[Q] Knuth on LATeX vs. PlainTeX ??
Heinz-Dieter Ecker
Since more than 10 years, I have been using only (modified) PlainTeX
for output, though -- unfortunately, I must say -- files of different
LATeX dialects as inputs. (I have $=\hbox{am}$ a one-person service
bureau producing books and scientific journals.)
This should be enough as a small hint to my not being very amused
about certain developments in the TeX world seen as a whole. Of
course, this is only MHO (and should and will not be of any influence
on these developments).
**** But --: can anyone give me (hints to) facts [published or not] on
Knuth's opinions towards the "LATeXization" of the TeX world, towards
the LATeX3 project, towards NTS? ****
Thanks in advance,
hde
--
HD [Heinz-Dieter] Ecker / Bonn [NOT Berlin], Germany
Ecker....@T-online.de / 022846...@t-online.de /
David Kastrup
Since more than 10 years, I have been using only (modified) PlainTeX
for output, though -- unfortunately, I must say -- files of different
LATeX dialects as inputs. (I have $=\hbox{am}$ a one-person service
bureau producing books and scientific journals.)
This should be enough as a small hint to my not being very amused
about certain developments in the TeX world seen as a whole. Of
course, this is only MHO (and should and will not be of any influence
on these developments).
LaTeX has been unifying work in the last time. While still suppoerting
old LaTeX2.09 documents, a lot of features where AMSTeX had been
necessary is provided now by AMSLaTeX. LaTeX itself has developed
immensely. This means, for example, that it is dead easy to refer to
Postscript fonts in various encodings in your document, that you can
write using your favourite input encoding instead of plain ASCII and
still have portable documents, that you can switch between various
font encodings (suitable for your typical languages to be processed)
even within one document and yet have an identical input syntax.
You simply cannot compare this with plain TeX where you are forced to
basically use one style file written by yourself for every document
you prepare, and have a lot of work getting things to work.
LaTeX might not be perfect, but it *is* the most active point of
user-oriented TeX development, and a *lot* of problems have been
addressed already when you are using it.
And this is from one who knows TeX intimately enough to get it to
produce a "This can't happen" error message with two lines of code:
\def\x#1{\if#1m\span\expandafter\x\fi}
\halign{&#\cr\expandafter\x\romannumeral260001 \cr}
Admitted, this is cheap.
--
David Kastrup Institut fuer Neuroinformatik, Ruhr-Universitaet Bochum
Email: d...@neuroinformatik.ruhr-uni-bochum.de Telephon: +49-234-700-5570
Robin Fairbairns
In article <4u91lr$a...@muller.loria.fr>,
Denis B. Roegel <roe...@loria.fr> wrote:
>In article <4u374r$g...@news00.btx.dtag.de>, Ecker....@t-online.de
>(Heinz-Dieter Ecker) writes:
>|> **** But --: can anyone give me (hints to) facts [published or not] on
>|> Knuth's opinions towards the "LATeXization" of the TeX world, towards
>|> the LATeX3 project, towards NTS? ****
>
>around TeX. I am sure he doesn't object to the use of LaTeX by many people.
>At TUG'95, he did even sign a copy of the LaTeX book, and this was a scoop!
Indeed so. We had a Q&A session at TUG'95, and (once we'd got him off
the business of TAoCP) he said things like
"I don't use LaTeX because I don't like big systems"
(cheers from the plain-TeXers at the back of the hall)
"I don't understand why so few people modify TeX for their own
purposes"
"I really couldn't say whether I prefer the Omega or the e-TeX/NTS
approach to the future"
(what would you *expect* him to say, given that there were papers from
from both camps at the meeting?)
The one topic on which there was very little meeting of minds was
semantic markup. As far as one could determine, Knuth thinks very
little of the idea, and many people there (involved in such things as
alternative presentation strategies) were all for it.
If you don't understand *why* semantic markup is an issue, I suggest
you seek out a copy of TV Raman's paper in TUGboat 16#3. Raman (who
is blind) remarks "one of the great things about TeX is that it can
produce predictably acceptable results in mathematics; it's hard to
understand why it's not taught to blind schoolchildren". However, if
the said schoolchildren are to read each other's work, they need
something like Raman's system, and _that_ relies on semantic markup.
--
Robin (the man with no voice) Fairbairns r...@cl.cam.ac.uk
U of Cambridge Computer Lab, Pembroke St, Cambridge CB2 3QG, UK
Home page: http://www.cl.cam.ac.uk/users/rf/robin.html
David Kastrup
In article <4u9tpk$j...@muller.loria.fr> roe...@loria.fr (Denis
B. Roegel) writes:
In article <m2bugnf...@wun.neuroinformatik.ruhr-uni-bochum.de>,
d...@neuroinformatik.ruhr-uni-bochum.de (David Kastrup) writes: is
from one who knows TeX intimately enough to get it to |> produce a
"This can't happen" error message with two lines of code: |>
\def\x#1{\if#1m\span\expandafter\x\fi} |>
\halign{&#\cr\expandafter\x\romannumeral260001 \cr} |> |> Admitted,
this is cheap.
Yes, but isn't this a bug in TeX ? Shouldn't TeX say something like
!Too many \span's
and not
!This can't happen
?
Perhaps just read in TeX, the program, what Knuth has to say about
it. He thinks that such excessive spans have about possibility
nil.
That's the main reasoning. The problem is that the data structures TeX
uses do not really support larger spans (if I remember
correctly). Strictly speaking, this might more properly cause a
TeX capacity exceeded
although there is no way you could ask a wizard to extend TeX in that
area (unless he is very wize).
If you feel that "This can't happen" should not be triggerable by a
simple program, however contrived, you just *might* get my vote for
it...
Michael Downes
Ecker....@t-online.de (Heinz-Dieter Ecker) writes:
Since more than 10 years, I have been using only (modified) PlainTeX
for output, though -- unfortunately, I must say -- files of different
LATeX dialects as inputs.
...
**** But --: can anyone give me (hints to) facts [published or not] on
Knuth's opinions towards the "LATeXization" of the TeX world, towards
the LATeX3 project, towards NTS? ****
In gkpmac.tex, the macros Knuth used for writing Concrete Mathematics,
he borrowed some of the LaTeX code for the `picture' environment.
I think Knuth did a fine job writing TeX the program but it doesn't
follow that his opinions about the way to write documents have any
extraordinary authority. He appears to like the idea of the
TeXnician/author: that the person doing the writing of a book or
whatever is proficient enough in TeX to produce custom macros for
whatever the material needs. He writes a custom TeX format on top of
plain TeX for just about every book that he does. Of course for such a
person it is easy to get the desired results for anything they write,
but for most authors who don't want to become TeX wizards in order to
produce their thesis or book, Knuth is not the best model to follow.
Knuth also appears to like using macros with delimited arguments to
facilitate writing and to make the TeX file look more `natural' on
screen. While this is pleasant and fun, it makes extra work later if you
want to produce, let's say, an HTML version of what you are writing, or
some other kind of conversion. A more consistent and bland syntax (like
LaTeX) makes general-use tools like latex2html usable for everyone who
follows the same syntax.
Some of the things Knuth does via TeX macros to save keystrokes are
probably better done using the capabilities of a smart editor. (No
single smart editor, not even emacs, can match the cross-platform
availability and uniformity of TeX, but the point is that using a hammer
to chop down a tree is not a good idea unless you really don't have an
axe nor any prospects of getting one.)
As a writer, I started out with plain TeX and loved it (The TeXbook
holds up remarkably well under many readings, I find) but after a while
I got tired of reinventing the wheel and switched over to LaTeX. I would
do things like write my own verbatim macros, then next week find that I
had overlooked a problem A (spaces at the beginning of a line, let's
say), and fix it, then next week find that I had overlooked another
problem B (?` and !` ligatures, let's say), and fix it, and so on, and
eventually I realized that all these nagging little problems I kept
stumbling over were ones that had already been dealt with by the LaTeX
verbatim environment.
Not that reinventing the wheel isn't a pleasant pastime, but it begins
to wear on you when you're trying to get your thesis printed for a
deadline and you suddenly discover that your clever verbatim macros have
a subtle bug that takes four hours to hunt down and fix.
Michael Downes
m...@ams.org
Richard J. Kinch
Denis B. Roegel (roe...@loria.fr) wrote:
: Maybe the time of Metafont and font development
: is yet to come, for instance with Omega.
It would appear that Yannis Haralambous himself has even retreated into using
Fontographer for his Omega Times and Omega Helvetica font design work.
Michael C. Grant
**** But --: can anyone give me (hints to) facts [published or not] on
Knuth's opinions towards the "LATeXization" of the TeX world, towards
the LATeX3 project, towards NTS? ****
If you're attempting to suggest that his opinions should significantly
affect the current course of LaTeX/e-TeX/NTS development, then I'd
have to disagree. To use an analogy which I hope connotes the proper
amount of respect: if physicists didn't reject Einstein's opinion that
"God doesn't play dice", out of some distorted form of respect, then
we wouldn't have quantum mechanics.
So if Knuth objects to LaTeX3, e-TeX, or NTS, then we should listen
respectfully but make up our own minds.
--
Michael C. Grant Information Systems Laboratory, Stanford University
mcg...@isl.stanford.edu <A HREF="http://www-isl.stanford.edu/~mcgrant">
------------------------------------------------------------------------------
"When you get right down to it, your "Long hair, short hair---what's
average pervert is really quite the difference once the head's
thoughtful." (David Letterman) blowed off?" (Nat'l Lampoon)
Steve Grathwohl
In article <4ua7i9$m...@lyra.csx.cam.ac.uk>, r...@cl.cam.ac.uk (Robin
Fairbairns) wrote:
> In article <4u91lr$a...@muller.loria.fr>,
> Denis B. Roegel <roe...@loria.fr> wrote:
> >In article <4u374r$g...@news00.btx.dtag.de>, Ecker....@t-online.de
> >(Heinz-Dieter Ecker) writes:
> >|> **** But --: can anyone give me (hints to) facts [published or not] on
> >|> Knuth's opinions towards the "LATeXization" of the TeX world, towards
> >|> the LATeX3 project, towards NTS? ****
> >
> >As far as I know, Knuth fosters all kinds of research and development
> >around TeX. I am sure he doesn't object to the use of LaTeX by many people.
> >At TUG'95, he did even sign a copy of the LaTeX book, and this was a scoop!
>
> Indeed so. We had a Q&A session at TUG'95, and (once we'd got him off
> the business of TAoCP) he said things like
>
> "I don't use LaTeX because I don't like big systems"
I think the question was "Why don't you use LaTeX", and Knuth's answer was
"I'm scared of big systems"
with a bit of intended humor. At least, that's my recollection.
> Robin (the man with no voice) Fairbairns r...@cl.cam.ac.uk
> U of Cambridge Computer Lab, Pembroke St, Cambridge CB2 3QG, UK
> Home page: http://www.cl.cam.ac.uk/users/rf/robin.html
OK, why are you the man with no voice?
--
Steve Grathwohl
Duke University Press
gr...@math.duke.edu
Donald A. Hosek
In article <4ua9el$b...@tesla.netline.net>,
Given that YH was really the only person doing serious work with MF,
it would appear that we could probably declare MF to have died as a
serious type development tool.
As I've pointed out earlier, it's not that the type design community
has never tried MF. They have, and the roles of MF-experimenters
include many of the most prominent names in the type world including
Fred Brady, Matthew Carter, David Siegel, Sumner Stone and Carol
Twombly. The final verdict was that it did not work the way that they
needed it to work: Visually and directly.
-dh
--
Don Hosek dho...@quixote.com Quixote Digital Typography
909-621-1291 fax: 909-625-1342 orders: 800-810-3311
For information about SERIF: THE MAGAZINE OF TYPE & TYPOGRAPHY,
http://www.quixote.com/serif/ or mail serif...@quixote.com
Donald Arseneau
In article <m2bugnf...@wun.neuroinformatik.ruhr-uni-bochum.de>, d...@neuroinformatik.ruhr-uni-bochum.de (David Kastrup) writes...
>produce a "This can't happen" error message with two lines of code:
I'll have the \span\span\span\span\span\span\span\span\span and \span.
Donald Arseneau as...@reg.triumf.ca
David Kastrup
In article <m2bugnf...@wun.neuroinformatik.ruhr-uni-bochum.de>, d...@neuroinformatik.ruhr-uni-bochum.de (David Kastrup) writes...
>produce a "This can't happen" error message with two lines of code:
I'll have the \span\span\span\span\span\span\span\span\span and \span.
Sorry, Sir, but this can't happen. Mis\haligned requests like that
are put straight into t\hbox.
