Re: [Boost-testing] Can we improve throughput on our testing?

From: Jeff Garland (jeff_at_[hidden])
Date: 2005-06-16 23:38:00

On Wed, 15 Jun 2005 09:25:38 -0500, Douglas Gregor wrote

Couple random thoughts:

> In truth, I'm actually hoping that we can add compilers/platforms 
> after the release. Once we have a clean slate (= no unresolved 
> issues), it's easier for us to keep it at a clean slate now that we 
> get daily feedback on our changes. If someone expresses an interest 
> in getting toolset X to work, we add it as a release compiler and 
> clean up the mess.

It will be interesting to see if the new testing infrastructure can actually
help make this happen.  There's some subset of us that would like to see the
latest CVS reliably build most of the time.  Also, we should immediately get
the authors of the newly accepted libraries that didn't make 1.33 started on
getting into CVS.  I'd say it takes 2-4 weeks minimum for a new library to get
added and integrated. Anyway, if we keep the latest CVS building making a
release would be a 3 week instead of a 3 month process....
 
> As for GCC 2.95.3... I never know what to do about that compiler. 
> I've heard that it's still used by lots of people, but I haven't 
> seen any evidence of that myself.

I don't see much evidence of 2.9x these days.  Linux, mingw, etc have all
upgraded to 3.x so it's only the folks with 'legacy projects' that are stuck.
Anyway, going into 1.34 I think it's time to drop all the old bad compilers. 
We can debate which ones are old and bad, but VC6 and 2.95 are on my list. 
One day I'd like to be able to clean my code of ugly legacy compiler hacks...

Jeff