[boost] Comeau/como-win32 regression test results

From: Beman Dawes (bdawes_at_[hidden])
Date: 2003-05-21 10:47:17

Today's Win32 regression tests are up at 
http://boost.sourceforge.net/regression-logs/cs-win32.html

The Comeau tests are particularly interesting now that we are getting close 
to a final configuration, since this compiler is based on the EDG front-end 
which is known to be highly standards compliant. The tests were run with 
the --a option, which has several effects:

* Enables void returns, allowing several tests (lambda, etc) to pass.

* Causes a few tests which use Microsoft headers to fail with errors which 
actually should be just warnings with --a option. This has been reported to 
Comeau.

* Detects various problems with Boost code (sometimes as errors, sometimes 
as warnings). For example:

     - Detects some regex problems that previously only Metrowerks was
       detecting.

     - Detects some date-time spurious semi-colons after function blocks.

     - Reports a curious 'no operator "<" matches these operands' in
       binary_search_test. Any opinions on this?

* Shows what looks like a Boost config problem (see date-time testclock, 
for example) which is causing boost::int64_t not to be defined even though 
long long support is present. Help needed on this one!

* Produces lots of warnings in lambda: 
"C:\boost\site\boost/lambda/detail/operators.hpp", line 123: warning #76-D: 
argument to macro is empty BOOST_LAMBDA_BE(operator+=, 
arithmetic_assignment_action<plus_action>, , const, reference_argument)

Is that a legitimate warning? If not, should we turn off 76-D?

--Beman