Sorry for the lousy joke,
Heinz-Dieter Ecker
[Denis B. Roegel, Robin Fairbairns, Steve Grathwohl, David Kastrup,
Denis, David, Donald Arseneau, Michael Downes, Michael C. Grant]
Let me, please, try to emphasize my main point again.
I am no author [though I could have become one :) ], but a member of
the (hopefully not too) "opposite camp," namely someone working on
other people's (LA)TeX source files to produce books and journals.
(Not too many, because it's only a one-man-show.) For my output, I use
only PlainTeX. On the input side, there are Word-and-whatsoever-files
(many), PlainTeX files (few), and LATeX files (many).
In the publishing world, the concept of "camera-ready production" is
not the only possible way. And --- without arrogance --- I maintain
that any book/journal I produced, also those made out of LATeX files
which seemed to be "camera-ready," has been improved through my work.
Camera-readiness is ok as a concept, of course --- and be it only out
of reasons to publish scientific results as quick and inexpensive as
possible. But there is more than that.
In my eyes, Knuth's original concept, resulting in TeX, plain.tex, and
the TeXBook, is a better concept to meet both ends:
-- to give authors the possibility to produce either "beautiful"
manuscript (and data source file) or "beautiful" camera-ready printed
pages,
*AND*
-- to give publishers (or service bureaus) an open programming system
to produce "beautiful" books etc.
LATeX seems to overemphasize the camera-ready aspect. You'll say,
"Well, it's not our problem how our files can be used in the
publishing business --- our final goal is control over the whole
process 'from brain to printed, or HTML ..., output'." And you could
say, people out of the publishing business speaking against camera
readiness merely are scared to lose their jobs. This might well occur,
but that's another topic.
LATeX brought ACCEPTANCE to TeX, and user-friendliness. But, IMHO, it
did not bring process-friendliness in the non-camera-ready publishing
world. And authors should be interested in the *whole* process.
LATeX's many dialects and included sub-files make things --- often
unnecessarily --- complicated, I think.
Authors not always are able to produce any camera-ready output
publishers would like them to deliver. (Apart from this: accepting
today's possibilities to bring thoughts into a formal order does not
necessarily have to mean that authors should attach more importance to
these formal aspects than to the contents of their thoughts.)
Couldn't you agree that things would be much easier with some sort of
TeX file (with correct mathematics), marked-up in one way or another,
but as "unformatted", simple as possible?
My question for Knuth's opinion on this was more or less a question of
interest. The given "citations" from TUG '95 seem to fit to my
impressions of what Knuth (also) had in mind when writing TeX. In my
opinion, his original concept was as "beautiful" as TeX itself.
But: things run as they run (certainly, there is a corresponding
American/English saying unknown to me), and perhaps my thoughts go
into a totally wrong direction.
Two additional catchwords:
-- Do you understand that I personally like von Bechtolsheim's
approach very much?
-- What do you think of "PlainTex and the Art of Text Maintenance"?
And, at last, a question for LATeX gurus: Is there a way to produce a
LATeX file with all the cross-referencing "expanded"? You remember:
I~re-edit any LATeX file to be acceptable as PlainTeX file, which
means that --- apart from re-writing all the arrays aso --- I have to
replace all the "equation~(\ref ...)" by "equation~(17)" or so.
Thanks for listening.
Sebastian Rahtz
-- to give publishers (or service bureaus) an open programming system
to produce "beautiful" books etc.
LATeX seems to overemphasize the camera-ready aspect. You'll say,
"Well, it's not our problem how our files can be used in the
publishing business --- our final goal is control over the whole
process 'from brain to printed, or HTML ..., output'." And you could
say, people out of the publishing business speaking against camera
readiness merely are scared to lose their jobs. This might well occur,
I don't buy your arguments. I work on authors' (La)TeX files for a
publisher, not on whole books, but relatively short articles. If I
had to do a 500 page book, maybe i would have the time to spend
massaging a new format each time, but for a 10 page article which
needs to be turned around in a few hours, LaTeX is *vital*. it is the
only system which allows a trivial change to switch the document style
to that needed for the job. The author can concentrate on the markup,
I can concentrate on the typesetting style, and we communicate by
means of common markup. The point of LaTeX is not the actual styles
that come with it, but the agreement on a markup scheme which is rich
enough for real-life material.
Don Knuth writes books, long complicated ones which take years. His
approach is perfect for that. Others of us use his excellent tool,
TeX, in a different way for different purposes. There isnt any conflict.
Sebastian
David Carlisle
In article <4ubctt$h...@news00.btx.dtag.de>
Ecker....@t-online.de (Heinz-Dieter Ecker) writes:
[Denis B. Roegel, Robin Fairbairns, Steve Grathwohl, David Kastrup,
Denis, David, Donald Arseneau, Michael Downes, Michael C. Grant]
Let me, please, try to emphasize my main point again.
I am no author [though I could have become one :) ], but a member of
the (hopefully not too) "opposite camp," namely someone working on
other people's (LA)TeX source files to produce books and journals.
(Not too many, because it's only a one-man-show.) For my output, I use
only PlainTeX. On the input side, there are Word-and-whatsoever-files
(many), PlainTeX files (few), and LATeX files (many).
In the publishing world, the concept of "camera-ready production" is
not the only possible way. And --- without arrogance --- I maintain
that any book/journal I produced, also those made out of LATeX files
which seemed to be "camera-ready," has been improved through my work.
Camera-readiness is ok as a concept, of course --- and be it only out
of reasons to publish scientific results as quick and inexpensive as
possible. But there is more than that.
In my eyes, Knuth's original concept, resulting in TeX, plain.tex, and
the TeXBook, is a better concept to meet both ends:
-- to give authors the possibility to produce either "beautiful"
manuscript (and data source file) or "beautiful" camera-ready printed
pages,
*AND*
-- to give publishers (or service bureaus) an open programming system
to produce "beautiful" books etc.
LATeX seems to overemphasize the camera-ready aspect.
From this point on I have difficulty following your argument.
The main reason point about the design of LaTeX is that it provides
a means of marking up the logical nature of the document *without*
regard to its visual appearance. In this respect it has the many of
the same aims as SGML but unlike a pure SGML system it also allows
direct access to the formatter (TeX) so of course it is possible to
produce `LaTeX' documents that consist mainly of explicit space and
font change commands. Perhaps you have the misfortune to receive
these type of documents, which has coloured your impression of LaTeX?
You'll say,
"Well, it's not our problem how our files can be used in the
publishing business --- our final goal is control over the whole
process 'from brain to printed, or HTML ..., output'." And you could
say, people out of the publishing business speaking against camera
readiness merely are scared to lose their jobs. This might well occur,
but that's another topic.
LATeX brought ACCEPTANCE to TeX, and user-friendliness. But, IMHO, it
did not bring process-friendliness in the non-camera-ready publishing
world. And authors should be interested in the *whole* process.
Although one aspect of LaTeX is to make it more `user friendly' that
is actually more of a side effect. The main point is to produce a
standard set of markup commands for documents. (Something plain TeX
does not attempt). ie it is specifically aimed at you!
LATeX's many dialects
There have only ever been 3. Latex 2.09, latex 2.09 +NFSS, and
LaTeX2e. The first two are obsolete, the second one was never
`standard'.
and included sub-files make things --- often
unnecessarily --- complicated, I think.
An necessary evil I'd say but there is always a balance between
increased complexity of the system, and increased functionality, so
different people will draw the line in different places.
Authors not always are able to produce any camera-ready output
publishers would like them to deliver.
Do they ever:-)
(Apart from this: accepting
today's possibilities to bring thoughts into a formal order does not
necessarily have to mean that authors should attach more importance to
these formal aspects than to the contents of their thoughts.)
Couldn't you agree that things would be much easier with some sort of
TeX file (with correct mathematics), marked-up in one way or another,
but as "unformatted", simple as possible?
My question for Knuth's opinion on this was more or less a question of
interest. The given "citations" from TUG '95 seem to fit to my
impressions of what Knuth (also) had in mind when writing TeX. In my
opinion, his original concept was as "beautiful" as TeX itself.
The TeX book makes it perfectly clear that you should not be using
`plain TeX' to make documents. The intention is that you create a
custom format (which you may, with todays faster machines just \input
rather than making a new .fmt file) that defines the markup suitable
for the document to hand.
This is fine if a) you have the knowledge to create such markup
commands, and b) you don't need to send the thing electronically to
anyone else. For many authors though this is not an option.
If you like (if you put constraints on your authors to stick to, say,
the LaTeX book) you can view LaTeX as a descriptive **language** and
ignore the implementation. Thus in a LaTeX document you know sections
and subsections are marked with \section and \subsection. So you can
implement those any way you want. (Most people find it convenient to
use the LaTeX commands for defining section headings, but you don't
have to: you can go \def\section{..any TeX code..} if you know what you
are doing. An extreme example of this (if my memory is correct) is one
of the journals that accept revtex (LaTeX + revtex style)
articles. they do not even use TeX but just take the markup and
convert it to some proprietary typesetter codes.
If you accept `plain TeX' then you have *no idea* how the document
will be marked up, it may be \sect or \section or \S or whatever.
This means if you want to change the markup you have to read and
understand the authors macro definitions. So it is *much harder*
for a publishing house to accept `plain TeX' input than LaTeX.
To overcome this, many journals that accept plain based submissions
essentially define their own set of markup commands, this is OK for
the journal, but a pain for the author as every journal uses an
essentially different markup language so moving documents from one
form to the other is a pain. LaTeX offeres the hope of a standard set
of conventions for document markup which makes document portability a
possibility.
And, at last, a question for LATeX gurus: Is there a way to produce a
LATeX file with all the cross-referencing "expanded"? You remember:
I~re-edit any LATeX file to be acceptable as PlainTeX file, which
means that --- apart from re-writing all the arrays aso --- I have to
replace all the "equation~(\ref ...)" by "equation~(17)" or so.
You could do this in TeX but I'd not bother. Just run the document
through LaTeX, then all the expansions are in the aux file. a line
\newlabel{Sec:class+packages}{{2}{4}}
means that \ref{Sec:class+packages} is `2' and
\pageref{Sec:class+packages} is `4'
so you can edit the document with sed or emacs or perl or ....
Thanks for listening.
hde
--
HD [Heinz-Dieter] Ecker / Bonn [NOT Berlin], Germany
Ecker....@T-online.de / 022846...@t-online.de /
David
David Kastrup
*AND*
So this is where I want to chime in: plain TeX, IMO, is *not* really
suited for collecting various documents and trying to make one book
with a unified outlook. This is because in plain TeX, you specify how
you want things to look, and not the structure of the text
itself. This means, for example, that itemized lists will use just
that item marks the text author thinks useful. When doing a plain TeX
text, you'll specify very precisely what you want to have as results
(quite often even changing the output routine). Plain TeX is nice if
you are doing your complete book design as well as the writing (what
Knuth does) and rejoice in it.
Typical instructions in plain TeX look like "get water from the tap
into the kettle, put it on the stove, light the stove (pressing the
knob until the fuse locks), wait till the water boils, put a filter on
the thermos, put approx. 1 heaped tsp. of coffee per cup in (plus
one), pour as much water on as to soak the powder, wait 30 seconds,
pour the rest of the boiling water on, turn off the stove". Now while
this is rather guaranteed to give quite good coffee, it is rather
non-portable. Take a kitchen with electric stove, for example.
In LaTeX I say "gimme a cup of coffee". While the quality might be
worse, at least I can fine-tune the coffee-making (for all drinkers),
and not need to change the instructions for every customer. And quite
a few customers know nil about making good coffee.
So as a starting point from lousy document designers but authors, I'd
very much prefer LaTeX source (yes, source!!!!! not camera-ready
copy!!!) so as to be able to press my design decisions onto the text
as easy as possible, of course reworking the sections where the author
goofed.
LATeX brought ACCEPTANCE to TeX, and user-friendliness. But, IMHO, it
did not bring process-friendliness in the non-camera-ready publishing
world. And authors should be interested in the *whole* process.
You must be kidding. I can typically mix dozens of articles into one
document, even with different input encodings and style conventions in
LaTeX with moderate work in LaTeX. This is near to impossible with
plain TeX, unless your document is very plain.
LATeX's many dialects and included sub-files make things --- often
unnecessarily --- complicated, I think.
And plain TeX's lack of anything built in makes merges about
impossible, and usually with non-uniform looking results.
Authors not always are able to produce any camera-ready output
publishers would like them to deliver.
You are mixing things up. Just because it is possible to demand
camera-ready output with LaTeX does not mean you have to do it (or
even, should do it). That people are thinking of this at all is not a
bad thing for LaTeX in itself. As long as you are not deluded into
thinking that a sensible publisher will leave all decisions to his
authors...
