP1 and 2 using /dev/mem. The
diff is attached. The data in the MCBSPLP_DRR_REG seems not to change
continously, but at least after every boot I have a different value
there.
But there is nothing suspicious. The registers can only be read if the
device is active.

And of course the full patch, I have now set up a proper git branch for
that. The patch is against 3.2

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: mcbsp.diff
Type: text/x-patch
Size: 1110 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/d3aa451a/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: si4721.diff
Type: text/x-patch
Size: 20189 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120530/d3aa451a/attachment-0001.bin>

From jose.charters at free.fr  Wed May 30 20:48:17 2012
From: jose.charters at free.fr (Jose CHARTERS)
Date: Wed, 30 May 2012 20:48:17 +0200
Subject: [Gta04-owner] GTA04
Message-ID: <4FC66B71.9090205@free.fr>

Hello,

Sorry, I know that this list is in english. But I need to meet somebody 
to help me. And I live in France.

Now I will write in french.

Bonjour,

J'ai achet? un GTA04, et je l'ai re?u, il y a un mois.
J'ai install? une debian ? partir du site suivant : 
http://projects.goldelico.com/p/gta04-rootfs/
Les binaires sont dans le site : http://download.goldelico.com/gta04/

Cela a fonctionn?, nickel.

Trouvant que la Debian n'?tait pas adapt? pour une utilisation avec le 
doigt et pour un t?l?phone, j'ai voulu mettre une QTMoko.

Et depuis, mon GTA04 ne s'allume pas du tout. J'ai chang? la ?SD, remis 
une debian, rien a faire, le GTA04 reste d?sesp?rement ?teint.

Je me sens seul et j'ai besoin d'une aide concr?te. Je me demande si il 
n'y a pas un probl?me hardware mais je n'ai pas le moyen de m'en assurer.

Aussi, je souhaite rencontrer un d?tenteur de GTA04 qui fonctionne 
correctement, pour refaire l'installation avec lui, m'assurer que ma 
batterie est op?rationnelle, v?rifier mes ?SD,etc... Peut ?tre ai je 
rater une manip que je ne vois pas.

J'habite pr?s de Paris, si donc une personne est pr?te ? me rencontrer 
pour faire ces manips, je l'en remercie ?norm?ment.

J'ai hate de pourvoir utiliser mon GTA04.

Jos? Charters

From christoph.mair at gmail.com  Wed May 30 21:27:15 2012
From: christoph.mair at gmail.com (Christoph Mair)
Date: Wed, 30 May 2012 21:27:15 +0200
Subject: [Gta04-owner] GTA04
In-Reply-To: <4FC66B71.9090205@free.fr>
References: <4FC66B71.9090205@free.fr>
Message-ID: <CAHvJMeE5TX5g90Z3wOizcyO34waf7rONOVLvsBewKTSvf2j72Q@mail.gmail.com>

Hi,

Neither do I live in france nor do I speak the language, so the
following is just guesswork from with the help of google translate:

Do you own simple measurement equipment to measure voltages (any cheap
multimeter will do the job) or maybe you own the serial debug cable?
With the multimeter you could measure the battery voltage. It should
be above 3.7V. If it isn't, try to a "recovery charge" (you can also
try this without measuring anything):
- make a fresh SD card with the binaries from qtkoko or debian
- insert the card and the battery
- connect the device to a USB host (laptop, computer, ..)
- wait at least for half an hour
- the GTA04 should charge the battery very slowly and at some point in
time it tries to power on. You will notice that the display backlight
switches on.
- if there is not enough charge to boot your OS, it will switch off
and try again. Just leave it there for a few hours and it should work
again.

To test if you bricked the USB socket, connect the GTA04 to a USB host
and then pull the battery. The display should start flickering. At
least my GTA04A3 board behaves like this. Maybe something changed with
the A4 version.

Of course you could also try to insert a known good BL-5C Nokia battery.

If you own the serial debug cable you will get a lot more debugging options...

Hope this helps somehow.

