lout equations, general commentary.

[David Feuer]

I have just started looking at lout, which seems to have some
nice features, but it seems to me that its syntax is extremely
verbose, its mathematics unstructured, and its command namespace
far too flat.
        It seems to take quite a bit of typing to get even a simple
equation.  One of the great advantages of TeX is that even
complex equations can be created without slowing the typist too
much.  Lout's equations don't seem so friendly.  The added
verbosity is quite acceptible in a high-level math language like
ML, in which the equation must include abstract meaning as well
as formatting, but seems excessive in a pure formatting
language.  I also think that lout's (like TeX's) equation
language is not nearly structured enough.  It does not seem to
care about the intrinsic grouping of the equation, ending up with
an equation that looks good on paper, but is not very meaningful
internally.  While I could easily be mistaken, it does not seem
that the logical components of an equation are boxed together.
        There seem to be many top-level macros for doing similar
things.  For instance, there is a separate macro for each
paragraph style, though fortunately they are all based on a
parameterized primitive.  I think it would be much nicer if the
command namespace were partitioned.  For instance,
parstyle.ragged
parstyle.normal
and so on.

        I really like lout's "objects".  They are far, far better than
TeX boxes.  I also like the diagramming tools, which seem to be
one of its stronger features.  The table support also looks
really good.  TeX alignments are rather limited.  I suppose that
each formatting system will have certain strengths, depending on
the needs of the author.

-- 
This message has been brought to you by the letter alpha and the
number pi.

David Feuer
address@hidden
address@hidden
Open Source: Think locally; act globally.
Check out humbolt.geo.uu.nl/lists