TeXhax Digest Thursday, June 11, 1987 Volume 87 : Issue 45 [SCORE.STANFORD.EDU]TEXHAX45.87 Editor: Malcolm Brown Today's Topics: Getting ASCII files out of TeX AM fonts and dvi format TeX 2.01 and LaTeX for VMS Are there device drivers which use pk format? BIBTEX reference generation in GNUemacs mf source for am fonts... tpic ? Probable bug in CMR fonts: lowres italic letter 'l' Xerox 9700 drivers Metafont and the LN03 dvi2ps for TOPS-20 Graphics facility for LaTeX wanted More on camera-ready conference proceedings Toshiba driver ---------------------------------------------------------------------- Date: Fri 29 May 87 09:38:26-EDT From: John Gourlay Subject: Getting ASCII files out of TeX To: brink@sushi.stanford.edu, texhax@score.stanford.edu Ed, I wrote a program several years ago that reads a dvi file and produces an ascii "doc" file that could be sent electronically to someone without TeX. Of course, it doesn't do anything useful with math, and my version is out of date with respect to tangle and weave. Nevertheless, it's circulating on the grapevine under the name dvidoc. I think Nelson Beebe brought it up to date and made it generally available. ------------------------------ Date: Sat, 30 May 87 22:31:45 PDT From: mackay@june.cs.washington.edu (Pierre MacKay) To: DHOSEK@hmcvax Subject: AM fonts and dvi format Unless you have access to a DEC20, the source for am fonts will not do you any good. The METAFONT used for AM was a totally different animal, running in SAIL, and using compiled subroutines rather than macros for repetitive tasks. I had access to a DEC20 until recently, but used CM as much as possible as soon as it came out. (METAFONT79 also ran on a DEC10, but I get the impression that it ran only on Donald Knuth's DEC10.) I don't know what to advise you if you are stuck with DVI files and no sources. Otherwise, you might as well convert to CM. Incidentally, the problem comes up in reverse if you are doing real phototypesetting. When we lost the use of a DEC20, we also lost the use of one of the very few remaining Alphatypes still running TeX output. I have since discovered that we were the only place running large amounts of output in CM fonts at a reasonable speed---not all that fast at that. You can get CM into Autologic machines, but only by downloading the fonts one by one at the price of a throughput of something like 15-25 minutes a page. I have a 190 page book in CM fonts that is presently an orphan, waiting for a typesetter. Autologic claims that it will some day --- HO HUM get around to CM fonts, but don't hold your breath. This summer we plan to try Varityper, which at least allows a bit mapped approach as a stopgap, so that something can be done while we try to work out a way to integrate CM into the rest of Varityper's offerings. At the moment, Varityper sounds far more cooperative than any other firm has ever sounded. There is at least room for hope. DVI format. The blocking specifications you mention are local machine-dependencies, and are not part of dvi format as described in (e.g.) the web listing of dvitype. DVI format knows nothing of blocking. It is a continuous stream of bytes from start to finish, and the only concession to machine architecture is the final padding bytes which guarantee that the file will contain a bytecount divisible by four. If you find differences in blocking factors. they have nothing to do with the version of TeX, but only with the accidents of implementation. DVI format has not changed in any way since well before TeX 1.0 was released. I have recently found myself answering a number of such queries, resulting from the occurrence of machine induced extra padding in various forms. People who do ports need to be very careful about this. A dvi file must end with dvi padding bytes---not with nulls, or trailer records, or terminator characters. The imposition of blocking factors, and other such extraneous features vitiates the whole purpose of a standard format, since it means that genuinely standard drivers cannot read files which claim to be dvi files. If, for instance, it is necessary to pad out a dvi file into multiples of 1024 bytes, then the padding should be DVI pad characters, even if there have to be 1023 of