TeXhax Digest Wednesday, April 29, 1987 Volume 87 : Issue 31 TEXHAX31.87 Editor: Malcolm Brown Today's Topics: Re: TeX vs. Mac centered line Re: V87 #28: dvitool - anything new? Re: TeX: alignment question ftp of TeX BibTeX wishes and browsing TeX and Mac :-} Re: TeX vs. Mac :-) AMS-TeX for MS-DOS PCs Re: TeXhax Digest V87 #30 shell script to print dvi file on a PS printer Alignment (also alinement) TeX for Italy (TeXhax Digest V87 #29) How to create preloaded versions of TeX and METAFONT. ---------------------------------------------------------------------- Date: 21 Apr 87 13:11 EST From: HALLETT JEFFREY A Subject: Re: TeX vs. Mac To: texhax-request@score.stanford.edu This is in reply to Paul Davis' speculation on the "ultimate" DTP system for the Mac. Tex produces super output, I don't think anyone will argue with that, provided that they never used pictures. I hate to say it, but TeX is extremely unfriendly to writers who use illustrations. Fortunately, many of the Mac TeX packages provide a facility for putting in pictures drawn in MacPaint/Draw, etc or putting in PostScript code (produced by some of the better draw-ers, eg. MacDraw, Cricket Draw). If there were a drawing program that produces TeX, well that would be a different matter, but the less masochistic of us use something like Interleaf or MacDraw, use a /special and then something like dvi2ps or dvi2imp to get pictures. Anyway, that is off the point. TeX falls short of DTP systems in one respect: it uses boxes. To REALLY make it in the Mac world, a system would need to be able to wrap text about ANY arbitrary region. Some of the newer DTP systems like ReadySetGo 3.0 and some new WPs like WriteNow and FullWrite do this, but as Paul pointed out, their text doesn't look as nice as TeX's does. I don't mind using \'s with special control commands; sometimes WYSIWYG can be limited by its own straight-forwardness. WYGIBTWYS (what you get is better than what you see) has its place and TeX is in that group. I will say "yes, TeX will become the ultimate DTP system" if it can be enhanced by putting text in an arbitrary region (eg. a ball of text) and wrap about an arbitrary figure boundary. On top of that, would be a nice mouse driven way to lay out the text/graphic regions without fiddling with \oddsidemargin and other funky things. Whew, sorry about the length. Jeffrey Hallett Software Technology Program General Electric - Corporate Research and Development ------------------------------ Mail-From: BEETON created at 21-Apr-87 21:06:18 Date: Tue 21 Apr 87 21:06:18-PDT From: Barbara Beeton Subject: centered line To: darrell@[128.54.0.204] this is not the most elegant way to solve your problem, but it should give the desired effect. \setbox0=\hbox{long text line to be centered} \newdimen\TestDimen \TestDimen=\hsize \advance\TestDimen by-\wd0 {\leftskip=.5\TestDimen \rightskip=\leftskip \parfillskip=0pt \noindent\unhbox0\break \null\hfill -- right-shifted text\endgraf} since a two-l \hfill is used in the second line, the \parfillskip=0pt is probably redundant, but the \endgraf almost certainly isn't -- you definitely must end this "paragraph" before you leave the local group. other possible methods for handling this involve building the paragraph in a \vbox using \wd0 as the \hsize, then centering the whole thing; however, keeping your entire text in paragraph mode gives better control over baselines of this relative to surrounding text. -- barbara beeton (bnb@xx.lcs.mit.edu) ------------------------------ Date: Tue, 21 Apr 87 20:13:52 PST From: jwm@renoir.Berkeley.EDU (Jeff Mc Carrell) To: texhax@score.stanford.edu Subject: Re: V87 #28: dvitool - anything new? The texgroup at Berkeley has been distributing dvitool version 2.0 for several months now as part of our TeX support software (see TeXhax V87 #22 for a brief description of our dvi->Xerox software). Jean-Francois Lamy (lamy@ai.toronto.edu) writes: > The version of dvitool we have has "questionable features" (it sometimes > omits lines and cannot handle pages bigger than 8.5 by 11 inches). It > sort of fits like a square peg in the current Suntools environment. Since dvitool 1.0 (which I assume Lamy has) was developed in the bygone days of SunWindows 1.0, it isn't surprising that it doesn't fit well in today's much changed Sun 3.2 environment. > We could live with these limitations, but a version that would support PK > or even GF fonts would be most desirable. Inquiries at > vortex-dist@berkeley.edu fell into a black hole. > a) is that the right address? No, the correct address is: dist-vortex@ucbvax.Berkeley.EDU or ...