RE: [Axiom-developer] axiom opportunity

[Page, Bill]

On Tuesday, April 25, 2006 8:32 AM David MENTRE wrote:
> [Open Office ...]
> On that side, a reasonable approach would be to add a new Axiom
> output format suitable for computer parsing (for example as sexp).
> One could use a standardized format, like MathML, but I am afraid
> of its complexity (that's said, that does not prevent us to look
> at how MathML has defined its building blocks).
> 

I think jsMath as an alternative to MathML has continued to
evolve and mature very rapidly:

http://www.math.union.edu/~dpvc/jsMath/changes.html

It uses just a subset of LaTeX embedded in HTML.

For use in a graphical user interface based on web browser
technology I still consider jsMath quite superior to MathML
in ease of use, compatibility and even performance.

> Regarding initial Tim question: OOo has a documented API but
> might not be suited to create new User Interface components.
> Moreover, its output is quite awful[1].

OpenOffice is a very good alternative to Microsoft Office and
I think that is really about as far as it goes... :( At work I
make this recommendation to everyone who comes to me with a
problem invoving Microsoft Word. But I think serious authors
who use a lot of mathematics expect more than what OpenOffice
and Word can provide. So at most I think, we should try only
to make it relatively easy for someone to import a document
containing mathematics generated by Axiom into OpenOffice. I
really can not imagine actually using OpenOffice as the primary
user interface.

> For the 30 years horizon, I would much prefer invest my time
> in an easily deployable interface (i.e. HTML/Javascript) and/or
> a innovative and good looking interface (i.e. a combination of
> TeXmacs and the one Tim is always giving as example, with
> drag&drop). OOo is not one of them.
> 

I agree. TeXmacs is a very interesting platform that has a
lot of potential but it needs more work done on the Axiom side
to really exploit it fully. TeXmacs is a bit slow and it's user
interface is a little weird (especially compared to what Microsoft
Windows users are used to), I think it can be acceptable for
a lot of potential users of Axiom. Plus it has interfaces for
many other open source (and even some commercial) math packages.
But one trouble with TeXmacs is that the developer base is very
small and it's source code is quite idiosyncratic - which makes
it hard for new developers to make even simple changes.

I personally still like the HTML/Javascript option the best.
One reason it is attractive to me is because it scales well
from desktop to online collaborative web environments. In 30
years time I am convinced that more and more research and
development will be done using online collaborative tools.
It is just not possible for individuals to build tools more
complex than Axiom is right now using the tools developed
30 years ago. It most be possible for people to work together,
fluently share and pool their work. They need new and more
powerful tools that they trust which will hopefully grow out
of the work on source code version control systems today.

Which reminds me to mention:

  https://launchpad.net

Take a look at:

  https://launchpad.net/products/?text=axiom

and

  https://launchpad.net/distros/ubuntu/dapper/+source/axiom

This system has some important features that Savannah and
SourceForge do not have (yet). See for example the online
support for translation, "new generation" revision control
system, specification tracking system and bounty incentives.

Regards,
Bill Page.