RE: Mauve test question

[Jeroen Frijters]

Archie Cobbs wrote:
> Jeroen Frijters wrote:
> > I don't think we should have a file based on black listing 
> tests that
> > don't work, I'd like to have a file that simply lists all 
> tests that are
> > known to be good. There are two reasons for this, the most important
> > being that I want to see who adds (or removes) a test to 
> the known good
> > list and why (and the second that I hate scripts ;-)).
> 
> The problem with that approach is that if someone adds a new test
> to Mauve, it doesn't automatically get added to our "white list".

That's a feature, not a bug! In practice new tests often get added that
don't yet run without failures (and this is the right thing to do). So I
strongly believe we should work with a white list.

Regards,
Jeroen