Case sensitivity, line labels and the future

In fact, I think C Y's proposal is rather clever, though differently
implemented. The discussion reminds me a lot of the discussion about
labels, so those who followed might know my thoughts already. I want to
make three points, (1a), (1b) and (2), where (1a) and (1b) are *really*
important and (2) is only important. However I can live with (most)
decisions to deal with (2). This is not so with the other cases. 

Please note that (1a) does *not* mean that there is anything to be done 
right now. It is only a question of future direction. (However a very 
important one)

(if I annoy you, stop reading NOW!)

Martin

(1a) I think the *most* important issue is to decide, who is our clientel.  
There are quite a lot of CAS out there, some free, some open source, some
commercial. I know

Axiom/Aldor, Gap, Mupad, Maple and Mathematica.

Only the last two are commercial, Mupad is a little in-between. Also, it 
seems that Mupad is somewhat a crossover between Axiom and Maple. (But I 
confess, I don't know it too much)

Axiom is on the way to become THE powerful free CAS, but they are not
there yet (they say that 80% of the algebra builds meenwhile, which is a
lot). As far as I can judge, it is *very* well done and will be *very*
powerful. I'm certain, that there will be *lots* of things that axiom will
be able to do and we can only dream of. Note that they also have efforts 
to create a 'user friendly' interface.

What I've seen of Mupad suggests that Mupad is a little easier to learn 
and not quite as powerful, but still very versatile.

I don't know Gap well enough to say much about it.

We all know Maple and Mathematica.

--------------------------------------------------------------------------

What strikes me, is that the only feature all the others lack - save maybe 
axiom/aldor, is a *real*, standardized programming language. Although 
Aldor is probably not standardized, however it is very well designed and 
maybe as powerful as lisp, a little more specialised, but certainly not as 
widespread as lisp.

--------------------------------------------------------------------------

In short: I argue that our real future is the integration of maxima and 
lisp. That means: Provide interfaces and documentation to make it easy to 
call routines maxima from lisp.

(1b) improve the math. I think there are some 200 math bugs out there.

(2) Case sensitivity and line labels: I think that user's choice is a good
thing, and in my opinion there is a very simple way to achieve it:

All contributed stuff must be case insensitive and must not use line 
labels in order to be accepted. In fact this is very easy to check: when 
receiving the source, demand that it's all in lower case...

The idea of CY to convert E into capital_e internally is bad, I think, 
because there might be a variable around which is (shit happens) called 
capital_e.

But there is no danger of confusion, if all contributed stuff is case 
insensitive.

Thanks, Martin