Mailing-List:	contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Subscribe:	<mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive:	<http://sourceware.org/ml/cygwin/>
List-Post:	<mailto:cygwin AT cygwin DOT com>
List-Help:	<mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/ml/#faqs>
Sender:	cygwin-owner AT cygwin DOT com
Mail-Followup-To:	cygwin AT cygwin DOT com
Delivered-To:	mailing list cygwin AT cygwin DOT com
Date:	Mon, 13 Sep 2004 15:10:29 -0400 (EDT)
From:	Igor Pechtchanski <pechtcha AT cs DOT nyu DOT edu>
Reply-To:	cygwin AT cygwin DOT com
To:	Jan Nieuwenhuizen <janneke AT gnu DOT org>
cc:	cygwin AT cygwin DOT com
Subject:	Re: Installation hangs because asks for input from terminal (Attn: tetex maintainer, setup maintainer)
In-Reply-To:	<87d60qm23w.fsf@peder.flower>
Message-ID:	<Pine.GSO.4.61.0409131502380.13303@slinky.cs.nyu.edu>
References:	<41416E12 DOT 30103 AT math DOT md> <Pine DOT GSO DOT 4 DOT 61 DOT 0409101300420 DOT 5749 AT slinky DOT cs DOT nyu DOT edu> <87llfenmgx DOT fsf AT peder DOT flower> <Pine DOT GSO DOT 4 DOT 61 DOT 0409131355080 DOT 13303 AT slinky DOT cs DOT nyu DOT edu> <874qm2nigl DOT fsf AT peder DOT flower> <Pine DOT GSO DOT 4 DOT 61 DOT 0409131423110 DOT 13303 AT slinky DOT cs DOT nyu DOT edu> <87d60qm23w DOT fsf AT peder DOT flower>
MIME-Version:	1.0
X-Scanned-By:	MIMEDefang 2.39

On Mon, 13 Sep 2004, Jan Nieuwenhuizen wrote:

> Igor Pechtchanski writes:
>
> > Not really.  As Max said, the environment in which the postinstall scripts
> > run is the same as the one you'd get when you run them on the command line
> > (albeit from a non-login /bin/sh).  That environment will contain all of
> > the Windows-set environment variables.  If the tools require that some
> > variables be unset (because they might throw the tools off, for example),
> > the variables have to be unset in the postinstall script -- setup.exe
> > can't be expected to know all of the variables potentially conflicting
> > with all of the packages.
>
> No, of course.  That's why I'd like to have an empty (rather than a
> cleaned) environment, isn't that possible?  That sounds cleaner than
> fixing all scripts?

Most scripts don't need an empty environment.

> FWIW, the way teTeX works, most variables set in texmf.cnf are
> overridden by the environment; it would be next to impossible to
> guarantee a successful teTeX environment after logging in if variables
> from the WINDOWS are allowed to silently enter the Cygwin
> environment...

Well, if you install teTeX with an empty environment, and then attempt to
run it from a user shell (which *will* inherit the Windows settings),
wouldn't it still be broken?

(Yes, you can play games with /etc/profile.d scripts, but that can only
take you so far...)

BTW, as CGF said, you can't have a completely empty environment, or
Cygwin won't work.  I recall there was a thread a while ago about patching
"env -i" to keep some of the necessary variables, but I just checked env
on my system, and looks like that thread didn't result in any changes to
sh-utils.

Why not just unset offending variables for now?  You can kill anything
starting with "TEX", for example, with something like

eval "`env | sed -n '/^TEX/s/=.*$/=/ p'`"

HTH,
	Igor
-- 
				http://cs.nyu.edu/~pechtcha/
      |\      _,,,---,,_		pechtcha AT cs DOT nyu DOT edu
ZZZzz /,`.-'`'    -.  ;-;;,_		igor AT watson DOT ibm DOT com
     |,4-  ) )-,_. ,\ (  `'-'		Igor Pechtchanski, Ph.D.
    '---''(_/--'  `-'\_) fL	a.k.a JaguaR-R-R-r-r-r-.-.-.  Meow!

"Happiness lies in being privileged to work hard for long hours in doing
whatever you think is worth doing."  -- Dr. Jubal Harshaw

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/

- Raw text -