Best regards,
  Christoph


2012/5/30 Jose CHARTERS <jose.charters at free.fr>:
> Hello,
>
> Sorry, I know that this list is in english. But I need to meet somebody to
> help me. And I live in France.
>
> Now I will write in french.
>
> Bonjour,
>
> J'ai achet? un GTA04, et je l'ai re?u, il y a un mois.
> J'ai install? une debian ? partir du site suivant :
> http://projects.goldelico.com/p/gta04-rootfs/
> Les binaires sont dans le site : http://download.goldelico.com/gta04/
>
> Cela a fonctionn?, nickel.
>
> Trouvant que la Debian n'?tait pas adapt? pour une utilisation avec le doigt
> et pour un t?l?phone, j'ai voulu mettre une QTMoko.
>
> Et depuis, mon GTA04 ne s'allume pas du tout. J'ai chang? la ?SD, remis une
> debian, rien a faire, le GTA04 reste d?sesp?rement ?teint.
>
> Je me sens seul et j'ai besoin d'une aide concr?te. Je me demande si il n'y
> a pas un probl?me hardware mais je n'ai pas le moyen de m'en assurer.
>
> Aussi, je souhaite rencontrer un d?tenteur de GTA04 qui fonctionne
> correctement, pour refaire l'installation avec lui, m'assurer que ma
> batterie est op?rationnelle, v?rifier mes ?SD,etc... Peut ?tre ai je rater
> une manip que je ne vois pas.
>
> J'habite pr?s de Paris, si donc une personne est pr?te ? me rencontrer pour
> faire ces manips, je l'en remercie ?norm?ment.
>
> J'ai hate de pourvoir utiliser mon GTA04.
>
> Jos? Charters
> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner

From andreas at kemnade.info  Wed May 30 22:38:47 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Wed, 30 May 2012 22:38:47 +0200
Subject: [Gta04-owner] graphics speed drop in 3.3 and 3.4 kernels
In-Reply-To: <3450316.xdxilV0xF8@rp-core>
References: <3450316.xdxilV0xF8@rp-core>
Message-ID: <1338410327.3981.873.camel@localhost>

Hi,

On Wed, 2012-05-30 at 11:00 +0200, Radek Polak wrote:
> Hi,
> after some time on 3.3 and 3.4 kernel i installed now 3.2 kernel and i was 
> quite surprised about the graphics or maybe overall speed difference.
> 
> 3.2 kernel is much faster. You can easily check this on debian squeeze:
> 
> 	apt-get install mplayer
> 	wget http://activationrecord.net/radekp/pub/impossible_is_nothing2.mp4
With 3.2 wget gives 100KByte/s, with 2.6.32 wget gives 200-600KByte/s.
But thats another story...

> 
> 	cat > ~/.mplayer/config << __END__
> vo=fbdev2
> ao=alsa
> [default]
> afm=ffmpeg
> vfm=ffmpeg
> vf=scale=480:640
> sws=0
> __END__
> 
> 	mplayer impossible_is_nothing2.mp4
> 
> With 3.2 kernel it plays nice, with 3.3 and 3.4 i get like 1 fps.
> 
I get more. Graphics look fluent with 3.4, 3.2 and 2.6.32.

gta04:~# uname -a
Linux gta04 3.4.0-gta04+ #63 PREEMPT Mon May 28 08:42:14 CEST 2012
armv7l GNU/Linux

with the default config except the rootfs.

But what mplayer displays looks a bit crazy. The A:-Number is running
like hell, the V:-Number runs like 1 per second.
A:245075.0 V: 233.4 A-V:244841.578 ct: 23.339 5836/5836 63% 34%  5.2%
5776 0 

> So i wonder if i did some mistake or everyone has this problem. Testing 
> results and ideas are welcome.
> 
I have a GTA04A4 with the DM37xx processor, is the 3.4 only fast with
that one?

Greetings
Andreas Kemnade



From neilb at suse.de  Wed May 30 23:29:56 2012
From: neilb at suse.de (NeilBrown)
Date: Thu, 31 May 2012 07:29:56 +1000
Subject: [Gta04-owner] QtMoko patch to show pressure from BMP085 device
In-Reply-To: <87aa0qqca1.fsf@neil-laptop.ossau.uklinux.net>
References: <87likbwgst.fsf@neil-laptop.ossau.uklinux.net>
	<1959700.PdKRYJD7rX@rp-core>
	<878vgbw6d8.fsf@neil-laptop.ossau.uklinux.net>
	<2974371.6dpNV8Oylk@rp-core>
	<874nqzvsnh.fsf@neil-laptop.ossau.uklinux.net>
	<20120530102232.0ed6e55b@notabene.brown>
	<87ehq2qdmd.fsf@neil-laptop.ossau.uklinux.net>
	<CAHvJMeFgK1kq1GvAhcnTFHyT0T8sR02teeXN6eoAXDxqRy9+HA@mail.gmail.com>
	<87aa0qqca1.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <20120531072956.625a593f@notabene.brown>

On Wed, 30 May 2012 10:54:46 +0200 Neil Jerram <neil at ossau.homelinux.net>
wrote:

> Christoph Mair <christoph.mair at gmail.com> writes:
> 
> >> root at neo:~# find /sys -name "*pressure*"
> >> /sys/devices/platform/omap/omap_i2c.2/i2c-2/2-0077/pressure0_input
> >> root at neo:~#
> >
> > There should be at least be a /sys/bus/i2c/devices/2-0077/.. (or
> > similar) path too.
> 
> You're right (of course, and as Radek has also found):
> 
> root at neo:~# cat /sys/bus/i2c/drivers/bmp085/2-0077/pressure0_input 
> 92895
> 
> So I've learned something else new: "find" apparently isn't reliable for
> finding everything under /sys.

