We're trying to reduce the number of external modules we depend on, and the subject of the OrderedSet modules came up.
I've put together a PR here - https://github.com/pykickstart/pykickstart/pull/325
I wasn't able to figure out exactly why we switched to using OrderedSet, the commit wasn't clear as to why it was needed. Internal to the add() function I can see no reason for it, the add and exclude lists are independent of each other so the ordering shouldn't matter when doing the set operations.
The only thing I can think of is dnf, via anaconda, and it really shouldn't care either.
So I dropped it, turning it back into a regular set, and I just sort the results when storing them so that they'll always have a consistent order.
Does this look ok to everyone? Can anyone think of a reason not to do this? Have I missed something obvious here? :)
Please comment on the PR so the discussion can all be in one place.
Thanks,
Brian
anaconda-devel@lists.fedoraproject.org