Re: tlug: Office suite for use under Linux

["Stephen J. Turnbull" <turnbull@example.com>]

>>>>> "jb" == Jonathan Byrne <jpmag@example.com> writes:

    jb> Well, yes and no.  Any major word processor today offers the
    jb> ability to have different style formats for different
    jb> sections, so you just choose the one you want, much like doing
    jb> it manually in TeX.

This is fine, of course.

As a matter of taste, I like seeing the logical divisions, ie, I
dislike WYSIWYG on principle.  One reason why I like LaTeX and higher
level additions, and SGML, is that by adding more abstraction you can
achieve "What You See Is The Structure Of Your Thought", ie,
WYSITSOYT, without annoying details.  The only universal advantage
that has over WYSIWYG, of course, is the number of letters in the
acronym is greater.  But I _like_ it.

    jb> If I'm writing a book (not something most people do very
    jb> often), I can set up style sheets for the various parts, type
    jb> it, and forget it.  The sytle sheet will take care of the
    jb> formatting automatically.  I will admit to having never
    jb> bothered doing any but the most basic style sheets, but they
    jb> were essentially point and click.  I'm not sure how long it
    jb> would take to do a more complex one, but in the worst case
    jb> imaginable, it could take no longer than learning to do it in
    jb> TeX.

You _can_ set up style sheets, basic ones are faster than learning TeX 
macros or LaTeX commands and parameters, but doing a complex one is
much harder than doing it in TeX because the point'n'click mechanism
makes it hard to take care of exceptions, like too long line length
and so on.   As time goes on, that will improve, but for the moment
TeX is much better.

This is important because it encourages the user to go back and change
the fonts or whatever, and ignore the style sheet.  It's really hard
to tinker locally with a TeX command; this encourages the writer to go 
back and fix the original definition.  I like that focus.  I think
text formatting languages emphasize the reader; high-end
word-processors the writer's artistic creativity.

What I would like to see is a Tsukuba-dai mandated thesis style sheet
that Word would _enforce_.  Then the last objection is moot.  But I
don't really see this happening.

If you need to read long documents in a foreign language, then a good
style sheet really helps.  Word strongly encourages ad hoc design.
That is only slightly edged out by the horrors of block copy.  It's
bad enough to read an illiterate abstract, but then you have to read
the same words again in the intro, body, and conclusion.  This has
nothing to do with either writing or reading in a foreign language, of 
course.  However, the toys that Word provides do tend to encourage
spending time on exciting visual presentation that would be better
spent editing the text.

Manuscripts written in TeX tend to be more uniform in style, have
bland visual presentation, and better thought out text.  Not to
mention properly formatted equations.  Guess which sensei likes?  :-)

As for ease of use, [X]Emacs + AUC-TeX + LaTeX solves that in the long
run.  Word has it all over TeX for "sit down and be productive," of
course.  I can't speak for Lyx, but the idea is excellent.

    jb> There is very little that can be done under TeX today that
    jb> cannot be done with similar or equal quality by a high-end
    jb> word processor (and I'm not just talking Mac, Windows, OS/2
    jb> here; Applix Words rocks, and it runs on Linux :-) ).

Circular paragraphs?  Or any other algorithmic shape you happen to
like.  :-)  TeX is Turing complete.  (Of course, the recent rash of
Word viruses is proof that so is Word, I guess.  Never heard of a TeX
virus.  Have you? :-)

    jb> One of the few areas where this may not be true (or may; it's
    jb> something I have no experience with) is typesetting
    jb> mathematics.  While MS Word's equation editor is apparently
    jb> suitable for most equations, this may be an area in which
    jb> things like TeX still excel.  But as I say, I have no
    jb> experience in that area, and could be completely wrong, and MS
    jb> equation editor could be utterly horrible; perhaps someone who
    jb> does know about it can comment on this :-)

MS's equation editor as of two years ago knew very little about
mathematics, and it is still very easy (as of the Jan 1998 round of
Master's theses in my department) for users to get it to produce
really ugly math.  Naive users also tend to know words like "subset"
or "in", and thus tend to select the correct symbol when using TeX,
where they are invoked by name.  They often make mistakes (sometimes
intentionally - "well, they look the same to me and that one's
prettier") in selecting from a palette.

But this is getting to be a moot point, as Mathematica knows how to
produce both good looking TeX and good looking Word, and many students 
are using Mathematica because it improves the accuracy of their
symbolic computations.

This still has its disadvantages for professionals, who write
variations on a theme, because Mathematica's editor stinks, and the
LaTeX it produces can't be edited.  (It's like the HTML tables
produced by most HTML assistant-type applications.)  But we can expect 
that in the long run equations will normally be produced by
Mathematica or similar applications, and the capabilities of Word and
LaTeX will be equally moot.


---------------------------------------------------------------
Next Nomikai: 15 May Fri, 19:30 Tengu TokyoEkiMae 03-3275-3691
Next TLUG Meeting: 13 June Sat, Tokyo Station Yaesu gate 12:30
Featuring Stone and Turnbull on .rpm and .deb packages
---------------------------------------------------------------
a word from the sponsor:
TWICS - Japan's First Public-Access Internet System
www.twics.com  info@example.com  Tel:03-3351-5977  Fax:03-3353-6096