UKTeX V88 #22 Friday 15 July 1988 unix tex tape TeX2.9 on UNIX SysV Re: UKTeX V88 #21 TiB vs BiBTeX re: Crashing LaserWriter (UKTeX21.88) TEX and device drivers File transfer Postscript symbiont for VMS Re: UKTeX V88 #20 TeX2.9(2) ctex on VAX Aston TeX database --------------------------------- Editor Peter Abbott We have been without the mail system for two days whilst the two DEC discs were replaced with a single (larger capacity) drive. During the startup period some messages may have been delivered to the `black hole'. Please resubmit if necessary. Files updated for TeX 2.93 are trip.tex in public.texvms.tex.trip and public.unix.tex82.ctex.trip trap.mf in public.mffiles and public.unix.mf84.cmf.trap Also in public.texstatus two files from Barbara Beeton message_006.09_jul_1988 message_007.10_jul_1988 Latest TeXhax in the Archive is #63 Latest TeXmag in the Archive is V2N4 --------------------------------- From: David Brightly To: abbott @ aston Date: Fri, 8 Jul 88 18:49 GMT Subject: unix tex tape Message-Id: <08 JUL 1988 19:21:37 WC@UK.AC.DARESBURY.NNGA> Comments: hi peter it occurs to me that potential customers of your tape copying service might like to know what theyre getting! my impression is that you dont have the facilities to un tar the tape, so here's the top level README file. The following directories should be present: compress - from Usenet undump - from Usenet VAXbinaries and SUNbinaries - some binaries for those who take the risk of using other peoples binaries rather than compiling from source. No guarantee that they are up to date. tex82 - the TeX82 distribution babel - the start of some foreign language stuff mf84 - WEB Metafont for Unix utilityfonts - mf source for various METAFONT operations cmfonts - mf source, tfms, pl, some gf and a little pxl LaTeXfonts - the same for the special LaTeXfonts (no pxl, however) amsfonts - Cyrillic and new math symbols from the American Mathematical Society, compiled with old Metafont and converted to gf. Also gf files, tfm files for the Euler fonts. +++Editor - We do have DECshell and a copy of the tape has been added to the archive under [public.unix] but this listing does make it simple to find out what is available +++ --------------------------------- Date: 8 Jul 88 19:40:47 BST (Fri) Message-Id: <8807081940.Ag16169@uk.ac.stir.cs> X-Mailer: Ag [V1.0-Alpha] From: Sam Nelson To: info-tex@aston.mail Subject: TeX2.9 on UNIX SysV I chickened out of the `big pull' from Aston and sent Don Ward a cartridge (thanks again Don). Build completed trouble-free on Suns, but still a problem with the interpreter's switch statement on AT&T 3Bs (must get the new compiler...). If anyone's interested/hasn't figured the problem out for themselves, there are too many labels (300) in the switch statement in tex8.c. I had the same problem with Pat Monardo's stuff last year. Unfortunately in the TeX2.9 case, tex8.c is a web2c-generated file, so it really needs a mechanical fix rather than the semi-mechanical procedure I'm using at present. When I figure one out, either I'll post it here, or it'll be so far down the road that no-one will have the 3.X AT&T C compiler any more. Sam. PS: Is it only me that thinks the TeX distribution is getting too large for comfort these days? The compilation, etc. time is pretty trivial compared to the disk admin time, I'm finding... --------------------------------- Date: Sun, 10 Jul 88 20:57:39 GMT From: Sebastian Rahtz To: ABBOTTP@uk.ac.aston Subject: Re: UKTeX V88 #21 I have clashed with Philip Taylor before about LaTeX, but I cannot resist a reply to his anti-LaTeX message > First, let me state my biases: I dislike LaTeX, and think TeX superb. > Now, facts: LaTeX is a `higher-level' language which sites on top of > TeX; it is, in fact, a `format', in Knuth's sense of the word. I am > told that it is much simpler for naive users to use. In practise, > it sits not only on top of TeX {\it qua} TeX, but also on top of > a (modified version of) Plain.TeX, which is Knuth's own medium-level > `format'. Because Plain has been modified to make Lplain, one can I think that Philip, like many people, assumes that 'plain.tex' is somehow sacrosanct because we all have it; I would contend that Knuth wrote plain as a demonstration of what could be done, not a bible. 