{decvax,ihnp4}!ucbvax!dist-vortex There is some other net site called vortex that mailers get confused over, so we prepended the "dist-" to reduce ambiguity. > b) is there a PK/GF previewer available anywhere? dvitool v. 2.0 supports PK fonts, and arbitrary page sizes (though that code hasn't been tested thoroughly since most of the work we do around here is 8.5 X 11), and page caching, and rebindable keys, and a much cleaner user interface than v. 1.0, and on-line help pages, etc. It compiles under Sun OS version 3.0 or later; it underwent substantial beta-testing before release and seems to be quite stable. The distribution comes with everything needed to install it, including appropriately sized CM fonts. Sending a request note to dist-vortex will get you a description of the software and a release form to be signed. There is a nominal fee for the distribution to defray our costs. Jeff McCarrell dvitool author jwm@Berkeley.EDU ...!ucbvax!jwm ------------------------------ Date: Tue, 21 Apr 87 22:53:37 PST From: To: TeXHAX@score.stanford.edu Subject: Re: TeX: alignment question From: "Don Hosek, Editor, TeXMaG" Subject: Re: TeX: alignment question To: Darrell%beowulf@sdcsvax.ucsd.EDU, TeXHAX@Score.STANFORD.EDU X-VMS-To: IN%"Darrell%beowulf@sdcsvax.ucsd.edu", IN%"TeXHAX@Score.Stanford.EDU" Here's my solution to the problem... I use similar macros for return addresses, &c., in letters and so forth. \def\quotation#1#2{{% An extra group to keep changes to \tabskip local \tabskip=0pt plus 1fill \halign to\hsize{##&\hfil##&##\cr &\cr &\crcr}}} \quotation{``Iyam what Iyam, and that's all what Iyam.''}{---Popeye the Sailor Man} This TeXnique might also be useful to the person who is trying to count the parameters in his list--rather than determine the longest length by hand, leave it all up to halign (what a nice macro). -Don Hosek BITNET: DHOSEK@HMCVAX ARPA: DHOSEK%HMCVAX.BITNET@Forsythe.Stanmford.Edu ------------------------------ Date: Wed, 22 Apr 87 09:34:01 est From: Michael McGuffey To: texhax@SCORE.STANFORD.EDU Subject: ftp of TeX Is there a site on the Arpanet where a current VMS version of TeX, Metafont, LaTeX, etc., source code may be ftp'ed? -mike Michael J McGuffey mcguffey@unc.CSNET decvax!mcnc!unc!mcguffey ------------------------------ From: harrison@mahogany (Michael Harrison) To: TeXhax@score.stanford.edu Subject: BibTeX wishes and browsing Date: Wed, 22 Apr 87 07:33:28 PST We have no solutions for your wishes directly, but there is a related need that is being addressed. The VorTeX emacs macros now have a new feature which allows you to focus on a \cite{foo} in a source files and displays the corresponding reference from some .bib file. (This is an ``smart operation'' in that all sorts of obvious questions are figured out by the macros. (This was programmed by Pehong Chen.) Another step is to merge browsing into our previewers, dvi2x and dvitool, so that selecting a reference allows a display of the (formatted) bibliography. (This is not implemented yet.) ------------------------------ Date: Wed, 22 Apr 87 08:54:17 PST From: To: texhax@score.stanford.edu Subject: TeX and Mac :-} Date: Wed, 22 Apr 87 10:55 EDT From: Subject: TeX and Mac :-} To: texhax@score.stanford.edu X-Original-To: texhax@score.stanford.edu, RICH A shot from the hip in response to Glenn Vanderburg: The quote from Lamport you cite against visual interfaces is slightly slanderous. There is no reason in the world why format changes in visual interfaces must be laborious -- though, often, in current implementations, they are. Consider Microsoft's Word 3.0 for the Mac, though, as a hint of how to do things better. 'Style sheets' can be used to specify the formats of various portions of a document. Later reformatting can be done simply by making a change to the style sheet. I don't think Word does automatic equation numbers, but if it did, it probably would provide you with a style sheet in which you could specify Arabic versus Roman numerals. (ROMAN equation numbers? Talk about stupid typography!) "Ah," you tell me, "but that's the point. There ISN'T a style sheet for equations. Style sheets give you a very limited set of options, and you're stuck with that. In TeX you can automate all sorts of things the designer of the system never thought of, because he's given you access to everything." OK, but I never said the Word interface was a panacea. I can imagine a system in which not only styles are definable, via style sheets, but style sheets themselves are definable: you can put in options like Roman equation numbers, sinusoidal