For A and B Parents, should A == B ever differ from A is B? This came up in tracking down a p-adics bug, but there's at least one place that assumes the equivalence of these conditions. The place I'm thinking of is in sage.categories.hom_set.Hom, where it checks a cache to see if the Homset has been computed already. Finding a key in a dictionary uses ==, but we require an "is" relation for domains and codomains of maps.
My opinion is that we should strive for (A == B) iff (A is B), so the Hom function is not a bug as written, and instead I just need to make either one function smarter or another function dumber in the p-adics code. But I thought I'd ask sage-devel, and raise the issue so that people are aware of this constraint on the __cmp__ method for Parents. David P.S. To find an example of two parents which currently compare as equal but are not the same object, do sage: R = Zp(5, print_sep="|") sage: S = Zp(5, print_sep=":") sage: R == S True sage: R is S False -- To post to this group, send an email to sage-devel@googlegroups.com To unsubscribe from this group, send an email to sage-devel+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/sage-devel URL: http://www.sagemath.org