My question for Knuth's opinion on this was more or less a question of
interest. The given "citations" from TUG '95 seem to fit to my
impressions of what Knuth (also) had in mind when writing TeX. In my
opinion, his original concept was as "beautiful" as TeX itself.
Knuth is interested in producing artful gems. Of course, he won't
delegate his document design decisions to third parties. He is an
artist.
But I think you are wrong in demanding that all your authors should be
artists as well, and still produce artworks that you can easily
incorporate into one common frame done by yourself.
But: things run as they run (certainly, there is a corresponding
American/English saying unknown to me), and perhaps my thoughts go
into a totally wrong direction.
[...]
And, at last, a question for LATeX gurus: Is there a way to produce a
LATeX file with all the cross-referencing "expanded"? You remember:
I~re-edit any LATeX file to be acceptable as PlainTeX file, which
means that --- apart from re-writing all the arrays aso --- I have to
replace all the "equation~(\ref ...)" by "equation~(17)" or so.
Ouch ouch ouch. Better write an own macro package which will properly
deal with cross references, or one change in text will cause you to do
a lot of ugly editing all over the place (that's the reason this
construct is there in the first place).
You might (while you are at your purist plain approach) want to look
into the eplain and the lollipop formats. One or the other, I believe,
has something like that built in.
Dr E. Buxbaum
Ecker....@t-online.de (Heinz-Dieter Ecker) wrote:
>LATeX brought ACCEPTANCE to TeX, and user-friendliness. But, IMHO, it
>did not bring process-friendliness in the non-camera-ready publishing
>world. And authors should be interested in the *whole* process.
>
>LATeX's many dialects and included sub-files make things --- often
>unnecessarily --- complicated, I think.
My understanding of the point behind LaTeX was that it allows the author to concentrate on the
logic structure of the text, rather than on the appearance. Under LaTeX I simply state that I
want to begin a new chapter, all the tedious bits necessary to do so are then handled
automatically: begin a new page, typeset the title in a larger font, put a distance between
title and following text, create an entry into the table of contents and so on.
As this process is controlled by exchangeble style files, a publisher can adopt a particular
"company" style by writing such a style file. I personally for example use the script styles
rather than the original LaTeX document styles, because they produce nicer output. However, the
structuring commands used in the text are the same, the different outputs are created by
replacing the word "article" in the preamble with "script_s".
In plain TeX the author would handle all the intricacies himself in a non-standardized fashion,
which should make the life of a publisher much harder.
This appears to be the main advantage of LaTeX, appart from the greater user friendliness. How
do you transform plain TeX files from one style to another?
Anselm Lingnau
Im Artikel <4ubctt$h...@news00.btx.dtag.de> schrieb
Heinz-Dieter Ecker <Ecker....@t-online.de>:
> LATeX brought ACCEPTANCE to TeX, and user-friendliness. But, IMHO, it
> did not bring process-friendliness in the non-camera-ready publishing
> world. And authors should be interested in the *whole* process.
>
> LATeX's many dialects and included sub-files make things --- often
> unnecessarily --- complicated, I think.
With LaTeX2e, the situation on the `many dialects' front was greatly
improved. No longer do you need a custom LaTeX just to use Times Roman
fonts! Also, with most of the functionality of AmSTeX being available
in LaTeX, the need for AmSTeX (as a separate `dialect' of TeX input)
has been greatly reduced. From this point of view, the convergence
to LaTeX is certainly a good thing.
> Couldn't you agree that things would be much easier with some sort of
> TeX file (with correct mathematics), marked-up in one way or another,
> but as "unformatted", simple as possible?
Not at all. This is a step backwards towards the time when authors would
submit typewritten manuscripts. The whole *point* of a system like LaTeX
is to keep the service bureau from having to second-guess the author as
to the logical structure of a manuscript. For example, I mark up my
itemized lists via \begin{itemize} ... \item ... \end{itemize} rather
than explicit indentations and bullets *because* I want to make things
easy on the service bureau -- who don't have to turn my custom mark-up
for itemized lists into their custom mark-up. One of the advantages of
LaTeX is that you can walk into any reasonable bookshop and take your
pick from various introductory books, rather than learn a new markup
language for every service bureau you (are forced to?) do business with.
As an author, I would certainly prefer spending my time thinking about
my *content* rather than learning the markup language of the week; with
LaTeX, there is one language that I have learned years ago, so now I can
concentrate on actually producing material that's worth being published!
In this sense, LaTeX with its reasonably standard markup system is
certainly a step into the right direction. Judging from my past
experience as an editor of proceedings and journals, the problem with
many LaTeX files is that many authors insist on peppering the input with
stuff like \\, \vspace{...}, ..., not that they're written in LaTeX in
the first place. What you want, and what you should ask your authors to
provide, is a LaTeX file that contains purely logical markup after the
\begin{document}; any special stuff should be declared in the preamble.
Such a manuscript gives a designer maximum leeway to generate custom
layout from a generic input -- doing the same in plain TeX would require
lots more work (including reinventing tons of stuff that already exist
in LaTeX), thus slowing things down and making them more expensive.
> And, at last, a question for LATeX gurus: Is there a way to produce a
> LATeX file with all the cross-referencing "expanded"? You remember:
> I~re-edit any LATeX file to be acceptable as PlainTeX file, which
> means that --- apart from re-writing all the arrays aso --- I have to
> replace all the "equation~(\ref ...)" by "equation~(17)" or so.
I would write a Perl script to do that if the idea didn't appear so
disgusting to me. The whole *point* of the cross-referencing is to make
stuff like that unnecessary in the first instance.
I used to think that plain TeX was superior to LaTeX for anything but
the most vanilla, run-of-the-mill stuff. However, I have learned a lot
about LaTeX over the years and am now firmly convinced otherwise. The
efforts of the LaTeX3 team and a lot of other contributors have made
LaTeX into a reasonably standardized and very powerful typesetting
vehicle which can perform most anything one is likely to encounter (not
only) in scientific publishing -- and lets you do so in a couple of
lines because so much stuff is available pre-cooked, stuff that would
require hours of coding and debugging in a patently difficult and
non-obvious macro language if I were to custom-build it in Plain TeX as
needed. LaTeX may not be as lean and mean as DEK's plain-based formats,
but then most authors are not CS geniuses either.
Anselm
--
Anselm Lingnau ......................... lin...@tm.informatik.uni-frankfurt.de
There are two ways of constructing a software design: One way is to make it so
simple that there are obviously no deficiencies, and the other way is to make
it so complicated that there are no obvious deficiencies. --- C. A. R. Hoare
Sophie Frisch
In article <m2bugnf...@wun.neuroinformatik.ruhr-uni-bochum.de>,
David Kastrup <d...@neuroinformatik.ruhr-uni-bochum.de> wrote:
>[ with LaTeX, you can ] switch between various
>font encodings (suitable for your typical languages to be processed)
>even within one document and yet have an identical input syntax.
>
>You simply cannot compare this with plain TeX where you are forced to
>basically use one style file written by yourself for every document
>you prepare, and have a lot of work getting things to work.
In my experience, it is more work to fight off LaTeX's attempts
to force its ideas about how your document should look like on you,
than to start out defining your own format in plain TeX.
To be fair, if you just want any kind of acceptable looking text on
paper (and you are sure you will never have any more specific needs),
LaTeX is probably your best bet.
If, however, you have a precise need for a particular layout,
it is much easier to start defining it from scratch in plain TeX
than to try to modify one of LaTeX's predefined formats.
>And this is from one who knows TeX intimately enough to get it to
>produce a "This can't happen" error message with two lines of code:
Precisely because you are an expert, you should not abuse your authority
to spread misinformation, such as the implication that in plain TeX
you could not use exotic fonts in funny encodings and still get
portable code. (You would of course include all font and encoding
specific things in a format, not a style file.)
I really think it is irresponsible to scare beginners off from
plain TeX and to direct them to "easier" LaTeX without warning them
that it is actually harder to acheive a specific layout of your choice
in LaTeX than in plain TeX, because you have to understand a much bigger
system.
Sophie Frisch
Institut fuer Mathematik C / Technische Universitaet Graz / Austria
David Kastrup
In article <4ufe51$a...@blah.math.tu-graz.ac.at>
fri...@blah.math.tu-graz.ac.at (Sophie Frisch) writes:
In article <m2bugnf...@wun.neuroinformatik.ruhr-uni-bochum.de>,
David Kastrup <d...@neuroinformatik.ruhr-uni-bochum.de> wrote:
>[ with LaTeX, you can ] switch between various >font encodings
(suitable for your typical languages to be processed) >even within
one document and yet have an identical input syntax. > >You
simply cannot compare this with plain TeX where you are forced to
>basically use one style file written by yourself for every
document >you prepare, and have a lot of work getting things to
work.
In my experience, it is more work to fight off LaTeX's attempts to
force its ideas about how your document should look like on you,
than to start out defining your own format in plain TeX.
Once you have crystallized the essence of your fight into a style
file, you are spared this fight for making existing documents look
like it. This is definitely less likely to be valid for plain TeX.
If, however, you have a precise need for a particular layout, it is
much easier to start defining it from scratch in plain TeX than to
try to modify one of LaTeX's predefined formats.
Particularly as a beginner you are much more likely to end up with a
solution that will take quite a bit of work to transfer to other plain
TeX documents. As plain TeX has about non-existent logical markup, it
is rather hard to do effect changes on a document not designed to be
versatile.
With LaTeX atleast you have a chance to *know* a priori what LaTeX
constructs suck, and can fix them once and for all.
While it is easier in plain TeX to fight the system until one
particular document looks just like you want, it is much easier in
LaTeX to take care that you have to do the fight just once for a
particular document look.
If, like Knuth, you are out to design every book you write from
scratch, making it look unique and perfectly suited to the contents,
and are rather sure you will *never* want the same style again, well,
perhaps plain TeX is for you.
Precisely because you are an expert, you should not abuse your
authority to spread misinformation, such as the implication that in
plain TeX you could not use exotic fonts in funny encodings and
still get portable code. (You would of course include all font and
encoding specific things in a format, not a style file.)
A format is a format, and it's complete. Who does the work of merging
in all requirements of me into one format?
In LaTeX, I can type \"a, and it means in old font encodings \accent
with appropriate code, and in newer ones a character glyph, and in
Postscript fonts the particular character slot the postscript font
happens to have there. I can even type it in using my national
keyboard, and it will mean just what is necessary in any font.
In short, a bunch of specialists has worked hard to get a unique and
easily accessed interface to fonts in LaTeX. In plain TeX, I am free
to do the same, but certainly will, usually, just do the small subset
I need for the work currently at hand.
I really think it is irresponsible to scare beginners off from
plain TeX and to direct them to "easier" LaTeX without warning them
that it is actually harder to acheive a specific layout of your
choice in LaTeX than in plain TeX, because you have to understand a
much bigger system.
Personally, I think it is irresponsible to point people which want to
create small documents with specific visual typically document-unique
layout (like posters and so on) to *any* TeX-related system. It is
true that you can probably fight plain TeX sooner into producing a
unique document than LaTeX, but both are rather a bit too much work
for almost anyone for those tasks, let alone a beginner. (Yes, I do
everything using TeX and its macro packages. I do my posters, however,
in LaTeX, because it is much harder to scale postscript fonts to fit
the paper size in plain TeX. In LaTeX I can use a few packages for
that which are reasily delivered with teTeX. Still, I consider myself
obstinate for doing everything by TeX, and are probably not the
fastest one with that).
With LaTeX, at least the beginner can crank out *useful* documents in
the course of a week. With plain TeX, this is simply not
possible. Using either, you need a lot of experience to be able to
complete cast your visions into a TeX document, be it plain or
LaTeX.
The problem with beginners in LaTeX is thaht they usually try to
struggle with \vspace and \\ and things to achieve some effects which
should better be encapsulated in some style (and most probably are
already). It's a doc problem.
If you solve it via the wizard approach, here are the two
possibilities:
plain TeX wizard:
Oh, yea, that's interesting to do. Come back in a few hours
-- two days later--
Uhm yeah, right, this was not as easy as it seemed. I had to do this
other problem from this other user, you know...
-- another day later--
here, this should do what you want.
(thing does not work in the document, but can be cudgeled into use
pretty easily, or not at all).
LaTeX wizard:
Why didn't you use package foohbar? (foohbar is installed at some
distant point of the galaxy where you would not have looked). It does
just that.
-- a day later you have found the docs --
-- another day later you have understood them --
(thing does not work in the document, but can be cudgeled into use
pretty easily, or not at all).
David Carlisle
> In my experience, it is more work to fight off LaTeX's attempts
> to force its ideas about how your document should look like on you,
> than to start out defining your own format in plain TeX.