left margins, and dangerous bend signs. Why not? I guess what I'm saying is, the alleged dichotomy between visual design systems and logical design systems is either, depending on how you define each, nonexistent or irrelevant. Irrelevant because, if I understand Paul Davis's question, what is wanted is not a different way to approach the document design -- visual as opposed to logical -- but a different way to tell the computer about the logical design one has chosen. One that doesn't require you to keep track of a huge vocabulary of control sequence keywords. One that lets you see the effect of a change NOW, not ten minutes from now. One that doesn't "mode you in" by making you laboriously enter and exit an editor, a processor, a DVI converter, back to the editor, and so on (and on and on). One that lets you SEE the logical organization of your document better than TeX does. (OK, here's the left bracket -- now where the $~&*@# did I put the matching right bracket? Assuming I did). And one that lets you use a different interface to the same system if you hate mice and windows. If logical design encourages more sound typography and better writing, and is more flexible, than visual design, then a transparent, friendly, probably visual INTERFACE approach to logical DESIGN should do the same in spades. I use TeX in spite of its unfriendliness, because it does what I want better than anything else I know of. Don't tell me "if it's easy, it's no good". You don't get poor results by doing things the easy way; you get poor results by doing things the wrong way, no matter how much or how little effort you expend. You get good results by doing things the right way -- and if you can make the right way easier, people will be more likely to do things right. THAT was the idea behind TeX; it's made typesetting easier; if it can be taken another step further, it should be. End flame. And so it goes. Rich Holmes Bit: Rich@SUHEP ------------------------------ From: riedl@purdue.edu To: TeXhax digest Subject: Re: TeX vs. Mac :-) In-Reply-To: Your message of Mon, 20 Apr 87 22:07:53 CDT. Date: Wed, 22 Apr 87 11:16:44 EST Glenn Vanderburg states the reasons that most of us prefer TeX to available WYSIWYG systems. The strongest argument is the lack of flexibility in these systems which makes it difficult to maintain a consistent style throughout a document or to change the style of a completed document. The fundamental criticism, then, is that existing WYSIWYG systems are not extensible whereas LaTeX is. But it is no contradiction for a WYSIWYG system to have an extension language that is as powerful as the TeX macro system, along with a means of associating textual units with styles defined in this extension language. Remember, the builders of WYSIWYG systems have never promised that "what you see is all you've got"; that's a statement from the critics. The essential WYSIWYG characteristic is that the system can perform the formatting almost as fast as a user can type and can display the formatted output on a write-many medium (e.g. terminal). "What you see is all you've got" is a valid criticism of current systems, but let's not restrict our imagination to what works today. John Riedl {ucbvax,decvax,hplabs}!purdue!riedl -or- riedl@mordred.cs.purdue.edu ------------------------------ Date: Wed, 22 Apr 87 09:26:14 PST From: To: texhax@score.stanford.edu Subject: AMS-TeX for MS-DOS PCs Date: Wed, 22 Apr 87 09:27 PDT From: David J. Buerger, Santa Clara Univ. P.C. Ce... Subject: AMS-TeX for MS-DOS PCs To: texhax@score.stanford.edu X-Original-To: texhax@score.stanford.edu Can anyone tell me where I can obtain AMS-TeX for MS-DOS PCs? I understand that Addison-Wesley is hoping to come out with such software in the future. Does anyone sell or share AMSTeX for the PC today? David J. Buerger P.C. Center, Santa Clara University dbuerger@SCU.BITNET ------------------------------ Date: Wed, 22 Apr 87 09:55:28 PST From: Lynn Ruggles Subject: Re: TeXhax Digest V87 #30 To: TeXhax@score.stanford.edu Although I agree with what you said in your comments about TeX, you have got one detail wrong. The reason Tex was written was that professional typesetters were doing things the *hard* way (and getting it right), but the process was very time consuming and expensive. Knuth decided to write Tex because he thought 1) it was possible and 2) it would be easier to get a computer to do the nitty gritty positioning. Without TeX, the second edition of his Volume 2 would have cost about ------------------------------ Date: Wed, 22 Apr 87 12:13:12 PST From: