My delay with the answer is explained by the fact that I have deliberately waited for a day to hear everyone's opinion.
On Mon, Apr 2, 2012 at 11:50 PM, Ronan Lamy <[email protected]> wrote: > Le lundi 02 avril 2012 à 21:28 +0100, Tom Bachmann a écrit : >> On 02.04.2012 15:30, Sergiu Ivanov wrote: >> > In conclusion, I cannot see how my ideas fundamentally contradict the >> > approaches evoked in this thread. Therefore, I will try to pose the >> > dual question: do you think the current Ring class is well-suited for >> > a future implementation of ring theory? I hope a definite answer to >> > this question will be more reachable :-) >> > >> >> I think the only reasonable answer is "yes, but ...". > > For me, the answer is clearly no. Hm :-) My opinion is, as you would guess, "No". >> The real problem with this discussion is, imho, that you are trying to >> propose a "perfect" framework, without any specific examples to test >> "perfectness" against. No matter how much we talk this over, trying to >> design code for such a complex system (at least "all of ring theory", >> you seem to have even more in mind) on the drawing board is futile (in >> my opinion). > > I don't think Sergiu is trying to implement all of ring theory. He only > needs to implement the basics of the language of ring theory, which is > much easier. Currently, there's no way to represent "let A = (E, +, .) > be a ring" in sympy. Fixing that is a prerequisite for implementing any > part of ring theory. Totally right. My only goal is to make my implementation of the Groebner walk depend on classes which will be easily substituted by the eventual implementation of ring theory. And yes, I basically only need the "let A = (E,+,*) be a ring" thing, and similar stuff for ring ideals and polynomial rings. >> Moreover, this proposal seems embedded in your gsoc plans, where it >> really does not fit: there is no need or justification for trying to >> write a "perfect, all-encompassing object oriented framework" in order >> to implement the groebner walk stuff (at the most, I think, this >> requires a very specialised framework - most of which I believe can >> already be found in domains/). > > You could say that for any kind of high-level framework. It's > theoretically possible to program everything in assembler, so why do we > bother with Python? Well, the answer is obvious: it makes a lot of > things much easier. I am rather in pursuit of foundation classes for a proper implementation of ring theory, which I am absolutely not going to do in this summer. Ronan explains my own goals and reasons very well. Now, the little brother of the question I have posed yesterday: do you find the framework of the classes I have sketched in the infrastructure part of my proposal acceptable? Tom, I think I understand your points, but I'd still like to ask you whether you find it acceptable to have the classes I have suggested in SymPy? Sergiu -- You received this message because you are subscribed to the Google Groups "sympy" group. To post to this group, send email to [email protected]. To unsubscribe from this group, send email to [email protected]. For more options, visit this group at http://groups.google.com/group/sympy?hl=en.