LaTeX has very few ideas about document design. It comes with some
`standard styles' that do enforce on particular design, but that is
another matter.
If you want to set up a design in LaTeX you don't have to modify one
of the standard classes if you don't want to, you can design a new
class from scratch. I fail to see how it could be easier to do this in
plain TeX than in LaTeX.
David
Victor Eijkhout
In article <4ufe51$a...@blah.math.tu-graz.ac.at> fri...@blah.math.tu-graz.ac.at (Sophie Frisch) writes:
> In my experience, it is more work to fight off LaTeX's attempts
> to force its ideas about how your document should look like on you,
> than to start out defining your own format in plain TeX.
It's a matter of degree. Yes, in plain TeX you can easily write
output={ something }
whereas in LateX you'd have to fight page after page of
output routine code.
Otoh, sooner or later you will need almost every single
feature in LaTeX, and programming them yourself in plain
TeX is an awful amount of work.
So I'd summarise the state of affairs:
plain TeX: low threshold for user customisation, long way
to go if you need sophisticated features.
LaTeX: high threshold, but most sophisticated features
are already build in, so customisation is quick once you
know where to look and what part of the code to tackle.
Victor@btdt
--
405 Hilgard Ave ................................. `We are in danger of getting
Department of Mathematics, UCLA ............. government by the clueless, over
Los Angeles CA 90024 ................. a place they've never been, using means
phone: +1 310 825 2173 / 9036 ....... they don't possess' [John Perry Barlow]
http://www.math.ucla.edu/~eijkhout/
Sophie Frisch
In article <m2686sd...@dol.neuroinformatik.ruhr-uni-bochum.de>,
David Kastrup <d...@neuroinformatik.ruhr-uni-bochum.de> wrote:
>[...] As plain TeX has about non-existent logical markup, it
>is rather hard to do effect changes on a document not designed to be
>versatile.
Well, so plain TeX forces you to decide upon a logical structure
for your text, and then to write macros that allow changing content
while retaining structure, and to change form while retaining
structure and content.
Having to do these things yourself helps you clarify your own
ideas about the logical structure of your text and the form
you want to give it.
Admittedly it takes a while to overcome the worst beginner's
clumsiness; but then you are preparing for a lifetime of publishing
(unless you plan to perish) and it doesn't matter much whether
the software you use takes 2 weeks or 2 months to learn.
LaTeX seems to assume that every text is structured as a tree
with levels of sections, subsections and subsubsections, et c.
like Wittgenstein's Tractatus.
I do not see why this should be so:
This tree structure is unlikely to be the structure of the ideas
you are trying to express, and it is not well representable
in the essentially linear medium of a sequence of pages on paper.
I have only recently grasped from the postings of various LaTeX
gurus that LaTeX sees itself as a document formatting language.
This is probably the essential difference to plain TeX, which
sees itself as a typesetting language.
I still think that a significant group of people are better
off with plain TeX, in particular scientists writing their own
papers, who have a pronounced opinion about how these papers should
look like on paper.
Anybody who likes to use a mechanical camera instead of a fully
automatic one, or vi instead of emacs, should give plain TeX a try.
TeX has this nice look and feel of a tool created by a mathematician-
computer-scientist for the use of other mathematicians and computer
scientists (or maybe just for the subset who like to have control
over the appearance of their papers on paper).
While TeX makes you feel the need for a library of macros, LaTeX is
not just that, but a "front end" to TeX, creating a totally
different feeling.
As an example, size-switching macros that change the size of all your
textfonts simultaneously, when you have a paragraph of small print,
are desirable and useful.
However the LaTeX solution to this is to have the big, complicated
(and admittedly artful) machinery of NFSS always churning away in the
background, just to save you the additional trouble of specifying both
bold and italic if you want to emphasize a word in a bold sentence.
As a side effect it creates the totally inappropriate illusion
that different design parameters of a font are (or should be)
independently changeable. Considering how many design decisions and
how much work are necessary for the creation of a single face
of a well-designed font, this is like going to a library and expecting
to pick up a book by specifying author, subject matter, number of pages,
language and publishing date independently.
>In LaTeX, I can type \"a, and it means in old font encodings \accent
>with appropriate code, and in newer ones a character glyph, and in
>Postscript fonts the particular character slot the postscript font [..]
But it is easy to redefine the accent macros to do just that, you
do not have to swallow all of LaTeX if you want that bit of
functionality.
And as soon as you use fonts with glyphs and ligatures that are
absent from the T1 encoding you have to do all the encoding related
stuff anyway.
Oh well, I guess I got carried away.
Sophie
Sophie Frisch
In article <udviesmqz6.fsf@vummath>,
David Carlisle <carl...@ma.man.ac.uk> wrote:
>If you want to set up a design in LaTeX you don't have to modify one
>of the standard classes if you don't want to, you can design a new
>class from scratch. I fail to see how it could be easier to do this in
>plain TeX than in LaTeX.
It's easier in plain TeX because you don't have to understand the
additional 7000 lines of code you need to know when building on top
of the LaTeX system.
Sophie
David Carlisle
> It's easier in plain TeX because you don't have to understand the
> additional 7000 lines of code you need to know when building on top
> of the LaTeX system
well to do it right you may find yourself writing a similar amount of
code on top of plain TeX, which isn't so easy either:-)
Don't get me wrong plain TeX is very useful, I use it quite often
myself. It's just that I wouldn't use it for typesetting a document.
But there is no need to disagree, each to his own as they say.
If you use vi rather than emacs then really there is no way to reason
with you to do something logical like use LaTeX rather than plain:-)
However to comment on something you said in your other post. (about NFSS)
> just to save you the additional trouble of specifying both
> bold and italic if you want to emphasize a word in a bold sentence.
No you've missed the point of the font scheme. the fact that you can
get bold italic is a useful (to some people) side effect. the main
point is that the fact that there is a scheme at all means that a
journal (or the author for that matter) can consistantly switch the
fonts in a document to the fonts used in house with just a few
declarations at the top.
This is just not possible in a plain TeX document. You have to really
read through the whole document each time working out what fonts are
loaded for what purpose, and what you want to change them to. In a
setting were you are trying to handle dozens or possibly thousands of
articles from different authors it is just not feasable to do this
while accepting `arbitrary' plain TeX documents.
David
Donald A. Hosek
In article <4uhmft$g...@blah.math.tu-graz.ac.at>,
Sophie Frisch <fri...@blah.math.tu-graz.ac.at> wrote:
>In article <udviesmqz6.fsf@vummath>,
>David Carlisle <carl...@ma.man.ac.uk> wrote:
>>If you want to set up a design in LaTeX you don't have to modify one
>>of the standard classes if you don't want to, you can design a new
>>class from scratch. I fail to see how it could be easier to do this in
>>plain TeX than in LaTeX.
>It's easier in plain TeX because you don't have to understand the
>additional 7000 lines of code you need to know when building on top
>of the LaTeX system.
It's easier in LaTeX because you have a solid framework on which to
build your work.
I started out with plain TeX and converted to LaTeX for a number of
reasons:
- I noticed that I was slowly but surely putting together a patchwork
simulation of LaTeX in plain TeX. As I needed each feature, I would
reinvent the wheel.
- For training purposes, plain TeX is a mess. To produce a vaguely
decent looking output document, the user has to know a *lot* of
plain TeX arcana. What's more, one basic knowledge base can be used
to produce books for a variety of publishers and to produce articles
for a number of journals.
LaTeX has its problems. Its frameworks are underdocumented and
oftentimes lack critical entry points resulting in things like copying
three or four macros to make a one-line change. But I don't see that
as necessarily inferior to the alternative of creating frameworks from
scratch.
Tom Scavo
In article <4uhm8b$g...@blah.math.tu-graz.ac.at>,
Sophie Frisch <fri...@blah.math.tu-graz.ac.at> wrote:
>
>I still think that a significant group of people are better
>off with plain TeX, in particular scientists writing their own
>papers, who have a pronounced opinion about how these papers should
>look like on paper.
Usually, the author doesn't have much say about what the paper
will utimately look like in print. S/he sends it off to the
editors and they take over from there. Better to concentrate on
the content in this case rather than waste time on layout and
design. Books, on the other hand, are a different matter,
especially if the author is responsible for everything,
including design. Few book authors are capable of doing that,
however. In my experience, publishers prefer LaTeX source
because it's easier to work with.
Cheers,
--
Tom Scavo
mailto:trs...@syr.edu
http://web.syr.edu/~trscavo/
Kin Yan Chung
In article <4uhm8b$g...@blah.math.tu-graz.ac.at>,
Sophie Frisch <fri...@blah.math.tu-graz.ac.at> wrote:
>David Kastrup <d...@neuroinformatik.ruhr-uni-bochum.de> wrote:
>>[...] As plain TeX has about non-existent logical markup, it
>>is rather hard to do effect changes on a document not designed to be
>>versatile.
>Well, so plain TeX forces you to decide upon a logical structure
>for your text, and then to write macros that allow changing content
>while retaining structure, and to change form while retaining
>structure and content.
Doesn't that assume that one actually tries to write macros instead of
treating (plain) TeX as a fancy typewriter? I know of people who use TeX
in just this way.
>Having to do these things yourself helps you clarify your own
>ideas about the logical structure of your text and the form
>you want to give it.
In my experience, the same is necessary to fully utilise the features of
LaTeX. After a while I realised that I needed to write "robust" LaTeX
documents so that any changes in notation could be achieved by simply
changing one definition in the preamble.
>Admittedly it takes a while to overcome the worst beginner's
>clumsiness; but then you are preparing for a lifetime of publishing
>(unless you plan to perish) and it doesn't matter much whether
>the software you use takes 2 weeks or 2 months to learn.
I would think that there is a difference between publishing and authoring.
Personally, I prefer to spend more time as an author than to worry about
technical aspects of typesetting. If you have to worry about how something
is going to look, then perhaps there is a better way to explain the idea?
>LaTeX seems to assume that every text is structured as a tree
>with levels of sections, subsections and subsubsections, et c.
>like Wittgenstein's Tractatus.
>I do not see why this should be so:
>This tree structure is unlikely to be the structure of the ideas
>you are trying to express, and it is not well representable
>in the essentially linear medium of a sequence of pages on paper.
This would apply only to the standard document classes (and probably most
of the available document classes too). I would assume that one can write
classes for any type of textual structure.
>I have only recently grasped from the postings of various LaTeX
>gurus that LaTeX sees itself as a document formatting language.
>This is probably the essential difference to plain TeX, which
>sees itself as a typesetting language.
This is a good reason for authors to prefer LaTeX over plain TeX.
>I still think that a significant group of people are better
>off with plain TeX, in particular scientists writing their own
>papers, who have a pronounced opinion about how these papers should
>look like on paper.
I don't know about that. Certainly there are many people who are better
off with plain TeX, but what about those who insist on using underlining
for emphasis despite this being a typesetting "no-no"?
>Anybody who likes to use a mechanical camera instead of a fully
>automatic one, or vi instead of emacs, should give plain TeX a try.
Really? I use vi almost exclusively (even in DOS I run a vi clone!) as an
editor, and although I use a fully automatic camera I often over-ride its
settings. Contrary to your suggestion, I stay away from plain TeX as much
as possible. It's simply too much work. If I wanted absolute control over
how my document looked like, I would probably bypass plain TeX and work up
from iniTeX (or is that virTeX).
--
Kin Yan Chung (kin...@math.princeton.edu) | Sydney _--_|\
Math Department, Princeton University, Princeton NJ 08544 | 2000 / \
WWW Home Page: http://www.princeton.edu/~kinchung | \_.--._/
I've got vi worked out.... :w :q :wq! ZZ ^Z ^D ^[[1 ^H ^C | v
Benjamin Bayart
roe...@loria.fr (Denis B. Roegel) writes:
In article <4uaqgh$p...@nnrp1.news.primenet.com>,
qui...@Primenet.Com (Donald A. Hosek) writes:
|> In article <4ua9el$b...@tesla.netline.net>,
|> Richard J. Kinch <ki...@netline.net> wrote:
|> >Denis B. Roegel (roe...@loria.fr) wrote:
|> >: Maybe the time of Metafont and font development
|> >: is yet to come, for instance with Omega.
|>
|> >It would appear that Yannis Haralambous himself has even
|> >retreated into using
|> >Fontographer for his Omega Times and Omega Helvetica font design work.
|>
|> Given that YH was really the only person doing serious work with MF,
|> it would appear that we could probably declare MF to have died as a
|> serious type development tool.
