daily-feed vs. SRCREV

[Rod Whitby]

Julian Chu wrote:
> On Wed, Jul 30, 2008 at 10:46:34PM +0930, Rod Whitby wrote:
>> Thanks for the explanation.  I would humbly suggest that buildhost only 
>> build the asu branches then, so that you focus user testing by the 
>> public on the development that is intended to be supported into the 
>> future by Openmoko.
> 
>    Actually, the buildhost didn't build ASU branch at all.
> We did that work on another machine in Taipei.

Hmm - so you *don't* want the public to be testing ASU yet?

>   Sorry, I am not 100% understand the last setences, could you please
>   explain that more ?

Sure.

My understanding is that ASU is the distribution that Openmoko intends 
to support into the future (as opposed to supporting 2007.2 into the 
future).

To get the best testing of the distribution you are supporting, you want 
to have the largest number of people testing the exact same 
distribution, so you get consistent and reproducible bug reports.

To do this, you really want to be building with sane-srcrevs and without 
moko-autorev on the buildhost that you make accessible to the public (as 
opposed to the other internal-only machine in Taipei).

All my opinion, of course, but I do this on the same sort of scale (5 
firmware distros three of which are based on OE, over 20K people on the 
mailing list, millions of firmware and package downloads) for the 
nslu2-linux.org project ...

-- Rod