No.  One of my gripes about sysfs - it isn't easily discoverable.

Possibly best way to find a good name is to start with the 'subsystem' link.

gta04:/# cd /sys/devices/platform/omap_i2c.2/i2c-2/2-0077
gta04:/sys/devices/platform/omap_i2c.2/i2c-2/2-0077# ls -l subsystem
lrwxrwxrwx 1 root root 0 Jan  1 11:03 subsystem -> ../../../../../bus/i2c


so start looking in /sys/bus/i2c.  For 'bus' you need to look in a 'devices'
subdirectory.  For 'class' you don't.

gta04:/sys/devices/platform/omap_i2c.2/i2c-2/2-0077# cd /sys/bus/i2c/devices
gta04:/sys/bus/i2c/devices# ls
1-0048	1-004a	2-0041	2-0048	i2c-1  i2c-3
1-0049	1-004b	2-0045	2-0077	i2c-2

2-0077 looks likely:

gta04:/sys/bus/i2c/devices#  ls -l 2-0077
lrwxrwxrwx 1 root root 0 Jan  1 11:05 2-0077 -> ../../../devices/platform/omap_i2c.2/i2c-2/2-0077
gta04:/sys/bus/i2c/devices#  ls -l 2-0077/.
total 0
lrwxrwxrwx 1 root root    0 Jan  1 11:05 driver -> ../../../../../bus/i2c/drivers/bmp085
-r--r--r-- 1 root root 4096 Jan  1 11:05 modalias
-r--r--r-- 1 root root 4096 Jan  1 11:05 name
-rw-r--r-- 1 root root 4096 Jan  1 11:05 oversampling
drwxr-xr-x 2 root root    0 Jan  1 11:05 power
-r--r--r-- 1 root root 4096 Jan  1 11:05 pressure0_input
lrwxrwxrwx 1 root root    0 Jan  1 11:03 subsystem -> ../../../../../bus/i2c
-r--r--r-- 1 root root 4096 Jan  1 11:05 temp0_input
-rw-r--r-- 1 root root 4096 Jan  1 11:03 uevent

Yes, that is it.  so  
    /sys/bus/i2c/devices/2-0077/pressure0_input

is a canonical name (as has already been discovered).

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120531/2f2be42d/attachment.bin>

From neilb at suse.de  Thu May 31 03:00:44 2012
From: neilb at suse.de (NeilBrown)
Date: Thu, 31 May 2012 11:00:44 +1000
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <20120530082349.7f88e606@kemnade.info>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
Message-ID: <20120531110044.03a7f05b@notabene.brown>

On Wed, 30 May 2012 08:23:49 +0200 Andreas Kemnade <andreas at kemnade.info>
wrote:

> Hi,
> 
> On Tue, 29 May 2012 23:11:01 +0200
> Christoph Mair <christoph.mair at gmail.com> wrote:
> 
> > 
> > > But my earlier mentioned wifi performance problem with 3.x, which does
> > > not exist in 2.6.32, is not better with the libertas_tf driver.
> > > Could the SD/MMC host driver be the bottleneck?
> > 
> > I tried to debug the MMC interface and found out that the chip
> > interface runs at 25MHz. I've read somewhere that the module should
> > support up to 50MHz but I could not convince the module to use this
> > speed. Anyways, 25MHz with a 4 bits wide interface should be enough
> > for everyone [TM].
> >
> Hmm, that transfers the 640KByte, which are enough for everyone in
> approx 50ms. Maybe people can live with that.
> Does the data get collected from there in time? Are interrupts
> correctly handled?
> 
> > Unfortunately I could not find out why the chip does not use higher
> > data rates than 5MBit/s. I've seen very short periods where iwconfig
> > displayed higher data rates (up to 48MBits/s) but the value returned
> > immediately to 5MBits/s when I called iwconfig again (a few times).
> 
> The libertastf driver adjusts the bit rates to higher values but
> that does not lead to higher data rates.
> 
> But why does the standard wifi driver in kernel 2.6.32 does not
> have the problem? Does the improved power management in 3.x play a
> role here?
> The most interesting question is probably when the problem did start.
> We know:
> 2.6.32: no problem
> 3.2, 3.3, 3.4: problem
> 
> Is that problem only with GTA04A4+Kernel3.x or also with GTA04A3+Kernel3.x?
> What about the first 3.x kernels?  
> How to bisect that problem? Do some minimal gta04 porting for the kernels,
> so that only the most important devices work? 
> 
> Greetings
> Andreas Kemnade

I can confirm that I see the same slowdown.

