libtool/mail/docs

From nobody Wed Oct 14 17:14:44 1998
X-From-Line: rms@santafe.edu Mon Jun 01 19:52:46 1998
Return-Path: <rms@santafe.edu>
Delivered-To: gord@trick.profitpress.com
Received: (qmail 15232 invoked from network); 1 Jun 1998 19:52:45 -0000
Received: from unknown (HELO bambam.m-tech.ab.ca) (127.0.0.1)
  by 127.0.0.1 with SMTP; 1 Jun 1998 19:52:45 -0000
Received: from sfi.santafe.edu (gateway [10.0.0.1]) by bambam.m-tech.ab.ca (8.8.5/8.6.9) with SMTP id MAA32739 for <gord@m-tech.ab.ca>; Mon, 1 Jun 1998 12:09:11 -0600
Received: from wijiji.santafe.edu by sfi.santafe.edu (4.1/SMI-4.1)
	id AA03877; Mon, 1 Jun 98 11:55:41 MDT
Received: by wijiji.santafe.edu (SMI-8.6/SMI-SVR4)
	id LAA04106; Mon, 1 Jun 1998 11:55:40 -0600
Date: Mon, 1 Jun 1998 11:55:40 -0600
Message-Id: <199806011755.LAA04106@wijiji.santafe.edu>
From: Richard Stallman <rms@santafe.edu>
To: gord@m-tech.ab.ca
In-Reply-To: <86hg27twsh.fsf@trick.profitpress.com> (message from Gordon
	Matzigkeit on 30 May 1998 12:53:50 -0600)
Subject: Re: libtool manual comments
Reply-To: rms@gnu.org
References: <199805240500.XAA01237@wijiji.santafe.edu> <86hg27twsh.fsf@trick.profitpress.com>
Xref: trick.profitpress.com mail.libtool:1476
Lines: 23
X-Gnus-Article-Number: 1   Mon Nov  2 17:18:09 1998

      Regarless, it needs to have a light touch on the option
    namespace, since it forwards any unrecognized options to the
    underlying compiler.  This is so that people can pass arbitrary flags
    that libtool doesn't know about.

Sorry, I don't follow the logic.

Long-named options are the GNU standard, so every a GNU program should
provide a long-named version of every option name.

If you think there is some practical reason why libtool should not support
long-named versions of its own options, would you please spell it out?
I don't see why it would cause any problem.

     RMS> In section 5.3.1 there is a table of environment variable names,
     RMS> that should be @table @code.  Section 12.4 has one too.

    Actually, these are not tables, they are lists of `@defvar' blocks.
    What would you recommend in this situation?

I'd recommend @table @code.  We don't use @defvar for environment
variables.