Re: [Axiom-developer] axiom opportunity

[C Y]

--- Antoine Hersen <address@hidden> wrote:

> > I think we need to start positioning ourselves to be the computer
> > algebra system for these linux systems. Which implies we spend
> > some time trying to get axiom working in conjunction with things
> > like Open Office.

Tim, do you mean have Axiom available as a "plug-in" to OpenOffice, or
use the OpenOffice mathematical editing environment in our own
interface?

> > Open Office has a component that creates mathematical equations
> > but does little else.
> 
> Koffice has one too, that might be easier to extract than  touching
> OpenOffice.

I doubt we want to extract components from these setups for a
stand-alone interface, although perhaps they are more modular than I
think.
 
> QT/KDE could also be a good choice for a cross platform UI has open
> source version are available for Linux/Win/Mac

If we want to take the shortest target to a good looking cross platform
Math GUI I would agree QT is a logical choice, and in that case Koffice
might indeed have useful code for our purposes once they port to QT4. 
(QT3 is not free on Windows, and apparently has some problems with
printing on Linux.)  The thing to keep in mind is that KOffice, by
design, builds off of the KDE enviornment.  I doubt we want to depend
on that for a working GUI, although I certainly have no objections to
taking advantage of it if it is available.  We might be able to use the
display and IO components of the math editing though.

For the record, there exists a (very) basic set of bindings for QT/Lisp
using CFFI:  http://members.aon.at/lispolos/  Of course GCL doesn't yet
support CFFI so this isn't of immediate use, but it might be a
worthwhile direction to pursue down the road.  I would love to see an
expansion of this work be a SoC project.

Camm, if you're listening, how much work is involved with getting CFFI
functional on GCL?  Would it need ANSI mode?

Anyway, I agree that we need to first define what we want an interface
to do/be.  Also, if I do a little crystal ball gazing, I think what
will eventually happen is that one or more open source CASs will sprout
a wide variety of interfaces integrating them into office suites,
editors, science/educational programs, etc.  EMaxima was in effect an
integration of Maxima into LaTeX documents, and I found that quite
useful, so I suspect we will see many of these types of things develop
gradually over time.

> > We could use this component as a front end
> > and overtake its current use by adding a combination of tools
> > under the math component.
> >
> > We could combine open office math with the lex editor (for literate
> > programs) and axiom.

Uh - do you mean the Leo editor?

> I did not have success for googling for "lex editor" can you post a
> url please ?
> 
> > Someone could study the various components and
> > do a high level design document that would coordinate combining the
> > components. Are there any 'software designers' in on the mailing
> > list?
>  
> Kind of. But did we decided on a set of desired capability ?

I would argue we're just at the beginning of that process.  Also, there
are some things I would worry about - for example, how does one do
interactive plotting in OpenOffice?  What about automatic line
breaking?  We still have no truly robust solution for line breaking (it
is a Hard Problem in general) and while we have a variety of ways that
work for most common cases I'm not sure how to use even those in
OpenOffice.

> > A widespread adoption of an axiom/OO math tool would also raise the
> > issue of B-natural rather quickly since people will certainly
> > complain about user interaction issues.

Very true.  But don't we need an open Aldor in order to be able to push
B-natural forward?  I thought we had reached the decision that a
language improvement was needed to do B-natural correctly? 
 
> Axiom/OO sound strange, I think there is a plugin to use Mupad with
> Ms Office anybody has used this system ?

Can't say that I have - I take it you're thinking of this? 
http://www.sciface.com/main.php?menu=2&ID=66760

It certainly looks interesting.

> > Trying to work with open office as a front-end will certainly
> > force axiom's user interface to mature rapidly.
> 
> There will be the problem of  Latex, that might be a lot of works
> to male OO parse Latex.

Maxima had someone who did work on this - it looks like the better way
is to teach Axiom to output OpenOffice math expressions.  I have no
idea how much work this would be.

Cheers,
CY

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com