With 2.6.32, it reports a Bit Rate of 24Mb/s and a link quality of 39/100,
and I get about 1.1 to 1.4 MB/s throughput.

With 3.4, it reports 5.5Mb/sec (and won't let me change it) with a link
quality of 63/70 (!) and I get about 150 kB/s throughput.

So it does look like something is broken in 3.4.
The difference between the hw kernel and 3.4 is 

 34 files changed, 5633 insertions(+), 3319 deletions(-)

which is quite a lot:

 ../gta04-hw/drivers/net/wireless/libertas//assoc.c |only
 ../gta04-hw/drivers/net/wireless/libertas//assoc.h |only
 ../gta04-hw/drivers/net/wireless/libertas//scan.c  |only
 ../gta04-hw/drivers/net/wireless/libertas//scan.h  |only
 ../gta04-hw/drivers/net/wireless/libertas//wext.c  |only
 ../gta04-hw/drivers/net/wireless/libertas//wext.h  |only
....
 ./drivers/net/wireless/libertas/cfg.c              | 2142 ++++++++++++++++++++-
....

Maybe scan and assoc were merged into cfg ??

The difference between vanilla 2.6.32 and the hw kernel are about as big:

 35 files changed, 3888 insertions(+), 4293 deletions(-)

Between vanilla 2.6.32 and 3.4 is 

 38 files changed, 7449 insertions(+), 11722 deletions(-)

So lots of changes all over the place.
I don't think there is any real chance of usefully bisecting this.
One possible approach would be to revert all the libertas changes
from hw and see if that compiles, runs, and is slow.  If it was,
you could then add all the changes back in a new branch and run
git-bisect on that to see where it gets fixes.
I don't have a lot of hope for this though.

You would probably be better off posting a question to some relevant mailing list.
Daniel Drake <dsd at laptop.org> seems to be the main contributor, so 'cc' him.
Dan Williams <dcbw at redhat.com> is listed as the maintainer.
libertas-dev at lists.infradead.org and linux-wireless at vger.kernel.org 
seem to be the relevant email lists.

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120531/a230033b/attachment.bin>

From andreas at kemnade.info  Thu May 31 08:14:35 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Thu, 31 May 2012 08:14:35 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <20120531110044.03a7f05b@notabene.brown>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
	<20120531110044.03a7f05b@notabene.brown>
Message-ID: <20120531081435.54820aca@kemnade.info>

Hi,

On Thu, 31 May 2012 11:00:44 +1000
NeilBrown <neilb at suse.de> wrote:

>
> I can confirm that I see the same slowdown.
> 
> With 2.6.32, it reports a Bit Rate of 24Mb/s and a link quality of 39/100,
> and I get about 1.1 to 1.4 MB/s throughput.
> 
> With 3.4, it reports 5.5Mb/sec (and won't let me change it) with a link
> quality of 63/70 (!) and I get about 150 kB/s throughput.
> 
> So it does look like something is broken in 3.4.
> The difference between the hw kernel and 3.4 is 
> 
>  34 files changed, 5633 insertions(+), 3319 deletions(-)
> 
> which is quite a lot:
> 
> [,,,]
> 
> You would probably be better off posting a question to some relevant mailing list.
> Daniel Drake <dsd at laptop.org> seems to be the main contributor, so 'cc' him.
> Dan Williams <dcbw at redhat.com> is listed as the maintainer.
> libertas-dev at lists.infradead.org and linux-wireless at vger.kernel.org 
> seem to be the relevant email lists.
> 
I asked google a bit:

http://www.spinics.net/lists/linux-wireless/msg52355.html

To cite a bit:
> What SDIO host controller? ?Does the host controller support interrupts
> and 4-bit mode? ?We've always been able to get good performance out of
> laptop host controllers (Ricoh mostly) and the largest source of
> performance issues has always been quirky hardware or non-optimized SDHC
> drivers. ?Are there any errata for your SDHC that you're aware of?

