Re: [boost] Intel added to Win32 regression tests

From: David Abrahams (dave_at_[hidden])
Date: 2002-07-25 08:32:44

Why don't you just fix the #ifdef checks? Seems easier and more beneficial
to me...

-Dave

----- Original Message -----
From: "Beman Dawes" <bdawes_at_[hidden]>
To: <boost_at_[hidden]>; <boost_at_[hidden]>
Sent: Thursday, July 25, 2002 9:20 AM
Subject: Re: [boost] Intel added to Win32 regression tests

> At 07:50 AM 7/25/2002, David Abrahams wrote:
>
>  >From: "Beman Dawes" <bdawes_at_[hidden]>
>  >
>  >> I've added Intel to the Win32 regression tests.
>  >>
>  >> There are a lot of iterator related failures.  I'm not sure why that
> is;
>  >> Intel uses the Microsoft library, yet the same tests pass with the
>  >> Microsoft compiler.
>  >>
>  >> If someone can figure out the fix for the Intel iterator problem, it
>  >would
>  >> move quite a few tests from fail to pass.
>  >
>  >In general it's due to checks in the test .cpp files which look for
>  >BOOST_NO_TEMPLATE_PARTIAL_SPECIALIATION which should instead be looking
> at
>  >something that tells them whether iterator traits work for pointers.
>  >
>  >HTH,
>
> Ah! Yes, I remember now. You previously suggested using STLport as the
> library. I tried briefly, but got stopped by the lack of an STLport
> makefile for Intel 6.0. The older Intel makefiles don't work. I checked
the
> STLport forum, and someone else reported the same problems.
>
> If someone has a .mak that works for STLport 4.5.3 and Intel 6.0 on Win32
> that they would like to share, please send it to me privately.  The
STLport
> folks would also appreciate it, I'm sure.
>
> --Beman
>
>
> _______________________________________________
> Unsubscribe & other changes:
http://lists.boost.org/mailman/listinfo.cgi/boost
>