them. A driver written to the standard can handle 1023 DVI pad characters (It is decimal 223 or 233, I can't remember which) with ease. It cannot handle even one trailing null character, at least not if random file positioning is used. Pierre A. MacKay TUG Site Coordinator for Unix-flavored TeX ------------------------------ Date: Fri, 29 May 87 14:07:22 PDT From: ucdavis!clover!hildum@ucbvax.Berkeley.EDU (Eric Hildum) To: texhax@score.stanford.edu Subject: TeX 2.01 and LaTeX for VMS Is there a machine which has the latest versions of TeX and LaTeX for VMS that can be reached via FTP? If so, what are the procedures for getting the programs? Thank you, Eric ------------------------------ Date: Sat, 30 May 87 18:06:21 +0200 To: TeXhax@score.stanford.edu From: mcvax!ruuinfvax!piet@seismo.CSS.GOV (Piet van Oostrum) Subject: Are there device drivers which use pk format? I have modified dvi2ps to use pk files rather than pxl files (actually the pxl code is still in, but the choice is made by #ifdefs). I also made some other changes: 1. I used the proper positioning algorithms (almost Knuth) 2. I have the program guess how much virtual memory the laser writer has left. If the VM is low or if the characters are big, characters are no longer downloaded, but sent individually. This allows us to print any document without troubles 3. Use of TEXFONTS environment variable 4. I incorporated (and improved) changes by Les Carr (University of Southampton, England) to use builtin Postscript fonts. I have also plans to upgrade the latter item (I have some mods by Jean-Francois Lamy (University Toronto) which are more universal but it is more work to integrate them with my own changes. Meanwhile we are moving to a new machine ... Anyone who likes can get my version. I think it would be a good idea to consolidate the different dvi2ps versions and put them in a central location. ------------------------------ Date: Fri, 29 May 87 18:31 PDT From: Mabry Tyson Subject: BIBTEX reference generation in GNUemacs To: texhax@score.stanford.edu There was a BIBLIO emacs library for TOPS-20 emacs that allowed somewhat formatted input of bibliography entries (it knew what were the legal/optional/ required parameters). Does anyone know of a similar facility for GNUemacs? Our secretary has a large number of books, etc to enter into a bibliography and it would be easier if there were some way to help automate it. ------------------------------ Date: Sun, 31 May 87 04:32:05 EDT From: Ray Hirschfeld Subject: mf source for am fonts... To: U12921%UICVM.BITNET@FORSYTHE.STANFORD.EDU I have the am sources in on xx.lcs.mit.edu. You can nab them via anonymous ftp. Any .mf file in that directory is for the am fonts; the cm sources are in . You'll also need the old metafont to run them off; there's a twenex copy in oldmf.exe, but I don't have the source (written in sail, I believe). Use of these fonts is probably officially discouraged. I wouldn't advise relying on their continued availability. Ray ------------------------------ Date: Mon, 1 Jun 87 10:28:03 -0100 From: mcvax!corto.inria.fr!shapiro@seismo.CSS.GOV (Marc Shapiro) To: texhax@score.stanford.edu Subject: tpic ? Reply-To: mcvax!inria.inria.fr!shapiro@seismo.CSS.GOV I have heard of a pic-like front-end to TeX/LaTeX. This would allow to create pictures in TeX texts using high-level commands similar to troff's 'pic'. Does anybody know where to find this wonderful thing? Marc Shapiro INRIA - bat. 11 B.P. 105 78153 Le Chesnay Cedex France tel.: +33 (1) 39-63-53-25 e-mail: shapiro@inria.inria.fr or: ...!mcvax!inria!shapiro ------------------------------ From: Oliver Schoett To: TeXhax@SCORE.STANFORD.EDU Reply-To: schoett%infovax.informatik.tu-muenchen.dbp.de@RELAY.CS.NET In-Reply-To: Subject: Probable bug in CMR fonts: lowres italic letter 'l' Date: Sun, 31 May 87 19:53:02 -0100 I noticed that in lowres italic fonts (81 dpi for VaxStation), the italic letter 'l' has a broken stem and thus becomes very hard to read. I tracked the problem down to what appears to be a bug in ITALL.MF, "Italic letter l": In the code fragment adjust_fit(if monospace:u#,u# else: 0,0 fi); pickup fine.nib; x2-.5stem=hround(1.5u-.5stem); x4=w; hook_out(2,3,4); % closing hook pickup