A Sage NSF proposal to the Computational Mathematics Program

[William Stein]

Hi,

Me and several people have been putting together a proposal to this NSF program:

http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=5390

I currently have a grant from them for Sage that will expire soon, and
would like another one :-)
This time, we're applying for 4 Sage Days workshops a year, with a
focus mostly in science, engineering, applied and numerical
computation, since that's where I think Sage most needs to improve.
The proposal is due Nov 30. Here is a draft of the main narrative
of the proposal:

http://wstein.org/home/wstein/tmp/compmath.pdf

Here's the original latex (with an hg repo):

http://wstein.org/home/wstein/tmp/compmath.tar.bz2

Any sort of feedback would be appreciated. I'm sure there are a lot
of dumb typos, etc., but feel free to tell me about them.

William



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

[ulfarsson]

Hi. I read through and made some comments into
the PDF with Skim on Mac OS X. I'll send it to
you via email.

Henning

[William Stein]

---------- Forwarded message ----------
From: John H Palmieri <jhpalm...@gmail.com>
Date: Sun, Nov 22, 2009 at 7:03 PM
Subject: [sage-grants] Re: A Sage NSF proposal to the Computational
Mathematics Program
To: sage-grants <sage-...@googlegroups.com>

On Nov 20, 4:24 am, William Stein <wst...@gmail.com> wrote:
> Hi,
>
> Me and several people have been putting together a proposal to this NSF program:
>
>      http://www.nsf.gov/funding/pgm_summ.jsp?pims_id=5390
>
> I currently have a grant from them for Sage that will expire soon, and
> would like another one :-)

I've posted a new pdf file here:

http://www.math.washington.edu/~palmieri/Sage/compmath.pdf

I've tried to take into account the comments that people have provided
(William forwarded them to me), although I haven't necessarily
followed all suggestions precisely.  Please take a look at the new
version and send more comments, posting them here or sending them to
both me and William.

Thanks,
 John

[David Ketcheson]

I'm excited to see a major focus on numerics and PDEs in Sage. I'm
thinking of using Sage to teach numerical methods for PDEs here at
KAUST next semester.

-David

[rjf]

From the proposal


... and which has sophisti-
cated interfaces to nearly all other mathematics software, including
Mathematica, Maple,
MATLAB and Magma. ...


Maxima just gets no respect. :)


Most of the facilities mentioned are already in Maxima.

And why is Cython much more than a Python to C translator? (This is
not sarcasm. I honestly have no idea that it was more.
I thought it was, if anything, less.)

"venerable" Maxima is mentioned once, suggesting that the only thing
it can do is symbolic integration and numeric integration.
Actually, while Maxima includes library access to Fortran methods, it
is far inferior to what could be done in numeric integration,
as demonstrated by recent Mathematica versions. You would hardly get a
hint that 75% of the sage-support messages are about Maxima.


Maybe what is needed is a Fortran to Python translator.

In the context of computational mathematics, it sounds more
interesting than "NSF: please send us $31,000 so our student Herman
Bartleby can spend 5 years of his life exploring the Klotacherry
variant of the Murtelbaskov-Lnikcjly conjecture.
We need $1,000 for a computer, and $30,000 more so Bartleby can
continue to rent his garret and buy the thin gruel he uses as the
basis for his diet. "

I think that if NSF sent the proposal over to computer science and
engineering, it might not get a great reception, but it is hard to
predict such things.

RJF

[William Stein]

On Mon, Nov 23, 2009 at 8:28 AM, rjf <fat...@gmail.com> wrote:
> From the proposal
>
>
> ... and which has sophisti-
> cated interfaces to nearly all other mathematics software, including
> Mathematica, Maple,
> MATLAB and Magma. ...
>
>
> Maxima just gets no respect. :)
> Most of the facilities mentioned are already in Maxima.

I wish...

> And why is Cython much more than a Python to C translator?  (This is
> not sarcasm. I honestly have no idea that it was more.
> I thought it was, if anything, less.)

It is much, much more than that. So now you know.

> "venerable" Maxima is mentioned once, suggesting that the only thing
> it can do is symbolic integration and numeric integration.
> Actually, while Maxima includes library access to Fortran methods, it
> is far inferior to what could be done in numeric integration,
> as demonstrated by recent Mathematica versions. You would hardly get a
> hint that 75% of the sage-support messages are about Maxima.

No they aren't.

Moreover, there isn't anything in the proposal that uses Maxima at
all. The proposal is about numpy/scipy/PDE/linear algebra/algebraic
topology/group theory/differential algebra/pynac; this is all
independent from Maxima.

> Maybe what is needed is a Fortran to Python translator.

:-)

> I think that if NSF sent the proposal over to computer science and
> engineering, it might not get a great reception, but it is hard to
> predict such things.

As I mentioned at the top of this thread, the proposal is to the the
computational mathematics program, which is part of DMS = "Division of
Mathematical Sciences".

-- William

[Simon King]

On Nov 23, 4:38 pm, William Stein <wst...@gmail.com> wrote:

> On Mon, Nov 23, 2009 at 8:28 AM, rjf <fate...@gmail.com> wrote:
> > From the proposal
>
> > ... and which has sophisti-
> > cated interfaces to nearly all other mathematics software, including
> > Mathematica, Maple,
> > MATLAB and Magma. ...
>
> > Maxima just gets no respect. :)

At the risk of writing the obvious:
There are two types of Sage's interfaces to "nearly all other"
software: To software that is distributed with Sage (Gap, Singular,
and, yes, Maxima), and to proprietary CAS. Aparently, the sentence
from the proposal is about the latter. So, I don't think there is
reason for the developers of Gap, Maxima or Singular to be upset.

Gap, Maxima and Singular *are* mentioned in the proposal (I must admit
that I find it a bit odd that the Maxima library is the only one that
is qualified as "venerable", though... :)

However, a concise list of software projects onto which Sage builds
seems missing to me (or perhaps I overlooked it). Perhaps, the phrase
"Already, Sage combines several hundred thousand lines of new code
with over 5 million lines of code from other projects." on the first
page could be extended to something like:

"Sage provides the capabilities of a wide range of open source
mathematical and non-mathematical software, such as GAP, JMol, Maxima,
Numpy, Pari/GP, R, Scipy, Singular, Tachyon and Twisted, to name just
a few of them. In addition, Sage provides features that were not
available in open source before, based on several hundred thousand
lines of new code."
(Hope that the "not available before" bit is correct...)

Cheers,
Simon

[William Stein]

On Mon, Nov 23, 2009 at 9:40 AM, Simon King <simon...@nuigalway.ie> wrote:
> On Nov 23, 4:38 pm, William Stein <wst...@gmail.com> wrote:
>> On Mon, Nov 23, 2009 at 8:28 AM, rjf <fate...@gmail.com> wrote:
>> > From the proposal
>>
>> > ... and which has sophisti-
>> > cated interfaces to nearly all other mathematics software, including
>> > Mathematica, Maple,
>> > MATLAB and Magma. ...
>>
>> > Maxima just gets no respect. :)
>
> At the risk of writing the obvious:
>  There are two types of Sage's interfaces to "nearly all other"
> software: To software that is distributed with Sage (Gap, Singular,
> and, yes, Maxima), and to proprietary CAS. Aparently, the sentence
> from the proposal is about the latter. So, I don't think there is
> reason for the developers of Gap, Maxima or Singular to be upset.
>
> Gap, Maxima and Singular *are* mentioned in the proposal (I must admit
> that I find it a bit odd that the Maxima library is the only one that
> is qualified as "venerable", though... :)

Maxima started in the 1960s, whereas Singular, Pari, GAP are from the
1990s, right? Venerable = "accorded a great deal of respect, esp.
because of age"

> However, a concise list of software projects onto which Sage builds
> seems missing to me (or perhaps I overlooked it). Perhaps, the phrase
> "Already, Sage combines several hundred thousand lines of new code
> with over 5 million lines of code from other projects." on the first
> page could be extended to something like:
>
>  "Sage provides the capabilities of a wide range of open source
> mathematical and non-mathematical software, such as GAP, JMol, Maxima,
> Numpy, Pari/GP, R, Scipy, Singular, Tachyon and Twisted, to name just
> a few of them. In addition, Sage provides features that were not
> available in open source before, based on several hundred thousand
> lines of new code."
> (Hope that the "not available before" bit is correct...)

