[Python-Dev] Proposed change to logging.basicConfig
Matthew Woodcraft
matthew at woodcraft.me.uk
Tue Mar 29 22:02:45 CEST 2011
More information about the Python-Dev mailing list
Tue Mar 29 22:02:45 CEST 2011
- Previous message: [Python-Dev] Proposed change to logging.basicConfig
- Next message: [Python-Dev] cmp= & key= (Re: Proposed change to logging.basicConfig)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Terry Reedy <tjreedy at udel.edu> wrote: > I am bothered by mutually exclusive parameters. This is one reason I was > glad to see cmp eliminated from list.sort. Quick: what happens if one > passes both cmp and key to list.sort? There are three reasonable > possibilities. As far as I can read, the answer is not documented.# > # Experiment with 2.7 shows that cmp wins. Though too late to change, I > consider this the worst choice of three. I think an exception should be > raised. Failing that, I think key should win on the basis that if one > adds a 'new-fangled' key func to an existing call with cmp (and forgets > to remove cmp), the key func is the one intended. Also, the doc clearly > indicates that key is considered superior to cmp. Neither 'wins': cmp is applied to the output of key. I agree that it would have been worth documenting this explicitly. -M-
- Previous message: [Python-Dev] Proposed change to logging.basicConfig
- Next message: [Python-Dev] cmp= & key= (Re: Proposed change to logging.basicConfig)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
More information about the Python-Dev mailing list