Christian Lohmaier, 16-03-2007 09:49:
On Fri, Mar 16, 2007 at 11:11:36AM +0900, Maho NAKATA wrote:
From: Shaun McDonald <[EMAIL PROTECTED]>
Subject: Re: [qa-dev] Automated GUI Testing: What is a 'global' CWS and how do
we test it?
Date: Thu, 15 Mar 2007 13:14:52 +0000
> Now that the issue has been raised, lets look at the process to see
> if we can improve it.
If you have some patches to fix for MacOSX, you can raise
as issue like `Build breaker for MacOSX : to be nominated cws mingw03'
and fix in this cws. This is possible, pjanik is fixing issues in such a way.
This is just my personal opinion, we (OOo community members outside
the Hamburg team) should left nomination process and RE process for
the Hamburg team. It will take several milestones to fix MacOSX porting
issue (maybe FreeBSD issues as well), but it is not a big issue of course.
Not a big issue?
Mac port has been the first one that had a milestone that build without
any patches where linux still needes some because of minor things here
and there). And built fine for more than the last 10 milestones.
Mac port is there already. It is up to the other ports. (speaking of the
X11 one)
So you're saying: Just break the build, we don't care at all?
There are some things entirely different in this thread.
One of them is: "I don't know how to find either analyze the info
available from buildbots (even what the build color means)".
Another one is: "I know that mingwport03 breaks the build for Mac OS X
and I don't care at all".
Neither Sun or OOo have enough resources (or interest from Sun) to fix
every Mac OS X bug someone finds on a CWS.
What was done here? Mac OS X port was broken because of fixing issues
with the MinGW port.
Should the Mac OS X port to raise an issue, create a CWS, fix it and
don't care if it break the MinGW port?
Maybe Sun doesn't like it because although MinGW is not the official way
of building OOo on Windows it's on a Sun supported platform.
I know very well the difference between officially supported ports,
experimental ones and non-officially supported ones working very well.
Mac OS X is stable, as a port, and vibrant, as a project.
The OOo community should care about breaking the Mac OS X build the same
way it cares for Linux.
I believe a red flag for Solaris would buy Sun attention and that
wouldn't have been nominated.
We should discuss the process of QA'ing for Mac OS X. Don't to enforce
anything, but would be good to have Mac OS X users willing to help with
the QA of some global CWS's. Off course, if had been granted (from the
developer, qa-responsible or defined by us) that his work would be taken
into account and introducing issues for Mac OS X would be a stopper for
its nomination.
Are you really (Co-)Lead of QA?
Well Christian, he's a bit busy this month and I believe he has
overlooked at this issue (for instance, that the integrated CWS was not
a big issue, even a bug in the product at all).
Christian, is there a Issue Tracker issue for this or have you only
commented at EIS?
[...]
That's why at this stage, nomination process for global CWS must leave
for Hamburg team. In short, we cannot handle, and responsible for such
a big cws.
So because it is big, you accept that the build will break for one of
the platforms?
We are responsible for the overall quality of the product, so it means
we are (in some way) responsible for these such a big CWS.
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
