TeXhax Digest Sunday, March 1, 1987 Volume 87 : Issue 14 TEXHAX14.87 Editor: Malcolm Brown Today's Topics: Latex bugs (?) + advice please TeX for HP-UX macro for ``time line''? DVI Driver for Talaris 800 BibTeX problems LaTeX notes new utilities. Comments by an out-going site administrator Doublespaced documents Page Description Language Standards Effort double column macro ---------------------------------------------------------------------- Date: Fri, 20 Feb 87 9:39:15 GMT From: trh%ukc.ac.uk@Cs.Ucl.AC.UK To: texhax@score.stanford.edu Subject: Latex bugs (?) + advice please I mailed this a week or so ago -- it didn't appear in volume 11 and I've solved one of the problems I had. I'd appreciate it if you'd distribute this rather than the original. Could somebody please identify whether the following are bugs in LaTeX or just misunderstandings by a naive user:- We are using Latex version 2.09 (18/12/85) and Tex version 2.0. In all cases the documentstyle is book. I can e-mail full texts to anyone who wants them. 1. Labelling chapters causes peculiar behaviour use \chapter{This is a chapter \label{chap}} and have at least two pages of text following. There is usually some error message like `missing control sequence inserted' to start with carrying on leads to an overful hbox containing a very garbled page header of the form CHAPTER1 title \NEWLABEL11\NEWLABEL1............. Further perserverence leads to the error messages being placed in the dvi file along with lots of \newlabel constructions. This all goes away if the \label is removed from the \chapter. 2. Starting an enumeration at something other than one in non-default format bombs out. e.g., \begin{enumerate} \setcounter{enumi}{4} \roman{enumi} \item junk \end{enumerate} produces `Latex error - something's wrong -- perhaps a missing \item'. Forcing latex to carry on produces a iv label in a mangled enumeration. 3. Can't change the format of an item tag inside a newtheorem e.g., use \newtheorem{exercise}{\small \bf EXERCISE}[chapter] then \begin{exercise} \begin{enumerate} \roman{enumi} \item should be a roman tag \end{enumerate} \end{exercise} produces a label 1. While I'm here can somebody please help with the following problem How do I temporarily switch off the output of figures? I want to define a newenvironment along the lines of \newenvironment{codeburst}{ switch off figure output; output code} { switch figure output back on} I've tried setting totalnumber to zero, textfraction to 1, topfraction to zero and both topnumber and bottomnumber to zero all without success. Setting totalnumber to zero and printing the current setting returns a null string rather than 0. Many thanks in advance for any assistance. Tim Hopkins, Computing Laboratory, University of Kent, Canterbury CT2 7NF Kent U.K. { trh@ukc.UUCP trh%ukc@ucl-cs.ARPA na.hopkins@su-score.ARPA } ------------------------------ Date: Fri, 20 Feb 87 04:47:35 PST From: Reply-To: MHAMALAI%FINFUN.BITNET@forsythe.stanford.edu To: texhax@score.stanford.edu Subject: TeX for HP-UX Is there a version of TeX for HP 9000/320 (HP-UX) workstations? What about a DVI driver for the high resolution bit mapped displays used in these machines? Matti Hamalainen ------------------------------ Date: Fri, 20 Feb 87 15:08:00 EST From: James Alexander To: texhax@su-score Subject: macro for ``time line''? Is there anyone out there who has a macro to make a ``time line?'' I would like to have tex read a file which consists of entries of the form (no problem sorting it chronologically or otherwise processing it) and produce a vertical line of some preassigned length with tick marks so that the top of the line represents the first date (or #1 in the macro call) and the bottom represents the last date (or #2). Down the line, with vertical spacing mimicking time intervals, the dates and events are printed horizontally out to the right. One problem is to do something intelligent when two or more of the dates cluster too closely (e.g. two events on the same date). I can see the general idea; I can also foresee many texnical details. Alternately, one could write a preprocessor in C or Pascal. Before I dig in (or give up), has anybody done it, or have any suggestions? ------------------------------ Date: 21 Feb 87 13:31:00 EST From: "IAN EVANS" Subject: DVI Driver for Talaris 800 To: "texhax" Reply-To: "IAN EVANS" Does anybody know if there is a PUBLIC DOMAIN DVI Driver for the Talaris 800 laser printer for VAX/VMS? I don't wish to spend $2700 on the Talaris TeXsupport and FontManager software! Any information appreciated. Ian Evans ARPA: evans@scivax.ARPA Phone: (301) 338-4756 Snail: Space Telescope Science Institute 3700 San Martin Drive Baltimore, MD 21218 ------------------------------ Date: Sun, 22 Feb 87 21:04:28 EST From: Stephen Gildea To: spqr%sot-cm@chevoit.ncl.ac.uk In-Reply-To: Sebastion Rahtz's message of Sun 8 Feb 87 23:07:54 GMT Subject: BibTeX problems BibTeX uses the sort.label entry variable as the key to sort on. It sorts by the text label only because in the function calc.label (in the natsci and alpha styles) sort.label is set to that. You should modify calc.label to prepend the author's name (or whatever you want to sort on) to sort.label, then BibTeX will sort first by author, and then by the other info in the label. < Stephen mit-erl!gildea gildea@erl.mit.edu ------------------------------ From: lamport@decwrl.DEC.COM (Leslie Lamport) Date: 22 Feb 1987 1954-PST (Sunday) To: TeXhax@Score.Stanford.edu Subject: LaTeX notes From: trh%ukc.ac.uk@Cs.Ucl.AC.UK To: texhax@score.stanford.edu Subject: Latex bugs(?) + help please 1. Labelling chapters causes peculiar behaviour... This seems to be a bug. Until it's fixed, don't put a \label command in the argument of a sectioning command--put it right after the command instead. 