This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author rhettinger
Recipients Akuli, Jim.Jewett, ShubhamKJha, Yuan, pablogsal, rhettinger, serhiy.storchaka, stutzbach
Date 2020-07-28.14:44:05
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1595947445.31.0.894165429158.issue41409@roundup.psfhosted.org>
In-reply-to
Content
> If deque does not fully support the MutableSequence interface,
> should not it be un-regitered as MutableSequence?

You could unregister it, but for such a minor variance, it isn't worth it.  We're not unregistering sets and frozenset from Set even though there are minor variances in the operators. 

The ABCs are quite limited in what they do. The underlying machinery is mostly about determining whether a method is present or not.  Concrete classes are free to override, extend or raise NotImplemented.  For example, a Counter is still a mapping even though fromkeys() raises NotImplemented and update() has different semantics than Mapping.  

Jelle has already adapted TypeShed to match the concrete class, so no actual user issue remains.
History
Date User Action Args
2020-07-28 14:44:05rhettingersetrecipients: + rhettinger, stutzbach, Jim.Jewett, serhiy.storchaka, Akuli, pablogsal, ShubhamKJha, Yuan
2020-07-28 14:44:05rhettingersetmessageid: <1595947445.31.0.894165429158.issue41409@roundup.psfhosted.org>
2020-07-28 14:44:05rhettingerlinkissue41409 messages
2020-07-28 14:44:05rhettingercreate