From: loth AT gec DOT net (Burton Radons)
To: Nate Eldredge <nate AT cartsys DOT com>
Cc: djgpp-workers AT delorie DOT com
Subject: Re: Regex patch
Date: Wed, 08 Jul 1998 13:02:24 GMT
Message-ID: <35a75ba0.37737978@mail.cow-net.com>
References: <35a68f9e DOT 9230585 AT mail DOT cow-net DOT com> <35A2F990 DOT 1D612365 AT cartsys DOT com>
In-Reply-To: <35A2F990.1D612365@cartsys.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 8bit
Precedence: bulk

On Tue, 07 Jul 1998 21:46:09 -0700, you wrote:

>-Wtraditional is a particular annoyance, since it seems there's some
>difference in the type of a numeric constant between K&R and ANSI C.  I
>don't know the details, but I suspect it will mean adding `U's to a lot
>of numbers.

Mainly const problems, actually.  Only two Us.

>Btw, this doesn't *have* to be done; there's a perfectly good way, via
>the Makefile, to just make GCC shut up.

I have been having bad luck with submitting modifications.

The changes that Nate's list of warnings requires have been
implemented.  Does anyone else have any surprise problems, before I
resubmit the full patch?  (I think it would be wise, since I both went
through the files and modified again (I made some changes that were
more than compiler-scale in the first), and want to avoid the
indiscretion in backslashes completely)

As you say, we could just get rid of the warnings with flags.  My
patch does things at such a minor scope that it produces the same
effect -- I'm just not letting the compiler guess.  Both paths are
open -- DJ?

I like the warning flag option best.  Or, rather, I find it's
advantages make it logically tolerable.

- Burton Radons, loth AT gec DOT net
Vancouver Island, British Columbia, Canada