That's a good idea.

William

[John H Palmieri]

On Nov 23, 8:28 am, rjf <fate...@gmail.com> wrote:
> From the proposal

[snip]

> And why is Cython much more than a Python to C translator?  (This is
> not sarcasm. I honestly have no idea that it was more.
> I thought it was, if anything, less.)

Look at the section on "Cython and Numerical Computation" for some
ways in which Cython extends Python's capabilities.

> You would hardly get a hint that 75% of the
> sage-support messages are about Maxima.

Did you mean "sage-flame" here? :)

--
John

[John H Palmieri]

On Nov 23, 9:55 am, William Stein <wst...@gmail.com> wrote:

> On Mon, Nov 23, 2009 at 9:40 AM, Simon King <simon.k...@nuigalway.ie> wrote:


> >  "Sage provides the capabilities of a wide range of open source
> > mathematical and non-mathematical software, such as GAP, JMol, Maxima,
> > Numpy, Pari/GP, R, Scipy, Singular, Tachyon and Twisted, to name just
> > a few of them. In addition, Sage provides features that were not
> > available in open source before, based on several hundred thousand
> > lines of new code."
> > (Hope that the "not available before" bit is correct...)
>
> That's a good idea.

I've incorporated that; the link I posted before now goes to the new
PDF file. There is also a copy of the Mercurial repository here:

http://www.math.washington.edu/~palmieri/Sage/compmath.tar.bz2

--
John

[Jaap Spies]

rjf wrote:
> From the proposal
>
>
> .... and which has sophisti-

> cated interfaces to nearly all other mathematics software, including
> Mathematica, Maple,
> MATLAB and Magma. ...
>
>
> Maxima just gets no respect. :)
>
>
> Most of the facilities mentioned are already in Maxima.
>

[...]

You are so funny. You remind me of the sour old men in the public of
the Muppet show. Complaining all the time, but not looking what's
really happening.

Exactly four years ago we had a discussion on sci.math.symbolic. You
predicted a short life for Sage, being an "elephant with feathers"!

But see! Sage is doing very well. I will not predict the extinction
of lisp based CASes, but in the Darwin Year we all may consider
the evolutionary principles of "the survival of the fittest".

Sage is definitely not the mastodon with feathers, doomed not to survive.
Sage is very much alive and kicking.

Four years ago I had a proposal: "Shall we join the party". It still stands.
Maybe you can put sarcasm aside and open your eyes (and mind) for the the
real value of Sage:

> "Sage provides the capabilities of a wide range of open source
> mathematical and non-mathematical software, such as GAP, JMol, Maxima,
> Numpy, Pari/GP, R, Scipy, Singular, Tachyon and Twisted, to name just
> a few of them. In addition, Sage provides features that were not
> available in open source before, based on several hundred thousand
> lines of new code."

(cited Simon King)

Best wishes,

Jaap Spies

[kcrisman]

>
> You are so funny. You remind me of the sour old men in the public of
> the Muppet show. Complaining all the time, but not looking what's
> really happening.

Statler and Waldorf.

Also, unless we predict the immediate demise of Sage's integration,
(hopefully soon) summation (review this please!), symbolic solving,
and several other very important things, we shouldn't predict the
demise of Lisp-based systems, even in the Darwin Year.

But kidding aside, maybe when the discussion has reached this point it
should move to sage-flame? Just throwing it out there.

- kcrisman

[Simon King]

Hi William!

On 23 Nov., 18:55, William Stein <wst...@gmail.com> wrote:
[...]

> Maxima started in the 1960s, whereas Singular, Pari, GAP are from the
> 1990s, right?   Venerable = "accorded a great deal of respect, esp.
> because of age"

My impression was that the word "venerable" (used *twice* in the
proposal) by itself is very old-fashioned and rather odd in a CAS
context. I would certainly not use the corresponding German word
"altehrwürdig" to refer to a computer algebra system or to an NSF-
funded winter school. To my ears, it has an ironic undertone.

But I am not a native speaker...
Cheers,
Simon

[John H Palmieri]

I think it sounds perhaps slightly odd, but I don't catch any irony in
it. I mean, we know something about the history and people involved
in Maxima and Sage, as well as various discussions about Maxima vs.
Sage, so it may be hard for us to read anything about both Sage and
Maxima without paying very close attention to what words are being
used and whether they are fraught with extra meaning.

I personally don't know anything about William's experience with the
Arizona Winter School beyond what he wrote in the proposal, but
presumably he liked it when he was a grad student, since he helped to
co-organize it recently. So I interpret the use of the word
"venerable" positively in that case, and I don't think there's much
ambiguity. I think it's reasonable to expect most readers (who don't
have any knowledge of past Sage-Maxima discussions) to then read the
second instance of the word in the same way, positively.

The proposal is certainly not meant to insult or snub Maxima (or any
other open-source free software) in any way.

--
John

[William Stein]

That is true. In fact, I hope in the proposal to not insult or snub
non-free commercial software either.

William

[rjf]

On Nov 23, 8:38 am, William Stein <wst...@gmail.com> wrote:

> On Mon, Nov 23, 2009 at 8:28 AM, rjf <fate...@gmail.com> wrote:
> > From the proposal
>
> > ... and which has sophisti-
> > cated interfaces to nearly all other mathematics software, including
> > Mathematica, Maple,
> > MATLAB and Magma. ...
>
> > Maxima just gets no respect. :)
> > Most of the facilities mentioned are already in Maxima.
>
> I wish...

I seem to recall that Sage people had their knickers in a knot when
the interface to Maxima's eigenvalue package
changed. Of course you might believe that Maxima has no linear
algebra, and no ODE solvers and no ... whatever.
Maybe your wish came true when you weren't looking?

>
> > And why is Cython much more than a Python to C translator?  (This is
> > not sarcasm. I honestly have no idea that it was more.
> > I thought it was, if anything, less.)
>
> It is much, much more than that.  So now you know.

That's helpful. Forgive me for asking for information. I see,
somewhat later, that this has to do with adding type declarations.
Just the ticket. To me is suggests that Python is inappropriate for
numerical work -- for which C is more appropriate. And so the veneer
of Python is insufficient. Of course using (optional) declarations in
Lisp helps some compilers to produce better numerical code. Some of
that code generation (depending on the Lisp system) is even done with
C as an intermediary. And has been done that way since, oh, 1984 or
so.

