Message-ID: <393D3932.AC26565E@softhome.net>
Date: Tue, 06 Jun 2000 20:47:30 +0300
From: Laurynas Biveinis <lauras AT softhome DOT net>
X-Mailer: Mozilla 4.72 [en] (Win98; U)
X-Accept-Language: en
MIME-Version: 1.0
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
CC: djgpp-workers AT delorie DOT com
Subject: Re: Testers wanted: a fix for GCC header problem
References: <3939239A DOT 18870 DOT 336AAF AT localhost> <393BE1CA DOT B06CAC0F AT softhome DOT net> <200006061748 DOT UAA20575 AT alpha DOT netvision DOT net DOT il>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Reply-To: djgpp-workers AT delorie DOT com

Eli Zaretskii wrote:
> > I'm afraid to make the conclusion that DJGPP headers should reuse GCC
> > sentinel system. This would be an internal change, so FAQs should
> > remain minimized. I could take responsibility for making this change
> > and ensure the absence of problems.
> 
> I would really like to explore the other solution first: namely, to
> get the GCC maintainers to refrain from forcing their headers on
> compliant libraries.  

*BSD folks have tried it but ended in submitting patches for those
GCC headers for adding additional #ifndef _BSD_... GCC maintainers
accepted those patches with greetings and smiles from ear to ear.

OK, I will try to ask them what are the *technical* reasons for
their current policy. Let's hope it won't cause a flame war.

> It is simply _not right_ for a compiler to force
> a library to track changes in compiler's configury.  A reasonably
> complaint library should be able to proceed with its development
> without having to consult the latest GCC CVS tree.

DJGPP doesn't have much luck there because it uses less common
type definition machinery with <sys/djtypes.h> instead of #ifndef _FOO_DEFINED
... #define _FOO_DEFINED. (I don't mean it's a bad thing). 

Laurynas