I thought that YH wanted to have his Omega fonts out as quickly as possible,
since after all, we are all longing to use Omega and we would like to try
Unicode fonts. Maybe when Omega is more widespread, and that there
is some solid
basement, YH will continue his work on exotic Metafonts. After all,
he's paid for it, no ? :-)
Last time I talk about this with Yannis, he explained me that pure
unicode MF fonts will require years of work, and that at the moment he
just need a test-font to check the design of some strange chars and to
allow a wide community to use Omega.
So, nowadays, Yannis is working on PostScript fonts since they are
supposed to be ready more quickly, but he want to come back soon to
Unicode Computer Modern fonts done with MF.
So, IMHO, MF is not dead at all...
Regards,
Benjamin
--
Benjamin BAYART
1, rue de Beaumont
95560 Maffliers
France bay...@lep-philips.fr
David Kastrup
In article <4uhm8b$g...@blah.math.tu-graz.ac.at> fri...@blah.math.tu-graz.ac.at (Sophie Frisch) writes:
In article <m2686sd...@dol.neuroinformatik.ruhr-uni-bochum.de>,
David Kastrup <d...@neuroinformatik.ruhr-uni-bochum.de> wrote:
>[...] As plain TeX has about non-existent logical markup, it
>is rather hard to do effect changes on a document not designed to be
>versatile.
Well, so plain TeX forces you to decide upon a logical structure
for your text, and then to write macros that allow changing content
while retaining structure, and to change form while retaining
structure and content.
Having to do these things yourself helps you clarify your own
ideas about the logical structure of your text and the form
you want to give it.
Uhm, excuse me, but you were complaining that I did not want to
recommend using plain TeX to beginenrs with mathematical typesetting,
and now you are suggesting that one should be able, when using plain
TeX, to properly grasp logical structuring of the text one is about to
write, and then implement an appropriate macro package for doing so,
making sensible typographic deisions on the way.
Now I am not too far from considering myself a computer wiz of some
sort, and "of course" I have acquired the TeX book before any LaTeX
book, and I think a solid plain TeX knowledge useful even for LaTeX
users (when I want to know something about LaTeX, I look in the
source. The docs are beyond me).
But believe me, to properly start cranking out solid documents I used
LaTeX, and still use it. And I *really* want to see the beginner which
starts off best by designing a consistent document formatting system
for his needs.
Admittedly it takes a while to overcome the worst beginner's
clumsiness; but then you are preparing for a lifetime of publishing
(unless you plan to perish) and it doesn't matter much whether
the software you use takes 2 weeks or 2 months to learn.
You are, wrongly, assuming a personal identity between author and
publisher. This is rather rarely the case.
>In LaTeX, I can type \"a, and it means in old font encodings \accent
>with appropriate code, and in newer ones a character glyph, and in
>Postscript fonts the particular character slot the postscript font [..]
But it is easy to redefine the accent macros to do just that, you
do not have to swallow all of LaTeX if you want that bit of
functionality.
If I want it functioning for more than one encoding I have quite a bit
to do.
And as soon as you use fonts with glyphs and ligatures that are
absent from the T1 encoding you have to do all the encoding related
stuff anyway.
Oh yes. Have you read fntguide.tex? It is a matter of half an hour to
define a new encoding, tell LaTeX where all those glyphs and ligatures
are, and in about next to nothing you can set an existing text (or
passages of it) in the new font, and use it, if you want, without
further work in any future document (including other fonts) at will.
If you believe that you can crank out with plain TeX with comparable
ease and speed, congratulations. You are a *real* wiz.
Oh well, I guess I got carried away.
Same here.
Rainer Schoepf
In article <4uhm8b$g...@blah.math.tu-graz.ac.at>
fri...@blah.math.tu-graz.ac.at (Sophie Frisch) writes in response to a
posting by David Kastrup:
>In LaTeX, I can type \"a, and it means in old font encodings \accent
>with appropriate code, and in newer ones a character glyph, and in
>Postscript fonts the particular character slot the postscript font [..]
But it is easy to redefine the accent macros to do just that, [...]
I don't believe that. On second thought, maybe you do consider TeX
programming tasks easier than me...
Rainer
--
Rainer Schöpf
Zentrum für Datenverarbeitung A point of view can be a dangerous
der Universität Mainz luxury when substituted for insight
Anselm-Franz-von-Bentzel-Weg 12 and understanding.
D-55099 Mainz
Germany Herbert Marshall McLuhan:
<Sch...@Uni-Mainz.DE> The Gutenberg Galaxy
Robin Fairbairns
In article <76bughr...@roquefort.zdv.Uni-Mainz.DE>,
> But it is easy to redefine the accent macros to do just that, [...]
>
>I don't believe that. On second thought, maybe you do consider TeX
>programming tasks easier than me...
Given Rainer's track record, I would guess that he and Sophie aren't
comparing like with like :-)
Rainer will have thought about full generality, and that's difficult;
Sophie will have thought about what she needs `for this document',
which is indeed pretty simple. Both of these are `correct' answers to
a problem, but they're not the same problem.
LaTeX is *big* because it tries to be general (and very largely
succeeds). This makes it good for people who want to use it as a
basis for a markup specification.
Plain is *small* because it makes no pretence of generality. This
makes it an eminently desirable teaching system, especially for the
pedagogical purpose of learning about typesetting (it was what I
started to learn about typesetting with). It is, however, explicitly
*not* designed as a document mark-up language; the discussion in this
thread has largely been about document mark-up, and whether it's
better to design your own or to use a canned set (which originally
derives from LL).
On that simple question, the whole thread has become ridiculous.
There *are* dogmatists who maintain that they know what everyone else
in the world should do, but the thread seems to have degenerated into
"I like to do it this way" ... "Nyah nyah, I can do it easily this
way". I don't want Sophie to do it my way; I'm somewhat interested in
her understanding my point of view, but I believe we can all live
entirely happily in the same world, using the same underlying system.
--
Holger Uhr
Sophie Frisch wrote:
>
> It's easier in plain TeX because you don't have to understand the
> additional 7000 lines of code you need to know when building on top
> of the LaTeX system.
>
By the same reasoning, it would be easier to write computer programs in
assembler than in a high-level language. :-)
You don't need to understand the lines of code, you just need to
understand the effect, which is documented (at least it should be).
Holger
--
| Holger Uhr
| Universität-GH Paderborn Phone : +49 5251 60-6623
| Fürstenallee 11 Fax : +49 5251 60-6619
| 33102 Paderborn E-Mail: mailto:hu...@uni-paderborn.de
| Germany http://www.uni-paderborn.de/cs/Holger.Uhr.html
| How to become immortal: Read this signature tomorrow
| and follow its advice.
Victor Eijkhout
In article <4uhm8b$g...@blah.math.tu-graz.ac.at> fri...@blah.math.tu-graz.ac.at (Sophie Frisch) writes:
> I still think that a significant group of people are better
> off with plain TeX, in particular scientists writing their own
> papers, who have a pronounced opinion about how these papers should
> look like on paper.
Scientists may have an opinion about how papers should look
on paper, but shouldn't be allowed to. Actually, in tech reports
you can mess around all you want, but as soon as it becomes
a published paper or book someone else will decide the layout
anyway.
> TeX has this nice look and feel of a tool created by a mathematician-
> computer-scientist for the use of other mathematicians and computer
> scientists (or maybe just for the subset who like to have control
> over the appearance of their papers on paper).
If I may quote the master himself (oh boy, I must have read
the TeX book too many times. This took me all of 20 seconds to find):
``TeX is intended to support higher-level languages for composition
in which all of the control seqeuences that a user actually types
are macros rather than TeX primitives.''
To me this indicates a separation between a macro writer,
who is actually the implementer of a coherent *higher-level language*,
and the user. To blurr the boundary lines is a mistake, if you ask me.
> Oh well, I guess I got carried away.
We all have our hobby horses.
Victor
AES
> .........................................the discussion in this
> thread has largely been about document mark-up, and whether it's
> better to design your own or to use a canned set (which originally
> derives from LL).
There's a third case which is somewhat in between.
If you hope to do a "real book" someday, and meanwhile are churning out
rough drafts and class notes, you can just be sure to "macro-ize", i.e.,
mark up, your manuscript thoroughly in commonsense fashion with your own
custom choice of macro names for elements like \title, \section,
\subsection, \figure, \examplebegin, \exampleend, etc., etc.; and write
your own elementary definitions for these in plain TeX, not trying to
reinvent LaTeX, just to get some usable output.
(Note that since you invent these yourself, you don't have to spend any
time learning or agreeing with LaTeX's choices; and you're much more
likely to remember them. And, you only have to learn as much of plain TeX
as you really need.)
Then when you're ready to do the book, and the publisher's book designer
has designed a unique book design for you, you get a TeXpert to define
your macros to meet that spec -- which standard LaTeX might not be able to
do at all, or only with substantial difficulty -- and you get exactly the
custom book design you and your publisher want.
Mark Wooding
Holger Uhr <hu...@uni-paderborn.de> wrote:
> Sophie Frisch wrote:
> >
> > It's easier in plain TeX because you don't have to understand the
> > additional 7000 lines of code you need to know when building on top
> > of the LaTeX system.
> >
>
> By the same reasoning, it would be easier to write computer programs in
> assembler than in a high-level language. :-)
Err... it is, isn't it? (I'm being serious here: I really do find it
easier.)
--
[mdw]
`When our backs are against the wall, we shall turn and fight.'
-- John Major
michael buening
Sophie Frisch (fri...@blah.math.tu-graz.ac.at) wrote:
: In article <m2686sd...@dol.neuroinformatik.ruhr-uni-bochum.de>,
: David Kastrup <d...@neuroinformatik.ruhr-uni-bochum.de> wrote:
:
: Well, so plain TeX forces you to decide upon a logical structure
: for your text, and then to write macros that allow changing content
: while retaining structure, and to change form while retaining
: structure and content.
: Having to do these things yourself helps you clarify your own
: ideas about the logical structure of your text and the form
: you want to give it.
I decided to chime in not because I'm an expert at anything but because
I'm interested in this debate from a spectators POV. I guess I don't
know how things work in the math world but over here in Chemistry when
you want to publish a paper you don't have one iota of say on how your
paper looks when it's in the journal. You either follow what you are
told by the journal people or it doesn't get published. As a matter of
fact what you send into the publisher is absolutely *NOT* what is
published (except for, hopefully, the actual words).
And from this perspective vi is about all you really need to create
the text for your article (although it seems MS-Word is the one most
often used). You submit text in plain type doublespaced and with
no logical format what-so-ever. You submit tables on separate pages
in "camera-ready" format as well as any graphs or other images necessary
to explain your material. The publisher does all the rest of the work.
To a large extent this is a very good idea. When you read a paper in
Biochemistry you know how it's going to be laid out and you know it will
look good and be readable (again from a layout perspective...words are
still the authors). There is a small part of me that says I can do this
my own way that is perfectly readable and if you'd just put a copy of
my ready-to-read output in your journal all would be well.
: I have only recently grasped from the postings of various LaTeX
: gurus that LaTeX sees itself as a document formatting language.
: This is probably the essential difference to plain TeX, which
: sees itself as a typesetting language.
Again I'm not sure where the author needs the typesetting language
unless the author is also typesetter/publisher. I could publish my
own book if I wanted to (who'd read it?) and in that case I'd
be very interested in handling the format myself. But in publishing
scientific research I don't have these jobs for myself (and for me
personally I don't want them).
: I still think that a significant group of people are better
: off with plain TeX, in particular scientists writing their own
: papers, who have a pronounced opinion about how these papers should
: look like on paper.
See above...I have yet to meet a scientist who had any say on how his
paper looked when published...this goes for whole chapters in books as
well. Yes you present the tables and graphs and pictures as you want
them but that's it.
I have seen one case where a particular conference accepted abstracts
in LaTeX only. This is presumably so that once the author has suggested
a particular structure the details of that structure (things like
margin sizes and indentations for things and the like) would be
standardized by the meeting organizers. And certainly when one thinks
of presenting posters or papers at meetings then the layout does
become the job of the author. But in this case I don't see where
LaTeX or plain TeX is as good a choice as a layout intensive document
preparation package.
--
Mike Buening
Dept. of Chemistry and Biochemistry
University of Notre Dame
mbue...@argon.helios.nd.edu http://www.nd.edu/~mbuening
Ralph Schleicher
David Kastrup <d...@neuroinformatik.ruhr-uni-bochum.de> cited below with
"DK" writes:
DK> But it is easy to redefine the accent macros to do just that, you
DK> do not have to swallow all of LaTeX if you want that bit of
DK> functionality.
DK> If I want it functioning for more than one encoding I have quite a bit
DK> to do.
Depends on your knowledge of the matter (as usual).
