The future of openmoko as a community distribution

Koen Kooi koen at
Thu Feb 15 18:01:41 CET 2007

Hash: SHA1

Sean Moss-Pultz schreef:
> On Thu, 2007-02-15 at 12:52 +0100, Koen Kooi wrote:
>> I'm really curious where FIC/ wants to take the
>> distribution after the p2 launch.
> Some of this stuff I just can't talk about publicly yet. OpenMoko is
> really strategic for us. I should just leave it at that for now. When
> we're in together in FOSDEM, I'm more than happy to talk more. 
>> I've stated this before:
>> "Stage 3
>> Assemble an Ångström-moko-image containing a recent gtk
>> +/pango/cairo/glib and compare that
>> against the OABI and EABI moko-images. If it turns out certain parts
>> are slower compared
>> to the moko-images I'll inform the people involved (most likely the
>> cairo people and
>> selected gtk hackers like Federico) and try to improve those areas.
> Great!
>> Stage 4
>> Convince FIC to merge their openmoko-distro into Ångstrom, so the
>> community doesn't
>> fragment between the two distros. Since openmoko was based on Ångstrom
>> anyway, it boils
>> down to branding issues, but I'm sure we can work that out."
> Koen, please understand that for a company using open source branding is
> in some ways, more often than not, more important that the code itself.

I actually wasn't talking about changing any of the branding :)

>> Do you want to keep the OM distro in your own control, or will be it
>> (partially) community
>> run at some point?
> We are _really_ hoping that OpenMoko will become a community /
> commercial platform. Believe me, I have lots of interests in seeing the
> community directly influence and help set the direction of OpenMoko.

I've seen a number of devices/distributions that are in a bad shape because the company
kept doing its own thing, like Sharp and the SharpROM.
I agree it's pretty hard to set long term (e.g. 5 years from now) goals when you don't
know what the devices are going to look like in ~2 years.

> Where I think we (as a company) can offer the most to a community (and
> in the end -- the general public) is by building devices that are
> designed from day 1 to be used with open source software. Neo1973 is our
> first device. Don't let the first specs fool you. We're really serious
> about pushing hardware at the same rapid pace that software can be
> pushed. Right now we're just trying to get hardware out in the wild. So
> we start getting your feedback and make the next iteration far better as
> an open device.

So if I understand you correctly, Openmoko Inc. will be a foremost a hardware company with
a department that will be assisted by the community to deliver the software for that, right?

> Ultimately, this community / commercial cooperation is what I think will
> be needed to open the mobile industry. It just don't see it happening as
> a one-sided effort. 

As you can see with nokia/maemo there are almost no 3rd party developers for the based
system (e.g hildon-libs, hildon app framework), because nokia can publicly show a roadmap,
since that would hint at new and 'secret' hardware revisions.

>> Regardless of that, would you want to work together with the angstrom
>> distribution in OE
>> to contain the fragmentation a bit? 
> Oh you bet. Please prepare your thoughts on this. I will do the same and
> talk it over with our coreteam. Let's meet over some beers in a few
> weeks. 

To be clear: it is really not about changing branding, but about giving the OM
distribution a flexible, but solid 'base'. Right now angstrom is the only distribution in
OE that is proven to work for arm, x86 and powerpc, which gives you an idea how flexible a
distro can be. But lets save the real discussion for FOSDEM :)


Version: GnuPG v1.4.5 (Darwin)


More information about the distro-devel mailing list