Re: Release Schedule 4.4

2009-08-27 Thread Tom Albers
At Thursday 27 August 2009 01:00, you wrote:
 Am Mittwoch 26 August 2009 22:23:33 schrieb Tom Albers:
  This schedule will become final next saterday, unless someone objects.
 
 What is the reason for having about one week between tagging and release of 
 the beta versions but only one day for the RCs?
 
 Imho the delay for RCs should be higher than for betas. (e.g. two days for rc 
 and one for beta)
 
 Pierre
 

No, the normal packaging to release cycle is ~5 days. Only if we apply that 
normal cycle to rc's the rc's would not get any useful testing, as their 
interval is set to two weeks. So the RC's follow a package-and-release cycle, 
with minimal testing. 

I do think that we need to stress to people to only fix showstoppers from rc1 
stage. In the last cycle we saw that people were insanely backporting their 
fixes between rc1 and rc2, which made rc2 rather bad, causing us to insert the 
rc3. We can try to prevent that this time.

Toma___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: Release Schedule 4.4

2009-08-27 Thread Pierre Schmitz
Am Donnerstag 27 August 2009 10:12:22 schrieb Tom Albers:
 No, the normal packaging to release cycle is ~5 days. Only if we apply that
 normal cycle to rc's the rc's would not get any useful testing, as their
 interval is set to two weeks. So the RC's follow a package-and-release
 cycle, with minimal testing.

 I do think that we need to stress to people to only fix showstoppers from
 rc1 stage. In the last cycle we saw that people were insanely backporting
 their fixes between rc1 and rc2, which made rc2 rather bad, causing us to
 insert the rc3. We can try to prevent that this time.

Thanks, I think I got it now. If there are only minimal changes between RCs 
this should be fine.

-- 

Pierre Schmitz, http://users.archlinux.de/~pierre
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


About tagging 4.3.1

2009-08-27 Thread Maciej Mrozowski
About KDE 4.3.1 tagging.

There are a few bugs apparently still present in 4.3 branch, do you think they 
are worth being blockers?

https://bugs.kde.org/show_bug.cgi?id=202722
https://bugs.kde.org/show_bug.cgi?id=204605

Also there is some kdevelop crash that may have its origins in kdelibs
https://bugs.kde.org/show_bug.cgi?id=204670

(apart from that, one minor issue with Notes plasmoid not being able to revert 
font color to 'Use theme color')

-- 
regards
MM
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team


Re: About tagging 4.3.1

2009-08-27 Thread Maciej Mrozowski
On Thursday 27 of August 2009 21:50:44 Aaron J. Seigo wrote:

 they were in 4.3.0. they aren't blockers then, they aren't blockers now.
  it's more important to get fixes that -have- been made and do timely
  releases than wait for fixes that have yet to be made.

Well, both bugs were reported after 4.3.0 was released and because I test only 
4.3 branch I have no idea whether they were introduced after 4.3.0 
(regressions), just before or present for a longer period of time (a'ka 
known). You apparently mean it's now more important to backport trunk fixes to 
4.3 branch.
While I always assumed they're being backported immediately after being 
commited to trunk - this way it would give more time for testings and catching 
issues earlier, thus making releases in more timely manner than what you seem 
to propose - backporting stuff now and expecting testers to catch in last 
minuete every possible problem introduced, not talking about fixing remaining 
issues.

from announcement Dirk posted:
We have a KDE 4.3.1 tagging milestone tonight at 23:59 UTC. If there are 
urgent bugs that must be fixed before 4.3.1, please tag them in bugs.kde.org 
with the kde-4.3.1-blocker keyword.

-- 
regards
MM
___
release-team mailing list
release-team@kde.org
https://mail.kde.org/mailman/listinfo/release-team