Hi, I was at Sage Days 8 and I posted what we were doing in there to my blog:
http://ondrejcertik.blogspot.com/2008/03/sage-days-8.html here I'd like to put some points related to SymPy and Sage.calculus. We discussed quite a lot of it in Austin, but I'd like to have it in the mailinglist as well. CCing sage-devel too. I am interested in any comments, especially if you think that I am completely wrong. :) 1) The idea of SymPy is to have something that is easily extensible, easy to use/install, pure Python by default (possibly later with some optional parts in C), that together with numpy, scipy, ipython, matplotlib and other projects will provide all the necessary tools that one needs to do scientific (engineering) calculations. Most of our users are imho people using these tools anyway. We should adapt our page, that currently says "It aims to become a full-featured computer algebra system (CAS)", because a CAS for me is anything I may need in physics/engineering, but for mathematicians, CAS means a lot of different things too. Also CAS probably implies all the other things that go with it, like numerics (scipy+numpy), notebook (Sage or ipython1) etc. 2) People, that want the bigger picture, all in one solution, should use Sage, that includes all of these tools. Sage.calculus is currently based on maxima and Gary is currently rewriting the core of it in Cython, so that maxima is only called for difficult things, like limits and integrals. When this is done, then it'd be quite easy to slowly replace maxima part by part with cython+other parts of sage. What I personally would like to have is something that is not depending on the rest of Sage, because I like the tool doing one job and doing it well. I don't like all in one solutions, but rather separate libraries, leaving it on myself to assemble the final program out of it. Sage mission is more to be a viable alternative to Maple/Mathematica/Magma/Matlab, all of which are monolitic all in one solutions (and much bigger than Sage actually). That said, I think that in few years when Sage matures, I think people will be interested in making the parts more independent. But currently to make Sage.calculus work, you need to download either the binary (260MB), or 200MB of sources, do "make" and wait for couple of hours. With sympy, you download 1MB of sources, do "import sympy" and you are up and running. See also below for more arguments on this. 3) Other advantages of SymPy over Sage are: * small and BSD licensed, so you can include it in your own project without any restrictions (technical or political). * being pure python and having a simple design, you can easily debug it and fix it to make it work for your problems (please send your patches back to us:) * works natively on all platforms As Michael Abshoff has pointed out, Sage will be working on windows natively too, maybe even by the end of the year. Actually Microsoft finances the port, which I find really cool that MS is financing a truly opensource project (that among other things brings a lot of people to Python). Another point that Michael made, is that Sage will finally get to linux distributions too (Debian and others), thus being easy to install and use. Which definitely makes a lot of things easier. Nevertheless, when something has 1MB, it is a lot easier to deal with, I must stand on my point here. :) Ondrej --~--~---------~--~----~------------~-------~--~----~ 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 -~----------~----~----~----~------~----~------~--~---
