Date: Wed, 31 Jul 1996 19:30:03 +0200 (MET DST)
From: Mark Habersack <grendel AT ananke DOT amu DOT edu DOT pl>
To: Alexander Lehmann <lehmann AT mathematik DOT th-darmstadt DOT de>
cc: djgpp AT delorie DOT com
Subject: Re: MSDOS Unzip Program
In-Reply-To: <4tll12$1nsr@rs18.hrz.th-darmstadt.de>
Message-ID: <Pine.NEB.3.93.960731192319.8041B-100000@ananke.amu.edu.pl>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII

On 30 Jul 1996, Alexander Lehmann wrote:

>Alaric B. Williams <alaric AT abwillms DOT demon DOT co DOT uk> wrote:
>: Mark Habersack <grendel AT ananke DOT amu DOT edu DOT pl> wrote:
>: >3.) You have LFNs if you recompile the sources under DJGPP V2 and use it under
>: >Win95. Try unpack RHIDE sources without it! Or maybe use WinZip32 (almost 1MB
>: >of soft and no cmdline) - no thanks!
>
>: I want LFNs, but I really really really don't want win95 (IT SUCKS). 
>
>: This is a shame.
>
>: Perhaps there is a scope for a DOS TSR LFN driver.
>
>: You'd see SFNs in the DOS utilities... but would they still work?
>: Would a disk doctor program see them and think "Oh, LFNs" or would it
>: think "This is DOS 6.0, so we won't have any LFNs"?
The TSR could preserve LFNs transparently for "old" DOS programms by applying
the WIn95 algorithm for modyfying file names to be in 8.3 format. TSR also
could fake that Win95 are running by providing the INT 2F interface. This,
however, could be dangerous as TSR wouldn't be able to provide ALL services
available when Win95 is installed - and this could cause a big CRASH.

>
>An old disk utility would probably think the directory is corrupt (the
>same in Win95, that's why old dos program may not write to disk
Not quite so. LFNs are stored in Win95 as regular DOS directory entries only
the attributes field is set to a non-existent attribute combination (0xFF I
think). So most of the DOS disk utils (including NU 8 for example) would just
skip the "bad" entries. LFN aware utility, however, could recognize that the
old dir entry is being followed by one with "bad" attributes, and that means
we're dealing with LFN. As far as I know, the LFN chain ends with an entry
with some special attribute value (0xF0 I suppose, but don't quote me).

Greetz, Mark

/*******************************/
/** So here I am once more... **************************************/
/** When you grown up and leave your playground, where you kissed **/
/** your Prince and found your Frog - remember the Jester that    **/
/** showed you tears, a Script for Tears... ************************/
/*********************************************/