TeXhax Digest Friday, September 25, 1987 Volume 87 : Issue 78 [SCORE.STANFORD.EDU]TEXHAX78.87 Editor: Malcolm Brown Today's Topics: Undump for SUN OS 3.4 Unbreakable text in LaTeX? Re: Questions about fonts TeX bug -- version 2.5 LaTeX APA style, Epson fonts LaTeX Notes (Re: TeXhax Digest V87 #76) More comments about index processing 'Sucking' files on Decnet Porting TeX to MS-DOS PC's Latex/GKS/UIS Re: VMS Tektronix 4010 preview gftopxl.web dvitty Plain TeX: using scaled fonts avoiding missing half sizes. ---------------------------------------------------------------------- From: Jackie Damrau MATH Date: Fri, 18 Sep 87 16:15:31 MDT To: lanl!score.stanford.edu!texhax Subject: Undump for SUN OS 3.4 Has anyone gotten TeX to compile on SUN OS 3.4? We have had other problems in loading TeX on the SUN. In particular, we have modified undump to work (we think!!), however, some variables in the core file (that has not been undumped) have different values than in the undump TeX file. Whenever we run TeX with our modified version of TeX, it dumps core. Jackie Damrau, UNM Math Dept., Albuquerque, NM 87131, 505-277-4623 UUCP: damrau@ariel.unm.EDU BITNET: damrau@unmb.BITNET (actually, the .BITNET can be omitted if you prefer but it does make things very clear) ------------------------------ Date: Sat, 19 Sep 87 15:37:10 EDT From: John T Kohl To: texhax@score.stanford.edu Subject: Unbreakable text in LaTeX? In Scribe, it is fairly easy to insure that a group of text appears on a single page (if it will fit) by wrapping it with @begin(group) ...text... @end(group) In LaTeX, I don't see any easy mechanism to do this. I was able to use \parbox and do some disgusting line length hacks so it looked fine, but it seems to me that there should be a more elegant solution. John Kohl MIT/Project Athena ------------------------------ Date: Sun, 20 Sep 87 14:21:02 EST From: Chris Torek To: TeXhax@Score.Stanford.EDU, crm@cs.duke.edu Subject: Re: Questions about fonts In TeXhax digest V87 #74 Charlie Martin asks: >How can I get fonts at particular sizes when they are not part of >the distribution? Use Metafont to build a GF file in the appropriate size. If you need PXL files, use GFtoPX to convert the result. If you can read PK files, use GFtoPK to convert it (PK files are smaller and easier to read). Metafont, GFtoPX, and GFtoPK are in the current Unix TeX distribution. >2.1 Is there a PK-form (or other compressed form) driver available for >the Imagen? My drivers (Imagen and Versatec) handle PK, GF, and PXL files. They also use TFM `fonts' to print SliTeX fonts in any size without using any[*] disk space for fonts---this is done via a general mechanism, not a special kludge---or blanks or boxes in place of letters in fonts in missing sizes. If you ask for cmr10 scaled 1972 and it does not exist, you can configure the drivers either to complain and stop, or to complain and continue, using box or blank fonts. (Box fonts print much like the example in the TeXbook that explains how characters are just boxes to TeX.) If you have some special local font format, it should not be too difficult to add this to the set of recognised fonts. [*Well, not much space: it requires some directory entries.] You can get these via anonymous FTP to host mimsy.umd.edu. Use image mode (on Unix, give the `image' command to FTP) to retrieve the file `tex/ctex.tar.Z'. This is a Unix `tar' file. Unfortunately, there is no documentation on setting up the system. I can answer specific questions (`what goes in a fontconf file'), but not general ones (`how do I install this'). >2.2 While we're on the subject, how about an imagen driver that >includes options to print only selected pages? Ctex.tar includes dviselect (version 2.6; if you have version 1.x, yours probably produces bad DVI files if the input file has \specials). Dviselect is a DVI to DVI translator that selects pages. It can be used with any driver, since it produces another DVI file. Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 7690) Domain: chris@mimsy.umd.edu Path: uunet!mimsy!chris ------------------------------ Date: Sun 20 Sep 87 21:58:57-EDT From: Barbara Beeton Subject: TeX bug -- version 2.5 To: texhax@SCORE.STANFORD.EDU This message was just received from Don Knuth; another bug has been found. As soon as the distribution files have been updated, TeXhax will be informed. One day after I made version 2.4, Alan Guth of MIT Center for Theoretical Physics sent me a letter, which I just received after a trip to Europe. He found a real live bug, so the saga has not yet ended! (Error occurred if you had \ifdim immediately after "fil" or "fill" or "filll"; I didn't save the global variable cur_order in the expand routine. Yes, it's another instance of `global variable considered harmful'.) Anyway 2.5 is a nice round number. The total number of bugs/changes is now 333; that also has a certain symmetry. So the last bug has presumably been found. ------------------------------ Date: 17 Sep 87 15:15 +0800 From: Vincent Manis To: texhax@SCORE.STANFORD.EDU Subject: LaTeX APA style, Epson fonts I've only been following TeXHaX for a short period of time, so these are undoubtedly dumb questions. Still, I don't know the answers: 1) Where is there a source of a LaTeX APA (Amer. Psych. Assoc.) style? (If it's short enough, I'd appreciate it if someone would just email it to me.) 2) We have some dot-matrix printers here, which we'd like to drive as proof devices. To generate the .pxl files, should we: a) ask for them somewhere, or b) do it with mf? I know that there's a PD dvi-epson translator out there; could anyone point me at it? AtDhVaAnNxCE Vincent Manis manis@instr.camosun.bcc.cdn Camosun College ihnp4 | 3100 Foul Bay Road seismo |!ubc-vision!instr.camosun.bcc.cdn!manis Victoria, BC V8P 4X8 uw-beaver | (604) 591-1281 x480 manis%instr.camosun.bcc.cdn@ubc.csnet manis%instr.camosun.bcc.cdn%ubc.csnet@relay.cs.net Controversy: a dispute in which at least one of the participants is wrong. ------------------------------ From: lamport@src.DEC.COM (Leslie Lamport) Date: 19 Sep 1987 1709-PDT (Saturday) To: TeXhax@Score.Stanford.edu Subject: LaTeX Notes (Re: TeXhax Digest V87 #76) Patrick Tang writes: (1) Is it possible to draw circles with diameter much greater than 1/2" using the LaTeX picture environment? (2) Drawing a real oval using picture environment rather than the current \oval which produces just a rectangle with round corners. (3) A program that will translate postscript format (namely postscript from MacDraw) to LaTeX picture Environment? For (1) and (2), you can try the bezier document-substyle, which allows you to draw aribtrary quadratic arcs. (I don't know what a "real oval" is, so I have no idea how hard it is to draw one with such arcs.) The answer to (3) is "no". However, some implementations of TeX have \special commands for including PostScript commands. Nelson Beebe writes: How does one produce a table in which column entries are aligned on decimal points, when the number of figures following the point is variable, i.e. a table with a column like this: --------- x.x x.xx xx.xxxx xxx.x --------- The easiest way is to put `r@{.}l' in the argument of the tabular environment and enter the number 123.45 by typing `123&45'. One can define an environment in which the user types `123.45' by \catcoding `.' to the `&' character. However, this is rather tricky and precludes the use of `.' for anything but aligned decimal points within the environment. Leslie Lamport ------------------------------ Date: Sat, 19 Sep 87 06:14:58 PDT From: phc@renoir.Berkeley.EDU (Pehong Chen) To: TeXhax@score.stanford.edu Subject: More comments about index processing Nelson Beebe recently commented on another view of indexing by pointing out that the "Chen/Lamport MakeIndex is about 2600 lines of C" and that, by contrast, the Kernighan et al's approach to indexing processing is only 60 lines of awk, including comments and blank lines. I have seen the awk book Nelson suggested and was also amazed by the examples given there. The same feeling has recurred every time I finish an awk program. Awk