lout-users
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Preaching DSSSL (Re: using LOUT)


From: Mike Dowling
Subject: Re: Preaching DSSSL (Re: using LOUT)
Date: Tue, 2 Sep 1997 12:34:58 +0200

>>> On Tue, 02 Sep 1997 03:01:18 -0400, Paul Prescod <address@hidden> said:
>> I am a newcomer to Lout myself, but it doesn't take very long to realize how
>> much superior it is to {La}TeX.  Just browse through, say, Leslie Lamport's
>> latex book, and the Lout manuals.  The first book is full of exceptions,
>> don't dos, and chicken wavings.

Paul> That's true, and its what attracted me to Lout over LaTeX, but it is
Paul> worth noting that a DSSSL user should never see either Lout or LaTeX so
Paul> Lout's usability advantages become substantially less overwhelming. On
Paul> the other hand, if Lout's usability allowed the Lout back-end to progress
Paul> more quickly than the TeX back-end, then that would be an advantage. If
Paul> it more reliably produced high quality output then that would be another
Paul> advantage.

I too once enthused about SGML.  The thing that I greatly disliked about LaTeX
is that they made an upgrade to LaTeX2e from a soup of files to be found at
some CTAN archive.  TeX is an oddball program, written in WEB, which serves
only to make it inefficient, and designed to be converted into PASCAL, a toy
programming language.  It writes a superfluous file (name.dvi) to disk instead
of writing a postscript file to stdout.  Silliest of all, you cannot simply
download a tar.gz file from CTAN; packages are always a soup of individial
files.  Ah, I thought, SGML is the solution.  It's a IEEE standard, and
upgrades will therefore be seldom, and hopefully more reasonable.

Having bought a couple of books on SGML, my first reaction was disappointment.
As far as I can see, and please correct me if I am wrong, SGML only caters for
run-of-the-mill documents.  If you want to write a letter or an essay, it's
fine.  But if you have very complex formatting requirements, then SGML lacks
the semantics for their creation.

As a mathematician, I have very complex, multi-line formulae.  As far as I can
see, the only way I can set them into type is by using LaTeX with the AMS
packages.  (I have not even found the equivalent of the AMS packages in lout!)
That leaves me using LaTeX by necessity for professional article, and lout for
the rest.

Where SGML comes into its own is with the Linux documentation project.  You can
convert the SGML documentation files automatically into postscript and read it
as a hard copy.  You can convert it into texinfo files, and view it online.  If
I remember rightly, you can convert it into HTML and man pages, but I won't
swear to that.

In short, the purpose of SGML would appear to be restricted to those documents
with simple structures which one would like to have in many, different formats.
Specialised stuff requires specialised type setters.

Cheers,
        Mike



reply via email to

[Prev in Thread] Current Thread [Next in Thread]