john@renoir.Berkeley.EDU (John Coker) To: texhax@score.stanford.edu Subject: shell script to print dvi file on a PS printer I have a relatively simple shell script below to convert and spool .dvi files to a PostScript printer using the dvi2ps converter. The reasons this approach is sometimes preferable to lpr -d is when one includes PostScript illustrations (with \special) or when wants to print multiple copies. This script causes multiple copies of the document to be printed, but only converted once (as opposed to lpr's -# option). Unfortunately, each page is printed several times and needs to be collated by hand. This is the -# option to dps, where is the number of times to print each page. No doubt, you'll have to change some of the system dependencies in this script. In particular, the name of the prologue file, the default printer name (as known to lpr) and perhaps the converter name. The specific dvi converter should not matter, as long as it doesn't output the prologue itself (which is a bad idea anyway). In that case, the script will need to be changed to emit a ``%1'' before the ``/#copies n def'' is included in the PostScript file going to the printer. John #!/bin/sh # # Print a .dvi file using the dvi2ps command directly. # # Since we do the conversion directly here, we don't # need to worry about the path names for included # PostScript files or other such nonsense, also we # get errors reported. # # (c) 1987 John Coker # University of California, Berkeley # john@renoir.Berkeley.EDU # LIBDIR=/usr/local/lib/ps PROFILE="$LIBDIR/psdvi.pro" PRINTER=${PRINTER-gp} export PRINTER PATH=/bin:/usr/bin:/usr/ucb:/usr/local export PATH USAGE="usage: `basename $0` [ -# ] [ -Pxx ] [ dvifile ]" trap 'rm -f $tmpfile $tmpinpt ; exit 1' 0 1 2 15 tmpinpt="/tmp/TP$$.dvi" tmpfile="/tmp/TP$$.ps" alist="-l" input= repeat= while test $# -gt 0; do case $1 in -#[1-9]*) # repeat count for printing repeat="`echo $1 | sed 's/^-#\([0-9]*\).*$/\1/'`" ;; -d) # throw away -d options; we know that already ;; -*) # save lpr options for later alist="$alist $1" ;; *) # this must be a file name to convert if test ! -z "$input"; then echo $USAGE >&2 exit 1 fi input="$1" ;; esac shift done if test -z "$input"; then input=$tmpinpt cat - > $input fi if test ! -r $input; then if test ! -r "$input.dvi"; then echo "Can't find $input or $input.dvi to print!" >&2 exit 1 fi input="$input.dvi" fi echo -n "Converting DVI output..." >&2 dvi2ps $input > $tmpfile if test $? -ne 0; then echo "DVI conversion failed; aborting!" >&2 exit 1 fi if test -z "$repeat" -o "$repeat" -le 1; then echo -n "Printing DVI output..." >&2 cat $PROFILE $tmpfile | lpr $alist status=$? else echo -n "Printing DVI output ($repeat times)..." >&2 echo "/#copies $repeat def" | cat $PROFILE - $tmpfile | lpr $alist status=$? fi if test $status -ne 0; then echo "DVI printing failed; aborting!" >&2 exit 1 fi echo " done." >&2 exit 0 ------------------------------ Date: Wed, 22 Apr 87 13:51:05 PST From: To: texhax@score.stanford.edu Subject: Alignment (also alinement) Darrell Long writes (Digest V87 #30) > I have a question about alignment. I want to have a centered line > with a right-justified line below it whose right boundry is that of the > centered line. Make sense? Here's an Example: > > Tout est pour le mieux, dans le meilleur des mondes possibles. > --- Dr. Pangloss (Voltaire) Here are just three ways to do just that. \quotethree is probably the most flexible way among the three. As seen in last example. \def\quoteone #1#2{\centerline{\vbox{\hbox{#1}\hbox{\phantom{#1}% \llap{#2}}}}} \def\quotetwo #1#2{\centerline{\vbox{% \halign{##&##&##\cr&\cr&\hfil #2&\cr}}}} \long\def\quotethree #1{\hbox to\hsize{\hfill\vbox{% \halign{##&##\cr#1\crcr}}\hfill}} \quoteone{% Tout est pour le mieux, dans le meilleur des mondes possibles.% }{--- Dr. Pangloss (Voltaire)} Tout est pour le mieux, dans le meilleur des mondes possibles.% }{--- Dr. Pangloss (Voltaire)} \bigskip \quotethree{% Tout est pour le mieux, dans le meilleur des mondes possibles.