Bazaar focus for 2.1 and 2.2

[Michael Hudson]

Jelmer Vernooij wrote:
> On Wed, 2009-12-16 at 11:36 -0500, Francis J. Lacoste wrote:
>> On December 16, 2009, John Arbash Meinel wrote:
>>> Francis J. Lacoste wrote:
>>>> On December 15, 2009, Martin Pool wrote:
>>>>> I just had a good talk with James about what the Bazaar team could do
>>>>> to help UDD move forward.  We are making progress on some particular
>>>>> bugs but the analysis feels a bit inchoate.  So my theory is that we
>>>>> will be more efficient if we pick a clearer focus to do first.
>>>>>
>>>>> We talked about:
>>>>>
>>>>> * vcs imports - very visible so could be good, but not a pressing
>>>>> problem now
>>>> Well, the linux kernel import is still not working. And that's with the
>>>> recent fixes to bzr-git by Jelmer and the improved memory usage by John.
>>>> So there are things to improve there.
>>> So I think the kernel is probably good for visibility, it certainly
>>> isn't worthwhile from a "people are going to start using bzr to develop
>>> the kernel" sort of thing.
>> It's more than for visibility, having an import of the kernel is a 
>> prerequisite for doing udd and daily builds with it.
> 
> Would it perhaps be an option to have the import system only import e.g.
> only 1000 revisions at a time?

Well, yes, I guess that would be possible.  It would also be possible
for bzr-git or even bzr to do this as well -- currently we have real
problems with the amount of memory bzr serve processes use...

> This would make the memory leaks less of
> a problem, and it should make the scheduling of code imports a bit
> fairer, since large branches would not keep the system busy for a long
> time. 

It would also need some work so that we don't publish an import until
the import actually finishes, I think that would be quite confusing.

> The overhead of resuming an existing import should be relatively small.

Yeah.  Let's talk about this in Wellington if we don't get to it before
then :-)

Cheers,
mwh