TeXhax Digest Tuesday, August 23, 1988 Volume 88 : Issue 75 Moderator: Malcolm Brown Today's Topics: Registering MF names Registering MF names PCTeX on Novell Re: Boxed Tables index to issues 1-25 of texhax for '88 LaTeX question LaTeX question Finding the height and depth of maths stuff PostScript emulators for common printers PC-WRITeX v2.2 Indented first paragraphs for sections (re: v.88, n.72) Re: The best way to do pictures in TeX DVICAN for MSDOS? Tex novice needs installation advice LaTeX meets a different book designer ---------------------------------------------------------------------- Date: Sat, 13 Aug 88 10:34:42 CDT From: Don Hosek Subject: Registering MF names Hey, there's an idea, why don't we do like ATypI and register names for MF generated fonts (as well as others). I think a few things worth including are demanding that names are unique to 8 characters (no matter WHAT the Unix people might think, IBM is NOT going to go away, so we may as well face up to the fact that we're stuck with 8 character file names for perhaps the majority of TeX users--a recent survey of TeXMaG readers showed that more than half used TeX on the PC, and this is ignoring that segment of the TeX User base that has no network access). Also, let's see names of sub-files that don't conflict--as an example, in Sylvio Levy's Greek font, he names one of his files accent.mf which is OK, as long as you aren't ever going to try running MF on CM which also has a file accent.mf (this is an issue on systems without directory trees, or those that require that MF files be in some MFINPUTS directory or directories--or even if you just want to be able to have MFINPUTS point to each directory containing any useful MF sources--not an unreasonable request I'd say.). Any reaction on the registry idea? -Don Hosek P.S. I'm returning to California soon, so please call me DHOSEK@YMIR.BITNET ------------------------------ Date: Sun, 14 Aug 88 23:04:51 PDT From: mackay@june.cs.washington.edu (Pierre MacKay) Subject: Registering MF names I roughed out an idea for that a while ago. Some further criteria. Classes of fonts should begin with the same letter. (that is why I go on changing LaTeX circle* to lcircle* --- it makes all LaTeX fonts identifiable as beginning with l.) All washington fonts go out with a w prefix. All ams fonts start with a, except for euler, which start with e. Don Knuth's principle is that font names should attempt to be unique as regards the outer 6 characters of the name, i. e. that sssnnnnnnsss should be reducible to ssssss without conflict with any other similarly derived 6 character name. This is more broadly successful than something like the first eight character unique system that many programming languages use. I have tried to keep to it. It means that naming conventions have to be treated with some care. You want the i of "italic" where it will be picked up, in that convention. It may be that 6 characters will at last turn out to be too restrictive, but it is worth seeing how long we can go on with it. (It reflects a filename restriction on the SAIL machine at Stanford.) I agree that component files should not be ambiguously named. The establishment of filename conventions is only one side of the question, however. The registry, which might become a part of Tugboat will also have to address the question of fully spelled out naming conventions. I would have no hesitation in calling a font I was reasonably pleased with "Computer Bembo," or "Washington Bembo" but I would not challenge Monotype's right to "Bembo" plain and simple. It is both a legal and a moral question. Email: mackay@june.cs.washington.edu Pierre A. MacKay Smail: Northwest Computing Support Group TUG Site Coordinator for Lewis Hall, Mail Stop DW10 Unix-flavored TeX University of Washington Seattle, WA 98195 (206) 543-6259 ------------------------------ From: "Mary E. McClure" Subject: PCTeX on Novell Can anyone offer advice or share their experiences on setting up PCTeX (V2.1) on a Novell server? ------------------------------ Date: Fri, 12 Aug 88 13:15 EDT From: Subject: Re: Boxed Tables Although I am new to TeX, I believe I can solve the problem with "nubs"{on the sides of boxed tables. I don't believe it's a math problem with either TeX or your driver, but rather with your output device. I was a typesetter for a couple of years and we experienced this problem on $100,000 Compugraphic equipment.... You see, when lines are thicker than 1/8 point (or whatever the device's minumum line width is), somehow it compensates by using a couple of lines together. The "nubs" appear if the device "adds" these extra lines to the right or left (or up or down, as the case may be). Therefore, if you want a box 10 pica x 12 pica with a 2p rule, and the output device draws rules left to right, you must define a box (10pica-2p) x 12 pica. This is the only way to