Re: [boost] Boost regression notification (2006-06-24 [RC_1_34_0])

From: Timo Geusch (timo_at_[hidden])
Date: 2006-06-25 03:35:42

Douglas Gregor wrote:
> Boost Regression test failures
> Report time: 2006-06-22T16:53:09Z
>
> This report lists all regression test failures on release platforms.
>
> Detailed report:
>   http://engineering.meta-comm.com/boost-regression/CVS-RC_1_34_0/developer/issues.html
>
> The following platforms have a large number of failures:
>   sun-5.8
>   
<snip>
> |algorithm/string|
>   regex:  sun-5.8*
>   replace:  vc-7_1-stlport
>   
I had a look at this from a sun-5.8 perspective and it seems to me that 
the SUN failures are more a problem of compiler configuration than one 
of compiler capability. Using the v1 sunpro-tools.jam with the patches I 
emailed to this list recently, the algorithm/string tests both compile 
and pass. However, using the v2 sun.jam even with the C++ 'OPTIONS' 
seeded with the necessary basic flags as per my v1 patch, the 
compilation of several tests fail even though the same test does both 
compile and pass on v1.

There are still a few problems with the 5.8 compiler and the latest 
available patchset from SUN but in my experience they are not as bad as 
the regression tests suggest.

Unfortunately I don't know quite enough about the way Boost build v2 
works in comparison to v1 so I'm a tad stuck duplicating the v1 patch 
for v2. If anybody who knows v2 better could get in touch off-list and 
lend a hand, that would be great.

Regards,
Tio