'lplain' is as valid as 'plain' in that someone with TeX (the REAL TeX, not TeX with plain preloaded) will be able to run it. > never assume that something which works in plain TeX will also work > unchanged in LaTeX; it is better to think of LaTeX as a separate language, > and never mix the two. Style-file designers are not allowed this luxury; true. but anything that is in TeX without a format preloaded will work in TeX. Because macros that Knuth defined in plain don't work when you load 'lplain' and 'latex.tex' is not a criticism of lplain. > they really do have to be gurus, conversant with both TeX and LaTeX. dont tell me that casual punters can write macros in plain TeX any more that they can in LaTeX. serious work demands a guru whatever happens > Leslie L. makes it plain that anyone thinking of writing (or modifying) > a style-file should be prepared to wade through realms of LaTeX source. quite right too! do you want to amend something you haven't read and understood? > > In practice, I do not believe that there is anything one can do in LaTeX > which cannot also be done in TeX (with the exception of the pictures well thats obvious, they have the same base (basic TeX)!. The pictures are just a sophisticated use of TeX macros > which LaTeX permits through its extended graphics fonts); the converse TeX *per se* does NOT define fonts. This myth that 'normal' TeX preloads CMR should be exploded once and for all. CMR is a nasty American typeface which Knuth unfortunately tied into 'plain.tex'. Lets all forget it as soon as possible and use decent fonts. > I believe to be probably true: that one can do some things in TeX which > one cannot do in LaTeX (or perhaps, which one can do in TeX easily, > and only with extreme difficulty in LaTeX). I really cannot think of anything that cannot be done in LaTeX *be a qualified practioner* (as in plain TeX). It shouldn't be _easy_ to change styles! > > I also feel that in using LaTeX, one cedes to much control over > the layout to the style-file designer; because I believe that layout do you really want each user to affect a style? what is the point of a style if it can be changed? LaTeX is for people who *want* to stick to a style Now if someone wants to criticise the facilities provided in a basic 'initex' then I will listen..... --------------------------------- From: Steve Kille To: UKTeX@uk.ac.aston Subject: TiB vs BiBTeX Date: Fri, 17 Jun 88 09:42:03 +0100 I am trying to choose between the two of these (currently finding myself using both). Two clear issues are: 1) TiB uses a database compatible with refer(1). This is very useful in the UNIX world. 2) BiBTeX seems to be more portable. Has there been any previous discussion on this choice? Are there any other issues? Would anyone care to comment on the quality of output of thw two systems? Lamport also notes "facilities available to ease the task of making bib files". Does anyone have pointers to such things for UNIX? Steve +++Editor - This is the first time I have seen this one so it must be one of those that our system posted to the `black hole' +++ --------------------------------- Received: from princeton.edu by NSS.Cs.Ucl.AC.UK via Satnet with SMTP id aa01969; 10 Jul 88 7:28 BST Received: from icecream.Princeton.EDU by Princeton.EDU (5.58+++/1.67) id AA06550; Sun, 10 Jul 88 02:46:14 EDT Received: by icecream.Princeton.EDU (3.2/1.62) id AA03760; Sun, 10 Jul 88 02:47:05 EDT Date: Sun, 10 Jul 88 02:47:05 EDT From: jonradel < (Jon Radel)jonradel%edu.princeton.icecream@edu.princeton> Message-Id: <8807100647.AA03760@icecream.Princeton.EDU> To: info-tex@uk.ac.aston, l.whitaker@uk.ac.city.cs Subject: re: Crashing LaserWriter (UKTeX21.88) Sounds like you're using a method of getting the PostScript files to the printer that doesn't follow the file with a ^D. This is the end-of-job marker in PostScript and, indeed, one of its effects is to clear out virtual memory. As you've noticed, the printer will in effect insert its own ^D after a certain idle period. My suggestions in order of elegance and reverse order of ease: 1) Get a "real" device driver, spooler, what-have-you, that knows enough about PostScript to seperate jobs with a ^D. I know nothing about UNIX (I assume this is what you're running) printer resources, so I can't help you on this. 2) Get dvi2ps to insert a ^D at the end of the file to be printed. 