Re: [Axiom-developer] literate programming and Claerbout's Insight

[C Y]

--- Gabriel Dos Reis <address@hidden> wrote:

> On Sun, 29 Jul 2007, C Y wrote:
> 
> | Sure, as a commercial product.  Open source is a different
> | ballgame;
> 
> What are the concrete differences?

For one thing, there is no legal problem with fixing problems yourself
and distributing the results to the world free of charge.

> The algorithms have not gotten better, for example.  The
> documentation is no really better than what they had before; 
> windows support is worse.

The differences are not concrete yet - I consider the intangible ones
very important but perhaps that is not a shared opinion.

> well, very few people want to be on board of something with
> the audience "too early to tell", but surely casual users are
> encouraged to look elsewhere. We should not forget that experts, 
> and those doing CAS for living, once started as casual users.  In a
> heathly, growing community, there are more "novices" than experts.

Well, in my case I found Axiom only after being introduced to the idea
of CAS on more "friendly" systems - Mathematica and Maxima.  However, I
now understand at least in part the merit of Axiom's complexity -
correctness is not simple.  Perhaps this means Axiom will never be a
logical candidate for an "entry level system."

> | Axiom as an open source project has not existed for 2 decades, 
> 
> that is true, but you can rewrite history.  Axiom has more than 2
> decade history.  For practical purposes, Axiom as a open source
> currently -- for many -- is no better than what it was before.  Are
> the algorithms better?  Are the known, documented, bugs fixed?  

Certainly it is not better yet.  I believe it eventually will be, but
it will take a LOT of time to get there.  Open source is not magic - it
is an opportunity to create something better, no more and no less.

> | working through legacy code to update/modernize/document is neither
> | sexy nor quick.
> 
> Right!  And yet we want to attract people.

So as far as I can see the only thing to do seems to be to get through
the "grunt work" and make a system that will attract people.

> | CAS developers, at the moment - those interested in what a CAS CAN
> | be, rather than what they are today.
> 
> but those do CAS for living; they are working computational
> scientists. We seem to be ignoring them.

I would say we aren't ready for use cases where people want to simply
use the CAS without developing it in the process.  If that doesn't
appeal to "working computational scientists" I'm not sure what the
answer is - the ground work needs to be done.

Cheers,
CY


       
____________________________________________________________________________________
Take the Internet to Go: Yahoo!Go puts the Internet in your pocket: mail, news, 
photos & more. 
http://mobile.yahoo.com/go?refer=1GNXIC