prevent this that I can think of. In my mind's eye, I think the problem Darrell Long encountered can be corrected by adding the width of the line to his original dimensions. I have no way of knowing what code changes this would require, as I am aTeX novice. Perhaps someone else can forward theses changes . . . Virtually, Avril Katherine Belisle Phinney a.k.a. Kate :-) BITNET: PHINNEY@CTSTATEU CompuServe: 76420,152 "Learning is finding out what you already know. Doing is demonstrating that you know it." -- Messiah's Handbook * Reminders for the Advanced Soul ------------------------------ Date: Mon 15 Aug 88 14:53:50-PDT From: Barbara Beeton Subject: index to issues 1-25 of texhax for '88 the file texhax88.in1 contains an index to the first 25 issues of 1988. it is available for ftp, and i will be in touch with the principal texhax redistributors to make sure that they get copies. this index was prepared by christina thiele and me. we would appreciate any comments on the format, arrangement, or anything else that can help us make the next installment better. (we {\sl are} working on the next installment, albeit slowly.) -- barbara beeton ------------------------------ From: INHB%MCGILLC.BITNET@Forsythe.Stanford.EDU Date: MON AUG 15, 1988 18.06.55 Subject: LaTeX question I have a question for all those TeXperts out there. I am in the final stages of preparing a book using LaTeX. It has something in excess of 500 cross references and I have tested this and discovered that I get the message: LaTeX Warning: Reference `prodset' on page 2 undefined. [2] (label.aux ! TeX capacity exceeded, sorry [hash size=2500]. \@testdef ...\ifx \csname #1@#2\endcsname \@tempa \else \@tempswatrue... l.134 \newlabel{constantiso}{{}{1}} If you really absolutely need more capacity, you can ask a wizard to enlarge me. Here is how much of TeX's memory you used: 583 strings out of 1483 4463 string characters out of 13938 56049 words of memory out of 65535 2500 multiletter control sequences out of 2500 19568 words of font info for 76 fonts, out of 22973 for 100 31 hyphenation exceptions out of 307 I really absolutely need more capacity, so I am asking a wizard to tell me how to enlarge him. Date: Wed 20 Jul 88 21:37:42-PDT From: Eric M. Berg Subject: Instructions for FTPing TeX from Score I've created a file -REAME-.GETTING-TEX in the directory on Score. It contains an explanation of the lay-out of the directory tree on Score, and some brief instructions on using anonymous FTP, both originally written by Barbara Beeton and updated by me. Anyone interested in FTPing TeX files from Score should be encouraged to retrieve this file first, if they have any questions about what files are where. Since this file is in , anyone who logs in via "anonymous FTP" can find it without having to change directories. We seem to get a couple of inquiries a month about how to do this, and hopefully this file will make responding to them individually a thing of the past. Please let me know if you have any suggestions or comments. /Eric FINDING TeX FILES ON SCORE.STANFORD.EDU USING ANONYMOUS FTP Most systems on the Internet (this includes the Arpanet, NSFnet and some others, but not Bitnet or USEnet) support a File Transfer Protocol (FTP) program that enables a user to make a connection to another node on the network and copy or send files between the two connected nodes. If the remote node supports anonymous FTP, it is not necessary for the user to have an account at that node. An FTP session (run on a TOPS-20 system) looks something like this. @ftp score.stanford.edu ... FTP> cd < Please log in first, with USER, PASS and ACCT. LOGIN (user) anonymous Password: anonymous ... FTP> get -chrono-.dir web-chrono.dir [source file] [target file] PS:-CHRONO-.DIR.14 => WEB-CHRONO.DIR.1;P777700 !! [OK] FTP> exit ORGANIZATION OF THE TeX DIRECTORIES ON SCORE In the above example, we are retrieving the -CHRONO-.DIR file from the directory, the directory which holds TeX.WEB and its change files, TANGLE and WEAVE, and so forth. Most of the subdirectories contain chrono files, which list the files in that directory chronologically by date of last modification (most recently modified files first). These chrono files are regenerated automatically every weekend. The subdirectories listed below are the ones that become part of the standard TeX distribution tapes: .TFM and raster files of 300 dpi AMS fonts AMSTeX.TeX and related macros and documentation the BiBTeX system, for use with LaTeX .MF sources of the CM* fonts manuals, bug and errata lists and other documentation .GF font files various macro files and similar the LaTeX macros and style files a VMS-based device driver for the DEC LN03 printer MF.WEB and change files, ``fontware'' TRIP, TRAP and other test-related files the TeXhax archives items that have appeared in TUGboat TANGLE and WEB; TeX.WEB and change files