Re: [boost] RC_1_29_0 Win32 regression tests updated

From: Beman Dawes (bdawes_at_[hidden])
Date: 2002-10-08 15:45:08

At 02:52 PM 10/8/2002, Alkis Evlogimenos wrote:

 >On Tuesday 08 October 2002 11:08 am, Beman Dawes wrote:
 >> See http://boost.sourceforge.net/regression-logs/cs-win32-index.html
 >>
 >> Of all the reports, the one I find most useful is
 >> http://boost.sourceforge.net/regression-logs/cs-win32-diff.html, which
 >> shows test results that have changed since the RC_1_29_0 branch in a
 >larger
 >> font.
 >
 >Is the windows platform the official platform for the boost libraries?

No, there isn't any "official" platform. The Win32 tests are a useful first 
cut at detecting errors because they cover a number of the most popular 
compilers, including several used heavily on other operating systems.

 >> Dave Abrahams and I had some discussion yesterday, and both think we 
are
 >> within a day or two of final release. I'll query the list before the
 >actual
 >> release, to make sure all pending fixes have been committed.
 >>
 >> Between now and final release I'll rerun the Win32 tests whenever
 >RC_1_29_0
 >> commits build up in CVS.
 >
 >Are there any plans for testing more often

There are a couple of issues with the regression test setup that require 
manual intervention, and sometimes rerunning all tests from scratch. Once 
those issues get cleared, it will be easier to run the tests more 
regularly.

Other factors (people's work schedules, new regression test setup, etc.) 
have cut down the frequency of testing on other platforms.  All that is 
temporary rather than permanent.

 > and on more gcc versions for
 >example: 2.95.3 (most common) and 3.2 (latest stable)?

Highest Win32 priorities right now are adding Comeau and VC++ 6.0, based on 
comments I've gotten. Metrowerks has just announced another update, too, so 
that has to be applied soon.  Testing more gcc versions isn't a priority on 
Win32, partially because the other platform's tests tend to include a lot 
of gcc versions.

--Beman