[OHSW.org] FOSDEM 2018 and beyond

H. Nikolaus Schaller hns at goldelico.com
So Feb 4 17:01:55 CET 2018

Hi all,
FOSDEM is over and I am leaving for the airport in 10 minutes :(

FOSDEM was an as interesting experience as every time.
Many people to meet, many talks, (too) many topics in
parallel... You have to decide how to spend time and know
in advance that you miss 90% of important things.

Unfortunately I didn't notice any talk, booth, presentation,
material to mention OpenMoko. Except some people who did remember
when we came into talking. But as said I haven't been everywhere :)
So if anyone did see something, please share.

Anyways, I learned a lot of interesting and useful things for our
projects. E.g. why the FST01 crypto key was developed an
about about the status. And I established a contact to the
developer and gnupg.org. Or I learned about the architecture of
the video4linux2 subsystem which will help to fix the remaining
issues with the camera and omap-isp on the GTA04. Because I know
know better where to add debugging code...

Fortunately we had a really nice meeting of our tinkerphones
community on Saturday evening in the restaurant that I had proposed.

Wolfgang also brought some Replicant guys with him and Harald,
who was several times participant of our OpenPhoenux hard&software
workshop, also came with a group. So in total we were almost.
30 people chit-chatting about open source, software and hardware.
And about past and future.

Interesting for me was to see and touch in life the ZeroPhone and
a laser-MEMS based mini projector built into a smartwatch (a commercial
effort using some open source components) and some Arduino based
IoT/Home Automation components.

I could show around the final prototype of the Pyra-Handheld
and the GTA04 crypto key and the HF08 charger. Most interesting
for me was to get feedback for these projects. It turned
out that the crypto key should be smaller, mainly shorter, to
fit into some laptop. A solution could be to replace the USB plug
by PCB printed USB contacts. This also reduces cost and simplifies

And I was happy to share knowledge discussion how to try to move
an external USB port inside the battery bay of a smartphone.

Also, the idea of a Pyra-Phone found quite some interest. A Pyra-Phone
would be a simple reuse of the Pyra circuits, the OMAP5 CPU board,
and the (resistive touch) display but have a different main board
which glues everything into a simple sandwich in some (bulky) smartphone

It will leave out the keyboard and nubs and hinge, but otherwise
will have the same features (especially multiple ()SD slots and
USB host/otg ports with SATA) as the full Pyra. This seems doable
in let's say less than 3 months for a prototype and 6 months to build
a handful of already good working developer devices (with 3D printed

While this will not be a design from scratch which would allow
to dream about thousands of infeasible ideas, it will be a
down to earth approach to quickly get a newer and quite powerful
smartphone hardware in short time (OMAP5 dual core seems to be
at least as powerful as i.MX8M quadcore).

Project risk is extremely low since almost everything is already
on the table and mature. It is more an evolutional than a revolutional
approach. So there will be no capacitive touch in the beginning,
but believe me the resistive touch of the Pyra has carefully be chosen
by very active members of the gaming community. And they are very
critical when it comes to user interfaces.

Software would be everything we already have from the GTA04
(Letux OS): Debian, LXDE, XFCE, Enlightenment, QtMoko and Replicant
and it would not need much adaption or time for bring-up, since
the OMAP5 board is already debugged and well supported by the
mainline kernel... From that other software stacks are not far

There is only one big area for basic research: power management
of the OMAP5 and the components we have in this device. But we
have a lot of experience from OMAP3 work to have the proper tools
to work on it. And every bit done for the Pyra will help the Pyra
Phone and vice versa.

Regarding the closed user-space libs of the GPU, there are two
ideas for a far future of a Pyra-Phone: design another (mechanically
and pin compatible) CPU board using i.MX6/8 or some RISC-V-SOC when
a powerful one becomes available and cheap enough (my rough guess is
3 years). The Pyra community is also interested in such future
CPU board upgrades, so we all would benefit from collaboration.

Speaking about budgets, we have of course a problem... There
isn't any and a hen and egg situation. We need some budget to
build prototypes for developers and we need working devices
(at lest in the hands of developers) to get other to provide
funding for some mass production. Fortunately the barrier isn't
high (ca. 3000€) so that some donations to
would be a good start. The developer prototypes would be sold
at a prive level comparable to a Pyra. (see www.pyra-handheld.com).

So for me this FOSDEM 2018 visit was sort of ignition to try to
restart our activities to continue what OpenMoko did start
many years ago. As there seems to be a hidden or dormant
interest for these topics, maybe we can have presentations
next year which demonstrate several Pyra-Phones in operation?

What do you think? Would there be some developers here who
could and want to afford buying one of a handful of prototypes
of a Pyra-Phone? If yes, we can easily set up our own
preorder/crowdfunding campaign.

Let me finish this mail with expressing the hope that we will
revitalize these mailing lists. They form the core of
communication besides personal meetings like at FOSDEM. So
please actively use them just for discussions, even if you
have no time for other contributions.

happy hacking beyond FOSDEM 2018,

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.goldelico.com/pipermail/open-hard-software-event/attachments/20180204/12441d6a/attachment.asc>

Mehr Informationen über die Mailingliste Open-hard-software-event