Re: [boost] $egression tests -- no links for Win32 errors?

From: Beman Dawes (bdawes_at_[hidden])
Date: 2003-01-22 12:22:39

At 10:23 AM 1/22/2003, David Abrahams wrote:

 >Beman Dawes <bdawes_at_[hidden]> writes:
 >
 >> At 07:41 AM 1/22/2003, Jeff Garland wrote:
 >>  >
 >>  >Seems like somewhere along the way we lost the
 >> failure detail for at least
 >>  >the Win32 regression.  Any ideas when this part will be
 >>  >back?
 >>
 >> Look at 
http://boost.sourceforge.net/regression-logs/cs-win32-index.html
 >>
 >> The way the tests are being run hasn't changed, but the
 >> summary is only linking to the version without links.
 >>
 >> We need to figure out exactly which versions of the
 >> tables are most useful, and which the summary should
 >> link to. It's a work in progress:-)
 >
 >Why would anyone want the version without links, ever?

It was done that way originally because that was the version that was going 
to be checked into CVS and become part of the release. The links file 
wouldn't be included in CVS and releases because of size.

Since we are discussing just keeping all the versions on the SourceForge 
site, that would seem to eliminate the need for a version without links.

There are two other versions we might want to consider eliminating (KISS):

   Messages only - Shows status only for tests with warnings or failures.
   Failures only - Shows status only for tests with failures.

I never find myself looking at those. Does anyone else use them? The 
thought behind them was that as compilers (and libraries!) improved, it 
will get pretty boring looking at a large table that is all "pass" on most 
rows.

Thoughts?

--Beman