Thanks David, Unfortunately, I'm going to -1 this version for two reasons:
1. MUST should be SHOULD 2. Rather than "the representation", it should be something like "an Atom Entry document as it would appear in the collection's feed". If you disagree, I could post an alternative pace and folks could decide which approach is best. - James David Sklar wrote: > > Tim Bray wrote: >> On Mar 24, 2006, at 12:48 PM, Kyle Marvin wrote: >> >>>>> +1 on tightening the spec. However, I would suggest saying instead >>>>> that >>>>> if anything does come back, it should be the atom entry as it would >>>>> appear in the collection's feed (e.g. a non-editable representation). >>>> >>>> This is helpful in implementations (such as ours) in which server is >>>> annotating the entry with various extensions or attributes that the >>>> client doesn't provide; it is a useful optimization over having the >>>> client issue another request to the URL in the response's Location >>>> header to retrieved the annotated entry. >>> >>> +1. We see similar usability benefits in echoing back the entry on >>> POST as opposed to making the client retrieve it immediately. I'd be >>> -1 on any proposal that explicitly prohibited this. >> >> OK, looks to me like if someone writes a Pace to make this change, it >> probably will sail right through. Who's volunteering? -Tim > > I've put it up as "PaceEntryInPostPutResponse". Be gentle -- it's my > first Pace. > > David > >