tiny.nib; pos1(stem,0); pos2'(stem,0); z2'=z2; top y1=h; x1=x2; filldraw stroke z1e--z2e; % stem the last line should be changed to top y1=h; x1=x2; filldraw stroke z1e--z2'e; % stem Explanation: Points z2l and z2r are defined by hook_out(2,3,4) with fine.nib as the current pen, whereas the stem is drawn with tiny.nib. The correct points for this pen are computed by pos2'(stem,0). The problem in principle occurs in all resolutions in which fine.nib and tiny.nib differ; it is particularly prominent in lowres situations where fine.nib is 1 pixel while tiny.nib is zero. Disclaimer: I am writing this quickly from memory and don't have the time to recheck anything, including my terminology (I only learned a bit of Metafont a few weeks ago to be able to generate fonts). So don't act on this until the bug has been 'officially' confirmed or denied. I apologize for the sloppy posting, but I'm quite sure there is a bug and some work regenerating fonts might be saved if it is fixed in connection with the 'digit 8' bug just confirmed. Oliver Schoett Technical University of Munich, West Germany schoett@infovax.informatik.tu-muenchen.dbp.de safer address from ARPA: schoett%infovax.informatik.tu-muenchen.dbp.de@relay.cs.net ------------------------------ Date: Mon, 01 Jun 87 13:06:26 CDT From: "Don Hosek, Editor, TeXMaG" Subject: Xerox 9700 drivers To: TEXHAX@score.stanford.edu There are now two new 9700 drivers available which make the Xerox 9700 a *useful* TeX output device. Arbortext's DVIXER has been enhanced to allow dynamic selection of fonts on the DJDE, so the prior restrictions to font packages are removed. There is also a package called TeXrox from Texas A&M that not only deals with fonts sensibly, but allows landscape and reference card printing, as well as a few other features. -DH ------------------------------ Date: Mon, 1 Jun 87 13:25 EDT From: (John Lavagnino) Subject: Metafont and the LN03 To: texhax@score.stanford.edu X-Original-To: texhax@score.stanford.edu,beeton@score.stanford.edu, LAV We have discovered that the LN03 and the LN03-Plus don't print the same way: a font that looks fine on the LN03 will look lighter on the LN03-Plus. In fact it isn't necessary to download fonts to observe this: even the internal fonts look different. We've been badgering DEC about this for some time, and they have finally agreed that this is the case. The current story is that they ``made the pixels smaller'' on the LN03-Plus, ``to make it look more like a typewriter.'' Be aware, then, that a good mode_def for one will only be a poor approximation for the other. John Lavagnino (lav @ brandeis.bitnet) Systems programmer, Feldberg Computer Center, Brandeis University 415 South Street Waltham, MA 02254 USA (617)736-4594 ------------------------------ Date: Mon, 01 Jun 87 15:58:18 EDT From: gene Denzel Subject: dvi2ps for TOPS-20 To: Texhax Digest We need a Postscript driver for a DEC 2060 under TOPS-20. We have been offered an executable set of files, but for obvious reasons would prefer source. Does anyone know where such can be obtained? Thanks in advance. ------------------------------ Date: Mon, 1 Jun 87 16:53:28 +0100 From: wtho%cgcha.UUCP%cernvax.bitnet@BERKELEY.EDU (Tom Hofmann) To: texhax@score.stanford.edu Subject: Graphics facility for LaTeX wanted We use LaTeX for typesetting and need better graphics facilities for interactively producing diagrams. Does someone know about an inter- active graphics system that produces LaTeX commands for the picture environment. Or is there another solution for simply producing and modifying graphics for LaTeX? Tom Hofmann, CIBA-GEIGY AG, Scientific Computer Center, Basel, Switzerland E-mail: wtho@cgcha.uucp ------------------------------ Date: Mon, 1 Jun 87 11:13 EDT From: Reid Rowlett Subject: More on camera-ready conference proceedings To: texhax@SCORE.STANFORD.EDU X-VMS-To: IN%"texhax@score.stanford.edu" I appreciate Leslie Lamport's and William LeFebvre's comments on my questions about creating camera-ready proceedings articles (TeXhax Digest V87 #35 and #38). I thought someone might be interested in what I've learned since then. As Leslie Lamport suggested, I contacted SPIE (Society for Photo-Optical Instrumentation Engineers, an optical engineering