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120531/24ebbc2e/attachment.bin>

From neilb at suse.de  Thu May 31 08:31:52 2012
From: neilb at suse.de (NeilBrown)
Date: Thu, 31 May 2012 16:31:52 +1000
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <20120531081435.54820aca@kemnade.info>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
	<20120531110044.03a7f05b@notabene.brown>
	<20120531081435.54820aca@kemnade.info>
Message-ID: <20120531163152.2ac35d94@notabene.brown>

On Thu, 31 May 2012 08:14:35 +0200 Andreas Kemnade <andreas at kemnade.info>
wrote:

> Hi,
> 
> On Thu, 31 May 2012 11:00:44 +1000
> NeilBrown <neilb at suse.de> wrote:
> 
> >
> > I can confirm that I see the same slowdown.
> > 
> > With 2.6.32, it reports a Bit Rate of 24Mb/s and a link quality of 39/100,
> > and I get about 1.1 to 1.4 MB/s throughput.
> > 
> > With 3.4, it reports 5.5Mb/sec (and won't let me change it) with a link
> > quality of 63/70 (!) and I get about 150 kB/s throughput.
> > 
> > So it does look like something is broken in 3.4.
> > The difference between the hw kernel and 3.4 is 
> > 
> >  34 files changed, 5633 insertions(+), 3319 deletions(-)
> > 
> > which is quite a lot:
> > 
> > [,,,]
> > 
> > You would probably be better off posting a question to some relevant mailing list.
> > Daniel Drake <dsd at laptop.org> seems to be the main contributor, so 'cc' him.
> > Dan Williams <dcbw at redhat.com> is listed as the maintainer.
> > libertas-dev at lists.infradead.org and linux-wireless at vger.kernel.org 
> > seem to be the relevant email lists.
> > 
> I asked google a bit:
> 
> http://www.spinics.net/lists/linux-wireless/msg52355.html
> 
> To cite a bit:
> > What SDIO host controller? ?Does the host controller support interrupts
> > and 4-bit mode? ?We've always been able to get good performance out of
> > laptop host controllers (Ricoh mostly) and the largest source of
> > performance issues has always been quirky hardware or non-optimized SDHC
> > drivers. ?Are there any errata for your SDHC that you're aware of?

Interesting. However that email thread is nearly 2 years old, and the
mmc driver in 3.4 does appear to support 4-bit mode and interrupts.
("grep mmc1 /proc/interrupts" shows lots of interrupts anyway).

So I suspect that the issues today are very different to the issues then.

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120531/49963621/attachment.bin>

From neil at ossau.homelinux.net  Thu May 31 09:21:51 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Thu, 31 May 2012 09:21:51 +0200
Subject: [Gta04-owner] graphics speed drop in 3.3 and 3.4 kernels
In-Reply-To: <3450316.xdxilV0xF8@rp-core> (Radek Polak's message of "Wed, 30
	May 2012 11:00:24 +0200")
References: <3450316.xdxilV0xF8@rp-core>
Message-ID: <8762bcu86o.fsf@neil-laptop.ossau.uklinux.net>

Radek Polak <psonek2 at seznam.cz> writes:

> Hi,
>
> after some time on 3.3 and 3.4 kernel i installed now 3.2 kernel and i was
> quite surprised about the graphics or maybe overall speed difference.
>
>  
>
> 3.2 kernel is much faster. You can easily check this on debian squeeze:
>
>  
>
> apt-get install mplayer
>
> wget http://activationrecord.net/radekp/pub/impossible_is_nothing2.mp4

> With 3.2 kernel it plays nice, with 3.3 and 3.4 i get like 1 fps.

I can confirm that that video plays beautifully on QtMoko, which
currently has kernel:

root at neo:~# uname -a
Linux neo 3.2.14-gta04+ #1 PREEMPT Tue Apr 10 16:42:22 UTC 2012 armv7l GNU/Linux

I haven't tried a 3.3 or 3.4 kernel yet.

     Neil

From neil at ossau.homelinux.net  Thu May 31 09:28:58 2012
From: neil at ossau.homelinux.net (Neil Jerram)
Date: Thu, 31 May 2012 09:28:58 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <20120530175752.084d26a1@kemnade.info> (Andreas Kemnade's message
	of "Wed, 30 May 2012 17:57:52 +0200")
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
	<87mx4qqdvl.fsf@neil-laptop.ossau.uklinux.net>
	<20120530175752.084d26a1@kemnade.info>
Message-ID: <871um0u7ut.fsf@neil-laptop.ossau.uklinux.net>

Andreas Kemnade <andreas at kemnade.info> writes:

> Basically send something bigger over the network, exclude as much possible
> bottlenecks as possible.

