[Community] New cell phone design wanted?
ben_deering at jeepingben.net
Mon Jan 4 23:47:00 CET 2016
On Sun, 3 Jan 2016 22:56:53 +0000 (UTC)
James <wireless at tampabay.rr.com> wrote:
> Kero <openphoenux at ...> writes:
> > > >> So what is your desired feature list? (processor, ram, flash,
> Well, these replies do tell this commmunity is alive and desires some
> new toys to get excited about. So, if I may be bold, how does this
> sound to satisfy what I have parsed from the previous posts ???
> multicore arm64, maui gpu, 4 gig of ram, micro-usb & usb3, wifi ,
> audio i/0, Bluetooth, 5" touch screen, sim card socket, camera, gps,
> hardened linux kernel , embedded gentoo linux os, lxqt, (this is
> what I'd do). Openrc (no systemd). SeLinux . Fully encrypted
> As a gentoo device, it's pretty straight forward to install whatever
> software your want from portage, within the constraints of the
> resources. Just so you know, I have little interest in a low price.
> I'm interested in building a cell comm device that kicks ass, is
> secure and users control/modify it as they like. Have troubles with
> the stack, hire somebody or create what you want in the open. BSD
> style licenses as much as possible, i.e. entrepreneurs are most
> I'm interested in expanding the Rf interface capabilities (ham, CB
> etc) so that is still floating on the exact details. That may be an
> add on board that 'clips on the back; decided once I get the final
> list of hardware to support.
Wouldn't there be problems getting a piece of hardware like this
approved by the FCC? It would be better to follow the example of the
Apple 2 and design the phone to accept a (transmitting?) Software
Defined Radio module and sell it separately. The phone could have a
space for a PCB antenna that could be swapped out to a VHF or
UHF or whatever. Wouldn't CB require a huge antenna? The extra radios
have been the most exciting idea I have seen in this thread. The idea
of a transmitting SDR built into a hand-held computer that can run
gnuRadio, etc and can make phone calls is pretty cool.
If building a new form factor (doesn't have to fit a N900 or GTA02
case) make a few separate boards with some logical separation of
parts. This would allow bug fixes and even some incremental upgrades.
I think the correct form factor for an opensource smartphone would be
as close to the dimensions of the screen as possible, but about twice
as thick as the current generation of mass produced devices. This
would make room for hacks, a big battery, and would allow for a thick
Is there any option for a Free embedded GPU? It will be hard to
convince me to spend what an open source phone like this will cost if
I need binary blobs to make the GPU/WiFi work.
I disagree with the call for an ultra low-power display. My current
device can get 5-7 days of moderate usage on a charge with a 5.5"
1920x1080 screen. Since this new device would probably already be
thicker than mass produced phones, just plan to put a huge battery in
it (huge battery will help with compiling all that source from portage
too ;-)). A secondary e-ink display to give notifications would be cool
though. Perhaps having e-ink over a few capacitative buttons would be
good for those who don't want the 3 standard Android buttons.
To be hacker friendly, consider exposing an internal usb
port and the I2C bus on some sort of standard sockets. Soldering
to the testpoints on GTA02 and GTA04 is tricky. Really, if I wanted to
add sensors to A phone, I would build an external BT-LE device even if
I did have an easy way to tap the I2C bus.
I am not concerned about seeing the PCB layout. Even if it is released,
we are still trusting that it matches the device, just like we are
trusting that the schematics match the device. Much of the talk about
this (and previous devices) has focused on security, but I wonder if
that is a real concern. Prior to replicant, how many of us had an
unencrypted phone that loaded the GUI as root with no password? I'm
not aware of anyone implementing a trusted boot chain for GTA0*.
> What I need is a consensus. A rally cry, so I can go into board
> meetings with several semiconductor firms and pitch a revolutionary
> device for the secure entrepreneur to prosper. I am influential and
> have a long history of squeezing (chip) vendors:: so no worries on
> how to get this done. Besides, the current cell phone sales have
> stalled because nobody is building anything that excites folks. I
> have zero interest in building a sub $300 cell phone, that is an
> exercise in futility, imho. I'm more interested in the 'big cannon'
> cell phone that folks will pay for to get one and become excited
> about writing and hacking software for it. Somebody else can focus on
> the low end (sorry if this offends anyone). Codes that invoke
> lincense issue will become user installable, via the gentoo package
> management system known as portage 
> So what did I miss on the hardware and software? (please leave
> emotion, guilt and other such issues out of the discussion) I just
> what input on the killer cell phone package (hardware and software)
> that folks want (like SElinux on an embedded linux cell device).
>  https://grsecurity.net/
>  http://selinuxproject.org
>  https://wiki.gentoo.org/wiki/Portage
> Community mailing list
> Community at openphoenux.org
More information about the Community