Other directories of interest: what the name implies sources for the DVIIMP program (Imagen device driver) local TeX input files SOME ADDITIONAL COMMENTS Please note that the WEB Change files in do not necessarily correspond to the latest versions of the .WEB files. (That is, there is frequently a delay between when a .WEB file is updated and when the corresponding Change files are updated.) This is particularly true for Change files for operating systems other than TOPS-20 (e.g. VMS, CMS, etc.). Also note that there are no Un*x change files included on Score. Finally, for other TOPS-20 sites, the .EXE files for the TeX software are stored in the directory. Original version: Barbara Beeton (BEETON@Score.Stanford.EDU) Last updated: Eric Berg (EMB@Score.Stanford.EDU), 7/20/88 ------------------------------ Date: Mon, 15 Aug 88 16:51:56 CDT From: "Patrick M. Landry" Subject: LaTeX question Please send responses to bgg@usl.usl.edu or bgg%usl@relay.cs.net In the LaTeX \newtheorem command, is there any way to reference the caption? That is, if \newtheorem{env_name}{caption} is used, is there any way to say something like \env_namecaption to have LaTeX produce the caption? This is useful in if you have many theorem-like items in a paper, and you decide to change the name of the caption that is produced. I realize that that I could define my own command that defines a string, and then calls \newtheorem, but is there something already built in to LaTeX? ------------------------------ Date: 16-AUG-1988 12:28:27 GMT From: CHAA006%vaxb.rhbnc.ac.uk@NSS.Cs.Ucl.AC.UK Subject: Finding the height and depth of maths stuff David Strip of Sandia says >>> I am trying to build a strut to force TeX to build a large delimiter >>> to match the height of the corresponding delimiter that closes an >>> equation split over two lines. I build a vbox with the what seems to be >>> stuff of the appropriate height, then try to use its dimensions. These >>> dimensions don't seem to be right. Here's a code fragment (plain TeX) >>> \newbox\eqnbox >>> \setbox\eqnbox=\hbox{\hsize=100pt \vbox{ {$${ M\over F} $$}}} >>> \showbox\eqnbox >>> \newdimen\eqnht \newdimen\eqndp >>> \dimen\eqnht=\ht\eqnbox \dimen\eqndp=\dp\eqndp >>> \showthe\dimen\eqnht \showthe\dimen\eqndp >>>If you execute this, you get height and depth of 6.0006 pts. >>>What I am I doing wrong? How do I get the proper height and depth. I think the main cause of the problem is a(n understandable) misundertanding about the nature of \newdimen and \newbox; the first generates an actual \dimen variable (and can therefore be used directly) whereas the latter only generates a reference to a \box (and therefore must be used in conjunction with \box or one of its close relatives). Put simply, having said \newdimen \mydimen and \newbox \mybox, one subsequently uses \mydimen by itself, but uses (e.g.) \box \mybox. So, if we take David's example, \catcode `\> = \active \def >>>{\ignorespaces} >>> \newbox\eqnbox % OK >>>%\setbox\eqnbox=\hbox{\hsize=100pt \vbox{ {$${ M\over F} $$}}} % or perhaps \setbox\eqnbox=\hbox{$ \displaystyle {M \over F} $} % several changes >>> \showbox\eqnbox % OK >>> \newdimen\eqnht \newdimen\eqndp % OK >>>%\dimen\eqnht=\ht\eqnbox \dimen\eqndp=\dp\eqndp % No - use instead \eqnht=\ht\eqnbox \eqndp=\dp\eqnbox % three changes here >>>%\showthe\dimen\eqnht \showthe\dimen\eqndp % No - use instead \showthe \eqnht \showthe \eqndp % two changes here \end we might get closer to what he wants. (Note that \dp\eqnbox replaces \dp\eqndp --- I assume that was just a typo). ** Phil. ------------------------------ Date: 16-AUG-1988 12:30:18 GMT From: CHAA006%vaxb.rhbnc.ac.uk@NSS.Cs.Ucl.AC.UK Subject: PostScript emulators for common printers This is not a TeX question, but I wonder if anyone in the TeX community might know the answer: does there exist a PostScript emulator (or a PostScript converter) for Epson FX-80 or Diablo-630 ? ** Phil. ------------------------------ Date: Mon, 15 Aug 88 00:52 GMT From: Peter Flynn UCC Subject: PC-WRITeX v2.2 Version 2.2 of the PCWRITeX converter driver is now available. The major change is that it will now do *TABLES* !!! Use the IBM PC box-drawing character set and you get a $$\vbox{\halign...}$$. The double-line character set produces a ruled table: the single-line set produces unruled tables. Only simple tables so far. I call it T2K4 --- ``Tables To Kill For" --- eat your heart out, Lance! As I shipped v2.1 only recently, I am supplying this version on request only, as nothing else major has changed. I have sent copies to Ken Yap at Rochester and Peter Abbott at Aston. I have tried, unsuccessfully, to contact the list owners of ListServ at Heidelberg and Texas A&M: if anyone knows how to wake these guys up, tell me! Please reduce network traffic, and retrieve your copy from your local suppository rather than asking me for it. Peter Flynn ------------------------------ Date: Tue, 16 Aug 1988 16:56:41.61 CDT From: (George D. Greenwade) Subject: Indented first paragraphs for sections (re: v.88, n.72) With respect to Walter Olthoff's question regarding paragraph indentation after the section marker, we use a brief documentstyle option for this purpose. It is called INDSEC.STY (Indented Sections) and follows below: %%%%%%%%%%% start of documentstyle option INDSEC.STY %%%%%%%%%%% % Document style option "indsec" % -- usage: \documentstyle[...,indsec,...]