TeXhax Digest Friday, November 6, 1987 Volume 87 : Issue 90 [SCORE.STANFORD.EDU]TEXHAX90.87 Editor: Malcolm Brown Today's Topics: immoderate notes: printing problems with TeXtures 1.0 Re: Verbs vs Symbols for defining VAX/VMS TeX LaTeX Pascal Environment Verbs vs. symbols -- a solution TeXhax Digest V87 #86 RE: FTPing Unix-TeX Bibliographies texhax87.87 Is this bug? LaTeX Notes (Re: TeXhax Digest V87 #88) Texas Instruments Omnilaser 2115 Desperately seeking C-TeX... PostScript for Imagen printers Re: Some TeX recommendations ---------------------------------------------------------------------- Date: the present From: Malcolm Subject: immoderate notes: printing problems with TeXtures 1.0 %%% Has anyone else run into a problem printing TeXtures 1.0 with %%% the new LaserWriter drivers (version 5.0)? Whenever I attempt %%% it, the PostScript interpreter aborts the job, complaining %%% about the undefined command "TeXdict". Anyone else encountered %%% this, and, if so, found a work around? ------------------------------ Date: Fri, 30 Oct 87 10:02 N From: Lars-Gunnar Olsson Subject: Re: Verbs vs Symbols for defining VAX/VMS TeX As I have not seen it mentioned in this debate, I would like to point out the option to create your own command table. This is done by the command $ set command/table=sys$library:dcltables/output=mytable - [path]texsys[.cld] This will put all the verbs in dcltables.exe, together with those defined in texsys.cld, into mytable.exe. The verbs in mytable.exe can then be made available through the command $ set command/table=mytable The point is that this command actually takes much shorter time to execute than $ set command [path]texsys[.cld] as, in my case, texsys.cld defines 10-15 new verbs. If the .cld- file defines less than 3-5 verbs, then this might not save you any execution time. Note also that the mytable.exe-file will allocate diskspace of at least 250 block. There is a caveat with using the verbs defined by $ set command. Namely, if you want to use the commands in a subprocess, then mytable.exe must be installed. Lars-Gunnar Olsson ccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccccc Lars-Gunnar Olsson, Swedish University of Agricultural Sciences Computer Section EMC, Box 7072, 750 07 UPPSALA, SWEDEN Tel +46 18 172407 usenet: enea!eva.sunet.se!larsg bitnet: larsg%eva@max.uu.se CSnet: chalmers!enea!eva.sunet.se!larsg@relay.cs.net PSImail(X.25): psi%240200100206::eva::larsg ------------------------------ Date: Fri, 30 Oct 87 08:45:48 CST From: Douglas Jones Subject: LaTeX Pascal Environment I have not been happy with the TROFF preprocessor GRIND, nor with the TeX preprocessor patterned after it, Tgrind. Neither fits in very well with the LaTeX philosophy that text to be formatted differently should be put in an environment indicating the formatting constraints. In response to this problem, I have hacked together a preprocessor and an environment so that I can write code such as: \begin{pascal} procedure brick( biffy: real ); begin write( biffy ); end; \end{pascal} And embed this in my LaTeX documents. The preprocessor (I called LaGrind) marks the lexical class of each item in the Pascal text, while various style parameters tell the Pascal environment what fonts to use for the result. Style parameters include what fonts to use for each of the classes of lexical objects (reserved words, identifiers, comments, quoted strings), and what to do about indenting and blank lines in the text). LaGrind is currently implemented as an sed script under UNIX, so it is not particularly maintainable. I will gladly send it, a user's manual, and the environment definition to anyone interested. I don't claim that my implementation does it right, but rather, that it does almost the right things. I would enjoy seeing support for similar environments named after other languages, particularly modula, c, and Ada. Douglas Jonesf ^ whered' that f come from jones@cs.uiowa.edu (on CSNET, that is) ------------------------------ Date: Fri, 30 Oct 87 22:33 EDT From: "Eliot Moss, GRC A351B, x5-4206 30-Oct-1987 2227" Subject: Verbs vs. symbols -- a solution I, too, was dismayed at how long it can take to run CDU to install all the things I might use. So, I defined things so that CDU is called upon first use of a command, to install it. As an example, consider some arbitrary program called FOO. In my LOGIN.COM I do: $ foo :== @mycoms:foo.com and set up FOO.COM something like this: $ foo :== $ set command foo $ foo 'p1' 'p2' ... 'p8' Get the idea? This is readily generalized to use a common utility; I call it USE: $ foo :== @mycoms:use.com foo foo with USE.COM being like this: $ 'p1' :== $ set command 'p2' $ 'p1' 'p3' 'p4' ... 