Date: Mon, 31 Aug 1998 17:39:52 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
To: rjvdboon AT cs DOT vu DOT nl
cc: djgpp AT delorie DOT com
Subject: Re: problem when unzipping
In-Reply-To: <6se3a2$ogr$1@star.cs.vu.nl>
Message-ID: <Pine.SUN.3.91.980831172834.24142J-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Precedence: bulk


On 31 Aug 1998 rjvdboon AT cs DOT vu DOT nl wrote:

> Eli Zaretskii <eliz AT is DOT elta DOT co DOT il> wrote:
> > If you install DJGPP on Windows 9X, get another unzip program which
> > supports long file names on Windows (unzip386 doesn't) and
> > re-install everything from scratch.
> 
> Wouldn't it be an idea to put a newer (Info)zip/unzip in the FTP-dir
> at least in v2apps/ The newest versions can make use of LFN's.

This has come up before.  The problem is that DJGPP v2 programs are not 
self-sufficient: they need a DPMI server to run.  And CWSDPMI comes in a 
.zip file, so it's a chicken-and-egg problem.

A possible solution might be to bind UnZip with PMODE/DJ, but somebody 
needs to test that binary in all kinds of platforms supported by DJGPP, 
to see it doesn't blow, before we could suggest people to use it.  Also, 
PMODE/DJ doesn't support virtual memory, so we need to be sure UnZip 
doesn't use too much memory and doesn't fail on memory-tight machines.

Since nobody really complained about unzipping, there was no real 
incentive to mess with this.  But if you have enough motivation, please 
go ahead, I think that having an LFN-aware UnZip program ready for use is 
a good idea.

> but I
> don't now yet if we're allowed to do it (I think one should get
> permission from the Info-Zip team, at least I'd try to get it)

There's EPZIP (v2apps/epz100b.zip) that is free or GPL, I think.  You 
could use that instead of asking somebody for permission.

> I know LFN=y should be set

Not true.  What you must make sure is that LFN is NOT set to N, that's 
all.

But I think it would be easier to make a special version of the program 
that avoids loading DJGPP.ENV, so that it will always work with LFNs, 
unless the user explicitly sets LFN=n in the environment, in which case I 
think you should honor that.