{anystyle} % -- last update: 6 April 1988 - GDG % This style option produces an indented paragraph following section heads. %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% %---Redefine @startsection to always get indentation after section headings. %---NOTE: Use \noindent to over-ride automatic indentation, if desired. \def\@startsection#1#2#3#4#5#6{\if@noskipsec \leavevmode \fi \par \@tempskipa #4\relax \@afterindenttrue \ifdim \@tempskipa <\z@ \@tempskipa -\@tempskipa \relax\fi \if@nobreak \everypar{}\else \addpenalty{\@secpenalty}\addvspace{\@tempskipa}\fi \@ifstar {\@ssect{#3}{#4}{#5}{#6}}{\@dblarg{\@sect{#1}{#2}{#3}{#4}{#5}{#6}}}} %%%%%%%%%%% end of documentstyle option INDSEC.STY %%%%%%%%%%% Hope this helps. Unfortunately, ``bad style'' and ``required style'' often are synonymous! George D. Greenwade Center for Business and Economic Research THEnet: SHSU::BED_GDG Sam Houston State University Internet: BED_GDG@SHSU.BITNET Huntsville, Texas USA 77341-2056 Voice: (409) 294-1518 ------------------------------ Date: 17 Aug 1988 08:52:01-WET Subject: Re: The best way to do pictures in TeX From: alien William LeFebvre suggests generating pictures/graphics through TeX by using Metafont to generate a one-off font containing the figure. This is indeed possible; and Rick Simpson has been using *exactly* this approach for both line graphics and halftones. The advantage of this method is that there is no requirement for a larger-than-average TeX; but it does require that the DVI translater/output device can handle these large fonts. Those of us who have LN03/LN03+ laser printers cannot download glyphs bigger than about 60pt (firmware restriction, I believe), so this approach isn't viable. (Halftone-setting using Don Knuth's halftone font does work on an LN03, although both TeX and the DVI->LN03 translator need large memories; see TUGboat vol 8 no 3, where the pictures were generated in this way.) With respect to generating Metafont glyphs containing only part of the figure and tiling them on output, this also works very well. The approach was used by one Donald E. Knuth in his article on halftones in TUGboat last year (vol 8 no 2); I've done much the same thing with line graphics. I generate figures through a general-purpose line graphics package using a special `Metafont' driver; this takes the entire figure and converts it into a series of glyphs small enough for my LN03 to handle (actually, the device resolution and maximum tile size are Fortran PARAMETERS so it could, in principle, work on other devices with a silly size restriction). It also generates a file which can be \input to TeX to abut the tiles. Adrian F. Clark JANET: alien@uk.ac.essex.ese ARPA: alien%uk.ac.essex.ese@cs.ucl.ac.uk BITNET: alien%uk.ac.essex.ese@ac.uk Smail: Dept. of Electronic Systems Engineering, Essex University, Wivenhoe Park, Colchester, Essex C04 3SQ, U. K. Phone: (+44) 206-872432 (direct) ------------------------------ Date: 08/17/88 1322 From: Subject: DVICAN for MSDOS? Is DVICAN (Beebe's version) available for MS.DOS? We encountered some problems for the compilation of this program. Any help would be appreciated. Thanks in advance. Yves Bertheau EARN/BITNET: BERTHEAU.FRINRA72 ------------------------------ Date: Wed, 17 Aug 88 12:19 EST From: Subject: Tex novice needs installation advice Greetings: I need advice on how to i) obtain ii) mount LaTex (and, necessarily I suppose) Tex. I have not yet found a software vendor who sells for my new machine; and I really don't want to spend the money on public domain source code. The machine I am buying is an Iris 4D/70 Graphics workstation. The operating system is AT&T Unix V.3. I will have a C compiler and am, thus, most interested in C source code. I am not sure yet what sort of Laser Printer we have - but it will certainly have some standard sort of driver. Thanks very, very much. Amy Bug Till 1 Sept: Dept. of Chemistry, Columbia University, (212)280-5650 BE$BUG@CUCHEM BE$BUG@CUCHEM.CHEM.COLUMBIA.EDU BE$BUG@CUCHEM.CHEM.COLUMBIA.EDU After 1 Sept: Dept. of Physics and Astron., Swarthmore College BUG@SWATPRM (Actually, the former address will be O.K. indefinitely, as mail will be forwarded) ------------------------------ Date: Wed, 17 Aug 88 09:33 MDT From: "Gary McClelland, x8122, Psych CB 345" Subject: LaTeX meets a different book designer Given the recent whinings of the