Hi, I posted this message (see below) in response to this Sage-related thread on sci.math.symbolic that Jaap pointed out to me:
http://groups.google.com/group/sci.math.symbolic/browse_thread/thread/9ac833fc15d6f44a/eca285de0b86c5a7 It is mostly about Sage, so might be of interest to readers of sage-devel. On May 13, 3:50 pm, Dave <[EMAIL PROTECTED]> wrote: > 2 wrote: > > "Dave" <[EMAIL PROTECTED]> wrote > > >> I think comparing to Sage would be reasonable verification. > > > Certainly, what you say is correct, but to what extent has Sage penetrated > > the marketplace compared to Maple or Mathematica (or muMath which I also > > learned, or its successor Derive)? > > It's hard to see how you can use the word 'marketplace' when Sage is free. > > I've no idea of the relative number of users of Sage vs Maple or > Mathematica, [Disclaimer: I started the Sage project.] I do. There's probably about 100 times as many Maple/Mathematica users as Sage users. That said, Sage fits nicely into the Python "ecosystem", and Python probably has an order of magnitude more users than Maple/Mathematica. DETAILS: The Maple and Mathematica websites estimate around 1 million "users" each for their software. This is roughly consistent with the yearly revenue they must take in to support their staff. I estimate that Sage has maybe 10,000 users based on weekly download figures, mailing list memberships, etc. (for example sage-support) has 503 members. Also, when we won the trophees du libre in November last year there were 10,000 downloads of Sage in one weekend, so many people have tried Sage. Sage still has a very small niche of users compared to Maple/Mathematica. A full native Microsoft Windows port is critical to changing this. Also, it takes time -- it was exactly one year ago that the first serious native symbolic calculus functionality was added to Sage. Before then, Sage was functionality-wise much more targeted at numerical and exact linear algebra, coding theory, cryptography, group theory, abstract algebra, combinatorics, and graph theory, which are all extremely interesting and important areas, but probably don't yields millions of users (well, numerical linear algebra might). > > > I may have been done a disservice. I > > expended consider effort in mastering Maple. I paid tuition. I was taught by > > the creators (of Maple). Yet Maple is a commercial product that I cannot > > afford because of my lifestyle. Nor do I have an entitlement. I am tapped > > out when it comes to learning new stuff. And I cannot learn unless I have > > faith of efficacy. Hence, it is simpler for me to steal a copy of Maple than > > it is to learn Sage or any other CAS that is free or relatively cheap. It's perhaps too late for you. I started the Sage project is so that there will be less people who end up in your unfortunate situation. I understand you, since I was in much the same situation (w.r.t. Magma instead of Maple) four years ago. To get out, I didn't just have to "learn Sage", I had to write a big chunk of it, build a community, spend massive time writing grant proposals, etc... and this is only the beginning. It's a huge gamble on my part. I wish the previous generation had already done this (like with R and statistics), but they haven't in mathematics. So it's up to us. > The fact companies such as Google and Microsoft are supporting Sage > suggests to me that you should give it serious consideration. One > assumes those companies have faith in the software. Google really likes Python and tools like Sage that improve it's potential. Microsoft Research has people doing research that find the unique functionality of Sage (in those non-calculus areas mentioned above) *extremely* valuable to their research. That said, I do not think they "trust" Sage. *I* do not trust Sage. Anybody who blindly trusts any complicated mathematical software system is being foolish. It's critically important to come up with as many double checks on computations as you can, to be skeptical of any results, to compare computations done in one system with those in another, etc. A major reason that Sage has native support for interfacing with Maple, Mathematica, Matlab, Magma, Octave, etc., and converting objects to and from those systems is so that people can do calculations both in Sage and another system and easily compare the results. That said, at least with open source software you can look at and read any random part of the full source code. One advantage of this is that it will probably *reduce* your blind trust in the software. That's a good thing. > The fact you can't afford Maple, and there is a free alternative, would > suggest to me you have better reasons than most for trying it. Sage is > based on Python, which is in itself a language well worth learning. I definitely agree that that person should try Sage. Everybody should. Try it out. If you don't like it for a number of reasons, TELL ME ([EMAIL PROTECTED])! I really want to know! I might not be able to do anything now about the problems that keep Sage from being adopted, but I definitely can't do anything if I don't know about them. > I've not used it myself, and I will not until there is a port to > Solaris, but I am trying to help the Solaris port. Thanks. There absolutely will be a Solaris port. We care deeply about supporting that OS. > > I have no criticism to offer regarding Sage. Commercial success is a > > struggle, as is scholarship. I wish Sage the best of luck in the > > marketplace. Thanks. By the way, regarding Vladimir's motivations for testing CAS's, I speculate that he is disturbed by people blindly trusting Maple (and other math software), so he wants to emphasize that they should not do that. I suspect this, because in his writings he describes vivid images of "bad things" happening as a result of people trusting math software when they shouldn't. He also quotes repeatedly from the Maple marketing literature which often states quite forcefully that one should trust Maple. The Sage marketing machine, as it is, does not work like that. Two days ago I read through his 300-ish page pdf about Maple's flaws. Some of the flaws are actually quite interesting. For example, he points out that typing (2^4000)! into Maple *crashes* Maple immediately. I tried that with Maple 11's command line and sure enough on the system completely crashes (not just a memory error but a total crash): D-69-91-159-111:~ was$ maple |\^/| Maple 11 (APPLE UNIVERSAL OSX) ._|\| |/|_. Copyright (c) Maplesoft, a division of Waterloo Maple Inc. 2007 \ MAPLE / All rights reserved. Maple is a trademark of <____ ____> Waterloo Maple Inc. | Type ? for help. > (2^4000)!; bytes used=4000572, alloc=3865916, time=0.04 Execution stopped: Stack limit reached. D-69-91-159-111:~ was$ In the GUI interface, the same input causes Maple to run very slowly for several *minutes* then finally the maple kernel crashes. He evidently has reported this issue to Maple for over 5 years. This is exactly the sort of bug that if somebody reported it to the Sage list it would instantly appear as a blocker in our public bug list: http://trac.sagemath.org/sage_trac for all to see. And it would get fixed. In fact, we *did* have almost exactly the same bug, which two years ago a high school student found and reported, and which I fixed the next day. -- William http://www.sagemath.org -- William Stein Associate Professor of Mathematics University of Washington http://wstein.org --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
