[sage-combinat-devel] Re: Is it really needed to make the parent class of a category depend on its base ring?

Hi! On 7 Okt., 19:46, "Nicolas M. Thiery" wrote: > On Fri, Oct 07, 2011 at 10:38:49AM -0700, Simon King wrote: > > But here, I'd like to have the answer "True": > >   sage: Algebras(GF(3)).parent_class is Algebras(GF(5)).parent_class > >   False > > +1 For the record: So far, it seems to me that

Re: [sage-combinat-devel] Re: Is it really needed to make the parent class of a category depend on its base ring?

On Fri, Oct 07, 2011 at 01:30:49PM -0700, Simon King wrote: > I really don't see the problem. > > On 7 Okt., 22:21, Simon King wrote: > > So, why shouldn't parent_class rely on the generic pickling of a > > dynamic class? Well, I guess I should simply try and look how it blows > > up... > > I si

Re: [sage-combinat-devel] Re: Is it really needed to make the parent class of a category depend on its base ring?

On Fri, Oct 07, 2011 at 10:38:49AM -0700, Simon King wrote: > But here, I'd like to have the answer "True": > sage: Algebras(GF(3)).parent_class is Algebras(GF(5)).parent_class > False +1 > > Short summary: sharing those classes was my original intent; see the > > discussion around line 240 o

[sage-combinat-devel] Re: Is it really needed to make the parent class of a category depend on its base ring?

I really don't see the problem. On 7 Okt., 22:21, Simon King wrote: > So, why shouldn't parent_class rely on the generic pickling of a > dynamic class? Well, I guess I should simply try and look how it blows > up... I simply dropped the reduction data from the definition of a parent_class and el

[sage-combinat-devel] Re: Is it really needed to make the parent class of a category depend on its base ring?

Hi! dynamic_class calls the cached function dynamic_class_internal. The reduction data is currently part of the cache key. That's probably too inflexible: Different reduction data may easily return the same result, and thus the reduction data should not be part of the cache key. And I wonder: Is

[sage-combinat-devel] Re: Is it really needed to make the parent class of a category depend on its base ring?

Hi Nicolas, On 7 Okt., 19:10, "Nicolas M. Thiery" wrote: >         Hi Simon, > > On Fri, Oct 07, 2011 at 09:51:39AM -0700, Simon King wrote: > > Currently, we have > >   sage: Algebras(QQ).ParentMethods is Algebras(ZZ).ParentMethods > >   True > > but > >   sage: Algebras(QQ).parent_class is Alge

Re: [sage-combinat-devel] Is it really needed to make the parent class of a category depend on its base ring?

Hi Simon, On Fri, Oct 07, 2011 at 09:51:39AM -0700, Simon King wrote: > Currently, we have > sage: Algebras(QQ).ParentMethods is Algebras(ZZ).ParentMethods > True > but > sage: Algebras(QQ).parent_class is Algebras(ZZ).parent_class > False Our messages crossed. See my comment on #

[sage-combinat-devel] Is it really needed to make the parent class of a category depend on its base ring?

Hi! Currently, we have sage: Algebras(QQ).ParentMethods is Algebras(ZZ).ParentMethods True but sage: Algebras(QQ).parent_class is Algebras(ZZ).parent_class False I wonder if that is really necessary? Couldn't one rewrite the parent_class and element_class lazy attributes, such that the re