> > I know, for example, SuSE has such a build farm that is accessible by > > email (IE you email patches to it). > > > > If they were willing to let the patchapp submit emails (or xmlrpc or > > whatever), and there was a way for it to notify the patchapp about the > > results (xmlrpc or http post would be fine), that would be ideal.
On Thu, Jun 15, 2006 at 07:55:30PM +0200, Zdenek Dvorak wrote: > there are several problems with the idea. With the current setup, we > cannot make the testers available to public for security reasons. > Testing each patch takes several hours, so for larger-scale usage we > would need to dedicate more machines to the task than we do now. Also, > the testers need some maintenance -- 1-2 hours each month for me, but > this would obviously go up with the more intensive usage. I do not > think it would be possible for SuSE by itself to accomodate for these > demands. This is understandable. In any case, it would probably best to have a human in the loop before submitting patches to autobuilders, both for security reasons and as a sanity check, to avoid wasting resources on an unacceptable patch. Machine donors (maybe SuSE?) would authorize a small number of people to submit patches to the auto-builder after an initial review. I'm now starting to get a clue as to why Mike talked about a review before the auto-builder instead of after. Maybe the pre-review would just be a very cursory sanity check, with the real review to happen after it builds. > Perhaps in the cooperation with other companies interested in gcc, we > could put together enough hardware and developer time to realize this > idea (don't take it as promise of anything from the SuSE side, though -- > I have very little say regarding that). I think it might be good to > discuss it at gcc summit. If people like the concept, but there's a shortage of dedicated equipment, perhaps donations can be solicited.