3) Get dvi2ps to insert a ^D at the beginning of the preamble; there are some circumstances where this will cause problems for the printer, because, if I remember correctly, there is a short period after the ^D where to printer doesn't want to communicate. --Jon Radel jonradel@icecream.princeton.edu jonradel%icecream.princeton.edu@princeton.edu --------------------------------- Date: 12-JUL-1988 14:30:09 GMT From: CNBS09@UK.AC.STRATHCLYDE.VAXA To: abbottp@UK.AC.ASTON To: abbotp @ aston From: cnbs09 @ strath.vaxa (Paul McAlinden 041-552 4400 x 2499 ) Subject: TEX and device drivers I should be grateful if you could provide me with copies of the latest versions of the above. Our configuration is a VAX/11 750 under VMS 4.4, with an HP laserjet + for output. We also have a Benson plotter - exact model no. unknown, do you know of any drivers for this. Thanking you in anticipation. +++Editor - The archive occupies 2 reels of 2400' tape at 6250bpi in backup format and the working vms copy of TeX at Aston is another reel. FTP is available for files from the archive on Spock user PUBLIC password PUBLIC. There is no interactive access at present. Mail access to texserver@aston.spock (or edinburgh) are other access routes. +++ --------------------------------- Date: 12-JUL-1988 14:54:22 GMT From: REEDA@UK.AC.BHAM.VAX1 To: ABBOTT@UK.AC.ASTON The following was produced as a responce to an FTP request for a file from the tex archive. Can you assist please? alan reed 021 414 4750 From: FTP_Manager 12-JUL-1988 14:50 To: REEDA Subj: File transfer File transfer 9635 completed Status %FTPB80-W-RNEGRCV, transfer rejected by remote server Information : VAX/VMS FTP (80) Version 4.2-3 Invalid value in command: invalid attribute - filename "[PUBLIC.TEXVMS.TEX.LN_FONTS]CMS6.896PK". File open failure reason - directory not found. VAX/VMS FTP (80) Version 4.3a. Remote sitename UK.AC.ASTON.SPOCK Local filename ACC$CCCIU:[REEDA.ASTON-PUBLIC.TEXVMS]CMS6.896PK; Remote filename [PUBLIC.TEXVMS.TEX.LN_FONTS]CMS6.896PK +++Editor - This is a common problem and is included to help others. If you ftp a copy of [public]000directory.list you will find the correct directory. the .896pk files can be found in [public.fonts.ln_fonts] +++ --------------------------------- Date: Tue, 12 Jul 88 14:58:30 BST From: Ian Moor To: tex-info@uk.ac.aston.mail Subject: Postscript symbiont for VMS Message-ID: <8807121458.aa17504@ivax.doc.ic.ac.uk> Before we converted to Unix, we used to run a symbiont on our Vax, I still have the C sources, the copyright notice indicates they may be distributed. The symbiont ran o.k. on VMS 4.5, I modified it to allow the laserwriter to be switched between RS232 and appletalk between jobs. Any output from the laserwriter (postscript errors) is mailed back to the user at the end of the job. I can mail the sources to individuals or the archive if you prefer. Ian Moor +++Editor - We would like a copy for use at Aston and I can include it in the archive for general use. Please send the files. +++ --------------------------------- Received: Date: Wed, 13 Jul 88 12:04:23 GMT From: Sebastian Rahtz To: ABBOTTP@uk.ac.aston Subject: Re: UKTeX V88 #20 I had a few queries about Morgan's web2c translator, and I append his replies. It looks as if it would be worth getting a new copy - could you ftp it, and let me know when you get it? +++Editor - Not before TeX88. Has anyone else ftp'ed this files already? If so please let me know +++ thanks sebastian -------------------------------------- You can anonymous ftp the latest version from host ics.uci.edu. Take "pub/web2c_*" (3 files) using binary mode, then uncompress and untar them. There's also a file web.tar.Z which contains the .web source files. I've just received mail that Knuth has made yet another change to TeX. It was already on version 2.93; this brings it up to 2.93a (for some reason he didn't want to change the minor version number, even though his change causes the output from the trip test to change slightly---I don't know why). I have yet to incorporate this latest change into my distribution. -------------------------------------------------------------------- Both of the problems you have encountered have been fixed in the current release (2.17). Here are the fixes for the specific problems you encountered: a) The TEXINPUTS environment variable is fossilized after an 'undump' (though it works normally with a 'virtex'); Move the location of the call to setpaths() in itex.c to BEFORE the line if ( readyalready == 314159L ) goto lab1 ; instead of after. Actually, this change should be made in the file ctex.ch; the above refers to the generated C code. b) The 'web2c' program fails with a 'Syntax error' when converting Metafont. This is a bug in web2c.yacc. Make the following change: ------------------------------Output from diff-------------------------------- 362c362,364 < {sym_table[last_type].var_not_needed = TRUE;} --- > { if (last_type != -1) > sym_table[last_type].var_not_needed = TRUE; > } 365c367,369 < {sym_table[last_type].var_not_needed = TRUE;} --- > { if (last_type != -1) > sym_table[last_type].var_not_needed = TRUE; > } ----------------------------------------------------------------------------- Feel free to send me any other questions or problems. I'm appending a list of all the changes which have been made since release 2.0. Tim Changes since release 2.0 --------------------------------- Release 2.1: 1. Added a line to cmf.ch to increase max_in_open from 6 to 15. 2. Touched tangle.c etc. so that make wouldn't thinnk it needed to be remade. 3. Fixed a bug in dvityext.c which declared cp as char ** instead of char *. 4. Changed the makefile and instructions slightly to make it clearer how to compile tangle without having a running tangle available. 5. Made the default for Metafont be no windowing support or libraries, for maximum portability. Release 2.2 [originally 2.2.1]: 1. Added a case to fixwrites for the character string "','", to fix a problem in converting BibTeX. [EKrell] 2. Fixed problem in fixwrites with \" embedded in strings. [EKrell] 3. Changed path initialization logic in TeX: if an environment variable is present, use it. Otherwise, if an envar was set when TeX was undumped, use that value. Otherwise, use the default value specified in "site.h". Metafont already worked that way, so it didn't change. [bug discovered by EKrell] 4. Added experimental program fmtdump. Release 2.3 [originally 2.2.2]: 1. Fixed a bug in web2c.yacc which could index an array using negative subscripts. [Leonard Zubkoff] 2. Fixed an unsigned arithmetic problem in tangle. Contemplated rewriting web2c to handle expressions correctly in all cases. 3. Added yet another patch to fixwrites (again, for bibtex, and discovered by EKrell). Again contemplated rewriting web2c. Maybe next month... Release 2.4 [originally 2.2.3]: 1. Fixed a bug in fixwrites which caused it to introduce newlines after '{' characters in quoted strings. Release 2.5: 1. Changed release numbering scheme :-) to major number + generation number (e.g., 2.5). 2. Added Pierre's SITEDIR changes to the make files. 3. Added beta release of BibTeX in C. Release 2.6: 1. Installed the fixed tangle.c in the distribution area (with the toint() fix). 2. Added Apollo to the list of successful ports in the README file. 3. Added to the main Makefile so that CFLAGS could be set there and inherited by the subordinate Makefiles. 4. Added patches to web2c.{lex,yacc} to support Pascal pointers [ekrell], in preparation for MFware programs. Release 2.7: 1. Added SITEDIR stuff to top-level Makefile. 2. Added fix for another casting bug which (mostly) affected only C compilers with 16-bit int's. Release 2.8: 1. calledit() was invoking the editor without closing the files first. On some operating systems, this could be problematic, so I added a loop to close all the inputfile[]'s first. Changes for both TeX and MF. 2. Tftopl needed to use the testeof() routine instead of simply feof to test for end of file. I changed the eof() macro in texware/web2c.h to use (and define) test_eof(). Since test_eof() calls feof() first, it should be backwards compatible. 3. Improved make file for ctex and cmf. It creates more .o files, but less recompiling and relinking is now done. 4. Replaced input_ln in Metafont with the hand-coded C version from TeX. This fixed a bug which could cause MF sometimes to require two newlines instead of one. Presumably, this also speeded MF up slightly. 