Path: utzoo!attcan!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!mcsun!ukc!dcl-cs!aber-cs!athene!pcg From: pcg@cs.aber.ac.uk (Piercarlo Grandi) Newsgroups: comp.text.tex Subject: Re: TEX availability Message-ID: Date: 16 Sep 90 17:45:29 GMT References: <8260@jarthur.Claremont.EDU> <8263@jarthur.Claremont.EDU> Sender: pcg@aber-cs.UUCP Organization: Coleg Prifysgol Cymru Lines: 63 Nntp-Posting-Host: teacho In-reply-to: dhosek@sif.claremont.edu's message of 4 Sep 90 06:27:57 GMT On 4 Sep 90 06:27:57 GMT, dhosek@sif.claremont.edu (Hosek, Donald A.) said: dhosek> Commontex. Jon Radel reports that the PC version does not pass dhosek> trip. This is very strange -- the UNIX version does, but for minor differences in the last few digits of some floating point numbers etc... I think that the trip test itself is a disgrace however, because it essentially requires any "conforming" TeX implementation to have the same extremely baroque user interface as the Pascal/WEB one. Testing for conformity should be based on ensuring that the input language accepted is the same, and the DVI output produces similar-looking printouts. I had considered giving CommonTeX 2.1 a more terse and elegant user interface, but did not, because then the coveted (if absurdly defined) trip-test compatibility would have been lost. dhosek> In any event, it's only around v2.1 (or at least that's the dhosek> latest version Jon has been able to get working under MS-DOS dhosek> according to the last catalog I got from him). CommonTeX release numbers bear no relationship to Pascal/WEB TeX ones; release 2.1 is essentially equivalent to Pascal/WEB TeX 2.9. dhosek> I suspect the whole effort was quietly abandoned with the advent dhosek> of WEB2C. This would have been unfortunate -- the WEB2C confusion is a disgrace. Also, the much preferable route to get a C TeX from the Pascal/WEB one nowadays is to use Gillespie's p2c translator, which does a much better (simpler use, more readable output) job. Fortunately I think you are wrong -- I have read that the author (a meritorious person!) has actually upgraded CommonTex so that it no longer uses the same internal structure as Pascal/WEB TeX, and, to great advantage in flexibility and speed, actually allocates records as needed from the heap using malloc or something similar. I have a profound antipathy for Pascal/WEB TeX, and I liked CommonTeX 2.1, even if I ultimately stuck with troff because the input language is vastly simpler (I had thought that a more difficult input language than troff's was an impossible feat, until I read the TeXbook :-/), and is twice as slow as troff (another feat that I thought impossible, especially considering that all TeX implementations precompile their macro packages, while the ditroff I am using does not -- notice though that CommonTeX is actually faster than Pascal/WEB TeX by a good margin). Note that much of the slowness could be attributed to the lack of a floating point coprocessor in my machine, which probably impacts a TeX implementation much more than a troff one -- but then that a text processor be floating point intensive seems to be an original defect, to me. I would be very interested in any updated CommonTeX, especially if it so much faster as to be comptetitive with troff, or at least does not have hard coded limits to the size of its tables. -- Piercarlo "Peter" Grandi | ARPA: pcg%uk.ac.aber.cs@nsfnet-relay.ac.uk Dept of CS, UCW Aberystwyth | UUCP: ...!mcsun!ukc!aber-cs!pcg Penglais, Aberystwyth SY23 3BZ, UK | INET: pcg@cs.aber.ac.uk