I downloaded Radek's video onto my laptop, then used scp to transfer
onto the phone over Wifi and USB.

Over Wifi:

neil at neil-laptop:~/Downloads$ scp impossible_is_nothing2.mp4 root at 192.168.24.104:/media/card/
impossible_is_nothing2.mp4                    100%   18MB 113.3KB/s   02:45    

scp initially reported the transfer speed as around 2Mb/s, but then it
dropped steadily until settling at around 100Kb/s.

Over USB:

neil at neil-laptop:~/Downloads$ scp impossible_is_nothing2.mp4 root at 192.168.0.202:/media/card/
impossible_is_nothing2.mp4                    100%   18MB   3.0MB/s   00:06    

Does that help?

      Neil

From andreas at kemnade.info  Thu May 31 09:44:12 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Thu, 31 May 2012 09:44:12 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <871um0u7ut.fsf@neil-laptop.ossau.uklinux.net>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
	<87mx4qqdvl.fsf@neil-laptop.ossau.uklinux.net>
	<20120530175752.084d26a1@kemnade.info>
	<871um0u7ut.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <20120531094412.7aa9e074@kemnade.info>

Hi,

On Thu, 31 May 2012 09:28:58 +0200
Neil Jerram <neil at ossau.homelinux.net> wrote:

> Andreas Kemnade <andreas at kemnade.info> writes:
> 
> > Basically send something bigger over the network, exclude as much possible
> > bottlenecks as possible.
> 
> I downloaded Radek's video onto my laptop, then used scp to transfer
> onto the phone over Wifi and USB.
> 
> Over Wifi:
> 
> neil at neil-laptop:~/Downloads$ scp impossible_is_nothing2.mp4 root at 192.168.24.104:/media/card/
> impossible_is_nothing2.mp4                    100%   18MB 113.3KB/s   02:45    
> 
> scp initially reported the transfer speed as around 2Mb/s, but then it
> dropped steadily until settling at around 100Kb/s.
> 
That is the rate seen from sending side, so scp can first fill up local buffers,
which is of course fast, and then has to wait for the network. That is the explanation
for the 2MB/s at the beginning. But it is another slow wifi experience with 3.2.
I finally know I am not alone, as Neil Brown said it is time to ask upstream.

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120531/cad1e848/attachment.bin>

From neilb at suse.de  Thu May 31 11:05:48 2012
From: neilb at suse.de (NeilBrown)
Date: Thu, 31 May 2012 19:05:48 +1000
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <871um0u7ut.fsf@neil-laptop.ossau.uklinux.net>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
	<87mx4qqdvl.fsf@neil-laptop.ossau.uklinux.net>
	<20120530175752.084d26a1@kemnade.info>
	<871um0u7ut.fsf@neil-laptop.ossau.uklinux.net>
Message-ID: <20120531190548.742547e3@notabene.brown>

On Thu, 31 May 2012 09:28:58 +0200 Neil Jerram <neil at ossau.homelinux.net>
wrote:

> Andreas Kemnade <andreas at kemnade.info> writes:
> 
> > Basically send something bigger over the network, exclude as much possible
> > bottlenecks as possible.
> 
> I downloaded Radek's video onto my laptop, then used scp to transfer
> onto the phone over Wifi and USB.
> 
> Over Wifi:
> 
> neil at neil-laptop:~/Downloads$ scp impossible_is_nothing2.mp4 root at 192.168.24.104:/media/card/
> impossible_is_nothing2.mp4                    100%   18MB 113.3KB/s   02:45    
> 
> scp initially reported the transfer speed as around 2Mb/s, but then it
> dropped steadily until settling at around 100Kb/s.
> 
> Over USB:
> 
> neil at neil-laptop:~/Downloads$ scp impossible_is_nothing2.mp4 root at 192.168.0.202:/media/card/
> impossible_is_nothing2.mp4                    100%   18MB   3.0MB/s   00:06    
> 
> Does that help?

I think the key thing to be looking at is 
  iwconfig wlan0 | grep Rate

I always see 5.5 Mb/s with 3.4 and much higher numbers with the hw-validation
kernel.  Given that, the actual throughput becomes rather uninteresting.