2. Starting an enumeration at something other than one in non-default format bombs out. 3. Can't change the format of an item tag inside a newtheorem In both of these problems, Hopkins seems to think that \roman{enumi} changes the way the enumi counter is numbered. It doesn't; it simply prints the value of that counter as a roman numeral. Try looking up `counters' in the manual's index. From: mroth@afit-ab.ARPA (Mark A. Roth) Subject: latex style for proceedings I know that I have seen somewhere a document style for LaTeX which does proceedings format and on the first page leaves the space at the bottom left for the copyright, copying policy information. That's the `proc' style. This should be described in your Local Guide. If it's not, please complain to your local LaTeX distributor. Leslie Lamport ------------------------------ Date: Sun 22 Feb 87 21:27:04-PST From: Pierre MacKay Subject: new utilities. To: texhax@SU-SCORE.ARPA By popular request, the scripts prepared by Karl Berry from Sauter's routines for truesize fonts are now available from ward.cs.washington.edu FTP them using the fill pathname, you can't connect to the directory. THey are compressed, and owing to the limitations of the TOPS20 directory convention you will have to rename SAUTER.SHZ to sauter.sh.Z to decompress. PS:SAUTER.SHZ Font hackers with a Sun workstation should take a look at Brad Rullman's Fonttool. Shar file compressed is in PS:FONTTOOL.SHZ, and documentation is in PS:FONTTOOL.TEX Make sure you get with type tenex for the SHZ files. THe TEX file should be got with type ascii. Pierre MacKay TUG Site Coordinator for Unix Systems ------------------------------ Date: Sun, 22 Feb 87 14:02:38 est From: gropp-bill@YALE.ARPA Subject: Comments by an out-going site administrator To: texhax@su-score.ARPA I would like to make a few comments on some of the problems of being a site administrator, and a few suggestions on improvements to the way TeX is handled. My biggest problem as the Apollo/Aegis site coordinator was discovering changes to the "standard" TeX software. Changes to TeX itself have been both mercifully few (a tribute both to the writers and the policy of rewarding those who find bugs) and well announced. However, changes to TeXware (tangle, dvitype, etc) and macro packages have been poorly announced. In some case, I discovered that a new version was available only because someone on site had mailed a bug report to the author, only to be told it had already been fixed in an (unannounced) release. With the revival of TeXhax, there has been a medium for announcing new releases. However, this digest has quickly become choked with novice questions and long pieces of TeX code, and information for site administrators can get overlooked. I would like to suggest: A mailing list of site administrators. This should be both e-mail and Postal mail. ANY change to TeX, TeXware, or (major) TeX macro packages should be announced directly to this list within one week of being released. A LaTeX mailing list. Much of the mail in TeXhax consists of LaTeX questions. TeXhax itself should be resevered for the arcane in TeXware and the TeX system (which could include subtle questions about a LaTeX implementation, but not "how do you do this in LaTeX?"). Under no circumstances should new LaTeX releases be announced only in TeXhax or the LaTeX list. Given the vagaries of various mail systems and the fact that many site administrators have full time jobs in addition to managing TeX, every six months a list of the current versions of the current standard TeX system should be distributed. In addition, this would serve to keep all TeX systems up to date, and hence insure that users aren't fighting old and known bugs. Now that TeX is maturing, with a broad user base and a wide variety of support software, it is important to make certain that when "TeX" is offered, it is a complete, up to date system. This will require a little more central management of changes and distributions, but I believe the effort will be worth it. Bill Gropp ------------------------------ Date: Mon, 23 Feb 87 14:53:53 PST From: Reply-To: OGAWA%SLACVM.BITNET@forsythe.stanford.edu To: TEXHAX@score.stanford.edu Subject: Doublespaced documents I'm sorry that I have not much more to offer than an opinion, but here goes: I agree with Lamport that much needs to be changed in the way theses are formatted. I have had the experience of working with dozens of thesis writers from all over the country. In each case the style sheet for the thesis clearly compromised the appearance of the document because of choices relating to the assumption that the thesis would be typed by hand. In many cases the student was unable to make even rational changes in the thesis format in the face of rigid resistance on the part of university administrators, in one case even in spite of support by the student's advisor. One of the compromises is the requirement for double-spaced, 12 point size. This is quite a bit less readable than a book or a journal article, as we all agree. Leslie says ``The only good reason I know of for double-spacing is to provide output for a copy editor''. I might point out that traditional typesetters require 14 point double-spaced typescripts. It seems a great shame that theses and also many preprints produced by even institutions equipped with TeX seem to be done with archaic requirements in mind. My recommendation to those producing such documents with my macro package, Psizzl, is to produce two documents. This is the same as Leslie suggests. I feel that many have been able to produce satisfactory documents by following this suggestion. The process goes something like this: 1. You're writing a thesis or article for publication. You want those reading the thesis or the preprint to enjoy the same readability as if they were reading a book or journal article. 