5. Added veryclean entry to top-level Makefile. Release 2.9: 1. Fixed a major bug in my new makefile for initex/virtex. Release 2.10: 1. Fixed a minor bug in the makefile for inimf. Also, changed the CFLAGS and LDFLAGS to make a production version instead of a debugging version. 2. Changed TeX's buf_size to 2000 instead of 500, at the suggestion of a site where users have exceeded the 500 value (!). 3. Changed TeX to look for the LAST "." in a file name instead of the first for purposes of generating the \jobname (and thus, the dvi and log file names). For example, "my.file.tex" would produce "my.file.dvi" instead of "my.dvi". Release 2.11: 1. Add $(WLIBS) to virmf: line in MF makefile. Release 2.12: 1. Fixed LDFLAGS in bibtex/Makefile to be -s instead of -O. [Leonard Zubkoff] 2. You no longer have to uncomment-out the regfix line in the convert script for TeX. Instead, you set another flag (REGFIX) in the site.h file. [Leonard Zubkoff] 3. Fixed declaration of sprintf() for SysV in splitup.c; made splitup depend on ../site.h in the Makefile. [dean@cs.wisc.edu] 4. Fixed similar declaration problem in web2cdir/web2c.h. [dean@cs.wisc.edu] 5. Changed references to "cc" to "$(CC)" in all Makefiles. [dean@cs.wisc.edu] 6. Removed a unnecessary "#undef SYSV" in texware/tanext.c. [dean@cs.wisc.edu] 7. Changed all csh scripts to use sh instead. [dean@cs.wisc.edu] 8. Change remaining references in the makefiles to "rm" to $(RM), adding a definition where necessary. 9. Added dvitype to the defaults for the texware Makefile, since it's needed for the trip test. Release 2.13: 1. Removed declarations of strcpy(), etc., and replaced them with #include (string.h on BSD systems). 2. Changed around the way that SITEDIR works, adding -I$(SITEDIR) to the makefiles. All programs now simply include "site.h", and rely on the -I to determine which site.h is included. 3. Fixed a bug in bibtex.ch which caused problems when including some auxilliary files. [Tom Rokicki] Release 2.14: 1. Changed the SITEDIR handling for the MFlib/ subdirectory slightly, so that SITEDIR itself is used consistently in all the Makefiles. Now, the MFlib/Makefile refers to ../${SITEDIR}. 2. Fixed OPT (was OPTIM) make variable in cmf/MFlib/Makefile. [Leonard Zubkoff] Release 2.15: 1. Added some comments in README about clearterminal() and wakeupterminal(), for those porting TeX to other environments. 2. Fix the vs. mixup I introduced in 2.13. Release 2.16: 1. Change clearterminal() to use TIOCFLUSH ioctl for BSD environments in ctex/extra.c. [Joe Ramey] 2. Fixed a couple of declarations which were "char" but should have been "schar" in texware/weavext.c. [Dennis Page] 3. Removed spurrious #define BSD and #undef SYSV's from cmf/mfext.c. [Dennis Page] 4. Fixed some more declarations of sprintf() for SYSV compilation (declaration as int * instead of char * if SYSV is #define'd). I don't know if there are still problems on (some?) SysV.3 machines which declare sprintf() in . [Dennis Page] 5. Added a conditional for the execution of ranlib when building mflib.a: check to see if /bin/ranlib exists before trying to execute it. [Dennis Page] 6. Updated texdir/ctex.ch for tex.web version 2.93. Also, changed |make_name_string| so it didn't call |str_room|, per Knuth's directions. 7. Upgraded to mf.web version 1.5. 8. Retested the entire distribution in both AT&T and Berkeley environments. 9. Changed the type of zinputln() in cmf/mfext.c to be nothing (i.e., int), since it isn't typed in the other modules. [lint] 10. Changed the definition of bwritebyte in cmf/mf.h to cast the first argument of the putc() to (char). [lint] 11. Added a (void) cast to the one call to putc() in cmf/mfdisplay.c. [lint] 12. [not a fix] Two known minor bugs in bibtex remain, but it should be fixed soon. One bug is that when given the name of a non-existent file, bibtex may go into an infinite loop. The other only applies if the type schar is not char. Release 2.17: 1. It is now recommended for most 32-bit systems that a 32-bit version of TeX be built, and that preloaded versions of TeX not be made. See the README file for details. 