NeilBrown
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 828 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120531/cacc5212/attachment.bin>

From andreas at kemnade.info  Thu May 31 12:57:23 2012
From: andreas at kemnade.info (Andreas Kemnade)
Date: Thu, 31 May 2012 12:57:23 +0200
Subject: [Gta04-owner] Host AP mode
In-Reply-To: <20120531190548.742547e3@notabene.brown>
References: <1338324306.2543.655.camel@localhost>
	<CAHvJMeEWKbArweRBcwZMtGaUFJxEXjPuLfg_aEsx7JkjQFJFcQ@mail.gmail.com>
	<20120530082349.7f88e606@kemnade.info>
	<87mx4qqdvl.fsf@neil-laptop.ossau.uklinux.net>
	<20120530175752.084d26a1@kemnade.info>
	<871um0u7ut.fsf@neil-laptop.ossau.uklinux.net>
	<20120531190548.742547e3@notabene.brown>
Message-ID: <20120531125723.29eaeda8@kemnade.info>

Hi,

On Thu, 31 May 2012 19:05:48 +1000
NeilBrown <neilb at suse.de> wrote:

> On Thu, 31 May 2012 09:28:58 +0200 Neil Jerram <neil at ossau.homelinux.net>
> wrote:
> 
> > Andreas Kemnade <andreas at kemnade.info> writes:
> > 
> > > Basically send something bigger over the network, exclude as much possible
> > > bottlenecks as possible.
> > 
> > I downloaded Radek's video onto my laptop, then used scp to transfer
> > onto the phone over Wifi and USB.
> > 
> > Over Wifi:
> > 
> > neil at neil-laptop:~/Downloads$ scp impossible_is_nothing2.mp4 root at 192.168.24.104:/media/card/
> > impossible_is_nothing2.mp4                    100%   18MB 113.3KB/s   02:45    
> > 
> > scp initially reported the transfer speed as around 2Mb/s, but then it
> > dropped steadily until settling at around 100Kb/s.
> > 
> > Over USB:
> > 
> > neil at neil-laptop:~/Downloads$ scp impossible_is_nothing2.mp4 root at 192.168.0.202:/media/card/
> > impossible_is_nothing2.mp4                    100%   18MB   3.0MB/s   00:06    
> > 
> > Does that help?
> 
> I think the key thing to be looking at is 
>   iwconfig wlan0 | grep Rate
> 
> I always see 5.5 Mb/s with 3.4 and much higher numbers with the hw-validation
> kernel.  Given that, the actual throughput becomes rather uninteresting.

I guess, looking at the actual throughput is also important.
The libertastf driver gives sane rate behaviour but the actual throughput is
constant low.
I have seen higher rates in iwconfig with the libertas driver when connecting
to a ralink usb stick configured as ap when I played a bit with the hostapd.conf
there. I do not remember exactly what I have done, but throughput did not
increase.
I played a bit with the supported mode (b or g) and the rate settings.

So it is really important to look at both. But anyways, I will report that
issue to the wireless guys you proposed in an earlier mail. 

Greetings
Andreas Kemnade
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120531/75fcc9e4/attachment.bin>

From hns at goldelico.com  Thu May 31 17:36:28 2012
From: hns at goldelico.com (Dr. H. Nikolaus Schaller)
Date: Thu, 31 May 2012 17:36:28 +0200
Subject: [Gta04-owner] Status OpenPhoneux / GTA04
Message-ID: <C245F840-CF51-43ED-8AB2-01A8DB6C1716@goldelico.com>

Hi all,
end of the month - time to take some breath of air and
write a status update of the OpenPhoenux/GTA04 project.


1. Group Tour
is still in production (with problems). About 30% of the boards
have been produced but had failures. About 1/3 of those had
been without problems or have been fixed now. These units
have already been shipped last week (based on a first order
first served sequence).

The other boards are in rework. The production company
now thinks they have managed how to raise production yield.
But due to illness and school holidays in Bavaria, they
have not yet continued to work.

So please expect that it will become End of June for delivery...

As soon as these group tour devices have finally been shipped,
we can start planning the next version/release. But then with
keeping a better eye on the production yield and speed.

To avoid misunderstandings: the electronical design is
working very well, as the devices show that have already
been shipped. It is a pure production problem with the solder
joints of the chips on the PCB. No electrical or electronical
problems without a workaround.

It is like running a completely debugged software on
defective memory... Or having a 200mph sports car that
you can't use because some of the screws are loose.
But for each unit there are different ones so you have
to check them all.


2. Presentation at LinuxTag
Lukas and myself gave a well recognized presentation
for an audience of approx. 70 listeners.

And we had that "project meeting point", which is a booth
shared by several projects. We had very interesting
discussions there.

Presentation slides: 

http://download.goldelico.com/default/Presentations/20120526%20LinuxTag%202012%20-%20Openmoko%20is%20dead%20-%20long%20live%20Openphoenux.pdf


3. free 3D Graphics driver
one topic came up during LinuxTag: free 3D graphics
drivers. As you may know, the DM3730 has a integrated
PowerVR SGX530 GPU which is currently not even used.

