New F14-arm build os40 - Almost in sync with x86!

[Simon Schampijer]

On 08/23/2011 12:54 PM, Peter Robinson wrote:
> On Tue, Aug 23, 2011 at 10:08 AM, Simon Schampijer<simon at schampijer.de>  wrote:
>> Hi Peter,
>>
>> thanks for the new build! First of all, it does boot fine.
>
> Thanks for confirmation, was racing out the door!
>
>> On 08/23/2011 09:28 AM, Peter Robinson wrote:
>>>
>>> The please just exit the building already release!
>>>
>>> There's been quite a bit of churn since the last release so its likely
>>> possible I've missed some bits. I plan to cut another release Friday
>>> morning so let me know what I need to ensure is pulled in soon.
>>>
>>> Download from:
>>>
>>>   http://build.laptop.org/F14-arm/os40/
>>>
>>> Changes and notes from Martin's OS36 (including some from my OS1 test):
>>> - sugar 0.92.4
>>
>> As we can not push 0.94 to F14 in Fedora I will do F14 builds for our builds
>> and put them into my public repo.
>
> OK, please use Fedora spec files or rebase OLPC's on them as there's a
> number of bugs I've seen in the olpc ones that have long since been
> fixed.

Yes, will do.

>>> - abiword and pyabiword with network support
>>
>> works now indeed.
>>
>>> - pybox2d distro package included so Physics should work
>>
>> Physics is actually missing in the build (see activities diff below)
>
> Weird. Its using the same Activity bits and URLs etc that are in
> mainline OOB now so it should have been the same. I'll have time to
> poke more at OOB this week in time for the next build on Friday.
>
>>  From my list:
>>
>> sugar-presence-service is now the latest package and fixes [1].
>
> Awesome.
>
>> "hostname set too late; upsets avahi" [2]: this does not seem to have been
>> fixed. Even though a new olpc-utils version based on olpc-utils 1.3.0 has
>> been pulled in the hostname is not set correctly, it is set to localhost.
>> Hmmm...
>
> There's an issue with avahi builds in the builder due to bloody gtk3
> which I need to take a closer look at, I wonder if there's a fix in
> that?

avahi-0.6.27-2.fc14.i686 is what we have in the i686 builds
, and building that for ARM failed as I see in the koji logs. 27-1 is 
what we have in our ARM builds at the moment, 27-2 was a python 2.7 mass 
rebuild it looks like - not sure it has an effect on the above.

Maybe Daniel has a clue what causes the error as he has been doing the 
hostname fix?

Regards,
    Simon