thank you all guys for these useful responses. so, I see that sympy and sympycore and pynac are coming from different people. But I assume that you have a very wide view of the actual situation, and this is a real added value! I mean, getting to know pros and cons of each technology is something that comes only from experience, and I would be really thankful (like many others, I am sure) if you could share us this! as I told, the symbench wiki page was really useful to me, because it has shown the differences in terms of performances and stability and confidence of the different symbolic packages.
Trying to summarize, what I'd like to know (for example in the case of symbolic, but that's pretty general) like a normal user is: - which is the package with the brightest future (in terms of chance to be actively developed) - where is each package doing is best? - where is each package failing in doing its job? thank you very much PS: I still can't find any piece of information about pynac, is there anything on the web? Maurizio On Feb 22, 7:25 am, Ondrej Certik <ond...@certik.cz> wrote: > On Sat, Feb 21, 2009 at 7:57 AM, Maurizio <maurizio.gran...@gmail.com> wrote: > > > Hi Burcin, > > thank you very much for this! > > > I admit that I've never heard of Pynac. I'm really glad to have > > subscribed this group, otherwise I couldn't have discovered it! > > This brings me to a new question: how are user supposed to go through > > SAGE features learning? > > I supposed that the reference manual and documentation should have > > been more than enough, but now I more and more get convinced that > > there is a lot of material out there (especially on the WIKI) that > > should get much better advertisement to all the users. > > > If I hadn't posted this here, I probably won't have learned about > > Pynac for weeks, even though I am doing web research of documentation > > about SAGE symbolic features. Another very interesting reading for me > > was this Symbolic Benchmarck: > > >http://wiki.sagemath.org/symbench > > > I have to recommend this to anybody interested in symbolic with SAGE. > > Before of this, I thought that SAGE symbolic features (coming from > > maxima) were pretty good, but something from sympy could do a better > > job. Now I learn that sympycore (I thought they were the same thing) > > beats sympy by far in some cases, and that Pynac is even better > > sometimes! So, is it possible to enhance the differences between the > > packages, and directly use them where they are best? > > > I mean, if SymPy is so good with complex symbolic calculus, why isn't > > this information well written and advertised and exposed somewhere in > > the SAGE documentation? Is it true? And why isn't out there a tutorial > > where they advice me to use Pynac wherever I can (I mean, in all the > > currently supported features), otherwise switch to SymPy or SymPycore > > (for example for symbolic integral) and otherwise go with SAGE built- > > in maxima interface? Would this be possibly automated? > > It's because sympy, sympycore and pynac were all written by different > people and as of now they are not compatible. I don't like the > situation either, and if I manage to find funding for the summer, I'd > like to speedup sympy using Cython to be as fast as sympycore or pynac > and also make it working on top of pynac so that it could be used in > Sage by default. > > It needs couple weeks of hard work though, so someone needs to invest > time to do it. > > Do you think this would be a worthwhile project for Sage? > > Ondrej --~--~---------~--~----~------------~-------~--~----~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to sage-devel-unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://www.sagemath.org -~----------~----~----~----~------~----~------~--~---