There exist drivers and non-free binaries for user space
and firmware available through the BeagleBoard.org project
and TI. So it is not used just because nobody did look deeply
enough into the installation procedures.

Unfortunately these are "non-free" software. While for the
MALI GPU, there is a very active project to reverse engineer
and write a free driver. This got quite a lot of attention during
FOSDEM and LinuxTag.

For the PowerVR there was also a project proposal last
year and it was even made a high priority project by FSF.

So I tried to find out the current status.

Well, FSF forgot to kick off the project (they said they will
now take care of it). And there is no more progress than

http://libreplanet.org/wiki/Group:PowerVR_drivers
http://lkcl.net/powervr/sgx/ (list of reverse engineered PowerVR SGX USSE Opcodes)
http://elinux.org/Create_Open_Source_PowerVR_GPU_driver

And I got in contact with a handful of people interested
in this.

Since I would find it good to give us more freedom of choice
for 3D drivers, I offered to use the gta04-owner mailing list for
discussions. And maybe some of you are also interested
to contribute to this.


4. Nomenclature OpenPhoneux/GTA04

There may be some confusion what "GTA04" and "OpenPhoneux"
are and what makes them different.

We define:
* GTA*: the next generation motherboard(s), i.e. electronics
* OpenPhonux: the future "independent mobile handheld" project
   aiming at complete devices (i.e. GTA04 + case + components)

Currently, we run the domains www.gta04.org and
www.openphoenux.org. The Openphoenux.org home
page will be made more prominent and content rich soon.

This is done to become independent from the OpenMoko
brand and the www.openmoko.org domain whose future
and fate is completely uncertain.


5. OpenPhoenux Community

For the reasons mentioned above, we invite everybody to
subscribe to

http://lists.openphoenux.org/mailman/listinfo/community

which we hope will become the new community for those
interested in free and open and independent smartphone
platforms.

For some time there will of course be overlap, duplication
and friction... But this new list belongs to a project that
is active.

After a while we may be able to close the gta04-owner list
and switch over to openphoenux community.


6. GTA04 installation parties

One more idea was triggered by the LinuxTag discussions:

Who of you would be interested in organizing a GTA04
installation party in your region?

Some member of the GTA04 core team could try to come
and help installing the device, the display, the camera
and also give some introduction presentations for software
installations.


Happy flying with the OpenPhoneux,
Nikolaus


From martin.jansa at gmail.com  Thu May 31 20:48:38 2012
From: martin.jansa at gmail.com (Martin Jansa)
Date: Thu, 31 May 2012 20:48:38 +0200
Subject: [Gta04-owner] graphics speed drop in 3.3 and 3.4 kernels
In-Reply-To: <3450316.xdxilV0xF8@rp-core>
References: <3450316.xdxilV0xF8@rp-core>
Message-ID: <20120531184837.GU3138@jama.jama.net>

On Wed, May 30, 2012 at 11:00:24AM +0200, Radek Polak wrote:
> Hi,
> after some time on 3.3 and 3.4 kernel i installed now 3.2 kernel and i was 
> quite surprised about the graphics or maybe overall speed difference.
> 
> 3.2 kernel is much faster. You can easily check this on debian squeeze:
> 
> 	apt-get install mplayer
> 	wget http://activationrecord.net/radekp/pub/impossible_is_nothing2.mp4
> 
> 	cat > ~/.mplayer/config << __END__
> vo=fbdev2
> ao=alsa
> [default]
> afm=ffmpeg
> vfm=ffmpeg
> vf=scale=480:640
> sws=0
> __END__
> 
> 	mplayer impossible_is_nothing2.mp4
> 
> With 3.2 kernel it plays nice, with 3.3 and 3.4 i get like 1 fps.

It was playing nice for me on latest SHR image with 3.4 and
MPlayer2 2.0-379-ge3f5043 (C) 2000-2011 MPlayer Team

but always after first 20-30s it got stuck and I had to ^C mplayer2
MPlayer interrupted by signal 2 in module: sleep_timer

When playing it for 5th time, I also got something like 1 fps or less
(without changing anything).

But this can be some bug in mplayer2 binary I'm using..

Cheers,

> So i wonder if i did some mistake or everyone has this problem. Testing 
> results and ideas are welcome.
> 
> Regards
> 
> Radek

> _______________________________________________
> Gta04-owner mailing list
> Gta04-owner at goldelico.com
> http://lists.goldelico.com/mailman/listinfo/gta04-owner


-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa at gmail.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20120531/83350812/attachment.bin>