Re: Bug report: Compile with Microsoft and Intel compiler

[Earnie Boyd]

On 7:21:06 am 2005-04-29 "Eli Zaretskii" <address@hidden> wrote:
> >  Date: Thu, 28 Apr 2005 23:21:54 +0000
> >  From: "Earnie Boyd" <address@hidden>
> >
> > >  For the MS compiler maybe.  But that's not the model that GNU
> > >  autoconf uses.  They don't test for platforms, they test for
> > >  individual features.  In the UNIX world it's quite possible to
> > >  have a system where getenv() is there, but not strerror(), etc.
> > >  so a single #define for "STDLIB" is not expressive enough.
> > >
> >
> >  Which is why I created MSYS.  MSYS provides a minimal system to
> >  execute configure for a native target.
>
> Being able to run configure scripts on a non-Posix platform does not
> mean we should use this in building a basic tool such as Make.  Make
> is one of the first few tools one needs to bootstrap a GNU development
> environment on Windows, so it makes sense to try to do that with a
> minimal set of tools, relying on the underlying OS and its native
> tools as much as we can.
>

Bootstrapping a new environment is only needed once.  Once the system is
built one should be able to use it to build itself on the hosted platform.

I'm not suggesting that GCC be used, I'm suggesting that the configure
scripts can be honed to
 
$ MSYSTEM=VCC uname -s
VCC_NT-5.1

or to

$ MSYSTEM=BCC uname -s
BCC_NT-5.1

or whatever else one wants so that configure can identify what each system
has in a usual method as has been designed within the code without having
to hand code a config file.

Earnie

--
       MinGW - http://www.mingw.org/
        Wiki - http://www.mingw.org/MinGWiki/
  Bug Report - http://sourceforge.net/tracker/?group_id=2435&atid=102435
Submit Patch - http://sourceforge.net/tracker/?group_id=2435&atid=302435
  SF Project - http://sourceforge.net/projects/mingw
 Job Listing - http://sf.net/people/viewjob.php?group_id=2435&job_id=21643