If you are a plain TeX user, then you know why you are it. If you can
live with LaTeX, then keep on using it. The only thing I really would
appreciate is when all LaTeX users would learn saying `LaTeX' and not
`TeX' if they are talking about LaTeX; especially when asking for help.
It's time to bury this thread I think. @xref{holy wars, , , jargon,
Hackers' dictionary}.
DK> If you believe that you can crank out with plain TeX with comparable
DK> ease and speed, congratulations. You are a *real* wiz.
Thank you, but I already know that. ;-)
--
Ralph * http://www.UL.BaWue.DE/~rs/
Q: Why do plain TeX users stand up in the morning?
A: Because they see the light!
Piercarlo Grandi
>>> "quixote" == Donald A Hosek <qui...@primenet.com> writes:
quixote> In article <4uhmft$g...@blah.math.tu-graz.ac.at>,
quixote> Sophie Frisch <fri...@blah.math.tu-graz.ac.at> wrote:
>> In article <udviesmqz6.fsf@vummath>,
>> David Carlisle <carl...@ma.man.ac.uk> wrote:
carlisle> If you want to set up a design in LaTeX you don't have to
carlisle> modify one of the standard classes if you don't want to, you
carlisle> can design a new class from scratch. I fail to see how it
carlisle> could be easier to do this in plain TeX than in LaTeX.
frisch> It's easier in plain TeX because you don't have to understand
frisch> the additional 7000 lines of code you need to know when building
frisch> on top of the LaTeX system.
quixote> It's easier in LaTeX because you have a solid framework on
quixote> which to build your work. [ ... ]
quixote> LaTeX has its problems. Its frameworks are underdocumented and
quixote> oftentimes lack critical entry points resulting in things like
quixote> copying three or four macros to make a one-line change. But I
quixote> don't see that as necessarily inferior to the alternative of
quixote> creating frameworks from scratch.
All good points, but this whole debate is confronting just _two_ macro
packages; the one called 'plain TeX' (which isn't plain) and the one
called 'LaTeX' (which isn't flexible).
Thye are different approaches to writing macro packages for TeX.
However both are rather bad, actually very very bad.
"plain TeX" is excessively primitive; "LaTeX" is excessive, period.
What is really lacking in the TeX world is some serious alternatives. In
the troff world one has 'me', 'ms', 'mm' and others that offer different
choices ('ms' is a bit like plain TeX; 'mm' is a bit like LaTeX; 'me'
has no equal).
Perhaps if all the effort that has been expended in fixup engineering
and complicated madness like the NFSS had been spent into designing some
other macro package...
Of course there are other macro packages; "lollipop" comes to mind as a
very interesting alternative (a framework for macro packages more than a
macro package), but none have reached, apart from TeXinfo, a seriously
used status (and TeXinfo, for all its rigidity, is a very serious
alternative to both plain TeX and LaTeX).
In part this is due, let be fair to the insane nature of the TeX engine
and its excessively complex and often maddening ``features''. Still...
Sebastian Rahtz
Perhaps if all the effort that has been expended in fixup engineering
and complicated madness like the NFSS had been
spent into designing some
other macro package...
TeXing life, it has been the single biggest advance to achieving
decent markup
Of course there are other macro packages; "lollipop" comes to mind as very interesting alternative (a framework for macro packages more than macro package), but none have reached, apart from TeXinfo, a seriously
used status
quite. why didn't Lollipop succeed? because its author couldnt/didnt
want to support and develop it. it was a great idea, fine, but it never
matured into a serious product.
some of this discussion ignores the fact that one reason for LaTeX is
not just the kernel, but the fact that huge numbers of useful packages
have been written to go with it. i can build a new class file from
these components in a very short time (on a good day :-}), but in the
rest of TeX world there are very few seriously reusable components.
(and TeXinfo, for all its rigidity, is a very serious
alternative to both plain TeX and LaTeX).
in what sense is it a serious alternative?
sebastian
Sebastian Rahtz
I still think that a significant group of people are better
off with plain TeX, in particular scientists writing their own
papers, who have a pronounced opinion about how these papers should
look like on paper.
than camera-ready copy for preprints, thats great. unfortunately, this
makes their documents moderately useless in the electronic publishing
world.
sebastian
David Kastrup
In article <x0bugfw...@lochnagarn.elsevier.co.uk> sra...@lochnagarn.elsevier.co.uk (Sebastian Rahtz) writes:
Perhaps if all the effort that has been expended in fixup engineering
and complicated madness like the NFSS had been
spent into designing some
other macro package...
i simply cannot understand why people don't appreciate the NFSS. in my
TeXing life, it has been the single biggest advance to achieving
decent markup
Markup? Don't know. Its syntax is sometimes a bit playful, but the
basic working results (consistent accent and composite character
addressing independent of used fonts, mixable encodings, orthogonality
of font accessing packages, postscript font support) are only to be
dismissed easily if you insist on writing purely OT1-encoded texts
(why did you switch to TeX3 in the first place?), and don't want to do
this with a foreign keyboard.
While I had not been overly interested in NFSS at first, and did not
bother loading my LaTeX2.09 versions with it, it is a very good thing
that it is now the default.
Being able to just say
\usepackage{times,euler,beton,utopia,palatino} and what else, and have
a document change outlooks on the fly is a *very* important
achievement.
Now LaTeX no longer means cmr, and you can fit publishers' bills
rather effortlessly.
I don't mind a world of fonts intruding in my small TeX-based
world...
Robin Fairbairns
In article <vwjybjk...@osfb.aber.ac.uk>,
Piercarlo Grandi <p...@aber.ac.uk> wrote:
>All good points, but this whole debate is confronting just _two_ macro
>packages;
Indeed so.
> the one called 'plain TeX' (which isn't plain) and the one
>called 'LaTeX' (which isn't flexible).
What's inflexible about LaTeX? ... perhaps I'm missing something, but
my perception of it is "a framework for constructing macro packages,
with some example packages bundled.
There is, of course, the inflexibility that derives from it being a
_big_ framework and TeX being an intrinsically _small_ system (roll on
Omega), but...
>What is really lacking in the TeX world is some serious alternatives. In
>the troff world one has 'me', 'ms', 'mm' and others that offer different
>choices ('ms' is a bit like plain TeX; 'mm' is a bit like LaTeX; 'me'
>has no equal).
eplain? (not that I know the least thing about the *roff family.)
>Of course there are other macro packages; "lollipop" comes to mind as a
>very interesting alternative (a framework for macro packages more than a
>macro package), but none have reached, apart from TeXinfo, a seriously
>used status (and TeXinfo, for all its rigidity, is a very serious
>alternative to both plain TeX and LaTeX).
Err ... I agree lollipop is(was) interesting, but isn't it the
_entire_ basis for Victor's (really rather good) `TeX by Topic'?
Which makes it just as much of a `system' as LaTeX, except that it
only has one example document design. (Which design wins hands down
in contest with all the standard LaTeX ones, but that's hardly a test
of anything. ;-)
>In part this is due, let be fair to the insane nature of the TeX engine
>and its excessively complex and often maddening ``features''. Still...
Good old Piercarlo ... never to be part of anyone's fixed universe ;-)
TeX's fine -- if you're the sort of person who likes to balance things
on your nose, like a sealion.
Fortunately for the world at large, there are a lot of us.
Holger Uhr
Mark Wooding wrote:
>
> Holger Uhr <hu...@uni-paderborn.de> wrote:
> > By the same reasoning, it would be easier to write computer programs in
> > assembler than in a high-level language. :-)
>
> Err... it is, isn't it? (I'm being serious here: I really do find it
> easier.)
I don't, especially for larger programs.
I must admit that I'm not sure, if writing computer programs is really
comparable to writing typesetting macros, but it has been shown that the
average computer programmer writes the same number of lines of code in
the same time, no matter what language is used. So if you use a
high-level language, you end up with a shorter program (because the
high-level constructs do more work for you per line) and you are more
productive.
Greetings,
Joel D. Young
If your back is against the wall, and you turn and fight, then
you would be facing the wall.... No?
Kind of like using assembler... :-}
On 12 Aug 1996, Mark Wooding wrote:
> > assembler than in a high-level language. :-)
>
> Err... it is, isn't it? (I'm being serious here: I really do find it
> easier.)
Andrea Moro
On 12 Aug 1996 15:04:57 GMT, michael buening (mbue...@bach.helios.nd.edu) wrote:
>I decided to chime in not because I'm an expert at anything but because
>I'm interested in this debate from a spectators POV. I guess I don't
>know how things work in the math world but over here in Chemistry when
>you want to publish a paper you don't have one iota of say on how your
>paper looks when it's in the journal. You either follow what you are
>told by the journal people or it doesn't get published. As a matter of
>fact what you send into the publisher is absolutely *NOT* what is
>published (except for, hopefully, the actual words).
>And from this perspective vi is about all you really need to create
>the text for your article (although it seems MS-Word is the one most
>often used).
From another spectator's pov, it is perhaps useful to note that things
are very different in Chemistry then in other sciences. In Chemistry
papers get published very soon and don't circulate much before
publication, so you don't have to worry about how they look.. just send
your vi-edited ascii code and your tables to the journal.
In Economics is a completely different matter: papers get published
after two years (sometimes more), when they are already old. Authors
have to take care of their circulation before publication; for this
reason they care a lot how they look like and they don't really want to
print them using vi.
>--
>Mike Buening
>Dept. of Chemistry and Biochemistry
>University of Notre Dame
>mbue...@argon.helios.nd.edu http://www.nd.edu/~mbuening
--
..............................................................................
Andrea Moro (elm...@ssc.upenn.edu)
Department of Economics
University of Pennsylvania, Philadelphia, PA 19104 (USA)
<a href="http://www.ssc.upenn.edu/~elmoro/index.html">Andrea</a>
..............................................................................
Richard Kaszeta
>In Economics is a completely different matter: papers get published
>after two years (sometimes more), when they are already old. Authors
>have to take care of their circulation before publication; for this
>reason they care a lot how they look like and they don't really want to
>print them using vi.
In Mechanical Engineering (at least in Heat Transfer), this is also
the case. Your papers seem to get a fair amount of circulation long
before they are published. Additionally, many of the journals (for
example, Journal of Heat Transfer) require camera-ready copy from the
author.
Rich
P.S. Anyone out there working on a robust asme style file? Most of
the ones I've seen are rather crude hack jobs, and I'd like to do
something a little cleaner and more robust.
--
Richard W Kaszeta Graduate Student/Sysadmin
bo...@bofh.me.umn.edu University of MN, ME Dept
http://www.menet.umn.edu/~kaszeta
Victor Eijkhout
> > .........................................the discussion in this
> > thread has largely been about document mark-up, and whether it's
> > better to design your own or to use a canned set (which originally
> > derives from LL).
>
>
> There's a third case which is somewhat in between.
>
> If you hope to do a "real book" someday, and meanwhile are churning out
> rough drafts and class notes, you can just be sure to "macro-ize", i.e.,
> mark up, your manuscript thoroughly in commonsense fashion with your own
> custom choice of macro names for elements like \title, \section, [....]
> Then when you're ready to do the book, and the publisher's book designer
> has designed a unique book design for you, you get a TeXpert to define
> your macros to meet that spec -- which standard LaTeX might not be able to
> do at all, or only with substantial difficulty
A short time apart, a friend of mine and I each wrote a PhD thesis.
He used LaTeX, I used my own macros. He wrote his own style file,
I wrote my own macro package. No difference in concept.
LateX is barely more than the shell, the syntax. Underneath you
can tinker to your heart's desire.
Vic...@been.there.done.that
Victor Eijkhout
> >Of course there are other macro packages; "lollipop" comes to mind as a
Oh boy. There are still people who remember that thing.
> Err ... I agree lollipop is(was) interesting, but isn't it the
> _entire_ basis for Victor's (really rather good) `TeX by Topic'?
Thanks for the compliment, but not sure what you mean by this
`entire basis'. Could you rephrase your point?
Lollipop delivered actually *two* full scale documents: my PhD
thesis, and "TeX by Topic". They look entirely unlike.
> Which makes it just as much of a `system' as LaTeX, except that it
> only has one example document design.
Define 'system'. The main difference between Lollipop and LaTeX was
that Lollipop has a well-defined and extremely easy to use style
designers interface. Writing a style should be barely harder
than using the style to write a document. Was my idea, at least.
> (Which design wins hands down
> in contest with all the standard LaTeX ones, but that's hardly a test
> of anything. ;-)
:-) And thank you on behalf of Merry Obrecht. I have produced several
good-looking documents in TeX, if I may say so, and none of them were
designed by me. I'm fully convinced that that's the right way
to do things.
About Lollipop support: sorry, I simply haven't had the time.
It's a complicated package, and I have other interests than TeX.
