Re: [boost] como/EDG and compile tests

From: Beman Dawes (bdawes_at_[hidden])
Date: 2003-12-01 07:32:23

At 06:19 AM 12/1/2003, David Abrahams wrote:
 >
 >I just noticed that our como-win32 toolset (and I don't know how many
 >other EDG-based toolsets) are set to do link-time instantiation.  That
 >means we may be getting misleading results for compile and
 >compile-fail tests.  What should we do about it?

I'm in the process of removing como-win32 from the Win32 regression tests I 
run. It just takes too much time dealing with all its oddities. Also, the 
results never meant much because they were inevitably obtained with a 
patched compiler that didn't represent any particular como release.

If you decide to change any of the como-win32 toolset settings, you might 
want to run "before" and "after" regression tests to see what changes. At 
least in the past, changes in the toolset have often had unintended 
consequences.

--Beman