> > "venerable" Maxima is mentioned once, suggesting that the only thing
> > it can do is symbolic integration and numeric integration.
> > Actually, while Maxima includes library access to Fortran methods, it
> > is far inferior to what could be done in numeric integration,
> > as demonstrated by recent Mathematica versions. You would hardly get a
> > hint that 75% of the sage-support messages are about Maxima.
>
> No they aren't.

I have supplied the calculation already. Open up a sage-support
window. There are 3480 or so messages. Now search for "maxima" and you
will have a hit on 75% of them.

>
> Moreover, there isn't anything in the proposal that uses Maxima at
> all.  The proposal is about numpy/scipy/PDE/linear algebra/algebraic
> topology/group theory/differential algebra/pynac; this is all
> independent from Maxima.

You would think
"Unifying Mathematical Software for Scientists,
Engineers and Mathematicians" might have something to do with
Scientists and Engineers.
But my mistake. Apparently its purpose is to allow those other people
access to algebraic topology/group theory/differential algebra.
That's what unifying means to a mathematician ?

>
> > Maybe what is needed is a Fortran to Python translator.
>
> :-)
>
> > I think that if NSF sent the proposal over to computer science and
> > engineering, it might not get a great reception, but it is hard to
> > predict such things.
>
> As I mentioned at the top of this thread, the proposal is to the the
> computational mathematics program, which is part of DMS = "Division of
> Mathematical Sciences".

As you may or may not know, the various program directors at NSF are
free to "shop around" proposal that seem to them to be wrongly
categorized, interdisciplinary, or otherwise in need of others'
guidance.

RJF

>
>  -- William

[rjf]

Definition of venerable:

1. Commanding respect by virtue of age, dignity, character, or
position.
2. Worthy of reverence, especially by religious or historical
association.


For the non-native English speakers (and maybe others) in the group,
it may be useful to point out that
labeling something or someone "venerable" has a disrespectful tone, at
least to my ear.

If a piece of software is useful, or clever and continues to be
useful, one could simply say it is useful, long-lived, standard, ....
To say that
a piece of software is venerable is to imply that its major virture is
that it has lasted a long time -- perhaps too long a time--- and that
perhaps it is time for it to be replaced.

Would you, for example, refer to the venerable Todd Coxeter Algorithm
(1936) ? I don't know if there are better algorithms, but it is old.

If a paper referred to Euclid's venerable GCD algorithm, I would
assume the paper was about some improved method.

It is, of course, your NSF proposal, and you can say whatever you
wish.

RJF

[mark mcclure]

On Nov 23, 3:30 pm, William Stein <wst...@gmail.com> wrote:
> That is true.  In fact, I hope in the proposal to not insult or snub
> non-free commercial software either.

But William, just two days ago on sage-support you wrote:
"Let's put Mathworks out of business."
http://groups.google.com/group/sage-support/browse_thread/thread/b56450bab6d43bf7/a82a770db1abd6dc#msg_d72f94b7ff856c29

So I wonder, what is your true feeling on this matter?

To be clear, I am not attempting to be inflammatory.
I have been genuinely interested in Sage for a couple
years and have used it for real mathematics sporadically.
I could even imagine jumping into Sage development but I
have been *very* put off by the outright hostile stance
frequently expressed towards the commercial systems on
the sage discussion groups. As I mentioned in my
response on sage-support, The Mathworks has played a
central role in the development of numerical software
(both commercial and free) for many years. Thus, I
really feel that your desire to "put Mathworks out of
business" is misdirected, particularly in the context
of your current grant proposal.

Mark McClure

[Jaap Spies]

mark mcclure wrote:
> On Nov 23, 3:30 pm, William Stein <wst...@gmail.com> wrote:
>> That is true. In fact, I hope in the proposal to not insult or snub
>> non-free commercial software either.
>
> But William, just two days ago on sage-support you wrote:
> "Let's put Mathworks out of business."
> http://groups.google.com/group/sage-support/browse_thread/thread/b56450bab6d43bf7/a82a770db1abd6dc#msg_d72f94b7ff856c29
>

Ever heard of the word humor? Even without a smiley this is possible in
a message.

Jaap

[Alex Ghitza]

On Mon, Nov 23, 2009 at 01:04:25PM -0800, rjf wrote:

> > > Actually, while Maxima includes library access to Fortran methods, it
> > > is far inferior to what could be done in numeric integration,
> > > as demonstrated by recent Mathematica versions. You would hardly get a
> > > hint that 75% of the sage-support messages are about Maxima.
> >
> > No they aren't.
>
> I have supplied the calculation already. Open up a sage-support
> window. There are 3480 or so messages. Now search for "maxima" and you
> will have a hit on 75% of them.
>

Theorem: More than 186% of messages on sage-support are about the
usage of definite articles in the English language.

Proof: There are 16516 messages on sage-support. Search for "the",
you get 30800 results. The ensuing computation is left to the gentle
reader.




--
Alex Ghitza -- Lecturer in Mathematics -- The University of Melbourne
-- Australia -- http://www.ms.unimelb.edu.au/~aghitza/

[Jason Grout]

mark mcclure wrote:
> On Nov 23, 3:30 pm, William Stein <wst...@gmail.com> wrote:
>> That is true. In fact, I hope in the proposal to not insult or snub
>> non-free commercial software either.
>
> But William, just two days ago on sage-support you wrote:
> "Let's put Mathworks out of business."
> http://groups.google.com/group/sage-support/browse_thread/thread/b56450bab6d43bf7/a82a770db1abd6dc#msg_d72f94b7ff856c29
>
> So I wonder, what is your true feeling on this matter?
>
> To be clear, I am not attempting to be inflammatory.
> I have been genuinely interested in Sage for a couple
> years and have used it for real mathematics sporadically.
> I could even imagine jumping into Sage development but I
> have been *very* put off by the outright hostile stance
> frequently expressed towards the commercial systems on
> the sage discussion groups.

Let it be clear that the opinion expressed above is William's opinion
(maybe in a moment of frustration about the patent business?). I have a
lot of respect for Cleve and what he's done, and don't feel like my
purpose for involvement is to put Mathworks out of business. In fact
(as William has expressed elsewhere about commercial software), I'm
*glad* that Mathworks is dumping (millions of?) dollars into math
software development. That's a great thing for mathematics and
mathematics software. Patenting some of the (obvious?) things they've
patented makes me sad, though.

Jason

--
Jason Grout

[Tim Daly]

I believe Sage is a very useful piece of software for many people.
But I have had numerous discussions (the last one was at the NSF
workshop in Rhode Island prior to ECCAD) about funding open source
computational mathematics. The definitive statement seems to be:

"NSF will not fund software development that competes with
existing commercial software."

I pointed out that nowhere in Axiom's goals was there any sense of
competing with commercial software but that response was unheard.

But given the definitive reply from NSF
I am somewhat surprised that NSF funds Sage at all.

[Martin Rubey]

I, as a german speaking, can see no humor in that post you link to. And
I had difficulty "getting" William's humour in the past already several
times. (Even though I lived in a british family in Australia for one
year as an exchange student. My english is far from perfect, but I'd be
surprised if the english of *all* others on sage-devel is better than
mine.)

That said, I would like to make it clear that in no way I endorse
software patents, or as a matter of fact, any patents at all.

Change of topic :-)

Reading Section 8, I gather sage has no interest in FriCAS abilities in
differential algebra? I'd be interested to know, because I was going to
put in some effort, specifically, symbolic treatment of ADE's and
recurrences.

