Message-ID:	<324B3876.2A5E@cs.com>
Date:	Thu, 26 Sep 1996 19:14:14 -0700
From:	"John M. Aldrich" <fighteer AT cs DOT com>
Reply-To:	fighteer AT cs DOT com
Organization:	Three pounds of chaos and a pinch of salt
MIME-Version:	1.0
Newsgroups:	comp.os.msdos.djgpp
To:	Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>,
	Charles Sandmann <sandmann AT clio DOT rice DOT edu>
CC:	DJGPP Workers Mailing List <djgpp-workers AT delorie DOT com>
Subject:	Re: 'Cannot open'
References:	<Pine DOT SUN DOT 3 DOT 91 DOT 960926170346 DOT 17125H-100000 AT is>

Eli Zaretskii wrote:
> 
> On Thu, 26 Sep 1996, Mark Habersack wrote:
> 
> > When trying to run a DJGPP v2 program from a clean system booted from floppy
> > I have received the following error:
> >
> > 'SPEC=A:\COMMAND.COM: cannot open'
> 
> Your PATH is empty (or garbled).  Set it to any directory you like and it
> will work.  Charles Sandmann promised to correct this for the next DJGPP
> release (it's a bug in the stub).
> 
> (Funny, this problem: a year has passed with nobody reporting it, and
> just for the last fortnight 4 reports, including 1 from myself.)

Eli, I have a question about this particular error.  For my development
of djverify, I have successfully implemented changes to the stub to
cause
it to emit different error codes based on the error that occurred.  I
also
wrote a batch file which interprets these error codes and displays their
causes and solutions.

Is it a good idea to mention this problem in the batch file's report, or
can I assume that it will be corrected in a later DJGPP version?
Alternatively, should I wait to release djverify until Charles releases
the corrected stub, and use that instead of the v2.0 one?

If I use the old stub, then the batch file will have to be responsible
for
detecting the PATH error instead of the program, but the behavior will
be
consistent with the "normal" stub.

If I use the new stub, then my program would be able to run properly and
detect the missing PATH when it runs its own diagnostics.  This would
produce better output IMHO, but would not be consistent with previous
versions of the stub (which might be confusing to some users).

Finally, would Charles be interested in implementing those error codes
in
the stub so that my batch file could be used as a universal diagnostic
instead of merely for my djverify program?  I know that I have asked
this
before, but I'd like a more comprehensive opinion.  The code required is
less than trivial.  :)

Thank you for your input!

-- 
John M. Aldrich, aka Fighteer I <fighteer AT cs DOT com>

-----BEGIN GEEK CODE BLOCK-----
Version: 3.1
GCS d- s+:- a-->? c++>$ U@>++$ p>+ L>++ E>+ W++ N++ o+ K? w(---) O-
M-- V? PS+ PE Y+ PGP- t+(-) 5- X- R+ tv+() b+++ DI++ D++ G e(*)>++++
h!() !r !y+()
------END GEEK CODE BLOCK------

- Raw text -