nature "I wish LaTeX did it this way or that," I thought TeXHaXers might be interested in my recent experiences writing a statistics textbook with LaTeX. (Note: many of the complaints seem to come from students trying to meet archaic standards imposed on dissertations. While most of these rules are dumb, those of us who are old enough to have once thought that a correcting IBM Selectric with a symbol ball was the greatest invention of the 20th Century don't have all that much sympathy.) I used LaTeX for three reasons: (1) its math capabilities far exceeded anything else I could find; (2) I wanted students in my classes who had to suffer for several years with draft chapters to have something readable; and (3) it helps motivate me to spend those long days writing if I can see some pretty text come out at the end of the day. Actually, I started with TeX but it forced me to make too many design choices. Believing that authors ought to write and book designers ought to design books I was ecstatic to find LaTeX which had made so many of those design choices for me. With one exception I used LaTeX default standards for everything. The exception is that I wanted tables and figures to be labeled as "Exhibits" and numbered within a single series. It was trivial to find the figure and table definitions and model an exhibit definition on them. I had no expectation that the publisher would actually use my LaTeX-produced pages. Indeed, they sent my ''draft'' to one of their own book designers. There is no reason to presume that my book designer is any better or worse than the one(s) consulted by Lamport in designing LaTeX, so I was curious to see what changes would be made in my manuscript. On the whole, very little was changed. The format of the header and subsection headings and lower were used as they were. Section headings became fancier as 2.1 -------------------------- This is Section Heading but of course the infamous lack of indentation after major breaks which has caused much comment of late was retained. The first page of each chapter was changed the most. The "Chapter" was omitted and the number written much, much bigger than \Huge, with a line part way across the page under that. The title began in what would be the margin notes area of LaTeX sort of like this: 2 ----------------------- Chapter Title Broken Very Frequently Like This First line of text was here and if I pad enough there will be a second line of text. The designer fussed quite a bit with space, wanting a bit more white space around environments such as quotation and enumerate and around equations that had summation signs with limits. The biggest change concerned placing and captioning the exhibits. The captions are entirely in the space that would be allocated to margin notes in LaTeX. For small figures the caption is aligned flush with the top of the exhibit; for large figures the caption is aligned with the top of the following text. In each case there is a small arrowhead pointing to the exhibit. I think that would be difficult to do in LaTeX because of the conditional placement but I'm sure some wizard could do it. However, don't send me the code! All of this is the publisher's problem, not mine. I would prefer to start on the next writing task than fiddle getting LaTeX to duplicate this new design. As I look back over the changes made by the book designer and the copy editor, most of them can be characterized as abuses of the easy freedom that LaTeX gave me. Two examples: (1) I tried to use different fonts and typefaces to make too many distinctions (e.g., some things in bold, others in italics). Although I thought I was careful, they simplified my excesses a lot. (2) It was too easy to get too fancy with the equations. On p. 139 of the TeXbook it says, "Mathematicians tend to 'overuse' \over when they first begin to typeset their own work on a system like TeX." Although I thought I was being judicious with my \over's the copy editor cut out about half my uses of \over. All of this convinces me that Lamport was correct in making it relatively hard for the writer (as opposed to a hacker) to change basic design features. Given too easy access to other design features I would have mucked it up even more. ------------------------------ %%% %%% Concerning subscriptions, address changes, unsubscribing: %%% BITNET: send a one-line mail message to LISTSERV@TAMVM1.BITNET: %%% SUBSCRIBE TEX-L % to subscribe %%% %%% All others: send mail to %%% texhax-request@score.stanford.edu %%% please send a valid arpanet address!! %%% %%% %%% All submissions to: texhax@score.stanford.edu %%% %%% Back issues available for FTPing as: %%% machine: directory: filename: %%% [SCORE.STANFORD.EDU]TEXHAXnn.yy %%% nn = issue number %%% yy = last two digits of current year %%%\bye %%% ------------------------------ End of TeXhax Digest ************************** -------