Martin

[Gonzalo Tornaria]

On Mon, Nov 23, 2009 at 7:04 PM, rjf <fat...@gmail.com> wrote:
> That's helpful.  Forgive me for asking for information.  I see,
> somewhat later, that this has to do with adding type declarations.
> Just the ticket.  To me is suggests that Python is inappropriate for
> numerical work -- for which C is more appropriate.

Indeed, Python is inappropriate for programming anything which is
cpu-intensive. IMHO, were not for cython/pyrex, sage would be doomed.
Actually, I believe pyrex is one of the main reasons William settled
with python for sage.

> And so the veneer
> of Python is insufficient.  Of course using (optional) declarations in
> Lisp helps some compilers to produce better numerical code. Some of
> that code generation (depending on the Lisp system) is even done with
> C as an intermediary.  And has been done that way since, oh, 1984 or
> so.

The goal of Python/Cython is to code with *less* parenthesis than C.
Other than that, it's probably the same venerable idea.

Gonzalo

[Nick Alexander]

It's time for this thread to move to sage-flame. Please no more
messages on sage-devel.

Nick

[Jaap Spies]

Martin Rubey wrote:
> Jaap Spies <j.s...@hccnet.nl> writes:
>
>> mark mcclure wrote:
>>> On Nov 23, 3:30 pm, William Stein <wst...@gmail.com> wrote:
>>>> That is true. In fact, I hope in the proposal to not insult or snub
>>>> non-free commercial software either.
>>> But William, just two days ago on sage-support you wrote:
>>> "Let's put Mathworks out of business."
>>> http://groups.google.com/group/sage-support/browse_thread/thread/b56450bab6d43bf7/a82a770db1abd6dc#msg_d72f94b7ff856c29
>>>
>> Ever heard of the word humor? Even without a smiley this is possible in
>> a message.
>
> I, as a german speaking, can see no humor in that post you link to. And
> I had difficulty "getting" William's humour in the past already several
> times. (Even though I lived in a british family in Australia for one
> year as an exchange student. My english is far from perfect, but I'd be
> surprised if the english of *all* others on sage-devel is better than
> mine.)
>

Nothing to do with language. Knowing the real situation of Mathworks (Matlab) <-> Sage
makes me laugh!

Jaap

[mark mcclure]

On Nov 23, 4:56 pm, Martin Rubey <martin.ru...@math.uni-hannover.de>
wrote:

> Jaap Spies <j.sp...@hccnet.nl> writes:
> > Ever heard of the word humor? Even without a smiley this is possible in
> > a message.
>
> I, as a german speaking, can see no humor in that post you link to.  

I guess I didn't either. In the overall context of the post, the
statement
seemed to be made in anger.

> That said, I would like to make it clear that in no way I endorse
> software patents, or as a matter of fact, any patents at all.

I understand this sentiment and at times share in the frustration
arising
from software restrictions. As a practical matter, however, I
generally
use whatever software seems best suited for *me* to whatever
particular
problem at hand. Part of the question I am posing (which I think is
relevant to sage development) is "how does Sage expand its user
base to include many people (like myself) who don't feel
philosophically
bound to open source"? Sage cannot expand to the 10^6 level by
alienating these types of users.

Mark

[rjf]

On Nov 23, 1:33 pm, Alex Ghitza <aghi...@gmail.com> wrote:
> On Mon, Nov 23, 2009 at 01:04:25PM -0800, rjf wrote:
> > > > Actually, while Maxima includes library access to Fortran methods, it
> > > > is far inferior to what could be done in numeric integration,
> > > > as demonstrated by recent Mathematica versions. You would hardly get a
> > > > hint that 75% of the sage-support messages are about Maxima.
>
> > > No they aren't.
>
> > I have supplied the calculation already.  Open up a sage-support
> > window. There are 3480 or so messages. Now search for "maxima" and you
> > will have a hit on 75% of them.
>
> Theorem: More than 186% of messages on sage-support are about the
> usage of definite articles in the English language.

read about "noise words" and search, and why searching for such words
is
not a good idea.

Here is a noise word list for English,

about,after,all,also,an,and,another,any,are,as,at,be,because,been,before
being,between,both,but,by,came,can,come,could,did,do,each,for,from,get
got,has,had,he,have,her,here,him,himself,his,how,if,in,into,is,it,like
make,many,me,might,more,most,much,must,my,never,now,of,on,only,or,other
our,out,over,said,same,see,should,since,some,still,such,take,than,that
the,their,them,then,there,these,they,this,those,through,to,too,under,up
very,was,way,we,well,were,what,where,which,while,who,with,would,you,your,a
b,c,d,e,f,g,h,i,j,k,l,m,n,o,p,q,r,s,t,u,v,w,x,y,z,$,
1,2,3,4,5,6,7,8,9,0,_

Note that "maxima" is not a noise word. "The" is a noise word.

I find it hard, sometimes, to detect humor on this mailing list, so I
don't know if you really
are making a joke and are implicitly agreeing with me, or if you
actually somehow think
that searching for noise words is a proof of relevance.

Anyway, I think that it is clear that sage-support is often discussing
something
related ---as I said, for good or ill-- about maxima. Is there really
support for an
alternative view??

[rjf]

Is the topic of "how should a Sage proposal be written so that it is
funded by NSF" really something to be relegated to sage-flame?
I don't know how many other readers here have (repeatedly) served as
NSF reviewers or panelists evaluating proposals.

Based on my contributions to the writing of this proposal, I would
probably disqualify myself from reviewing it.
Lucky you :)

[William Stein]

Mission accomplished!

http://www2.warwick.ac.uk/fac/soc/pais/currentstudents/moduledocs/pgmodules/is/whatisinternationalsecurity/bush_mission_accomplished.jpg

William

[William Stein]

On Mon, Nov 23, 2009 at 1:04 PM, rjf <fat...@gmail.com> wrote:
>> > "venerable" Maxima is mentioned once, suggesting that the only thing
>> > it can do is symbolic integration and numeric integration.
>> > Actually, while Maxima includes library access to Fortran methods, it
>> > is far inferior to what could be done in numeric integration,
>> > as demonstrated by recent Mathematica versions. You would hardly get a
>> > hint that 75% of the sage-support messages are about Maxima.
>>
>> No they aren't.
>
> I have supplied the calculation already.  Open up a sage-support
> window. There are 3480 or so messages. Now search for "maxima" and you
> will have a hit on 75% of them.

That metric does not give 75%. This page shows that there are a
total of 16521 messages in sage-support:

http://groups.google.com/group/sage-support/about

Searching maxima yields 2620 results:

http://groups.google.com/group/sage-support/search?group=sage-support&q=maxima&qt_g=Search+this+group

We have

2620/16521 = 0.158586042007...

or "about 16%".

CAVEAT: I don't trust Google groups search results. Take the above
with a grain of salt.


William

[rjf]

On Nov 23, 3:49 pm, William Stein <wst...@gmail.com> wrote:

> On Mon, Nov 23, 2009 at 1:04 PM,rjf<fate...@gmail.com> wrote:
> >> > "venerable" Maxima is mentioned once, suggesting that the only thing
> >> > it can do is symbolic integration and numeric integration.
> >> > Actually, while Maxima includes library access to Fortran methods, it
> >> > is far inferior to what could be done in numeric integration,
> >> > as demonstrated by recent Mathematica versions. You would hardly get a
> >> > hint that 75% of the sage-support messages are about Maxima.
>
> >> No they aren't.
>
> > I have supplied the calculation already.  Open up a sage-support
> > window. There are 3480 or so messages. Now search for "maxima" and you
> > will have a hit on 75% of them.
>
> That metric does not give 75%.     This page shows that there are a
> total of 16521 messages in sage-support:
>
>    http://groups.google.com/group/sage-support/about
>
> Searching maxima yields 2620 results:
>

>    http://groups.google.com/group/sage-support/search?group=sage-support...

