Re: [tools-dev] Re: Comments on Mathias blog post about contributing

2009-09-23 Thread Mathias Bauer
Rene Engelhard wrote: > Hi, > > On Wed, Sep 23, 2009 at 11:39:49AM +0200, Michael Stahl wrote: >> or maybe we could simply have a "convenience" flag for configure like >> --official-build, that would simply behave as if the set of necessary >> non-default flags were given? > > That would need Su

[tools-dev] Re: Comments on Mathias blog post about contributing

2009-09-23 Thread Bjoern Michaelsen
Rene Engelhard openoffice.org> writes: > That would need Sun people actually touching configure when they > change their defaults. [...] Please read mst's mail again. Especially the last sentence. The problem is well- known, we know how to fix it (by using configure for all build): the only pro

Re: [tools-dev] Re: Comments on Mathias blog post about contributing

2009-09-23 Thread Frank Schoenheit, Sun Microsystems Germany
Hi Rene, > Last occurance of Sun not caring about configure at all It would make discussions easier if you would learn to stick to facts, and refrain from ungrounded accusations. If we (Sun) would "not care about configure at all", there would be much much more problems. Usually, whenever we cha

Re: [tools-dev] Re: Comments on Mathias blog post about contributing

2009-09-23 Thread Rene Engelhard
Hi, On Wed, Sep 23, 2009 at 11:39:49AM +0200, Michael Stahl wrote: > or maybe we could simply have a "convenience" flag for configure like > --official-build, that would simply behave as if the set of necessary > non-default flags were given? That would need Sun people actually touching configure

[tools-dev] Re: Comments on Mathias blog post about contributing

2009-09-23 Thread Michael Stahl
Caolán McNamara wrote: > On Mon, 2009-09-21 at 20:15 +0200, Sophie wrote: >> - Install sets by the bots are different from Sun builds that make some >> automated tests failing (why the quickstarter is available on linux >> builds? it took me some time to understand that I have to disable it first

[tools-dev] test granularity (was: [tools-dev] Re: Comments on Mathias blog post about contributing)

2009-09-23 Thread Frank Schoenheit, Sun Microsystems Germany
Hi Bjoern, > If a test checks for 10 things at once and in an atomic operation it is broken > by design anyway. That would need to be fixed thus eliminating the problem, > IMHO. This depends on the definition of "test", "test case", etc. - which might yield a longer discussion than is really appr

[tools-dev] Re: Comments on Mathias blog post about contributing

2009-09-23 Thread Bjoern Michaelsen
Frank Schoenheit sun.com> writes: > Don't think this is a good idea, since a test can be broken in different > ways. For instance, if your test checks 10 aspects, and one of them is > broken in MWS, you still want to know if the other 9 are okay in your > CWS. Otherwise, you'll notice a breakage