Having to finish TbT was a good incentive for at least getting
the package working in *some* form, but after that my
enthousiasm petered out. Sorry.
Victor.
Dr Yoshimasa Tsuji
In article <m27mr8q...@dol.neuroinformatik.ruhr-uni-bochum.de> d...@neuroinformatik.ruhr-uni-bochum.de (David Kastrup) writes:
|Typical instructions in plain TeX look like "get water from the tap
|into the kettle, put it on the stove, light the stove (pressing the
|knob until the fuse locks), wait till the water boils, put a filter on
|the thermos, put approx. 1 heaped tsp. of coffee per cup in (plus
|one), pour as much water on as to soak the powder, wait 30 seconds,
|pour the rest of the boiling water on, turn off the stove". Now while
|this is rather guaranteed to give quite good coffee, it is rather
|non-portable. Take a kitchen with electric stove, for example.
|
|In LaTeX I say "gimme a cup of coffee". While the quality might be
|worse, at least I can fine-tune the coffee-making (for all drinkers),
|and not need to change the instructions for every customer. And quite
|a few customers know nil about making good coffee.
There are bad styles in LaTeX as well as in TeX. All
depends on the authors competence. In the example above,
I -- not having spent time reading lplain.tex, no user
of LaTeX -- would also write "gimme a cup of coffee" and
put the details of making coffee in yet another macro
file. It is a bad idea to fill the text file with
unintelligible \controle_squences. The less the mark up,
the better. For example, a tab can be used as a
beginning of a paragraph with an indentation, thus
saving an odd looking white line. Or a full stop with two consecutive
spaces can be made the end of sentence while that with
one space as an unsplittable white space (avoid using
~). I wouldn't write e.g. \index{Joe Smith} in the text
for making indices at the end of the book. I would let a
filter program add \index{} to the each entry in yet
another file where items to be indexed are listed.
Whatever system you are using, the details have to be
kept away from the substance. You can concentrate on
writing articles with TeX if your friend is going to
write technically most difficult part of macros in
another file.
I think TeX is a typesetting language -- one for the
typesetters. Typesetters are not allowed to argue for
their own taste. They must work exactly as the designers
instruct them.
People here seem to have forgotten the difference
between the works of an author, editor, designer,
and the typesetter. LaTeX is not the language of
typesetters, surely. That is clear enough.
Cheers,
Tsuji
--
Dr Yoshimasa Tsuji
e-mail to: yts...@cfi.waseda.ac.jp
http://www.yt.cache.waseda.ac.jp
Robin Fairbairns
In article <EIJKHOUT.96...@jacobi.math.ucla.edu>,
Victor Eijkhout <eijk...@jacobi.math.ucla.edu> wrote:
>In article <4upsgi$9...@lyra.csx.cam.ac.uk> r...@cl.cam.ac.uk (Robin
>Fairbairns) writes:
>> Err ... I agree lollipop is(was) interesting, but isn't it the
>> _entire_ basis for Victor's (really rather good) `TeX by Topic'?
>
>Thanks for the compliment, but not sure what you mean by this
>`entire basis'. Could you rephrase your point?
"Lollipop as found on CTAN includes some styles for producing `real'
documents, and is therefore every bit as much a system as is LaTeX.
Packages like eplain don't offer that."
>Lollipop delivered actually *two* full scale documents: my PhD
>thesis, and "TeX by Topic". They look entirely unlike.
I knew about your thesis, but a momentary lapse of memory caused me
not to mention it in my post ;-)
Mark Wooding
"Joel D. Young" <jdy...@afit.af.mil> wrote:
> If your back is against the wall, and you turn and fight, then
> you would be facing the wall.... No?
Yes. That's the point. It's humourous.
> Kind of like using assembler... :-}
It depends on your mindset then, maybe. I find it a trivial matter to
write a few thousand lines of assembler; no more difficult than C. This
isn't relevant to TeX, although I'll gladly stand my corner in the war
currently raging across various comp.lang and other newsgroups.
Clark Gaylord
Kin Yan Chung (kin...@math.princeton.edu) wrote:
: as possible. It's simply too much work. If I wanted absolute control over
: how my document looked like, I would probably bypass plain TeX and work up
: from iniTeX (or is that virTeX).
plain TeX is for wimps. If you really want control over how your document
looks, write it directly in PostScript. :-)
Dr E. Buxbaum
fri...@blah.math.tu-graz.ac.at (Sophie Frisch) wrote:
>
>LaTeX seems to assume that every text is structured as a tree
>with levels of sections, subsections and subsubsections, et c.
>like Wittgenstein's Tractatus.
>I do not see why this should be so:
>This tree structure is unlikely to be the structure of the ideas
>you are trying to express, and it is not well representable
>in the essentially linear medium of a sequence of pages on paper.
>
My texts as a biologist are usually like that: Introduction (what is
already known), Question (what do I want to know), Methods (how did I do
it), Results (what was the outcome) and Diskussion (what does it mean),
all with subsections, subsubsections and paragraphs. I can not realy
imagine any essentially different way of expressing scientific material
(of course, literary works are a different kettle of fish).
>I still think that a significant group of people are better
>off with plain TeX, in particular scientists writing their own
>papers, who have a pronounced opinion about how these papers should
>look like on paper.
and get them changed to a journal style by some unsympathetic copy editor
anyway? For that purpose any old LaTeX style file is good enough! And if
you have such specific ideas, why not encase them in a LaTeX style file?
At least they are portable.
Clark Gaylord
David Kastrup (d...@neuroinformatik.ruhr-uni-bochum.de) wrote:
: of font accessing packages, postscript font support) are only to be
: dismissed easily if you insist on writing purely OT1-encoded texts
: (why did you switch to TeX3 in the first place?), and don't want to do
I don't give a tinker's damn for non-OT1. I switched to TeX3 (or, rather,
2e) because I felt I had no choice in getting support from this community.
This notwithstanding the fact that I see it as most benefitting Addison-
Wesley. The biggest reason to run 2e for me has been more consistent
package development, especially by way of \usepackage.
: Now LaTeX no longer means cmr, and you can fit publishers' bills
: rather effortlessly.
Until we have a plethora of *math-enabled* fonts, LaTeX WILL continue
to mean cmr. Mathtimes and Lucida are nice (I recently purchased the
latter) but we need a *lot* more.
--
Clark Gaylord
Dept of Aerospace and Ocean Engineering
Virginia Tech
Blacksburg, Virginia USA 24061
voice: 540/231-9061 fax: 540/231-9632
gay...@aoe.vt.edu http://www.aoe.vt.edu/~gaylord
Robin Fairbairns
In article <4v25fm$h...@solaris.cc.vt.edu>,
Tosh. When _I_ were a lad we had to write out the bits corresponding
to the pixels, and then glue each pixel dot down onto the paper
individually.
We _really_ understood what was going on then.
Kirk A. Stork
In article <yhuohk6...@styx.lep-philips.fr>, Benjamin BAYART
<bay...@lep-philips.fr> wrote:
...big snip...
> The *really* good thing will be unicode, but, as 8 bit mail is still
> not available everywhere, I suppose that unicode will be for 22nd
> century... :-(
>
pgp will convert anything to mail-safe text using the ascii armor option
(even the export version of pgp).
Benjamin Bayart
gay...@aoe.vt.edu (Clark Gaylord) writes:
David Kastrup (d...@neuroinformatik.ruhr-uni-bochum.de) wrote:
: of font accessing packages, postscript font support) are only to be
: dismissed easily if you insist on writing purely OT1-encoded texts
: (why did you switch to TeX3 in the first place?), and don't want to do
I don't give a tinker's damn for non-OT1.
Well... my english is really far from perfect, so, I'm not sure tu
understand very well all of the slangs used here :-(
But, is my understanding is not too bad, this really look like a pure
american way of thinking. T1 fonts, or whatever non-OT1 are *required*
for something like 99% of the languages around the world. The TeX
primitive \accent is just fine for silly things like a french word in
an english text, but a really bad thing for anything else.
What is a pure american way of thinking is to say ``I don't need T1
for my own personnal language, so it might not exist at all'', which
is crazy.
TeX is the only typesetting system I know that can typeset quite
everything if you give it the right tools: as Knuth stated in the
TeXbook, at least real accented fonts for european labguages, and
specific fonts for the others.
Even for a pure english text with even no foreign words, one would
need the dc fonts: at least for the MeV kerning as stated in an other
thread :-)
So, well, you don't need it for *your* documents in *your* language on
*your* computer, that's a fact. But TeX is also used because the
documents are exchangable from one computer/place to another. If
everybody, like you, just want to use only what oneself require, then
this will be lost. This is the reason why I think than the
standardisation of encodings is a good thing.
The more intelligente choices, IMHO, is something like that:
- cmr fonts for maths,
- dc fonts (real ones) when one use any european language, and virtual
ones when he just want to typeset a few foreign things,
- same thing for other languages/coding: to have an acces to the fonts
and a good support in the tool to allow anyboy to use them with only
few modifications, I meen: adding mf/tfm/pk files and no more.
The *really* good thing will be unicode, but, as 8 bit mail is still
not available everywhere, I suppose that unicode will be for 22nd
century... :-(
: Now LaTeX no longer means cmr, and you can fit publishers' bills
: rather effortlessly.
Until we have a plethora of *math-enabled* fonts, LaTeX WILL continue
to mean cmr. Mathtimes and Lucida are nice (I recently purchased the
latter) but we need a *lot* more.
Well, as certain people systematicaly refuse new things, I can
understand that quite nobody has written math fonts... :-)
Regards,
Benjamin.
--
Benjamin BAYART
1, rue de Beaumont
95560 Maffliers
France bay...@lep-philips.fr
David Kastrup
In article <20AUG199...@reg.triumf.ca> as...@reg.triumf.ca (Donald Arseneau) writes:
That is well and good, but Clark Gaylord's reaction was more specific,
and I deleted the extraneous (impertinent?) lines of David's to highlight
the remark on TeX3, implying that the only reason to use a decently
modern TeX is to enable 8-bit input. This is very "American" in its
narrow-mindedness. It is also out of place in a promotion of dc (or T1)
fonts because they would work just fine in TeX 2.x which always handled
8-bit fonts; it just lacked 8-bit keyboard input.
Sorry about my impertinence (which is worse in writ than in natura, I
guess).
But TeX did not really handle 8-Bit fonts well. In fact, it would have
been pretty difficult to make anything but the keyboard input (which
could be made to work via the typical translations) work: there was no
\uccode, \lccode, \mathcode, \sfcode and even \catcode for those
characters, \hyphenchar could be none of them (I believe), and
several other pretty hard limitations. Apart from being able to
typeset them (including ligatures) with \char"xx sequences, there was
not much that you *could* do with them.
The support was mainly on the level that you'd think necessary to
typeset from dingbats or other assorted font selections. Or you'd have
accessed the upper font positions just by ligatures, for example for
hand-written fonts (which require a lot of ligaturing). You'd
certainly would not want to typeset texts starting from those
positions.
It was *really* not just for trifles that Knuth was persuaded to
tackle TeX again in order to make it support 8-Bit fonts (and several
hyphenations patterns) properly.
Woody Jin
In article <EIJKHOUT.96...@jacobi.math.ucla.edu>, eijk...@jacobi.math.ucla.edu (Victor Eijkhout)
<m2686sd...@dol.neuroinformatik.ruhr-uni-bochum.de>
<4uhm8b$g...@blah.math.tu-graz.ac.at>
<76bughr...@roquefort.zdv.Uni-Mainz.DE>
<4umv0h$1...@lyra.csx.cam.ac.uk>
<siegman-1208...@aesmac.stanford.edu> wrote:
>A short time apart, a friend of mine and I each wrote a PhD thesis.
>He used LaTeX, I used my own macros. He wrote his own style file,
>I wrote my own macro package. No difference in concept.
Only if you don't ever need information interoperability.
Think about the situation, two or three people are writing a paper.
Think of another situation, someone else other than you might wish
to reformat your document (and you are not there, of course).
How do they know about your customized macro, and who cares ?
It is not the matter of which one is more powerful or who has
more brain cells to write documents in TeX. The real issue is that
you need to interchange/cooperate to write documents.
--
Woody Jin
Donald Arseneau
In article <yhuohk6...@styx.lep-philips.fr>, Benjamin BAYART <bay...@lep-philips.fr> writes...
>gay...@aoe.vt.edu (Clark Gaylord) writes:
> David Kastrup (d...@neuroinformatik.ruhr-uni-bochum.de) wrote:
> : (why did you switch to TeX3 in the first place?), and don't want to do
> I don't give a tinker's damn for non-OT1.
>Well... my english is really far from perfect, so, I'm not sure tu
>understand very well all of the slangs used here :-(
I don't know how the tinker, an itinerant repairman, enters into
this phrase, but it means the same without him.