2. To this end, ctex/extra.c has been modified so that virtex will attempt to load "&" automatically. So, for example, make a link from virtex to "latex", and from lplain.fmt to latex.fmt, and users can just run the command "latex" and virtex will automatically load latex.fmt. [Ken Yap] 3. Patch files to build a big (32-bit) TeX and also XeT-TeX are now included in the distribution. See the file README.PATCHES for details. [Pierre MacKay] --------------------------------- Date: Thu, 14 Jul 88 12:43:33 BST From: David Barfoot To: AbbottP@UK.AC.ASTON Subject: TeX2.9(2) Message-ID: <9ED03FEEF5F59A70@UK.AC.CAM.PHX> I've hunted around SPOCK for v2.9 TEX.WEB, but all I can find is v2.0. Where can I find v2.9 and the changes for v2.92 (as advertized in UKTeX) ? Thanks, dB +++Editor - Directory [PUBLIC.TEX] TEX2V9_2V92.UPDATE;1 TEXMFWEBPLAIN.UPDATE;2 Total of 2 files. +++ --------------------------------- Received: from logitek by kestrel.Ukc.AC.UK with UUCP id aa12528; 14 Jul 88 17:57 BST Received: by logitek.co.uk (smail 2.5 [x25]) id AA13172; 14 Jul 88 13:01:10 EDT (Thu) Received: by inmos.uucp (3.2/SMI-3.2) id AA14800; Thu, 14 Jul 88 17:34:36 GMT Date: Thu, 14 Jul 88 17:34:36 GMT From: David Shepherd Message-Id: <8807141734.AA14800@inmos.uucp> To: info-tex@uk.ac.aston.mail Subject: ctex on VAX I've just got ctex to run on one of our VAXes here at INMOS and found that it doesn't run at the advertised twice as fast as pascal speed! I expect that this is being caused by IO problems as VAXes have the capability of doing IO very slowly if you're not careful. Has anyone got ctex to interface with the IO requirements of VMS properly ? In particular, has anyone rewritten fmt.c as loading the state is *very* slow. I would like to avoid repeating work that is available somewhere else as I'm not really meant to spend my time hacking around with TeX sources here (so a hack TeX occasionally, I can handle it, I could give it up ...... if I wanted to). David Shepherd inmos ltd. PS ctex also compiles for the transputer and, once you've got the information, across PC bus into a B004, it runs pretty fast. --------------------------------- Date: 15-JUL-1988 11:19:07 GMT From: JM@UK.AC.RL.STAR To: abbottp@UK.AC.ASTON I would first like to say that I do enjoy TeXmag and have often found it useful. At the moment I am interested in generating an index with LaTeX and as you may know LaTeX dosen't do all the work for you - the LaTeX book advises you that a supplementary program is necessary. I thought the directory [PUBLIC.TEXVMS.TEX.IDX_GLO] advertised in your MAIL message distributed on May 20th looked a promising source for such a program. I tried to copy the files to my own directory using TRANSFER/CODE=FAST. When prompted for a username and password, I tried First NETUSR in response to both and then PUBLIC, but in each case these were rejected. Can you tell me how I can copy this directory? thank you, Jo Murray +++Editor - I suspect you were trying to ftp the files from aston (where the digest comes from). The address of the archive is aston.spock The username is PUBLIC and password PUBLIC +++ --------------------------------- Date: 15-JUL-1988 12:40:53 From: PHILIP@UK.AC.BBK.CU To: ABBOTTP@UK.AC.ASTON.MAIL Subject: Aston TeX database I would like to use the TeX database at Aston. Could you let me have the names of any 'introductory' files in [PUBLIC] that explain how your system works. I am used to the KERMIT and MICROS services at Lancaster and I assume your system is similar. Thanks, Philip Riebold Central Computing Services Birkbeck College Malet Street London WC1E 7HX PHILIP@BBK.CU +++Editor - We do not have kermit access to the archive only ftp and mail. Details of access are contained in the file listed below which you should `ftp' from aston.spock +++ --------------------------------- !! !! Files of interest [public]000aston.readme !! [public]000directory.list !! [public]000directory_dates.list !! [public]000directory.size !! !! Editor - I have a tape labelled TeX 2.9 LaTeX 2.09 Metafont 1.3 !! Unix 4.2/3BSD VAX SUN 2/3 Pyramid Seqeunt SYS V: 3B2 Tar 1600 bpi blocked !! 20 1 file dated 26 may 1988 (from washington.edu). !! !! I have the facilty to copy this tape for anyone who sends the following !! !! 1 2400 tape with return labels AND RETURN postage. !! !! Send to !! !! P Abbott !! Computing Service !! Aston University !! Aston Triangle !! Birmingham B4 7ET !! !! !! Replies/submissions to info-tex@uk.ac.aston please !! distribution changes to info-tex-request@uk.ac.aston please !! !! end of issue