From: Caio Tiago Oliveira <[EMAIL PROTECTED]>
Subject: Re: [qa-dev] Automated GUI Testing: What is a 'global' CWS and how do 
we test it?
Date: Mon, 19 Mar 2007 02:48:50 -0300

> I'm not requesting Sun QA'ing for Mac OS X, I'm just requesting the 
> nomination process should take into account the QA results for Mac OS X 
> if someone (from the community or a volunteer from Sun) does the job.

Okay. Assume you found some defects in a cws marked or ready for QA.
Then communicate with QA representable or/and members who has a commit
right to ask for integration of your patches. If rejected this might be a
problem. I belive this is not the case.
After marked as ready for QA, it depends. IMHO, once global cws
is `approved by QA' it is very hard to back to the status to `new'.
It wastes the time and resouce of Hamburg QA team.
As Jogi wrote, this is a hard task.
If you have objection for this process, then, MacOSX must be a supported
platform for nomination process. Please ask othres (Hamburg RE)
to implement it.

BTW: can somone forward a message from EIS when a created cws are global one
to warn other porters to check porting issues?

>  From my point of view, for everything I've read about it: a CWS should 
> not introduce new issues.

I agree. But for which platforms? You may want to add MacOSX here.

> Are you saying that Mac OS X, FreeBSD and everything other than GNU 
> Linux/Windows/Solaris should stay forever as second class platforms, 
> regardless their maturity level?

This is what I cannot decide. This is what the Hamburg RE can decide.
However, I think - it is very good idea to restrict primary platforms as
Windows/GNU/Linux and Solaris.

Of course, if resouce is available, +1 for MacOSX.

> > e. there are not so many unfortunate CWS like mingport03. 
> > If so many, we expect even stability and we cannot do porting!
> 
> Could you rephrase this?
> I believe I misunderstood you here.

Usually, there are not so many cws like breaking other ports. 
I belive OOo source code are very clean and CWS process is usually very good
.IIRC for FreeBSD, I have only one or two FreeBSD patches during OOo 2.1
to 2.2. For MacOSX, there are, but not so many in my impression, right?
I could be very lazy for porting. If so many global CWS are introduced at a
time, we cannot do porting. This might happen for early development stage
of 3.0. In such a case, we should wait for while and then do a porting and
stablize.

If MacOSX is primary platform for nomination process, every small build
breaker must be taken care carefully. But this is not what I can do...

[So - that's why I prepared MacOSX packages for new chart2. Checking
buildablity. of course other purpose as well.]

Another problem:
Hamburg team uses slightly different build system (do not use
configure). Sometimes we encounter build issues.

Still have a question?
-- Nakata Maho ([EMAIL PROTECTED])

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to