[3498] in SIPB bug reports
Re: TeX file cleanup in the sipb locker
daemon@ATHENA.MIT.EDU (Calvin Clark)
Wed Feb 10 16:11:35 1993
Date: Wed, 10 Feb 93 16:10:35 -0500
From: Calvin Clark <ckclark@mit.edu>
To: amgreene@Athena.MIT.EDU
Cc: jtkohl@Athena.MIT.EDU, jik@Aktis.COM, bug-sipb@Athena.MIT.EDU
In-Reply-To: [3497]
Reply-To: ckclark@mit.edu
[3497] daemon@ATHENA.MIT.EDU (amgreene@Athena.MIT.EDU) SIPB bug reports 02/10/93 08:01 (11 lines)
Subject: TeX file cleanup in the sipb locker
Date: Wed, 10 Feb 93 08:01:03 EST
From: amgreene@Athena.MIT.EDU
I'm afraid I'm no longer maintaining TeX stuff... I think
Calvin was the last one to touch it... :-)
- Rhu the Buck Passer :-)
--[3497]--
I have no problem with what jik did, considering that he needed the
space and I haven't gotten around to installing 3.141 and friends yet.
When I do that (I am not putting a schedule on it), there will probably
be some user-visible differences, but only to those who use
tex/latex/biglatex from the SIPB locker, which is not a heck of a lot of
people.
These are the plans, for your approval/disapproval. (Note that I am
likely to react to violent or irrational disapproval with something
along the lines or "bugger off; do it yourself." :-)
Some of this was discussed at a recent SIPB meeting:
1. All TeX stuff will be in a replicated tex-install volume, and
symlinks from the sipb locker will be created into that volume. This
includes all fonts, macros, format files, MF bases, and binaries.
2. LaTeX will have support for NFSS the New Font Selection Scheme, but
the support will not be on by default. You will have to use the
"newlfont" style to get NFSS if you want it. Hopefully, people will try
it out, because this method will become the default in future versions
of LaTeX.
3. AMSTeX/LaTeX will exist, and have as much support stuff as I can find
and throw in with it. This is primarily fonts, macros, and a format
file or two.
4. Metafont 2.71. It exists, it works, yippie hoo.
5. New version of dvips/xdvi.
This requires some explanation. Both dvips and xdvi can be configured to
build pk files if particular magnifications are not found. The question
I have for those who might care about this are:
a) Should this be enabled?
b) Should it be the default, or should a special flag
to dvips/xdvi be required?
c) Where should the new pk files be placed.
My guess is that it should be enabled, but probably not by default,
because there are many novice users who probably have their dotfiles so
screwed up that they can't find cmr10.300pk, and there is no need to
build fonts in that circumstance. Also, there are still some slow
machines out there, and running MakeTeXPK isn't exactly fast on, say, an
RT. However, I think that user should be able to build these fonts
automatically if they want to. That covers a) and b).
As for c), I was thinking about placing the new pk files in a
world-writable directory on the local disk of the workstation, perhaps
in /usr/tmp/pks, or something. The programs dvips and the SIPB xdvi
would look there automatically.
Now, of course, it we be ideal if somehow we could be notified when
someone needs to build a particular magnification, so that the
appropriate pk files can be installed in the regular directory, for
others to use. How should this notification occur? Should MakeTEXPK
print a message saying, ``Yo. Tell us about these fonts that you're
making, and we'll put them in the main directory so that you don't have
to do this again?''
The solution some sites use is to have a global world-writable directory
for this. This could be done easily on AFS by creating a directory with
system:anyuser (or authuser) rli access, and new pk files would be put
in there. Every once in a while, the maintainer would copy the files
from that temporary repository into the main directory, although both
directories would be searched by the dvips and xdvi. Is this a good
idea? Note that it provides users with a sort of a bitbucket, which is
subject to potential abuse. However, if the access were authuser, we'd
know who it was, and we can presumeably find out where they live. :-)
These are decisions I'd rather not make myself, but I'll make them
myself if no one else provides input. Note that I have almost all of
this stuff working. If you do this, you can use most of it on most
platforms. I haven't finished all the AMSTeX stuff yet, and I
haven't copied all of the random SIPB latex macros that have dubious
origins yet. (That's a lot more work that it seems. Maybe I won't
worry about trying to find the newest versions of everything. I mean,
who gives a willy if there's been an update to siam.sty?)
In any case, to test out, do this:
athena% detach sipb
athena% attach -t afs -m /mit/sipb -e /afs/sipb/project/tex/install
athena% set path = (/mit/sipb/bin $path)
Run anything in /mit/sipb/bin.
Nothing else should be necessary. Oh, yeah, I haven't done xdvi yet,
but that's just one patch.
Again, it's not finished yet; I just want people to know that it's not
complete vaporware out in lah-lah land, either.
-Calvin