is indeed a powerful and elegant language for rapid prototyping. But back to the problem of indexing, I hope people are not misled by just looking at the numbers: 2600 vs. 60 is indeed a very significant difference. No programmers would want their code to be superfluously long. So there must be a reason for MakeIndex to be ``long'', other than some intrinsic differences between C and awk in terms of their respective expressive capabilities. The reason is simple --- speed and functionality. In fact, earlier versions of MakeIndex were all written in awk/sed. But the most recent awk/sed version of MakeIndex is 3 to 5 times slower than the current C implementation, while the former's functionality is less than half of what the latter can do. We have also evaluated Kernighan et al's indexing package, make.index. Their claim is that by breaking the system down into small pieces, it is easier for the user to adapt the package to various situations not possible to envision in advance. Their approach, therefore, seems to follow the ``do it yourself'' metaphor: get a kit and assemble the final product yourself. The basic package covers only those features used by its authors. A third party user like you and me would have to modify their awk code if something different is desired. The design of MakeIndex is quite different. Our intention is to build a complete system by carefully analyzing the tasks involved in index processing. Parameters are derived to form a declarative table-driven style handling facility. All underlying data structures and processing mechanisms are transparent. Yet it is robust enough to handle different precedence schemes, ordering rules, etc. To use the system, a user needs not deal with the processing details. If the default is inappropriate, the only adaptation required is straightforward attribute specification. For MakeIndex to process troff documents, one only has to specify attributes for the input format and output style. For make.index to process TeX documents, however, one would have to hack their awk code, which would be difficult for casual users. Again, make.index's functionality is comparable to the awk/sed version of MakeIndex, which is less than half of what the current MakeIndex can do. I fully agree with Nelson Beebe that the elegance of awk is marvelous. But in the context of index processing the tradeoff is more than elegance. I'm not disputing the usefulness of make.index, nor am I advocating MakeIndex. It's just some comments for would-be indexer users to think about. Pehong Chen Computer Science Division UC Berkeley ------------------------------ Date: Sat, 19 Sep 87 08:58 N From: Reply-To: Subject: 'Sucking' files on Decnet To: TEXHAX@score.stanford.edu Is anybody outside there so kind who has PSPRINT and DVITOVDU from A. Trevorrow as in Tugboat vol 8 n.1 pag 42, on a Decnet link and is so kind to let me copy them all? Unfortunatly here in Italy we have no FTP. Thanks Max Calvani Decnet address 39003::fisica ------------------------------ Date: Mon, 21 Sep 87 15:12 EDT From: Subject: Porting TeX to MS-DOS PC's To: texhax@score.stanford.edu I'm getting ready to (attempt to) port TeX, MF, et.al. to the MS-DOS PC environment using MS-Pascal. If anybody has any pointers (change files, etc.) please mail me. My goal is to port the software and make it available to the TeXommunity for distribution costs only. Support would be unavailable from me. However, any public domain fonts, DVI drivers, etc. will gladly be included in the distribution. If you have anything that may be of use to someone running TeX on a PC, let me know and when I'm ready, I'll ask for it. Thanks, --mike "Isn't there some site out there who can take over mcguffey@muvms1.bitnet for WISCVM as the new BITNET/INTERNET gateway?" ------------------------------ Date: 21 Sep 87 23:26:00 EDT From: "ROBAX::KANG" Subject: Latex/GKS/UIS To: "texhax" Reply-To: "ROBAX::KANG" **** This question is sent to info-vax and tex-hax ***** September 21, 1987 Greetings, When writing papers with Latex, the worst part is