General discussion

[Jerzy Karczmarczuk]

Thomas Steffen has a particular opinion about TeX:


> i have been using lout for one project only, and a few minor files
> plus a new letter head style. my opinion is that lout is *much* easier
> than tex/latex, in just about every way. the expert guide is
> brilliant, the only thing needed is ...

...

> configuring tex is a nightmare. britle command, no syntax at all,
> files all over the place, unix scripts, c code, floating point
> emulation in latex code, sever limitations in a lot of fields... 

...

> i don't think that a serious text processing system can be a turn-key
> solution. it is powerful *because* you can configure it.

                           **           **            **

A long time ago Valeriy Ushakov put in his Lout archive a rhetoric
question: "why the CTAN collections keep Lout?" (a clear subsumption:
the TeX "mafia" has no reason to advertise its direct concurrents...)

--

Please.

I am *very far* from being a Lout expert. I just appreciate the courage
and the inventivity of Jeffrey Kingston. Lout seemed to me a very nice
alternative, *differently* structured than TeX. The *difference*, the
new approach was more important than just weighing:    better-worse.

Lout is not better than TeX. It is different. There are very nice things,
for example left arguments to operators, but the syntax is not the whole
issue. Yes, the configuration is simpler, but configuring the web2c 
collection of TeX support, with kpathsea, etc. is not so difficult, and 
anyway it is sufficiently portable to do it once by the installation 
authors. The issue is the following: the number of people engaged in the
maintenance of TeX is 100 times bigger than our little Lout world. Knuth
had and still has the benediction of AMS, and despite the fact that the
system is frozen, and the only changement will be the upgrading of TeX
3.14159 into \pi when DEK writes his final \end, there is plenty of evolutive
processes underway (LaTeX 3 project, and eTeX).

But TeX was born too early, and the discovery of Lout was for me a 
pleasure. I wanted several things from a new typesetting (*not* a
text processing...) system. I wanted a different programming constructs,
graphics, less anxiety when designing my own page output routines, etc.,
in general, something more modern. I got it, but after some experiences
I have still some doubts. I don't think that this "crisis" of Lout which
has been suggested and then refused by Jeffrey Kingston, Valeriy Ushakov
and others, is the problem of incomplete documentation, etc. I would love
to use more frequently Lout, but:

1. It is a bad surprise to find that the "verbatim" option invokes an
   external script (and so, it is system dependent). And I teach program-
   ming, so I need to include verbatim programs in my lecture notes.

2. PostScript graphics was *very* important issue for me, in fact I wanted
   to see finally a true integration between the typical text macros in
   TeX or Lout, and the graphic object *programming*. (I would like to see
   the true integration between TeX and MetaPost, if you know what I mean.)
   To write a list of numbers which will be simultaneously typeset, and
   displayed by PostScript (or whatever), or even animated.
   In Lout the communication between the procedures in Lout and PostScript
   is (at least for me) still quite difficult. It is easier (again: for me)
   to write some \specials in TeX (or to use PSTricks).

3. The transparency of the "digestion" process of Lout (as compared with
   TeX, where I feel utterly lost trying to find *when* all these
   \afterassignment_s, \futurelet_s etc. are executed and in which local
   context...) is much bigger.
   So, would it be possible to generate *almost directly* from Lout the
   PDF, RTF or HTML documents? 
   Personally I need PDF. Of course one can always distil the PS, but then
   adding links, etc. is cumbersome (and you have to buy a commercial Adobe
   software, Ghost will not do all that yet.)

   So, I use pdftex. 

3. This is discutable and personal, but I think that the administration of
   white space by Lout is less flexible than in TeX. Adding "glue" to Lout
   seems not to be easy, I failed to convert to Lout something which needed
   in TeX a simultaneous usage of \hfil and \hfill.
   
--

Such list of small - sometimes amusing, sometimes annoying - differences
between a system you know for more than 20 years, and something relatively
new, may be arbitrarily long. But we should pose some questions, which
even if cannot be solved immediately by the Implementor Himself, might
one day become useful.

There are *not* too many naive and simple questions on this list. If somebody
asks for a comparison between TeX and Lout, it remains unanswered. This is
bad, as "spreading the gospel" means to compare the darker outside with the
internal light. 

There are sometimes some touchy points in the discussion. Somebody wants to
do something quickly, and confesses that he has no time to learn Lout
thoroughly. Valeriy Ushakov responds with reproaches, saying that if the
other wishes to use some system competently, he must master it first.
You might say: true, isn't it? Of course. 

But such an answer from a guru is methodologically doubtful. Outer people will
choose this or that system in function of their own needs and time, and lack
of time is a genuine problem. So if I were *really* a bad guy (I am not!) I
would say: "Ok, forget this Lout business. They will accept you in their
sect when you learn by heart their Holy Scriptures first, otherwise they
will treat you as somebody immature, not ready to enter the Temple". This
is obviously not our aim.

My best wishes to all of you.

Jerzy Karczmarczuk
Dept. of Comp. Sci., University of Caen, France.