>
> We have
>
>    2620/16521 = 0.158586042007...
>
> or "about 16%".
>
> CAVEAT: I don't trust Google groups search results.   Take the above
> with a grain of salt.
>
> William

You are right! I was taking the number of MESSAGES and dividing by
the number of DISCUSSIONS.
So the average discussion has about 16500/3400 or 4.8 messages.
The count I was looking at was not the relevant one.
Thanks for examining the statistics. Frankly, I was surprised at the
75%, but didn't understand where I went wrong.


So only about a sixth of all support messages mention maxima. Given
that there are some 100 components to Sage
I think that is still quite a large percentage.

It may be worthwhile pondering Tim's comment...

"NSF will not fund software development that competes with
existing commercial software."

Here are two themes for proposals:

1. "we are doing this great new mathematics and the best way to do
this work is for us to write programs and run them."

2. "we are writing this program so that people (engineers,
mathematicians, students) will have a viable alternative to
Mathematica."

Which kind of proposal did NSF fund last time?

[Simon King]

Hi All!

On 24 Nov., 01:15, rjf <fate...@gmail.com> wrote:
[...]

> It may be worthwhile pondering Tim's comment...
>
>   "NSF will not fund software development that competes with
>    existing commercial software."

Indeed, that's irritating. What exactly does NSF mean?

Do they mean (1) "no fund for software that competes economically with
existing commercial software" (Sage doesn't, because it is free as in
beer), or do they mean (2) "no fund for software development that
competes quality-wise with existing commercial software"?

I could understand if NSF doesn't want to fund software developers
that want to make money by competing with commercial CAS. But I could
not understand if they denied their support to non-commercial high-
quality software projects.

Cheers,
Simon

[kcrisman]

>
> > It may be worthwhile pondering Tim's comment...
>
> >   "NSF will not fund software development that competes with
> >    existing commercial software."
>
> Indeed, that's irritating. What exactly does NSF mean?
>
> Do they mean (1) "no fund for software that competes economically with
> existing commercial software" (Sage  doesn't, because it is free as in
> beer), or do they mean (2) "no fund for software development that
> competes quality-wise with existing commercial software"?
>
> I could understand if NSF doesn't want to fund software developers
> that want to make money by competing with commercial CAS. But I could
> not understand if they denied their support to non-commercial high-
> quality software projects.

My understanding, from a number of similar conversations, is that NSF
doesn't feel comfortable funding proposals whose goal is to implement
something essentially in commercial programs.

However, if you are implementing an improvement to something that is
not software, and need to improve a certain piece of software in order
to do this (for instance, some curricular thingie), you can be ok. I
think Geogebra would be a case in point, but there are plenty of
others.

Or, if you are implementing something that really doesn't exist in any
other software, or not to that efficiency, you can be funded to do it
in an existing package. I presume this is how Sage has gotten most of
its funding thus far, though no one has directly responded to rjf's
question on this point. If I'm not mistaken, the abstracts of those
proposals are available on the NSF website?

By the way, apropos of computational mathematics, AMS has a new book
out on computational topology for beginners... sounds like it's time
to implement persistent homology? I didn't see that on the list.
Certainly it could be worth contacting some of those folks, unless
it's in Sage... http://chomp.rutgers.edu/ has GPL software with a
Python GUI, for instance, and presumably people like Robert Ghrist are
doing lots with this computationally...

- kcrisman

[Tim Daly]

Mathematica is a commercial, for-profit organization that is the current
market leader
in computational mathematics. The NSF has said they will not fund
software development
that competes in that market. Exactly what "competes" means is open to
interpretation.

Axiom's goals are all research and educational, which is exactly the
reason why NSF exists.
I hardly expect Mathematica to release a fully open, fully literate
version that can be used
as teaching material for computational mathematics so I don't see the
competitive aspects.

As I understand Sage's stated goal, the project IS intended to compete
with, and displace
the current commercial software products of Mathematica, Maple, Magma,
and Matlab.
(I may be wrong but that's my impression from reading the mailing list
and other sources).

So, given the premise that NSF will not fund competing software
and Sage is competing software
therefore..... Sage will get the lion's share of funding

Because logic is not a required subject.

I DO hope that NSF funds Sage heavily. I want you guys to be very
successful.
I want the next generation to HAVE a computational mathematics program
that is
open source because I fear that the failure of the commercial software
will bring
the whole subject into the dark ages.

Does Sage also compete with Axiom? Ummm, nobody cares, including me.

Tim

[William Stein]

Judge for yourself: http://wstein.org/grants/sage-06/

The only fixed rule I know about NSF is that they fund stuff that
"promotes the progress of science; advances the national health,
prosperity, and welfare; secures the national defense…" They also
specifically "Foster and support the development and use of computers
and other scientific methods and technologies, primarily for research
and education in the sciences."
See http://www.nsf.gov/about/glance.jsp

They are an organization of _people_ that continually reinvents
itself, subject to the guidelines at the above page, regularly gets
new directors, employees, etc. And they are generally relatively
honest and scrupulous about avoiding conflict of interest. I like
'em.

-- William

[Tim Daly]

Is there a plan for what happens if funding is not approved?
Do the servers continue? Are the students reassigned?
Do the Sage days continue? Does the code move to
sourceforge or github? Does this become a free-time only,
non-academic activity? Once Sage becomes a non-academic
"free and open source" project there is no hope of getting
NSF to pay attention again. (Axiom used to be funded at IBM).

Computational mathematics suffers from one critical flaw
(which I pointed out to the NSF). Someone in an academic
position only gets tenure track credit for published PAPERS.
Source code does not count and it is a net loss to write
working code if you are on a tenure track. This makes it
hard for someone to aspire to a tenure position and also
participate effectively in open-source computational math.

So if (when) Sage is no longer a "hot topic", funded, and a
top-of-the-interest-pile for new papers then the momentum
slows a bit. NSF funding is critical for keeping the conference
papers flowing. Without it there will be no papers and Sage
will become "venerable" software.

It is very difficult to run a large open source project with
no funding at all. Axiom costs me between $4000 and $5000
per year for hosting costs, equipment, personal travel, free
books (I buy old copies of the Axiom book and give them
away), etc. So while Axiom is "technically free-as-in-beer"
the reality is that it comes out of my personal budget.

I assume that Sage has similar direct and indirect costs.
If these are no longer receiving NSF and/or academic
funding (e.g. free server hosting) what becomes of the
Sage project?

Tim

[William Stein]

Hi Tim,

Thanks for airing your concerns and worries.

On Mon, Nov 23, 2009 at 8:12 PM, Tim Daly <da...@axiom-developer.org> wrote:
> Is there a plan for what happens if funding is not approved?

Yes.

> Do the servers continue?

I'm not asking for funding for servers. The actual hardware is
owned by UW (but paid for by the NSF), and all hosting of those
servers is graciously done by the mathematics department at UW. This
is an extremely valuable perk of my job, and of being at a university
that understands software development, perhaps due to the influence of
Microsoft, Adobe, and the many other local software companies here. I
sysadmin them.

