mailing-list for TeXmacs Users

[Henri Lesourd <address@hidden>]

Javier Arantegui wrote:

> Hello,
>
> I do believe that TeXmacs is getting older pretty fast. Maybe it's not a 
> technological problem, but definitively the look'n'feel is too old. Lately 
> I've been trying to migrate a couple of user to TeXmacs and I was unlucky. In 
> both cases they found the interface almost unusable.
>
>  
In this respect, the problem is that I'm not so sure
that simple solutions (e.g., using a standard widget
set like Qt, GTK, etc.) would really solve the problem.

You really have to dig into what people mean when they
say that the interface is "almost unusable", and chances
are that what you discover is that the answer is that a
lot of annoying details plague the life of users. As far
as the users I know react, the main problems (stated using
an approximate order of importance) are :

-> The emacs "chords" (i.e. keyboard shortcuts),
  which become more and more unknown & unintuitive
  for new users, because over the last years, whether
  you like it or not, more & more people become used
  to the Windows usual things in user interfaces. This
  item is probably one of the main points which leads
  to a feeling of an old-fashioned look & feel ;

-> Bad import/export to LaTeX (need to hack in the
  TeXmacs markup/write Perl scripts, etc., to solve
  such issues) ;

-> Crashes, which lead to losing your data ;

-> The File chooser widget (in fact, it is the main one
  which is really annoying : all the rest of the interface
  is made of menus which, in any case, look and behave
  somehow the same in any case) ;

-> Uncomfortable things, like, for example, how the page
  jumps when the cursor is located on the last line on
  the screen and when the user hits the "Down" key. In
  this latter case, the text should scroll up smoothly,
  and only one new text line be added on the last line
  on the screen ;

-> Etc.


To summarize, my feeling is that the problem of TeXmacs's
user interface is much more a problem of lots of little
details that should be carefully handcrafted (such details
being currently very insufficiently polished), than a problem
of one or two big things that you could just do and hop !,
the problem is solved.

The problem with details like that is that the people
which will handcraft them will probably have to dig
into lots of different details of different parts of
TeXmacs, instead of being able to work on an independent
part which would just need to be plugged into TeXmacs
once it's finished.


This is why my opinion is that in order to be efficient,
it is mandatory to concentrate first on having a **detailed
description** of what the problems are from the point
of view of user interfaces, and from the point of view
of the **users**.

i.e., OK, your users said that they "found the interface
almost unusable", but **why** ? Can they say which are
the precise things they feel are a problem ?

As for me, I will not deny that I had similar feedbacks
from my users, but when I digged further, I came with
the kinds of answers I listed above, --not with the kinds
of answers more technically inclined people most of the
time give when they see TeXmacs in the first place--.


To summarize, the fact that the interface **looks** old
fashioned is not the real problem IMHO. The real problem
is that it **IS** really hard to use for non-technically
inclined people (i.e., not the kind of people who would
immediately tell you that emacs chords is the best invented
thing in the world, if you see what I mean), and that
in order to solve this problem, it is much more important
to address all the little details which are a real suffering
for people who are not used to them (and should never be
forced to go along such a way) than to address the problems
of "look", which are much less important.

Of course, this is just my personal opinion, but it is
not completely devoid of some roots in the reality of
user's experience, even if on the other hand, I didn't
asked thousands of users (indeed, if possible, it would
be something like that which could be the more useful
in the 1st place).


Best, Henri