Sorry for the monologue, but here are some more observations:

The parenting situation is much improved if Hom sets depend on the 
domain/codomain on the nose and not just on their isomorphism class. This is 
probably more natural, too:

Hom(A,B) is Hom(A',B')    <=>    A is A' and B is B'

Now the uniqueness of the Hom set can be determined by id(domain) and 
id(codomain), which is well-defined for incomplete objects. Horay!

But not all is well. There some places where Cython classes are initialized 
from potentially incomple objectes during unpickling, for example in 
sage.categories.map. This is also going to break with incomplete objects, 
too. We probably have to record such situations, complete the unpickling, 
and then initalize the Cython objects. Any thoughts on that one?

-- 
To post to this group, send an email to [email protected]
To unsubscribe from this group, send an email to 
[email protected]
For more options, visit this group at http://groups.google.com/group/sage-devel
URL: http://www.sagemath.org

Reply via email to