drawing pictures. Has anyone out there come out a way of drawing pictures with a mouse, as is on a Macintosh), and meanwhile automatically generating a sequence of Latex commands (\put...) that corresponds to the pictures. I think such a package would be welcome and it is certainly a good project for Computer Graphics courses. I am runnig a MicroVaxII Workstaion and have GKS and UIS for graphics. If someone has a good suggestion, please contact me directly. --- Kang Sun Internet: kang%robax.decnet@venus.ycc.yale.edu (prefered) sun@venus.ycc.yale.edu (if the above doesn't work) sun-kang@yale.edu Bitnet: kang%robax.decnet%venus@yalecs.bitnet (prefered) sun-kang@yalecs.bitnet (if the above doesn't work) sun@yalevms.bitnet (for file transfer over JNET) ------------------------------ Date: Tue, 22 Sep 87 11:35 N From: Subject: Re: VMS Tektronix 4010 preview To: texhax@score.stanford.edu In TeXhax 87-76 I placed a message about a DVItoVDU program, contained in the Kellerman & Smith VMS TeX tape. This program can show your page lay-out and character shape on several terminals, the quality of the preview depending on the capabilities of the terminal, with best results on a VT640 (DEC) or an IBM(-compatible) PC with a VT640-emulating program. Since then I have received some questions about where to find this program, how to get the tape etcetera. I strongly urge everyone interested to contact Kellerman & Smith directly, since I do not know wether the DVItoVDU program is public domain or not. If it is, you almost certainly can get it from K&S for free, or at media cost. The address of K&S is: Kellerman & Smith, 534 SW Third Ave Portland, OR 97204 tel 503-222-4234 tlx 9102404397 Herman F. Vogt University of Groningen Computing Centre Earn/Bitnet: VOGT@HGRRUG5 PS Do K&S have access to TeXhax and, if so, what is their e-mail address? HFV ------------------------------ Date: Tue, 22 Sep 87 16:11:33 PDT From: mackay@june.cs.washington.edu (Pierre MacKay) To: TeXhax@SCORE.STANFORD.EDU Subject: gftopxl.web The directories on SCORE show a GFTOPXL.WEB that was last touched on 10-Jun-87, but which still seems not to address the problem raised by William LeFebvre some time ago. I enclose a context diff of the UnixTeX version which has caused no problems since it was Since the last substantive change is dated 1985, before LeFebvre reported the problem, I continue to assume that his code is genuinely needed. To be sure, we are still trying to wean users away from pxl files, but there are a lot of them still out there and much needed. Note that this is not a case for a change file, since it seems to be valid for all uses of gftopxl. Pierre A. MacKay TUG Site Coordinator for Unix-flavored TeX ------------------------------ Date: Tue, 22 Sep 87 16:14:10 PDT From: mackay@june.cs.washington.edu (Pierre MacKay) To: TeXhax@SCORE.STANFORD.EDU Subject: dvitty Can anyone put me in touch with Svante Lindahl, who wrote this. Or alternatively, point me at a source for the source? Pierre A. MacKay TUG Site Coordinator for Unix-flavored TeX ------------------------------ Date: 18-SEP-1987 18:51:01 From: CHAA006 <@NSS.Cs.Ucl.AC.UK,@cs.ucl.ac.uk:CHAA006@vaxb.rhbnc.ac.uk> To: texhax-request Subject: Plain TeX: using scaled fonts avoiding missing half sizes. Every now and them, I fall into the trap of setting a document with global \magnification\magstephalf, and then attempting to use a font (say) scaled \magstep3. The result is invariably a missing-font diagnostic at DVI time (I do not have fonts at *\magstep3-and-a-half). Not wishing to use explicit fonts sizes (after all, the \magstep's are there to avoid just such nasties), could anyone suggest a way of specifying the font such that its size _after_ magnification is (say) \magstep3. The (non-existent) construct that I want would be something like * \font\titlefont = cmcsc10 truescaled \magstep3 or * \font\titlefont = cmcsc10 at 10\magsize3 truept ** Phil. ------------------------------ %%% %%% 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 ************************** -------