Re: [Axiom-developer] RE: User interface,TeXmacs and Axiom

[Ralf Hemmecke]

> Axiom, in its current state, is much more for programmers and
> developers than for end users.  An Emacs inteface which helps those
> people seems like a logical step at this time.

Hmmm, at first I thought you are right. But what does it mean to have an 
axiom interface for developers? If you develop a package you write some 
Aldor (instead of SPAD) code. The only thing you would need is a support 
to load your package into axiom. But that is nothing. If you work with 
Axiom you are a user.

> An "official" GUI, of sufficient quality to be worthy of the Axiom project,
> is EXTREMELY non-trivial.

That should become part of the WishList and after a while it should 
develop into a project description where somebody could start to work on it.

> > So we should not force users to learn Emacs if they just want to
> > use Axiom.

> I agree.  However, the kind of interface the modern CAS user expects
> will take a LOT of non-trivial work, and there is not likely to be much
> of a concensus on the "best" way to do it.

Well, one thing I would start with is a description of what features we 
would like that GUI to have. The implementation or what tools/languages 
is then used to implement it is a completely different issue.
And maybe after we have a proper problem description, we should contact 
open source developers that are knowledgable in GUI building in order to 
implement our whishes. I think, just saying "we need a 'nice' GUI" is 
not enough. We could talk forever.

It seems that you are one of the people with much interest in GUI. Start 
a Wiki and try to collect opinions, wishes, ideas, links to relevant 
papers. That all should finally become a problem description.

Yes, we need something like that also for other subprojects.

Kai Kaminski made a kind of project description on the wiki before he 
started.

http://wiki.axiom-developer.org/AxiomUI

We should have this for any subproject, together with a list of all 
these subprojects.

> Also, before we do a GUI we would probably need to bring B-natural
> or something like it to maturity.

Well, I think a prerequisite of this is that Aldor gets support for 
reflections. And a prerequisite of that is that the aldor compiler 
becomes open source or at least that there is a developer at aldor.org 
who implements reflections.

Ralf