RE: [Axiom-developer] RE: tex4ht and jsmath

[Eitan Gurari]

 > > LaTeX2HTML uses a different approach than tex4ht for translating
 > > latex into hypertext, and I do think the axiom people should
 > > consider the option of using LaTeX2HTML instead of tex4ht.
 > 
 > Could you please explain why you think we should consider LaTeX2HTML
 > instead of tex4ht? It seems to me that tex4ht supports more LaTeX
 > then LaTeX2HTML in a more flexible manner. As the author of tex4ht
 > do you see some inherent limitation that is not obvious to me?

If the interest is to follow the model of planetmath.org with regard
to math representation, then there is no point to start from scratch
with tex4ht.  Instead one might just use their tools.  The
planetmath.org site uses bitmaps for math, and the developers of that
site for reasons I'm not familiar with decided to go with LaTeX2HTML.

My understanding was that some axiom people preferred to use jsmath
instead of mathml or bitmaps, and tex4ht was brought into
consideration to try doing the job.  I believe tex4ht is capable to
provide good outcome.  I successfully roughly tested the approach the
past few days with tex4ht on the testmath.tex file of the amsmath
society for non-normalized latex output.  The outcome can be viewed at

   http://www.cse.ohio-state.edu/~gurari/temp/testmath.html

by inspecting the html source of the file (or employing a text base
editor that shows CDATA sections).  For jsmath the high level math
configurations in use should be replaced with low level tex
configurations that abide to the capabilities of jsmath.  I currently
can't finish this last step because I don't know how to enable jsmath.

-eitan