> Are the students reassigned?

I'm not asking for funding for students.

> Do the Sage days continue?

Yes (see below for more). There would just be less of them.

> Does the code move to sourceforge or github?

No. Though if it weren't for their limited quotas we would already
have a mirror there.

> Does this become a free-time only, non-academic activity?

No. The Sage project has a broad base of support:
http://sagemath.org/development-ack.html

Also, there is some revenue from people buying the tutorial on
Amazon.com, donations from individuals, etc. People make
tax-deductible donations here:

https://secure.gifts.washington.edu/as_mathematics/gift.asp?page=make&Code=MATSAG

That money is extremely important, but we're not dependent on it.

Regarding students, there are a *bunch* (6 or 7?) of UW undergrads
working on Sage right now. Some are doing it for fun. Most are
getting paid some amount. Currently, the ones that are getting paid
are getting paid from an NSF grant from the Applied/PDE group in the
department---the UW math department is amazingly collegial, and the
different groups try to support each other and work together as a team
(rather than against each other). Just like we should *all* do as
math software projects.

> Once Sage becomes a non-academic
> "free and open source" project there is no hope of getting
> NSF to pay attention again.

From my perspective, Sage had no direct financial support for about
half the time it has been around. Despite the generous support the
project has received during the last 2-3 years, I think it is
definitely not dependent on support. Just to emphasize this, consider
that most release management for the last 5 months has been done by
Minh Nguyen (an undergrad in Australia) and Mike Hansen, and that was
all 100% volunteer work.

Sage is dependent on something: it has to continue to satisfy genuine
user demand and be a fun project to work on in order to remain strong
and flourish.

> Computational mathematics suffers from one critical flaw
> (which I pointed out to the NSF). Someone in an academic
> position only gets tenure track credit for published PAPERS.
> Source code does not count and it is a net loss to write
> working code if you are on a tenure track. This makes it
> hard for someone to aspire to a tenure position and also
> participate effectively in open-source computational math.

There is some truth to this, but in my experience it also
oversimplifies things. I have repeatedly seen exciting doors open to
young mathematicians as a direct result of quality coding and
computational work they have done. Also, in my experience some
computational work has had trouble being accepted by serious
mathematicians, partly because -- like classical Italian "proofs" in
Algebraic Geometry -- it violates accepted mathematical standards by
being built on top closed source commercial software. It's not
surprising that a very young field, relative to research math, which
is thousands (!) of years old, would have some growing pains.
Perhaps all our work on Sage, Axiom, Maxima, PARI, etc., is an
important foundation that will give young mathematicians a chance to
address this problem, just like the work of Bourbaki and others was in
making pure mathematics more rigorous and respectable. I decided
that instead of complaining, I would do work toward earning the
respect of the mathematical community. I am glad that you have also
put so much effort into doing the same (and you started years before
me).

Mathematica, Matlab, Maple, and Magma have the same place in genuine
quality research mathematics as non-rigorous proofs. They are useful
-- sometimes wonderful, inspiring, and incredibly useful (!) -- but
they aren't the real deal.

> So if (when) Sage is no longer a "hot topic", funded, and a
> top-of-the-interest-pile for new papers then the momentum
> slows a bit. NSF funding is critical for keeping the conference
> papers flowing.

I don't know what you mean by "papers". However, there could
certainly be Sage Days even without direct NSF funding -- just less of
them. To prove this, let's take the next 10 already planned Sage
Days (all in the next 9 months!), and in the interest of full
disclosure, I'll describe their funding sources:

# Sage Days 18 -- Cambridge, MA (December 1-5, 2009); (funded by the
Clay Math Institute and MIT)

# Sage Days 19 -- Seattle, WA (at Shuey House Mansion, January 16-20,
2010); (funded by the NSF)

# Sage Days 20 -- Marseille, France (February 22-26, 2010); funding: France

# Sage Days 20.5 -- Toronto, Canada (May 3-7, 2010); funding: Canadian govt.

# Sage Days 21 -- Seattle, WA (June?); funding: tentative, probably
PIMS = mostly Canada; otherwise, NSF

# Sage Days 22 -- Berkeley, CA (June 21 to July 2, 2010); funding by
MSRI (= partly by MSRI sponsoring institutions)

# Sage Days 23 -- Leiden, Netherlands (July 5-9, 2010); theme: Number
theory (funded by Holland)

# Sage Days 24 -- Linz, Austria (July 17-22, 2010); theme: (funded by
RISC = Austria)

# Sage Days 25 -- Mumbai, India (August 9-12, 2010); funded by India
(some national grant)

# Sage Days 26 -- Kaiserslautern, Germany (August 27-31, 2010); funded
by the German government

As you can see above, funding of these Sage Days is not dominated by
an NSF grant directly to me. It's shared between many countries,
including France, Germany, Austria, India, Canada, etc., and a private
foundation (Clay). Why? It's due to the grassroots effort of a
_lot_ of mathematicians all over the world who have decided to put in
a huge amount of effort writing their own small grant proposals for
conference funding. This model for funding is sustainable so long as
we all continue to work together as a team. The NSF proposal that
forms the main topic of this thread is simply a request to NSF to help
me do my part.

> Without it there will be no papers and Sage
> will become "venerable" software.

I hope someday Sage becomes venerable. Only RJF could twist
venerable = "calling forth respect through age, character, and
attainments; broadly : conveying an impression of aged goodness and
benevolence" into something negative....

> It is very difficult to run a large open source project with
> no funding at all. Axiom costs me between $4000 and $5000
> per year for hosting costs, equipment, personal travel, free
> books (I buy old copies of the Axiom book and give them
> away), etc. So while Axiom is "technically free-as-in-beer"
> the reality is that it comes out of my personal budget.
>
> I assume that Sage has similar direct and indirect costs.

There are no costs for hosting, as I mentioned above.

> If these are no longer receiving NSF and/or academic
> funding (e.g. free server hosting) what becomes of the
> Sage project?

It thrives, less.

-- William

[Robert Bradshaw]

On Nov 23, 2009, at 8:12 PM, Tim Daly wrote:

> Is there a plan for what happens if funding is not approved?
> Do the servers continue?

Yes.

> Are the students reassigned?

There aren't any students being funded by unconfirmed funding, and the
grant isn't about funding students either.

> Do the Sage days continue?

Whenever we can get funding for them. Many other sources have funded
(or helped fund) a Sage days or two. There wouldn't be as many though.
Given that Sage days are a major source of development (and the
reasons are social as much as anything else) it would slow things down.

> Does the code move to sourceforge or github?

Not for the foreseeable future.

> Does this become a free-time only, non-academic activity?

It already is for many people, but so few people are actually funded
to work on Sage to any degree that I doubt it would have a major
impact there.

> Once Sage becomes a non-academic
> "free and open source" project there is no hope of getting
> NSF to pay attention again. (Axiom used to be funded at IBM).
>
> Computational mathematics suffers from one critical flaw
> (which I pointed out to the NSF). Someone in an academic
> position only gets tenure track credit for published PAPERS.
> Source code does not count and it is a net loss to write
> working code if you are on a tenure track. This makes it
> hard for someone to aspire to a tenure position and also
> participate effectively in open-source computational math.

Yes, it is unfortunate that this is currently the case.

