Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!ub!rutgers!cmcl2!sbcs!max!rosalia
From: rosalia@max.physics.sunysb.edu (Mark Galassi)
Newsgroups: comp.sys.sgi
Subject: Re: TeX 3.0 is busted, get 3.1
Message-ID: <1991Jun9.191920.13047@max.physics.sunysb.edu>
Date: 9 Jun 91 19:19:20 GMT
References: <9106032156.AA20117@karron.med.nyu.edu> <luis.644517375@keckiris>
Reply-To: rosalia@max.physics.sunysb.edu (Mark Galassi)
Organization: Institute for Theoretical Physics, SUNY at Stony Brook
Lines: 24

In article <luis.644517375@keckiris> luis@rice.edu (luis soltero) writes:
>TeX 3.0 compiled and ran fine on a 4D/310 under IRIX 3.3.1
>using gcc.

>>Get 3.1. 
>>It works. 3.0 is badly busted. I had the same problem.

>>>Has anyone successfully built TeX 3.0 on an 4D/25?  I tried building it and
>>>it compiled o.k. (after two tries), but when I tried to run latex on the
>>>sample document sample.tex, it generated one word per page. 

If you look at the "site.h" file in the UNIX TeX3.0 distribution which
is right now on Labrea, you will see that there is a place where you
must #define or typedef "schar" to get signed chars.  The MIPS C
compiler will (ordinarily) keep chars unsigned.  I kept on getting one
word/page until I changed that.  After I changed it, it worked fine.

Also:  what's all this about 3.1?  I checked labrea.stanford.edu
(or whatever) and didn't see it a couple of weeks ago.

-- 
    {These opinions are mine, and should be everybody else's :-)}
	Mark Galassi		rosalia@dirac.physics.sunysb.edu
				rosalia@sunysbnp.BITNET