RE: 3.5: VERSION-CONTROL response codes

> From: [email protected]
> [mailto:[email protected]]On Behalf Of Clemm, Geoff
> Sent: Thursday, October 10, 2002 4:18 PM
> To: [email protected]
> Subject: RE: 3.5: VERSION-CONTROL response codes
>
>
> Could you motivate why we would want to do this?

Sure.

1) Consistency with HTTP: requests that create new resources should return a
201.

2) Consistence with deltaV (defining DAV:version-control-response as
optional response body)

3) Performance: I've seen many cases where after a version-control, a client
would like to access either the checked-in version or the VHR -- in which
case it currently needs an additional PROPFIND.

> (I'm not saying it's a bad idea, but it is additional
> text that would need to be added to the protocol definition,
> and we've got a lot of text already :-).
> In particular, this information is easily obtainable
> with a subsequent PROPFIND (and even a streaming PROPFIND,
> i.e. you can issue the PROPFIND without waiting for
> the VERSION-CONTROL to succeed).

It's still an additional request and doesn't come for free. In general, a
server will have to do another call to it's backend to get this information.