'p8' Of course, you're a little limited in terms of the number of arguments, but I have not really had any problems with that. Eliot Moss Ass't Prof. Univ. of Mass. ------------------------------ Date: Fri, 30 Oct 87 23:44:09 PST From: mackay@june.cs.washington.edu (Pierre MacKay) Subject: TeXhax Digest V87 #86 RE: FTPing Unix-TeX > It would help a great deal if the Unix-TeX site coordinator posted the > system for anonymous FTP over the Arpanet. I hope this won't sound too querulous, but I have to point out that the organization of the Unix-TeX distribution is still a one-person show, carried on in the spare time left over from membership in three academic departments. My primary concern *has* to be the organization of the tape, and that is trickier than it may seem. An innocent change in the TeX portion of a web file may affect hidden parts of a number of change files (we now do VAX, SUN, Pyramid, some Sequent, and 3b2) and tracking down all posible alterations takes a great deal of time. I do, as a matter of fact have two diffs for the upgrades 2.1 to 2.3 and 2.3 to 2.5, which I shall add to the end of this, but the marshalling and organization of diffs for every significant change just in the past 3 months would take more time than I can easily imagine sparing. The second problem is that disk availability here has been on a rollor-coaster track since the loss of WARD. I don't know from one forthnight to the next how much space there is going to be on ~ftp/pub and that makes it very difficult to arrange for the appropriate selection of material to go onto the ~ftp/pub directory. The main archive for the distribution is for the present on a machine that has no network connections whatsoever, (not even a serial port!), and anything that gets into a network machine has to be transferred by walk-net. I have, to my shame, sent out several buggy Makefiles in the past, because there was not enough spare disk to check compilation. At least by being on a non-networked system I can try things out before committing them to the distribution, but it all takes time. I hope eventually to get back to the moderately friendly lot of ftp-able files that I used to put on WARD, but WARD is gone, and there is no really equivalent resource right now. Even when there is, I will still have to give priority in what time is available, to the organization of the tape. I will never be able to promise that ftp files will be up to date. There are simply not enough hours in the day. If all people want is the latest WEBS etc., those are kept in the archives on SCORE, in the directories. Getting them from SCORE makes more sense than getting them at second hand from here. As I know to my distress, ftp does not always deliver perfect files, and every additional copying effort adds to the likelihood of significant error. I pick them up when I can, and compile them on as many distinct architectures as I can gain access to before I let them get into the UnixTeX distribution. It seems worth the effort (just discovered that although you can use -O to compile TeX on ULTRIX, you can't use it to compile METAFONT---hopelessly buggy assembler code results). extracting ftp-able files will always be secondary to that effort. But here are some diffs that should work with patch. In the 2.1_to_2.3 patch, watchout for the formatting change that may affect your particular change file. %% Pierre MacKay's submission is too long for digest distribution. %% It is available for FTPing under the name %% [SCORE.STANFORD.EDU]MACKAY.TXH %% A copy has been forward to the TEX-L list server for BITNET %% access. Malcolm Pierre A. MacKay TUG Site Coordinator for Unix-flavored TeX ------------------------------ Subject: Bibliographies Date: Fri, 30 Oct 87 17:59:49 PST From: Phil Windley Is there a program that uses the standard UNIX bibliography database (the ones produced by addbib and used by refer) to produce TeXable source? I know about BibTeX, but that seems to use a different format than the UNIX database. (I think it is more like Scribe's bibliographic db). Phil Windley Robotics Research Lab University of California, Davis ------------------------------ Date: Fri, 30 Oct 87 21:50:06 EST From: "David F. Rogers" Subject: texhax87.87 Is this bug? Re: From: CHAA006%VAXB.RHBNC.AC.UK@forsythe.stanford.edu Subject: TeX bug ? Undesireable fontname interaction with \bigl TeX gurus: is this a bug ? ** Phil. \font\big = cmr10 scaled \magstep5 $\bigl(x-s(x)\bigr)\bigl(y-s(y)\bigr)$ % example from p.146, ``The \TeXbook'' \end yields (Kellerman & Smith implementation) :- This is TeX, Version 2.0--0 on VAX/VMS (preloaded format=plain 87.5.8) 23 OCT 1 **test (SYS$USERDISK1:[CHAA006.TEX.BAE]TEST.TEX;24 ! Missing , inserted. \big l.2 $\bigl (x-s(x)\bigr)\bigl(y-s(y)\bigr)$ ? x No pages of output. I also received an error when running MicroTeX Version 1.5A1 on a real IBM AT. What has happened here is that you are redefining one of Plain TeX's macros which is subsequently used in \bigl and \bigr etc. I modified the test case as follows and it works. \font\biggg = cmr10 scaled \magstep5 \biggg $\bigl(x-s(x)\bigr)\bigl(y-s(y)\bigr)$ \end If you look in the file plain.tex usually found in the \tex\inputs directory, you will find the following macro definitions. \def\bigl{\mathopen\big} \def\bigm{\mathrel\big} \def\bigr{\mathclose\big} \def\big#1{{\hbox{$\left#1\vbox to8.5\p@{}\right.