New Open Hardware company

Rafael Campos methril at
Tue Jul 21 15:55:26 CEST 2009

On Tue, Jul 21, 2009 at 1:52 PM, Wolfgang
Spraul<wolfgang at> wrote:
> Shaz,
> (cross-posting to developer at, if you have more Qi-specific
> questions please ask me over there)
>> This is good news. Congrats and good luck. By the way how are you guys
>> planning to go along with OpenMoko efforts?
> Both Openmoko and Qi believe in free technology, so for all of us it's great
> if there are more players playing by the same rules.
> The same is true for gta02-core, the University of Sao Paolo, etc.
> I wish Openmoko success with Project B or any future projects that may be
> coming. Sean has reorganized the company, but I'm sure he will come back
> stronger and will have a few surprises for us! Also don't count out the
> FreeRunner yet, the development activity in this forum shows it.
> It remains the most interesting mobile Linux development device on the market
> right now, by far!
> Everybody - buy more of them and keep developing, the best pieces of what you
> are developing will find their way into free alternatives to the corporate
> consumer electronics giants we have around us right now.
> At the same time I hope Openmoko the company and community wish Qi Hardware
> well, because at the very least, even if we fail, we had some fun and
> contributed more bits and pieces to the pool of free technology.
> I think the breakthrough in this will happen for sure, only question is
> when.
>> Are the hardware efforts in parallel to openMoko community
> Openmoko published the schematics in PDF form, plus some parts of the
> electrical layout (netlist, footprints).
> gta02-core goes one step further and uses a free tool, KiCAD, for all
> electrical design.
> There are not many free EDA tools out there, so it was natural for Qi to
> follow Werner's leadership in selecting a tool. So Qi will use KiCAD as
> well, and our original KiCAD files will be live on the web from day 1
> (they are already,
> If Werner would change the tools he uses for gta02-core, Qi Hardware would
> seriously consider doing the same. We would not compromise on the degree of
> freedom, however (I doubt Werner would do that either :-))
> I think the most important is that we stay close to each other in the tools
> and file formats we are using, contribute to a common pool of libraries
> (, etc. There is _HUGE_ room for optimization in the hardware
> industry, the dregree of standardization is depressingly low for me.
+1 for Qi!!! :)
I love the idea to some good Electronic Engineers choose KiCAD and
test it in depth, this is going to give us a really OpenSource
alternative to all the propitary tools for electrical design.

> For mechanical design, it's even harder. Werner thinks HeeksCAD is the best
> candidate, so Qi is looking into that now. Our mechanical engineer Tully
> (ex-OM as well) has found some serious problems with HeeksCAD already, that
> he says prohibit him from doing effective work to prepare for plastic
> injection. Oh well...
> So on the mechanical tool, same thing, we want something free, and we will
> follow Werner's leadership or ask him for advice if we are ahead of gta02-core
> in some area.
This a surprise to me, as i didn't know this mechanical tool.
>> OpenMoko community decided hardware will be taken care of by your company
> Definitely possible. That's one reason we are staying close to the free tools
> and file formats Werner has chosen for gta02-core.
> Making a smartphone is a terribly hard job. I just come from 2 years of
> round-the-clock work of doing it. So I really appreciate that for now Werner
> keeps the momentum, the University of Sao Paolo (and actually 1 or 2 others)
> have shown up to try to produce them. More power to them!
> But later on, definitely, it is imaginable that Qi Hardware will produce free
> gta02-core phones. Talk to Steve about the roadmap.
How about the Qi roadmap? ;)

>> The software stacks are the same as much as I can comprehend from the thread.
> We try, but calling them 'the same' right now would not be correct.
> We will probably use u-boot, Linux kernel of course.
> For our own images, we will try to use OpenWrt initially, not OpenEmbedded.
> We try to get Debian running well, so basically we want to have Debian if you
> want full flexibility, and OpenWrt if you quickly want to cook up an image
> targeted at a particular use case or user.
> d-bus, Python.
> Above that it becomes less clear. One reason is that the first generation
> device has no RF. Once we have RF, I can tell you in general I prefer FSO over
> Android any time. More flexible, more maintainable, more modular.
I'd like to give that machine in OE.
I need to read Qi's devel ml.


And of course: GOOD LUCK!!


Rafael Campos
o0 Methril 0o

More information about the community mailing list