Currently, the cover relations of a poset are returned as lists:
sage: P = Posets(4)[3]
sage: P.cover_relations()
[[0, 1], [0, 2], [0, 3]]
This seems to be a bad choice, because lists are not hashable, and also somewhat inconsistent, because edges of graphs are returned as tuples.
The same goes for P.relations().
1) Is there a good reason for this choice, or is it just history?
2) Provided that I do the work, is there any chance that a ticket switching to tuples gets a positive review? A preliminary run shows only trivial doctest failures, that is, [a, b] vs (a, b).
(One non-intrinsic motivation is that it would make some user-visible sagemath code generated by FindStat slightly easier to understand for casual users.)
Martin