From xemacs-m  Mon Jan  6 23:20:42 1997
Received: from venus.Sun.COM (venus.Sun.COM [192.9.25.5])
          by xemacs.cs.uiuc.edu (8.8.4/8.8.4) with SMTP
	  id XAA02615 for <xemacs-beta@xemacs.org>; Mon, 6 Jan 1997 23:20:42 -0600 (CST)
Received: from Eng.Sun.COM ([129.146.1.25]) by venus.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id VAA26482; Mon, 6 Jan 1997 21:20:13 -0800
Received: from kindra.eng.sun.com by Eng.Sun.COM (SMI-8.6/SMI-5.3)
	id VAA04051; Mon, 6 Jan 1997 21:20:10 -0800
Received: from xemacs.eng.sun.com by kindra.eng.sun.com (SMI-8.6/SMI-SVR4)
	id VAA06182; Mon, 6 Jan 1997 21:20:10 -0800
Received: by xemacs.eng.sun.com (SMI-8.6/SMI-SVR4)
	id VAA13029; Mon, 6 Jan 1997 21:20:07 -0800
Date: Mon, 6 Jan 1997 21:20:07 -0800
Message-Id: <199701070520.VAA13029@xemacs.eng.sun.com>
From: Martin Buchholz <mrb@Eng.Sun.COM>
To: Steven L Baur <steve@miranova.com>
Cc: xemacs-beta@xemacs.org
Subject: Release schedule for 19.15 and 20.0 (how about January 20?)
In-Reply-To: <m2eng04z0i.fsf@altair.xemacs.org>
References: <m2eng04z0i.fsf@altair.xemacs.org>
Reply-To: Martin Buchholz <mrb@Eng.Sun.COM>
Mime-Version: 1.0 (generated by tm-edit 7.100)
Content-Type: text/plain; charset=US-ASCII

>>>>> "sb" == Steven L Baur <steve@miranova.com> writes:

sb> I would like to get both 19.15 and 20.0 released as soon as possible.
sb> I see the 19.15 release being a normal XEmacs release with all the
sb> usual binary distributions.  To make things easier for the binary
sb> builders 20.0 will be distributed Emacs-style source-only.  I also see
sb> less of a drive to have all the binary 19.15 builds in place by net
sb> release time, over the following week or two is good enough.

sb> We have five options on how the v20 distribution tree will look.

sb> 1.  Don't change anything.
sb> The problem is that we don't have full .elc compatibility between mule
sb> and latin-1 environments, and people choosing latin-1 will have to
sb> rebuild an arbitrary amount of the lisp directories in order to get a
sb> functional program.  I don't see this as a particularly viable
sb> option.

As I pointed out earlier, if we remove Japanese text from tm-edit.el,
it will solve the biggest byte-compile problem.  latex-math-symbol.el
can be left byte-compiled for mule - it won't work otherwise.

Here's a way to determine if there are any deeper (e.g. defsubsts)
differences between xemacs-mule + tm and xemacs-latin1 + tm: 

- build tm install dirs from the original tm source distribution using
  both xemacsen and then

- compare the generated .el's and .elc's, ignoring the generated
  `header' part of the .elc's, which contains timestamps.

I vote for option (1) if we can get a working tm + xemacs-latin1 using
.el's byte-compiled by xemacs-mule, AND, if we collectively make a
commitment to putting out a 20.1 soon thereafter (within 2 months,
say).  There will be enough of a chance to split XEmacs into pieces
then, esp. if we ever implement the packaging system.

sb> 2.  Distribute in 3 tar files, xemacs-20.0-common.tar.gz for all
sb>     common files, xemacs-20.0-mule-elcs.tar.gz for all mule lisp .elcs,
sb>     and xemacs-20.0-latin-1-elcs.tar.gz for latin-1 .elcs.
sb> Download the common tar, and pick one of the two .elcs tars to fit.

sb> 3.  Distribute in 2 tar files, xemacs-20.0-mule.tar.gz and
sb>     20.0-latin-1.tar.gz, each full distributions, pick 1.

sb> 4.  Distribute in 2 tar files, xemacs-20.0-mule.tar.gz (full mule
sb>     distribution), xemacs-20.0-latin-1-elcs.tar.gz.

sb> 5.  Distribute in 2 tar files, xemacs-20.0-latin-1.tar.gz,
sb>     xemacs-20.0-mule-elcs.tar.gz.

sb> I prefer option 2 the best.

sb> The tentative schedule looks like:

sb> Jan-11	beta 90
sb> Jan-15	beta 91 (if necessary)
sb> Jan-18	beta 92
sb> Jan-20	Net Release.

sb> We'll know very quickly whether a 19.16 release will be required.  If
sb> it goes anything like 19.14, there will be a 19.16.  I see 19.16/20.1
sb> coming out in short order, perhaps by the end of February.

Agreed.

Martin

