Date: Fri, 27 Apr 2001 20:35:42 +0300
From: "Eli Zaretskii" <eliz AT is DOT elta DOT co DOT il>
Sender: halo1 AT zahav DOT net DOT il
To: "Tim Van Holder" <tim DOT van DOT holder AT pandora DOT be>
Message-Id: <9791-Fri27Apr2001203542+0300-eliz@is.elta.co.il>
X-Mailer: Emacs 20.6 (via feedmail 8.3.emacs20_6 I) and Blat ver 1.8.9
CC: djgpp-workers AT delorie DOT com
In-reply-to: <CAEGKOHJKAAFPKOCLHDIAENKCCAA.tim.van.holder@pandora.be>
Subject: Re: ANNOUNCE: Fileutils 4.0 beta 2
References:  <CAEGKOHJKAAFPKOCLHDIAENKCCAA DOT tim DOT van DOT holder AT pandora DOT be>
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> From: "Tim Van Holder" <tim DOT van DOT holder AT pandora DOT be>
> Date: Fri, 27 Apr 2001 18:03:15 +0200
> 
> > There are two ways:
> >
> >   - set FNCASE=y before zip'ping the package.
> >
> >   - edit the *.mft files so that they have the right letter-case, then
> >     invoke zip with the manifest file as a response file, like this:
> >
> > 	   zip -9 fil40b.zip @manifest/fil40b.mft
> Hmm - I think recent versions of zip already store the exact case on LFN
> systems (I know my zip (2.3.3) has been doing so for a while (I last built
> my own end of '99)).

You are right for the case of files whose letter-case survives until
it gets to zip's application code, but this is a different situation.
File names such as TODO and NEWS are downcased by readdir (unless
FNCASE is set), because they fit into DOS 8+3 limits, so zip doesn't
know that they are in upper case.  That's what the two possible
solutions above are trying to avoid: the first stops downcasing in
readdir, the second avoids calling readdir in the first place.

> This is a bfd problem; it still uses a fixed-size stub area, so stubs that
> aren't exactly as long as the standard stub used by bfd won't work. IIRC
> this would be hard to fix, as most bfd routines expect to find the header
> info (the coff header, for DJGPP executables) at the start of an object.

Then perhaps we need to define a couple of new targets:
coff-go32-cwsdstub and coff-go32-pmodestub.

However, I wonder: does this mean that none of the Binutils programs
will work with these stubs?  Will `nm' and `strings' fail, for
example?