X-Authentication-Warning: hp2.xraylith.wisc.edu: khan owned process doing -bs
Date: Wed, 19 Jan 2000 21:02:55 -0600 (CST)
From: Mumit Khan <khan AT NanoTech DOT Wisc DOT EDU>
To: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
cc: Nate Eldredge <neldredge AT hmc DOT edu>, djgpp-workers AT delorie DOT com
Subject: Re: (patch) updated protoize patch
In-Reply-To: <Pine.SUN.3.91.1000119121342.9609e-100000@is>
Message-ID: <Pine.HPP.3.96.1000119205812.12013C-100000@hp2.xraylith.wisc.edu>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: dj-admin AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

On Wed, 19 Jan 2000, Eli Zaretskii wrote:

> > 
> > foo.sav might be more reasonable.
> 
> I agree: IMHO foo.sav is better.

That's what I meant, not foo.csav, sorry (getting used to 8.3 must be
hard!).

I've submitted my final changes to gcc-patches that include two things
that handle 8.3 filenames:
  
  - use foo.cX instead of foo.c.X
  - use foo.sav instead of foo.c.save

when __MSDOS__ is defined.

My patch does *not* contain the djgpp-specific code using $DJGPP magic
variable. DJGPP folks will have to take that one up once my last change 
is in.

> It would be even nicer if the code would test whether long file names
> are supported, and if so, use the original .save extension appended to
> the file name.  A call to pathconf (with the _PC_NAME_MAX argument) is
> all you need to check for long file names: if it returns 12, long
> names aren't available (if they are, it will typically return 255).

I agree, especially on systems that support pathconf(). Unfortunately,
it tells you nothing about how many "."s you can have, so you'll still
need __MSDOS__/__DJGPP__ and so on.

I'm done with what I had intended to do for protoize, at least for now.

Regards,
Mumit