group) concerning whether they would accept camera-ready copy on 8.5 x 11 (already-reduced) sheets, rather than on the larger sheets they provide, which are too large for our Imagen laser printer. I was told they would not. They suggested I use a two-column format, but have LaTeX (which they had never heard of) produce the columns indivdually, and paste them up. However, if I do this I lose LaTeX's page-breaking capability, and it's double-column features (I suppose I could use a smaller page size for the first two columns to accomodate the title, and the actual page size thereafter, but I hate to do this when LaTeX already knows how to handle two-column articles. What I can do instead is to tell LaTeX the actual dimensions I want, and then print the document twice on the laser printer (in landscape orientation), using different print margins so that first the top half and then the bottom half of the page is printed. I still have to paste up the results, but at least LaTeX has done the formatting for me, and the format exactly fits SPIE's specifications. I like William LeFebvre's idea of working at the final reduced size, and using \magnification=(1/reduction) to fit the larger sheets, but the need for the oddball fonts is a problem. Since (1 / 0.77) is close to magstep1 (1.2), I could approximate his idea by working at a slightly enlarged aspect which magnification=1200 would then fit onto SPIE's large sheets without using oddball font sizes. The only problem here is that I need a 9-point option for ARTICLE.STY to obtain approximately 11-point fonts at magnification=1200. Perhaps Leslie Lamport can give me direction on creating a 9-point option style by modifying ART10.STY (can I simply use the \small declarations and redefine all the sizing commands?). I should add that my goal is to produce the most readable document I can within the constraints SPIE has placed on me, not to make my document "look nice". Although I have no training in document design, I do not consider myself naive with respect to judging readability. What a document designer presumably understands that I do not is WHY one style is more readable than another, and HOW to select dimensions and fonts and so forth to make a readable style. Reid Rowlett Unisys Corporation Reston Technology Center Reston, VA ------------------------------ Date: Tue, 02 Jun 87 10:56:17 GMT From: vivian harrington Subject: Toshiba driver To: TEXHAX@score.stanford.edu Does anyone have or know of a driver for a toshiba 2001G printer or a similar driver that can be easily converted to drive the above printer for Tex running under VM/CMS 3.1. Desperate! Vivian Harrington (Advisory Service, University College Dublin) ------------------------------ Date: Tue, 2 Jun 87 15:50:30 PDT From: dan@sun-valley.STANFORD.EDU (Dan Rovner) To: texhax@score To whoever can explain this phenomenon: Below is a fragment of a Latex input file which does something that I wish Latex would do all the time: it puts a float in the middle of a paragraph. What is it about this file that causes this to happen ? How can I make this happen routinely ? An explanation would be greatly appreciated ! dan@sun-valley.stanford.edu \documentstyle[11pt,suthesis]{report} % \renewcommand{\bottomfraction}{.95} \renewcommand{\topfraction}{.95} \renewcommand{\textfraction}{.05} % \begin{document} \chapter{Minimal Parameter Constraint Approach} \label{chap6} Something to take up space. \newpage Figure~\ref{c6f14} displays tip position and parameter time histories for an experiment using the algorithms described. % \begin{figure}[htb] \setlength{\unitlength}{1.0in} \begin{picture}(6.0,6.5) \special{psfile=c6f14.p} \end{picture} \caption{Adaptation experiment with a 250 g load.} \label{c6f14} \end{figure} % The arm was initially unloaded and was commanded to slew to a position of -.1 m using a SISO pole placement controller designed offline. The arm gripped \end{document} ------------------------------ %%% %%% subscriptions, address changes to: texhax-request@score.stanford.edu %%% %%% submissions to: texhax@score.stanford.edu %%% %%% BITNET redistribution: TEX-L@TAMVM1.BITNET (list server) %%% %%%\bye %%% ------------------------------ End of TeXhax Digest **************************