Sage goal, reinterpreted

[rjf]

As some of you know, I have tried to figure out what Sage is really
about.

At the risk of immodesty, I provide a link to a 1999 paper of mine
http://www.eecs.berkeley.edu/~fateman/papers/RIMS.pdf

(Invited talk at Research Institute for Mathematical Science (RIMS),
Kyoto University, Conference on Program Transformation, Symbolic
Computation and Algebraic Manipulation, November 29 – December 1,
1999. Some material first appeared in a talk at THIRD IMACS
INTERNATIONAL CONFERENCE ON EXPERT SYSTEMS FOR NUMERICAL COMPUTING,
May 17-19, 1993
)

Although the title is "Generation and Optimization of Numerical
Programs by Symbolic
Mathematical Methods" that is not entirely what is it about. It talks
a lot about CAS system design (and is 28 pages long).

I have not re-read the whole paper, but I noticed that I predicted
various things that may have come to pass in Sage, regarding gluing
pieces together.

I also note that future builders of CAS will have different
priorities.

Indeed, I think I may have characterized Sage in this quote (from page
5).

"The goal is to wipe out competing programs."

As he dons his kevlar jacket and breathing apparatus.

http://www.labsafety.com/store/Safety_Supplies/Protective_Clothing/Flame-Resistant_Clothing/+-21614/

[William Stein]

2009/9/4 rjf <fat...@gmail.com>:

:-) This is the sage-flame mailing list, so no harm done.

Now, I'm pulling out:
http://zedomax.com/image/200610/zedomax_lego_flamethrower.gif

And since it is sage-flame, I feel justified in stating clearly that
the statement "The goal is to wipe out competing programs." has little
to do with why I started Sage and work on Sage. Everybody has their
own motivations for working on Sage. My *personal* motivation for
Sage is to create better technology than anything else out there. I
need something that is actually good enough to support my research
program.

In the 1996-1997 I used PARI, but found that really limiting,
especially because I do a lot of exact linear algebra. So I switched
to C++ and the corresponding libraries when I was a Berkeley grad
student in 1997. That was better for what I did, but still quite
frustrating, because of a lack of good libraries. So in late 1998 I
switch to Magma, which had a huge amount of relevant library code for
my research -- stability and language-wise though it was a step back
from even C++. I was pretty enthuesiastic about Magma until about
2003, when I started using modern open source tools for web
programming and scripting of parallel computing tasks, etc., and was
reminded just had behind Magma-as-a-language was then. For
arithmetic geometry and number theory research,
Magma-as-a-library-of-code was at least "a decade" ahead of all other
math software available on the *planet*, but the language and
developer infrastructure (e.g., closed source) was poor compared to
what was *possible*. Moreover, I suspected (correctly) that it
would stay that way at least for the next ten years because of John
Cannon.

Thus I started Sage, and my motivation is the same today as it was
then. It is to create technology that for certain applications is
superior to everything else.

That's my *personal* goal.

However, every developer has their own reasons and goals for being
involved in Sage, and I like to listen to other people's goals and
work toward achieving those to.

For example you -- rjf -- are definitely involved in Sage, have
certainly put a lot of effort into thinking about the project, have
genuinely contributed to it (in my opinion), and I suspect your
motivation is much different than mine. Which is fine.

William

>
> http://www.labsafety.com/store/Safety_Supplies/Protective_Clothing/Flame-Resistant_Clothing/+-21614/
> >
>



--
William Stein
Associate Professor of Mathematics
University of Washington
http://wstein.org

[William Stein]

2009/9/4 rjf <fat...@gmail.com>:
> (just to you)
> I read your note as ... your goal is to wipe out Magma/ John Cannon.

Then I should expand on what I said, since the above is actually not
my goal (hence I communicated badly). MAGMA isn't up to snuff to
support my research. There is nothing I can do to change that.
Wiping out Magma would not help with making Sage the tool I need for
my research. In fact, wiping out Magma would hurt my research,
because I wouldn't have the option to do the same calculation using
two independent tools. I have a vested interest in not wiping out
Magma. Sage may indeed render Magma irrelevant, but that is not
my goal.

> I don't really care;  [... snip since I'm applying onlist]
>
> I also don't care if you replace Maxima with something better.
>
> II do think it would be unfortunate if you displace  Maxima with
> something that is not better.

Thanks for explaining your motivation for contributing to the Sage
project. That is very sensible, and I appreciate it.

William