Sender: bill AT taniwha DOT tssc DOT co DOT nz
Message-ID: <359A07F3.6E49F4C4@taniwha.tssc.co.nz>
Date: Wed, 01 Jul 1998 21:57:07 +1200
From: Bill Currie <bill AT taniwha DOT tssc DOT co DOT nz>
MIME-Version: 1.0
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
CC: Hans-Bernhard Broeker <broeker AT physik DOT rwth-aachen DOT de>,
        Andris Pavenis <pavenis AT lanet DOT lv>,
        djgpp workers list <djgpp-workers AT delorie DOT com>
Subject: Re: Some notes about DJDEV202.ZIP
References: <Pine DOT SUN DOT 3 DOT 91 DOT 980701110656 DOT 12487D-100000 AT is>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Precedence: bulk

Eli Zaretskii wrote:
> 
> On Tue, 30 Jun 1998, Hans-Bernhard Broeker wrote:
> 
> > Maybe a little change to the gcc
> > sources, switching the order of searching, so that $LIBRARY_PATH
> > and $C_INCLUDE_PATH are searched after the version-specific
> > directories.
> 
> IMHO, making the DJGPP port behave differently from the Unix version
> is a bad idea.  It invalidates the description in the docs of the
> search process and the experience of those who use gcc on several
> platforms.

What about telling gcc to search in $DJDIR/{include,lib} instead of
/usr/{include,lib}?  This would make djgpp operation (IMHO) VERY
standard even if slighly skew-wiff.  Heck, isn't $DJDIR effectivly /usr
anyway?  It's got bin, include, src, tmp and lib haning off it and
contrib is sort of like /usr/local.

This seems to me to be the cleanest way of doing this as this is how gcc
was designed.  ie look in the version specific dir, then /usr/*.

Bill
-- 
Leave others their otherness.