\n@space$}}} Dave Rogers dfr@usna.arpa dfr@usna.mil -- domain addressing ------------------------------ Date: 31 OCT 87 10:31-N From: PHTGERMO%CNEDCU51.BITNET@wiscvm.wisc.edu I am searching for a translator able to convert a VAX/VMS WPS+ file to a TeX file. Any information about the existence or development of such a program will be appreciated. Thank you in advance. Jean Beiner Ecole d'Ingenieurs du Canton de Neuchatel Av. de l'Hotel de Ville 7 2400 LE LOCLE SWITZERLAND BITNET: PHTGERMO@CNEDCU51.BITNET ------------------------------ From: lamport@src.dec.com (Leslie Lamport) Date: 31 Oct 1987 0947-PST (Saturday) Subject: LaTeX Notes (Re: TeXhax Digest V87 #88) Siddhartha.Chatterjee@VLSI.CS.CMU.EDU writes: My officemate has a lot of program fragments in his thesis (which he is typesetting in LaTeX) which he wants typeset in tt font, with tabs, underscores, braces and other normal programming language symbols being obeyed. The catch is that he does not want to have to put formatting commands in the code itself (so a solution like the example in p.234 of the TeXbook is not acceptable). Does anyone have a set of macros which could be used to do this? Am I missing something? Except for tabs, it seems like the verbatim environment does exactly what he wants. Tabs are the bane of text processing; the best thing to do with them is to remove them from all text files and replace them by the appropriate number of spaces. It is trivial to write a program to do this is any number of programming languages. Nicholas Beser writes: I am using LaTeX to write several chapters in an computer engineering textbook. I have been breaking the chapters up using the \include file option since the chapters have many sections in them. The trouble with this method is that include files are forced to begin a new page. (Page 76 of the LaTeX book). Is there some way to disable this? The only option that I can think of is to write a preprocessor to expand all of the includes prior to runing LaTeX. The purpose of an \include is so that the pages produced by running any subset of the \include'd files are exactly the same as the corresponding pages produced by running the entire file. I defy anyone to come up with a way of achieving that in TeX or any other formatting language without requiring each \include'd portion to start on a new page. I don't know what Mr. Beser's exact constraints are, but I suspect that the real source of his difficulties is that he's worrying about page layout while writing, which is always a mistake. If he were content to see the exact form of every page only after completing a chapter, then he would probably be able to find a simple solution to his problem. Eric Hildum writes: I am trying to format a set of equations such that the terms of the equation line up into columns. Unfortunately, the terms contain subscripts of varying lengths, and I need to indicate missing terms (with centered dots). Currently, I am using the eqnarray environment, which provides me with the equation numbering I need, but does not give the alignment of the terms on the left hand side of the equation that I need. What Mr. Hildum needs is an eqnarray environment that allows the user to specify the number of columns and their alignment. I've had several requests for this; perhaps if I were doing it over again I'd provide one. Creating this kind of environment is beyond the abilities of mere mortals. (I was only able to create the array and tabular environments because Don Knuth provided some magic macros.) However, it should not be too hard for an experienced TeX hacker to modify the current eqnarray environment to create a new environment that produces any particular number of columns. Nico Poppelier writes: In several TeXhax Digests people have asked questions about the table and footnote mechanisms of LaTeX. Here's another one, concerning generating footnotes in a minipage environment. I encountered the problem when I was writing a paper with lots of tables. In one of the tables I wanted to have footnotes just below the table. So I used the following nest of environments ... The footnotes are numbered a, b, c, ... But then I wanted to refer to the same footnote more than once. The LaTeX book suggests using \footnotemark and \addtocounter{footnote}{..