This is the mail archive of the
mailing list for the Cygwin project.
Re: Question about writing reports
- From: Jonathan Hayward <jshayward at pobox dot com>
- To: cygwin at cygwin dot com
- Cc: mike at eolas dot com
- Date: Thu, 21 Feb 2002 08:15:41 -0600
- Subject: Re: Question about writing reports
- Organization: Eolas Technologies
> On Wed, Feb 20, 2002 at 06:26:17PM -0600, Jonathan Hayward wrote:
> > I posted an apparent bug report about gcc/Cygwin 1.3.9 aborting on
> > function prototypes that worked flawlessly under gcc/RH. I tried to
> > write a descriptive statement and included cygcheck output. It was
> > posted about a week ago, and I haven't seen a reply yet. Is there
> > further information I can post that will be useful in replying?
> Sorry but, what reply do you expect? The error message is very
> likely caused by a datatype not known by Cygwin. You changed the
> Makefile to define __WINDOWS__ and __WIN32__. That's not the best
> way to build a Cygwin binary since the package could use these
> defines to include different headers and such. When compiling for
> Cygwin, try the pure unixy way first. And you're talking about a
> header `crypt.h'. You know that a header with that name already
> exists in Cygwin, don't you?
Thank you for all these. I'll try to see what I can do without
__WINDOWS__, trying to avoid "crypt.h" conflicts, etc. Much appreciated!
(A four dimensional maze, stories, essays, artwork, and other things...)
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
Bug reporting: http://cygwin.com/bugs.html