[Bf-committers] [Bf-blender-cvs] SVN commit: /data/svn/bf-blender  tags/blender-2.69-release/blender: svn merge ^/trunk/blender -c60755 -c60766 -c60792 -c60796
montagne29 at wanadoo.fr
Mon Oct 21 15:33:07 CEST 2013
For the record, if we decide to do RC3 instead of release (which I'd
find more sensible now), I will also backport two oneliners - both are
full secure, and one is a fix for a new 2.69 feature (not sure this has
same status as regression?): 60857 and 60858.
On 21/10/2013 15:26, Sergey Sharybin wrote:
> I would strictly recommend stop backporting fixes for a not-a-regression
> but to release branch. 60774 is a fix for 36905 which is not a regression.
> It is NOT to be merged.
> And strictly speaking 60766 should have not been merged. Not as if is to be
> reverted, but next time one merged not-a-regression fix to release branch
> he'll get kicked. Seriously, we can not fix all the bugs in release, we
> could only fix regressions. And having bunch of changes happening after tag
> just makes things more complicated.
> Also the fact that we've got a regression in RC2 comparing to RC1 (#37100)
> needs to be investigated further. I would not expect regressions in RC(N)
> comparing to RC(N-1).
> Assuming that Shinsuke mentioned 2 more revisions (even one-liner) to be
> merged, i would call it RC3 today, NOT a release.
> Brecht, do you agree with having RC3 today?
> P.S. Splash is to be changed again it seems.
> On Mon, Oct 21, 2013 at 3:10 PM, Thomas Dinges <blender at dingto.org> wrote:
>> My point here is, that we agreed on doing proper proper Release
>> Candidates and Release now.
>> I don't mind a small one-liner, in this case we could risk it.
>> But since RC2 we had 60845, 60871 and 60875 with backports...also not
>> all of them seem to be really showstopper bugs?
>> I know we have a lot to handle this week, but I see no reason to rush
>> the release. If we can't handle it properly this week, do it next week.
>> Am 21.10.2013 14:57, schrieb Thomas Dinges:
>>> Yesterday we talked about just 1 or 2 one-liners, and now it's getting
>>> In this case we maybe should consider doing a RC3...
>>> Also, we should only backport regressions. Is 60774 a regression?
>>> Am 21.10.2013 14:41, schrieb Antony Riakiotakis:
>>>> Hey Campbell, could you merge 60774 as well?
>>>> Bf-blender-cvs mailing list
>>>> Bf-blender-cvs at blender.org
>>> Bf-committers mailing list
>>> Bf-committers at blender.org
>> Bf-committers mailing list
>> Bf-committers at blender.org
More information about the Bf-committers