}, but this does not produce the desired result inside a minipage. In the LaTeX manual, the subentry "footnote counter for" of the index entry "minipage enviroment" refers to page 91, which contains the following: The mpfootnote counter numbers footnotes inside a minipage environment. Leslie Lamport ------------------------------ Date: Fri, 30 Oct 87 13:35:36 -0100 From: mcvax!ruuinfvax!piet@uunet.UU.NET (Piet van Oostrum) Subject: Texas Instruments Omnilaser 2115 Does anyone have experience with TI's Omnilaser 2115. This is a Postscript Machine, supposedly white on black. Specific questions: 1. Can I use the 'normal' metafont fonts or must they be rerun with other mode_defs and so on. 2. If the fonts must be regenerated, can I still use these with the Apple LaserWriter (we will also keep our LW, and I don't want to have two sets of fonts)? Alternatively, is there e set of parameters that will work for both? 3. Any other known problems (or benefits)? Please reply to me directly. If necessary, I will give a resume. Piet van Oostrum Telephone: +31-30-531806 Dept of Computer Science UUCP: ...!mcvax!ruuinfvax!piet University of Utrecht Budapestlaan 6, P.O. Box 80.012 3508 TA Utrecht, The Netherlands ------------------------------ Date: 10/31/87 2316 From: Subject: Desperately seeking C-TeX... In a recent TUGboat issue, some advertising was done about a LaTeX style "repository" at cs.rochester.edu. The directory listing, among many other interesting things, mentioned the source files for Pat Monardo's Common TeX in C. I was very disappointed to find that these files had disappeared when I tried to transfer them. I sent a message to Ken Yap, in charge of the latex style collection. He answered me that these files were too big to keep on line, that they were not precisely part of a LaTex style collection, and that many TeX users had them now, so that somebody should be able to send them to me. So : is anybody willing to send me these files? Or does anybody know a place where they are stored and from where I could get them? If nothing is possible through the network, I would be glad to send a tape to anybody willing to fill it... ------------------------------ Date: Sat 31 Oct 87 20:32:07-PST From: Eric M. Berg Subject: PostScript for Imagen printers I've heard the rumor that a postScript upgrade (they probably don't call it that) for the imagen will be available around the first of the year. It is a fact, not a rumor, that Imagen will soon be supporting PostScript for their printers. For existing printers, this will require both a hardware upgrade and new software. Contact your Imagen sales rep. (or Imagen directly) for details of price & availability. Eric M. Berg Computer Facility Graduate School of Business Stanford University ------------------------------ Date: Sat 31 Oct 87 21:09:20-PST From: Eric M. Berg Subject: Re: Some TeX recommendations Date: Tue, 27 Oct 87 08:57 EST From: Jeffrey Mark Siskind Subject: Some TeX recommendations [....] 1) The LaTeX \today command is usually the wrong thing to use. What is really needed is a \lastFileWriteDate command which would be more appropriate but is not provided. [....] 2) [....] The version of LaTeX I am using on a Symbolics lisp machine allow me to \include{figures>figure1}. The problem is that if I transfer my files to another system, say Unix or TOPS-20, I must change my source file to process the document. [....] What is needed is a portable syntax for specifying pathanmes. Both of these suggestions really pertain to TeX itself, rather than LaTeX, as the original message recognized they might. Regarding suggestion 1), "TeX the program" includes a variable which contains the current date; each implementation of TeX has to include some (system-specific) code for getting the date from the operating system. There's probably no reason why TeX couldn't also have included variables to keep track of the last-write-date for various files; this would have been just another system-dependent piece of code that needed to be written for each new system TeX was ported to. However, TeX didn't, and that's not going to change. One possible solution would be to create a new LaTeX style which would print "\lastFileWriteDate" whereever LaTeX currently prints "\today", and then define "\lastFileWriteDate" in your document and update it each time the file was edited. (A sufficiently clever editor such as EMACS could probably be set up to do this automatically.) Regarding suggestion 2), one way to deal with this is to give just the actual file name as the argument to the "\input" command, and then change the definition of the TEXINPUTS: logical name (TOPS-20 or VMS) or environment variable (Unix) so that it points to the directories where your various source files live. This has the desired results of making the source files themselves system-independent, and lets the user tell the program where to find them at run-time. Eric M. Berg Computer Facility Graduate School of Business Stanford University ------------------------------ %%% %%% subscriptions, address changes to: texhax-request@score.stanford.edu %%% please send a valid arpanet address!! %%% %%% submissions to: texhax@score.stanford.edu %%% %%% BITNET redistribution: TEX-L@TAMVM1.BITNET (list server) %%% %%%\bye %%% ------------------------------ End of TeXhax Digest ************************** -------