[boost] Re: Comeau/como-win32 regression test results

From: Greg Comeau (comeau_at_[hidden])
Date: 2003-06-14 16:54:16

In article <323d01c31ff2$91a76800$c64f2697_at_bagio>,
Giovanni Bajo <boost_at_[hidden]> wrote:
>Beman Dawes <bdawes_at_[hidden]> wrote:
>>  >Well, Comeau in strict mode (--a) defines BOOST_DISABLE_WIN32, so it
>>  >should
>>  >never include any Windows header. That's because we know that Windows
>>  >headers can't be compiled in strict mode, can they?
>>
>> Greg Comeau has been working on getting the Windows headers to work better,
>> and has asked for error reports where there are still problems. It may be
>> that we will have to disable --a on certain compiles, but I don't want to
>> do that unless Greg says it is necessary. The primary advantage of adding
>> the Comeau compiler to the mix is to get its strict error checking, but
>> that would be defeated by turning off --a.
>
>I agree we should not turn --a off.

Ok, I was asking about this in .build..... and you may be
touching on a good part of my questions.

>Instead, which are the libraries which
>does not honor BOOST_DISABLE_WIN32 and still include windows headers? I
>think we should fix those. To do tests, we can comment the definition on the
>macro within comeau.hpp, but since it's enabled now, I guess those libraries
>should be fixed.

One part I'm confused about is why it should only be addressed
with comeau?

Or are you saying that regex has not real windows.h requirement
anyway, so fix the libs, but for right now, define the macro
which turns controls whether windows.h is #include'd?
-- 
Greg Comeau/ 4.3.0.1: FULL CORE LANGUAGE, INCLUDING TC1
Comeau C/C++ ONLINE ==>     http://www.comeaucomputing.com/tryitout
World Class Compilers:  Breathtaking C++, Amazing C99, Fabulous C90.
Comeau C/C++ with Dinkumware's Libraries... Have you tried it?