>But, is my understanding is not too bad, this really look like a pure
>american way of thinking. T1 fonts, or whatever non-OT1 are *required*
>for something like 99% of the languages around the world.
That is well and good, but Clark Gaylord's reaction was more specific,
and I deleted the extraneous (impertinent?) lines of David's to highlight
the remark on TeX3, implying that the only reason to use a decently
modern TeX is to enable 8-bit input. This is very "American" in its
narrow-mindedness. It is also out of place in a promotion of dc (or T1)
fonts because they would work just fine in TeX 2.x which always handled
8-bit fonts; it just lacked 8-bit keyboard input.
>What is a pure american way of thinking is to say ``I don't need T1
>for my own personnal language, so it might not exist at all'', which
>is crazy.
Nobody said that. Do you not permit someone to say ``I don't need T1,
so I don't care about T1''?
Donald Arseneau as...@reg.triumf.ca
Victor Eijkhout
> >A short time apart, a friend of mine and I each wrote a PhD thesis.
> >He used LaTeX, I used my own macros. He wrote his own style file,
> >I wrote my own macro package. No difference in concept.
>
> Only if you don't ever need information interoperability.
Never say never.
In this case we were both taking (about to be) published articles
and reformatting them to make a self-contained thesis.
No interoperability needed.
> Think about the situation, two or three people are writing a paper.
In that case I'd use LaTeX, yes. But in other cases I don't.
> Think of another situation, someone else other than you might wish
> to reformat your document (and you are not there, of course).
If I write a book, I'll make damned sure that no one will
reformat my document !! :-)
> It is not the matter of which one is more powerful or who has
> more brain cells to write documents in TeX. The real issue is that
> you need to interchange/cooperate to write documents.
Too dogmatic, Woody. There is a time and place for everything.
Victor.
--
405 Hilgard Ave ................................. `[W]e don't usually like to
Department of Mathematics, UCLA ............. talk about market share because
Los Angeles CA 90024 .................... we're not going to share anything.'
phone: +1 310 825 2173 / 9036 .................. [Jim Cantalupo, president of
http://www.math.ucla.edu/~eijkhout/ McDonald's Int.]
Ken Smith
r...@cl.cam.ac.uk (Robin Fairbairns) writes:
>In article <4v25fm$h...@solaris.cc.vt.edu>,
>Clark Gaylord <gay...@aoe.vt.edu> wrote:
>>Kin Yan Chung (kin...@math.princeton.edu) wrote:
>>: as possible. It's simply too much work. If I wanted absolute control over
>>: how my document looked like, I would probably bypass plain TeX and work up
>>: from iniTeX (or is that virTeX).
>>
>>plain TeX is for wimps. If you really want control over how your document
>>looks, write it directly in PostScript. :-)
>Tosh. When _I_ were a lad we had to write out the bits corresponding
>to the pixels, and then glue each pixel dot down onto the paper
>individually.
>We _really_ understood what was going on then.
And, of course, if you _really_ want to have control of everything,
nothing beats writing your program in machine code - none of your
fancy assembler, thanks - and then punching your own cards in binary.
Ah, those were the days.
And you got oodles of tiny pieces of thin cardboard for use as
confetti at weddings, as a bonus!
>--
>Robin (the man with no voice) Fairbairns r...@cl.cam.ac.uk
>U of Cambridge Computer Lab, Pembroke St, Cambridge CB2 3QG, UK
>Home page: http://www.cl.cam.ac.uk/users/rf/robin.html
Ken Smith
--
Dr Ken Smith <k...@maths.uq.oz.au> | "God, we know you are in charge, but why
Department of Mathematics, | don't you make it slightly more obvious?"
The University of Queensland, | Archbishop Desmond Tutu, 1990
St Lucia, Qld. 4072. Australia. | (address to students at at West Point)
Piet van Oostrum
>>>>> kas...@mbay.net (Kirk A. Stork) (KAS) writes:
KAS> In article <yhuohk6...@styx.lep-philips.fr>, Benjamin BAYART
KAS> <bay...@lep-philips.fr> wrote:
KAS> ...big snip...
>> The *really* good thing will be unicode, but, as 8 bit mail is still
>> not available everywhere, I suppose that unicode will be for 22nd
>> century... :-(
>>
KAS> pgp will convert anything to mail-safe text using the ascii armor option
KAS> (even the export version of pgp).
You can do that even with standard stuff (MIME).
--
Piet van Oostrum <pi...@cs.ruu.nl>
URL: http://www.cs.ruu.nl/~piet [PGP]
Alexandre Suchkov
Victor Eijkhout (eijk...@jacobi.math.ucla.edu) wrote:
: If I write a book, I'll make damned sure that no one will
: reformat my document !! :-)
I see. And you are planning on handling all the logistics of publishing,
finding and paying for the press, doing all the marketting, bribing the
reviewers, etc? :-)
Robin Fairbairns
In article <4vf3jd$p...@solaris.cc.vt.edu>,
Alexandre Suchkov <gay...@aoe.vt.edu> wrote:
>Victor Eijkhout (eijk...@jacobi.math.ucla.edu) wrote:
>: If I write a book, I'll make damned sure that no one will
>: reformat my document !! :-)
>
>I see. And you are planning on handling all the logistics of publishing,
>finding and paying for the press, doing all the marketting, bribing the
>reviewers, etc? :-)
There _are_ publishers who allow you to work directly with the
editors/designers and to design your own macros to achieve the
requisite effects.
The one book of Victor's _I_ have is (a) distinctively designed, but
(b) shows no sign of a battle between publisher and author, while (c)
I happen to know it was written using macros based on Lollipop.
Dr Yoshimasa Tsuji
In article <4vf3jd$p...@solaris.cc.vt.edu> gay...@aoe.vt.edu (Alexandre Suchkov) writes:
|
|Victor Eijkhout (eijk...@jacobi.math.ucla.edu) wrote:
|: If I write a book, I'll make damned sure that no one will
|: reformat my document !! :-)
|
|I see. And you are planning on handling all the logistics of publishing,
|finding and paying for the press, doing all the marketting, bribing the
|reviewers, etc? :-)
The author and the publisher are completely different people.
The publisher usually decides the ultimate format of the
book (books they publish usually look very similar, not only the
binding, size, fonts they use, etc, but also the style of writing.)
The author need not to bother how their book will look
in the end. All he needs to do is to designate that this
piece is a footnote to this word, this is a foreign
word, etc. so that the publisher will deal with them
appropriately. (i.e. marking up is all the author needs
to do.)
If a camera ready thing is required, the publisher must
give the author the exact details, with which an ordinary
author will not be able to cope. If the publisher
supplies a TeX/LaTeX style file, the author ought to use
it without ever overriding anything.
If I am the new typesetter of a tiny publisher and wish to use
TeX, I need to obey all the detailed instructions about
our own house style from technical editors and previous
typesetters. Otherwise, the book won't be "our" book.
You will be better off writing the style file from
scatch and feeding it to INITEX, surely.
Good style is always built on a sound
advice from professional designers. If your designer
rejects Computer Modern, you must throw CM away, or
you need to find a designer who tolerates CM.
Steve Kroeker
Hello,
The Profs I had at MIT supplied "camera ready" TeX based output.
i.e. Intro to Algorithms by Tom Cormen, Charles Leiserson and Ron
Rivest.
The department there perhaps has some experience in this field.
Steve
Victor Eijkhout
> Victor Eijkhout (eijk...@jacobi.math.ucla.edu) wrote:
> : If I write a book, I'll make damned sure that no one will
> : reformat my document !! :-)
>
> I see. And you are planning on handling all the logistics of publishing,
> finding and paying for the press, doing all the marketting, bribing the
> reviewers, etc? :-)
Well, in fact I *did* write a book, and after the publisher
and I agreed on the design, I delivered the dvi files to someone
who printed them on a Linotronic, and they sent the camera-ready
proofs to the printer.
This only required
1/ the publisher and me to agree on the design
2/ me and the phototypesetter to have the same tfm files.
After that *they* paid *me* (in fact, even before royalties
I received a few bucks per page because I had saved them
the typesetting), they did the marketting, and they bribed
the reviewers :-)
Duncan Hothersall
<4vd6g4$4...@Masala.CC.UH.EDU> wrote:
>In article <4vd6g4$4...@Masala.CC.UH.EDU> wj...@cs.uh.edu (Woody Jin) writes:
>
>> >A short time apart, a friend of mine and I each wrote a PhD thesis.
>> >He used LaTeX, I used my own macros. He wrote his own style file,
>> >I wrote my own macro package. No difference in concept.
>>
>> Only if you don't ever need information interoperability.
No implementation of TeX, with whatever macros, can be
efficiently used as an information interchange format. If you
want repurposable documents, your best bet is to invest in an
SGML system. Then you can generate TeX, write your own macros,
use LaTeX or whatever you like to produce printed copy, and still
have a highly efficient interchange mechanism.
( http://www.sil.org/sgml/sgml.html )
Duncan
-------------------------------------------------------------
Duncan Hothersall Phone: +44 (0)131 451 3526
Edinburgh Business School Fax: +44 (0)131 451 3002
Heriot-Watt University Email: d...@ebs.hw.ac.uk
Edinburgh, EH14 4AS, UK URL: http://www.ebs.hw.ac.uk
-------------------------------------------------------------
AES
d...@ebs.hw.ac.uk (Duncan Hothersall) writes:
> >No implementation of TeX, with whatever macros, can be
> >efficiently used as an information interchange format. If you
> >want repurposable documents, your best bet is to invest in an
> >SGML system.
and t...@maths.tcd.ie (Timothy Murphy) replies:
> If you are talking about mathematical material,
> this is more or less complete nonsense.
Let's not raise the temperature unnecessarily here, guys.
This has the makings of an educational thread, from which
I'd hope to obtain two objectives:
1) Increase my understanding of SGML (I know TeX very well).
2) Try to promote the development of future standards for
the web especially, in which I'd not have to learn _any_
new syntax for coding math in my original source documents
in order to make them readable on the web (that is, browsers
should be able to decode TeX math, even if they don't somehow
"contain TeX', and there should be no new "html math syntax"
for encoding math that I have to learn).
Those are my personal objectives; fire away!
Timothy Murphy
d...@ebs.hw.ac.uk (Duncan Hothersall) writes:
>No implementation of TeX, with whatever macros, can be
>efficiently used as an information interchange format. If you
>want repurposable documents, your best bet is to invest in an
>SGML system.
If you are talking about mathematical material,
this is more or less complete nonsense.
--
Timothy Murphy
e-mail: t...@maths.tcd.ie
tel: +353-1-2842366
s-mail: School of Mathematics, Trinity College, Dublin 2, Ireland
Murray Adelman
> And you got oodles of tiny pieces of thin cardboard for use as
> confetti at weddings, as a bonus!
>
mutilate.
Murray
Robin Fairbairns
In article <siegman-2308...@aesmac.stanford.edu>,
AES <sie...@ee.stanford.edu> wrote:
>d...@ebs.hw.ac.uk (Duncan Hothersall) writes:
>
>> >No implementation of TeX, with whatever macros, can be
>> >efficiently used as an information interchange format. If you
>> >want repurposable documents, your best bet is to invest in an
>> >SGML system.
>
>> this is more or less complete nonsense.
>
Timothy is undoubtedly right. He might have expressed his remark
slightly less tersely...
> This has the makings of an educational thread, from which
> I'd hope to obtain two objectives:
>
> 1) Increase my understanding of SGML (I know TeX very well).
A bit of background to Timothy's remark.
1. There are various efforts at expressing maths in SGML. They are
almost none of the `repurposable' (if I understand the coinage
correctly), and some of them are ludicrously wrongheaded.
2. The commonest failing of the (public) SGML Maths DTDs[*] is to
assume that maths is a fixed entity, the whole of which can be written
down by a few well-intentioned people.
3. The second commonest failing is to assume that mathematical symbols
are in some way `characters'. So we find symbols such as \Gamma being
expressed (translated to LaTeX-ese) as \begin{greek}G\end{greek}, when
its relationship to the capital Greek letter is visual only.
4. For all that, mathematical texts expressed in any of these DTDs is
far harder to understand than in LaTeX (which is in its turn
marginally more verbose than plain TeX).
5. Some of the DTDs confuse concepts from various parts the TeX macro
world (for example, saying that they are "direct copies of LaTeX" when
they're nothing of the sort.
[*] SGML only specifies how you might specify the mark-up syntax of a
document. The DTD specifies the syntax that you actually write. (How
the resulting document looks is specified in yet another way.) SGML
is to the DTD as initex is to the macro packages we all actually use.
(Roughly ;-)