I: gcc ... -U_WIN32 ... may cause problems

Earnie Boyd earnie_boyd@yahoo.com
Fri Sep 11 15:13:00 GMT 1998

---Geoffrey Noer <noer@cygnus.com> wrote:
> On Thu, Sep 10, 1998 at 08:54:29AM -0700, Earnie Boyd wrote:
> [...]
> > Geoffrey, __CYGWIN32__ means "access to both Win32 API calls and
> > Unix/Posix API calls".  _WIN32 means "access to only Win32 API
> >  __MINGW32__ means "access to only Win32 API calls" (Actually
could be
> > dropped in favor of _WIN32).
> Although I understand why people might argue for your definitions,
> this is *not* how we've been using them in any of the tools shipped as
> part of the gnu-win32 releases.
> We've always used __CYGWIN32__ to conditionalize code inclusion on
> access to the Cygwin32 API (checking for the existance of Unixy
> calls).  We've used _WIN32 to check for accessibility of the Win32 API
> calls.  And finally, we've used _MSC_VER to identify code that's only
> intended for VC++.

Since the use of _WIN32 to conditionalize code for portibility started
prior to the cygwin project, I ask that you pick a new macro, say
_WIN32_API_, to use for your definition.  I shouldn't be forced to
have to be concerned with breaking headers if I undefine _WIN32 as the
original poster of this message had to do.

BTW, IMHO, your use of the _MSC_VER is asking for trouble.  Firstly,
_MSC_VER didn't exist in earlier versions of VC++.  Secondly, _MSC_VER
is used by MS to control version compatibility with their own product.
 Thirdly, they could change it's name just because they want to.

-        \\||//

Get your free @yahoo.com address at http://mail.yahoo.com

For help on using this list (especially unsubscribing), send a message to
"gnu-win32-request@cygnus.com" with one line of text: "help".

More information about the Cygwin mailing list