> So if (when) Sage is no longer a "hot topic", funded, and a
> top-of-the-interest-pile for new papers then the momentum
> slows a bit. NSF funding is critical for keeping the conference
> papers flowing. Without it there will be no papers and Sage
> will become "venerable" software.

This is one of the big differences between Axiom (at least the
impression I get about it) and Sage. Axiom's goal was to build a CAS,
and people published papers about building computer algebra systems
and about Axiom itself. Most people working on Sage are motivated by
getting the tools they need to do research (and teach). If you look at
the list of publications using Sage, by far the minority are actually
about Sage. People who use Sage to do cutting-edge research,
especially where there are no existing tools or code, will continue to
contribute as part of their research job without explicit funding for
Sage. Teachers will continue to improve Sage to make it a better tool
for their teaching. And some people will contribute just for fun.

> It is very difficult to run a large open source project with
> no funding at all. Axiom costs me between $4000 and $5000
> per year for hosting costs, equipment, personal travel, free
> books (I buy old copies of the Axiom book and give them
> away), etc. So while Axiom is "technically free-as-in-beer"
> the reality is that it comes out of my personal budget.
>
> I assume that Sage has similar direct and indirect costs.
> If these are no longer receiving NSF and/or academic
> funding (e.g. free server hosting) what becomes of the
> Sage project?

Good question. UW is proud to to be associated with Sage and provide
hosting support (they own the hardware that was specifically
purchased, via a grant, for Sage)--I don't see that changing anytime
soon. They've been hosting the open-source program "pine" for two
decades, and I bet that's not the only example. Other than that the
costs to keep going are low, though more funding will help us get to
where we're going faster.

- Robert

[Florent Hivert]

Hi There,

Speaking specifically about Mathworks, I just want to recall what happen
to MuPAD:

http://en.wikipedia.org/wiki/MuPAD

"In September 2008, Sciface was purchased by Mathworks and the Mupad code was
included in the Symbolic Math Toolbox add-on for MATLAB. On 28 September 2008,
Mupad was withdrawn from market as a software product in its own right."

See:

http://www.mathworks.com/support/faq/mupad.html

In the operation, several friend there lost their job (actually this in not
Mathworks fault and probably the result of Sciface being bought by Mathworks is
that some of them keep their job)...

I'm not sure anyone in this sharks market will care a cent about putting sage
down...Now, though I completely agree with sage goal, I don't feel very
comfortable with the motto:
"open source alternative to Magma, Maple, Mathematica, and MATLAB."

I clearly speaks about competition from the beginning...

Just my two cents...

Cheers,

Florent

[Harald Schilly]

On Nov 24, 7:20 am, William Stein <wst...@gmail.com> wrote:
> It's shared between many countries,
> including France, Germany, Austria, India, Canada, etc.,

Since I do statistics, I want to add that since this fall Sage is by
far no longer an us-only project. There is more activity from europe
than from north america. Therefore, I am sure more and more european/
eu agencies will fund Sage which would make Sage even more
international and independent of nsf!

H

[William Stein]

> Florent

How do you imagine that one of those sharks could "put Sage down"?
I'm genuinely curious, because I don't see how *they* can. The
obvious attacks that come to mind are:

* Sage is totally immune to what happened to Mupad, since the
copyright ownership of Sage is spread across hundreds of individuals.

* A direct FUD attack would likely raise Sage's profile (since
people would wonder why Ma* is suddenly so worried).

* A frivolous patent lawsuit. This could scare people and hurt
individuals. But is it a likely course of action by the Ma's, given
their past activities? Also, similar bullying against open source,
e.g., by SCO have not gone so well for the aggressors.

Now, having actually talked with people from at least Magma,
Mathematica, and Maple, I think the more likely attack is:

* Work their assess off to innovate and create a better product.

It's an honest "attack", and that's exactly what I would expect from
the people working at the Ma's. Perhaps, I'm young and naive, but I
think they are by and large good and honest people. I think this
sort of competition is an overall plus for end users, in the long run.

-- William

[Dr. David Kirkby]

Florent Hivert wrote:

> I'm not sure anyone in this sharks market will care a cent about putting sage
> down...

I suspect some of the employees at Mathworks probably welcome Sage, though they
would not say so officially.

> Now, though I completely agree with sage goal, I don't feel very
> comfortable with the motto:
> "open source alternative to Magma, Maple, Mathematica, and MATLAB."
>
> I clearly speaks about competition from the beginning...

If Sages motto was "to put Mathworks, Wolfram Research & Maplesoft out of
business", then I for one would not want to develop for Sage.

<joke> The only good thing about Wolfram Research going out of business would be
a dent in the ego of Steven Wolfram, but that is so large, I doubt anyone would
notice</joke>

What is so wrong about competition? In the UK, we have a 'competition commission'

http://www.competition-commission.org.uk/

to ensure there is competition. It is seen by many as a good thing.

There are two major aircraft manufactures - Boeing and Airbus. Each has an
incentive to produce better planes, which are more fuel efficient, to win
customers. Do you really think they would invest as much time in making planes
more fuel efficient, if airlines had no choice but to buy from them?

Would you not agree that competition between Intel and AMD forces each to spend
money on R+D to win sales from the other? Does that not have a result that CPUs
are better than they should be? Is the same not true of the producers of
graphics cards?

I know there was a discussion on here some time back that Mathematica's
PrimePi[] was a bit quicker than Sages. But some Sage developers were aware of a
faster algorithm. I'm not sure if that ever got implemented, but to beat
Mathematica in this way can only be a good thing. No doubt Wolfram Research will
respond.

Competition is seen by many as a good thing most, though clearly if you are the
leader in a market and customers have no choice but to go to you, then
competition is the last thing you want. You would probably try to buy the
competition, which is where the competition commission will step in, and
sometimes blocks this.

BAA has been forced to sell one of its London airports, as it delivers poor
service and has no real incentive to do better as customers do not have any choice.

There may be a 1001 reasons not to use Sage, but the fact it is competition with
commercial producers of similar software is not one of them.



Dave

[Dr. David Kirkby]

William Stein wrote:

> It's an honest "attack", and that's exactly what I would expect from
> the people working at the Ma's. Perhaps, I'm young and naive, but I
> think they are by and large good and honest people. I think this
> sort of competition is an overall plus for end users, in the long run.
>
> -- William
>

I suspect I'm quite a bit older than you (born in 1963), but I have the same
opinion. Competition is good.

[William Stein]

This is current active work in progress by two UW Undergraduates:
Kevin Stueve and Andrew Ohana.
See

http://trac.sagemath.org/sage_trac/ticket/7013

The last activity on that ticket was a day ago.

[rjf]

On Nov 24, 6:54 am, William Stein <wst...@gmail.com> wrote:
> On Tue, Nov 24, 2009 at 1:15 AM, Florent Hivert
>
>
>
>

> How do you imagine that one of those sharks could "put Sage down"?

they don't have to. they just ignore it.

> I'm genuinely curious, because I don't see how *they* can.  The
> obvious attacks that come to mind are:
>
>    * Sage is totally immune to what happened to Mupad, since the
> copyright ownership of Sage is spread across hundreds of individuals.

No, it's not. Say that William is refused NSF funding and other
sources of support dry up.
(Like Mupad's situation, I think). Someone says to him, isn't it
possible to come up with
a new really neat interface for Sage -- say a proprietary closed-
source Windows version,
for, say, a million dollars? And that can use the back-end free "sage-
stuff" without violated GPL
because sage is still there, separately, in the background. (I don't
want to get into GPL
discussion here, please).