2. You produce the thesis or article in the style that the university or publisher mandates. This may involve setting the \baselineskip to 5ex or something horrible like that. Also the basic font has to be chosen appropriately. All these ``compromised'' commands are segregated in a .sty file or other, so that they may be turned off easily. 3. You hand in the two (2) copies of the thesis, or send the one (1) typescript to the publisher and wash your hands. One copy of your thesis goes to the university library where nobody looks at it for a few years while it looses its timeliness. The other copy goes to Ann Arbor where it is microfilmed (now you know why it had to be done in 12 point, double-spaced). Hapless interested parties may request marginally readable copies for an exorbidant price. No, you don't receive a royalty. In the case of an article for publication, the typescript goes to the publisher and thence to the typesetter, who ultimately sends galley proofs back to you with many typographical (keyboard) errors. You correct these. Then your article is published in the journal of your choice, where it once again gathers dust. Why? Everybody who is interested in your work has already read your (beautifully formatted) preprint (see below in step 5). 4. Once the thesis or typescript is completed, you now produce the readable version of the document. This is done by turning off reference to the optional ``compromised'' command file and reverting to a .sty file or other that produces a readable document, such as \documentstyle{book} or, in the case of an article, \documentstyle{preprint} (not implemented in LaTeX, to my knowledge; correct me if I err). 5. Now the document is re-done (computer charges be d____d) and printed in quantity at your, or your institution's expense. Since the document was not formatted double-spaced, and was probably set in 10 point type, doublesided and all, this version is much shorter that the one produced in step 2. It is more readable and has up to 4 times fewer pages and is correspondingly cheaper to produce. Your institution's library sends out copies of the preprint to libraries of other institutions where it is read and xeroxed ad infinit'em. Some additional things to mention: the typescript you produced in step 2 also conformed to the requirement that all figures were collected together at the end of the document and all the figure captions were printed out on a separate page by themselves. Same for tables. This makes a typescript totally inappropriate as a preprint since the reader has to read your text while having to look at two other places each time a figure or table is called out. So you also choose not to produce your document in typescript form until the one time when it is necessary. All your drafts are produced in the preprint form, HA HA! Also, it is customary for the publisher of scientific journals to provide you, the working author, with preprints of your article in advance of the article's publication. When you receive these, you will gaze at them with a bit of satisfaction, since your own preprints were produced two to three months earlier and look much nicer. I won't go so far as to recommend that you use \tt\raggeedright in formatting the typescript in step 2, but I do want to imply by the above that with a bit of committment on your part as the working author you can produce a much finer document than that created by following the outdated rules, and further that the form of your document produced ``by the book'' in step 2 has not nearly the impact of the nicer version in step 5 simply because it is marketed more poorly. Those wishing to learn more about the Psizzl macropackage, in use here at SLAC, may contact me for a copy of SLAC-268, the manual. ------------------------------ Date: 25 Feb 87 11:10:00 EST From: "ROSENTHAL, LYNNE" Subject: Page Description Language Standards Effort To: "texhax" For those who are interested: ANSI X3V1, Information Processing - Text and Office Systems (Office Document Architecture and SGML) is starting to work on standards for Text Composition Semantics and Syntax (TCSS) and Text Presentation Metafile (TPM). TPM = Page Description Language. If interested the next X3V1 meeting is March 9-13, Golden Colorado at the Denver Marriott West. The X3V1 chair is Millard Collins, (214)556-7690, however, the work will be conducted in Task Group 8 -- information can be obtained from Jim Mason (615)574-6973 ------------------------------ Date: Wed, 25 Feb 87 12:34:58 PST From: mahoney@lbl-csa4.arpa Subject: double column macro To: texhax@score.stanford.edu I have heard of a double column macro that really works, i.e., switches between single and double successfully on a page. The Knuth macro balks at the second single-to-double switch. The working double column macro reportedly appeared in a TUGboat about two years ago. I would like to get a copy of it and wonder if you can help. THank you. Jeannette Mahoney ------------------------------ % %\bye % End of TeXhax Digest ************************** -------