\cr \hfil --- Dr. Pangloss (Voltaire)} \bigskip \quotethree{% \vbox{\hsize=20em\parskip=0pt\parindent=0pt Tout est pour le mieux, dans le meilleur des mondes possibles. Tout est pour le mieux, dans le meilleur des mondes possibles. Tout est pour le mieux, dans le meilleur des mondes possibles. Tout est pour le mieux, dans le meilleur des mondes possibles. Tout est pour le mieux, dans le meilleur des mondes possibles. \par \rightline{--- Dr. Pangloss (Voltaire)}}} Best regards, Medhi Widjaja ------------------------------ Date: Wed, 22 Apr 87 17:29:22 PDT From: mackay@june.cs.washington.edu (Pierre MacKay) Subject: TeX for Italy (TeXhax Digest V87 #29) The UnixTeX distribution has been sent to every country that has asked for it, without restrictions. The only difficulty with some countries (Italy is not alone) is that the tape may go into customs and the recipient not be informed. I had one tape returned to me after ten months sitting in customs, and the recipient had never been informed. Any shipment of Free Public Domain software should be sent out with large indications that it is just that, and that it has no commercial value. The recipient should be told by separate mailing that it is on the way. I do not know whether this is really necessary, but several European countries have asked me to wrap every tape in heavy anti-magnet shielding. I suspect the problem is more with the fields created by the door-closing motors in the baggage hatch than with the necessary postal procedures. Pierre A. MacKay TUG Site Coordinator for Unix-flavored TeX ------------------------------ Date: Wed, 22 Apr 87 14:53:14 PST From: To: texhax@score.stanford.edu Subject: How to create preloaded versions of TeX and METAFONT. Date: Wed, 22 Apr 87 16:51 CST From: Subject: How to create preloaded versions of TeX and METAFONT. To: texhax@score.stanford.edu X-Original-To: texhax@score.stanford.edu, BELL We have (finally) V2.0 of TeX and V1.0 of METAFONT. I am currently in the process of getting these running and generating enough of the pixel files for our LN03 to please people. I am confused about one thing, however. The versions we have are from Stanford (the CH files are dated (internally) May 1985 and July 1985, respectively, although the dates (per VMS) on the files are more recent(c. June 1986)). The problem I am having is how do you go about preloading a format/base? The versions we have say, This is METAFONT, Vax/VMS Version 1.0 (no base preloaded) and similarly for TeX. We have had the Kellerman & Smith implementation in the past which has on it a program called VirTeX which loads a format and creates a new executable which contains the format preloaded. Is there any trick to doing this? Obviously one can say ``&plain'' for plain TeX and METAFONT, and ``&lplain'' for LaTeX, but I would prefer preloaded formats of the programs which match the documentation in The TeXbook and The METAFONTbook. Can someone help? Ed Bell Dept. of Physics and Astronomy University of Kansas BITnet: BELL@UKANVAX SPAN: KUPHSX::BELL (7.220) ------------------------------ Date: Thu, 23 Apr 87 10:01:53 GMT From: CMI011%UK.AC.SOTON.IBM@ac.uk To: TEXHAX@score.stanford.edu in TeXhax 87.27, there are three notes about quoting program fragments in TeX documents. May I describe my solution to the problems when using LaTeX? We have three methods of program quote: a) a `proquote' environment that is a `quote' with tabbing set, and 8 predefined tabs, the whole thing to be in sans-serif. You have to do the rest (escaping special characters etc) yourself b) ^program and ^endprogram macros which set up a verbatim environment with everything in ^tt etc. This is for first drafts and when you are in a hurry c) the ^program macro actually has a parameter of language type (Pascal, C, Prolog etc), and you can utilize this by running the LaTeX source through a pre-processor which pipes the program quotes through a modified version of tgrind. If anyone doesnt have the latter, it puts keywords in bold, comments in italic, numbers lines etc etc, reading a file which contains information about the different programming languages (its on the Unix TeX tape) The advantage of this system is that it leaves pure readable code in your LaTeX source, which can be changed, read in from genuine program etc, and lets you have a simple version for when you cannot be bothered with tgrind. I would be interested to hear other people's reactions to the idea of pre-processors for TeX documents (life is too short to do it all in TeX.....) As for the question about a Lisp font, the solution adopted here by the Lisp guru is to use a special program in place of tgrind above, which gets the tabbing exactly right using a proportional font (Palatino is what is used by preference). Its quite pleasing aesthetically. anyone who wants the relevant programs, macros, let me know. they are not very difficult to write from scratch anyway - I recommend the Lisp one though. sebastian rahtz computer science, university, southampton, uk cmi011@uk.ac.soton.ibm spqr@uk.ac.soton.cm ------------------------------ %% %%\bye %% End of TeXhax Digest **************************