Re: 2.6.7 [was Re: [Axiom-developer] Tiddly Axiom Wiki]

[Bob McElrath]

Camm Maguire address@hidden wrote:
> My $0.02 regarding UI for browsing and graphics:
> 
> 1) I agree with Martin, we need an emacs interface option too.(!)  It
>    is the only tool in which I can get any real work done.

As long as we keep text/tex output and text input, I don't think this
will ever be a problem.  (e.g. keep the UI and Axiom as decoupled as
possible)

> 2) Please make sure we centralize on freely available (as in free
>    software, 'libre', not beer), widely distributed, maximally
>    portable, longstanding and lightweight solutions.  I have my own
>    questions regarding jsmath, mathml, and the like, mostly stemming
>    from my ignorance of them, and general suspicion of anything which
>    might make us depend on java (do these?).

Java != Javascript.  Perhaps we should be saying emcascript instead
(which is an ISO standard).  No one has suggested using java for
anything at all.  There is absolutely no reason to add another language
to the mix, except as may relate to creating user interfaces.
Emcascript is solely a scripting language used to bind actions to UI
elements in a web browser.  It is already built into all web browsers.
My tiddly-axiom-wiki example contains a large amount of emcascript code.
This is undesirable...but it's just a prototype...

It *could* be done in java, but I wouldn't touch it with a 10 foot pole.
Besides, if in java, then why not in lisp?  I agree on the libre point.
The technologies we are discussing (mozilla, emcascript, jsmath, mathml,
svg) are all libre or defined standards by recognized standards bodies
such as the W3C and ISO.

Someone has been spamming the AxiomUI page with anti-java remarks.
Dunno who.  I reverted them.  But this project has absolutely nothing to
do with java.

> I don't think anything
>    can match TeX in scientific user mindshare.  To the extent we can
>    lever this avenue for result presentation, I think we'll be set for
>    life.  I realize however that this might not be a total solution.

For the near future we will depend on tex output, and have it rendered
in the UI using jsMath.  In the future we should switch to MathML
(probably with a tex->MathML translator such as itex in the middle for a
while), because this will allow greater interaction with expressions.
(e.g. highlighting terms, etc -- which isn't possible with jsMath)

If I have anything to say about it, tex will *always* be an output
format.  I don't see myself writing physics papers with MathML anytime
soon.

--
Cheers,
Bob McElrath [Univ. of California at Davis, Department of Physics]

    "One of the best ways to get yourself a reputation as a dangerous citizen
    these days is to go about repeating the very phrases which our founding
    fathers used in the great struggle for independence." --Charles A. Beard

signature.asc
Description: Digital signature