This is not to say it is likely, since I do not see anyone offering
William a million dollars for this.
And there are other implausibilities in the hypotheses as well.

>
>    * A direct FUD attack would likely raise Sage's profile (since
> people would wonder why Ma* is suddenly so worried).

I imagine they are not worried, and they wouldn't bother. Did
Mathematica attack Mupad?

>
>    * A frivolous patent lawsuit.  This could scare people and hurt
> individuals.   But is it a likely course of action by the Ma's, given
> their past activities?  Also, similar bullying against open source,
> e.g., by SCO have not gone so well for the aggressors.

So far as I can tell, only WRI engages in legal threats. Sage has no
protection against suits,
but it could happen. Let's say that someone writing one of those 100
components happens
to have actually truly copied proprietary source code from somewhere.
(You know, a student
in a summer job, had access to <whatever>). Sage could be (quite
reasonably) forced to
remove (or rewrite) that piece. This might be a minor inconvenience.

>
> Now, having actually talked with people from at least Magma,
> Mathematica, and Maple, I think the more likely attack is:
>
>    * Work their assess off to innovate and create a better product.

I think that they do not actually have to create a better product if
they already have one,
but putting that aside, they can just spend more money on marketing.

>
> It's an honest "attack", and that's exactly what I would expect from
> the people working at the Ma's

Depends on how you feel about marketing. My impression is that Sage
people are
very happy to do marketing, which includes a range of activities from
announcements,
puffery, innocent or deliberately misleading statements.

.  Perhaps, I'm young and naive, but I
> think they are by and large good and honest people.  

I think you have had insufficient contact with the software marketing

people.

>  I think this
> sort of competition is an overall plus for end users, in the long run.

Actually, there is an alternative view, at least as follows:

Take the limited amount of funding from US and other governments, plus
charities/foundations (like Clay), and say you want to "build the
world's best
mathematical system [whatever that means]".
Which makes better sense to you:

1. For each of N persons who says "I want to build a competitive
system from scratch" give
that person 1/N of the available money and say "enter the
competition".

or
2. Encourage activity at the frontier: people who take the most
advanced of the existing systems and
push it further.

Now William may consider that he is doing (2). I consider that, to
the extent that he is encouraging
the rewriting of existing code (in Python) it is more like (1).
[Though re-using existing systems has
aspects of (2)]

RJF

[rjf]

On Nov 24, 6:55 am, "Dr. David Kirkby" <david.kir...@onetel.net>
wrote:
.....

>
> What is so wrong about competition? In the UK, we have a 'competition commission'
>
> http://www.competition-commission.org.uk/

Since Sage is given away, Sage makes no money when more people use
it. I think the analogy with
Boeing and Airbus is not as good as you suppose.

How do you feel about "competition" between 2 organizations attempting
to give away the same thing, more or less?

Does that improve the product?

(I have not followed in detail, but there may be an example in the
fork of Emacs code to several products. All free?)

Does the existence of "A new front end for Maxima written in --
tada-- python" competing with "A new front end for Maxima
written in --tada--X11" competing with .. TCL ...,
OpenGraphics, ... etc etc ... actually improve Maxima?
While such a competition is relatively harmless when it occupies one
person (who is an enthusiastic programmer in some
graphics package) for a brief time, and typically that person doesn't
know much about Maxima or math, and it involves
potentially interesting design issues that are still in turmoil in the
user-interface community --- it could become
unproductive if many people are encouraged to (say) spend their time
duplicating complicated algorithms, or worse, writing
naive versions of complicated algorithms -- that do not quite do the
hardest stuff -- . That kind of competition is not helpful, in my
view.


RJF

[rjf]

Oh great, another prime counting function, and you think that WRI will
be compelled to respond because Sage has some undergraduate project on
the topic.
Sure.

[Dr. David Kirkby]

rjf wrote:
>
> On Nov 24, 6:54 am, William Stein <wst...@gmail.com> wrote:
>> On Tue, Nov 24, 2009 at 1:15 AM, Florent Hivert
>>
>>
>>
>>
>> How do you imagine that one of those sharks could "put Sage down"?
>
> they don't have to. they just ignore it.

That's why searches on 'sage math' in Google bring up sponsored advertisements
paid for by Wolfram Research. Perhaps I am thick, but that suggests to me
Wolfram Research are not ignoring Sage.


Clearly though, Wolfram Research ignore Sage more than you, as it seems to be a
hobby of yours to attack Sage. Unless you do it in your free time, I'm surprised
your head of department does not suggest you devote your time to something more
constructive.

Dave

[William Stein]

On Tue, Nov 24, 2009 at 8:49 AM, rjf <fat...@gmail.com> wrote:
[...]

Many thanks for your comments. However, I'm concerned that your
level of rudeness is not appropriate
for this list.

-- William

[Harald Schilly]

On Nov 24, 6:05 pm, "Dr. David Kirkby" <david.kir...@onetel.net>
wrote:

> That's why searches on 'sage math' in Google bring up sponsored advertisements
> paid for by  Wolfram Research. Perhaps I am thick, but that suggests to me
> Wolfram Research are not ignoring Sage.

I don't know how they have tuned their campaign, but advertising works
like an auction, where you bid on words. My guess is that they bid a
high value on the word "math" and that "blah math", "math foo" and
"foo math bar" might trigger the same ads. So, it's not sage specific.
(also, don't forget, there is also the finance software sage ... wri
could be interested in searches for that one, too! ... although, in
general, you are not allowed to set any trigger keywords for
registered trademarks!)

h

[Dr. David Kirkby]

I do not claim to know how the ads work, but I have tried

1) foo math
2) bar math
3) accounting math
4) astronomy math

5) sage math
6) wolfram math

Only 5 and 6 generate the advertisements from Wolfram Research.

Dave

[William Stein]

The ad I get is usually for Mathematica "Home Edition" for $295. It
might be fun to make a spoof of their ad advertising Sage
"Professional Edition" for $0.

-- William

[rjf]

I do not have any special knowledge about why sage math leads to a
wolfram ad, but

perhaps Google has noticed that the pages that are about sage often
mention Mathematica.
Rather prominently, too.

The pages about foo don't mention Mathematica much.

Therefore sage math is in closer proximity to the target audience.

[Florent Hivert]

Hi William,

> > In the operation, several friend there lost their job (actually this in not
> > Mathworks fault and probably the result of Sciface being bought by Mathworks is
> > that some of them keep their job)...
> >
> > I'm not sure anyone in this sharks market will care a cent about putting sage
> > down...
> > Florent
>
> How do you imagine that one of those sharks could "put Sage down"?
> I'm genuinely curious, because I don't see how *they* can. The
> obvious attacks that come to mind are:

I wrote my e-mail in a kind of hurry. And my English is not good enough to
reflect with precision what I had in mind... Let me try to precise it more.

One of the main reason I considered the switch from MuPAD to sage is exactly
because Sage is now sufficiently big enough that I'm sure It will never be
controled by a private company. As far as I know (I'm too young to have
witnessed those event), MuPAD and maple as well started as a projet in some
university. And at some point they where brought by some private company. And
I know for a fact that MuPAD had been under attack by one of those company
long before they were fired out by the University of Paderborn.

> * Sage is totally immune to what happened to Mupad, since the
> copyright ownership of Sage is spread across hundreds of individuals.

And I'm very glad for that.

Cheers,

Florent