Re: Precondition add-must-be-new-label

Werner DonnĂ© <[email protected]> Sun, 16 Jul 2006 10:38:12 +0200
Newsgroups gmane.ietf.deltav
Organization Re
Message-ID <[email protected]>
Thank you Geoff. I wrongly interpreted the "set" operation as
replacing the label-name-set of a version with a singleton
containing the provided new label name.

Regards,

Werner.

Geoffrey M Clemm wrote:
>=20
> The add-must-be-new precondition only applies if DAV:add is specified i=
n
> the LABEL request (i.e. you want it to fail if the label is already
> applied to some version).
>=20
> You use DAV:set in the LABEL request if you want to move the label from
> one version to another.
>=20
> Cheers,
> Geoff
>=20
> Werner wrote on 07/14/2006 12:58:41 PM:
>=20
>>
>> Hi,
>>
>> The add-must-be-new precondition of the LABEL method makes it
>> impossible to move a label. Moving a label requires a remove
>> followed by an add, which are two round-trips. Instead of
>> having the precondition, the postcondition label-must-be-unique
>> would provide the same data integrity, while avoiding the
>> extra round-trip.
>>
>> An additional advantage is that the move would be atomic. A
>> remove followed by an add is not. The impact of a failure in
>> between can be dramatic.
>>
>> Regards,
>>
>> Werner.
>> --
>> Werner Donn=E9  --  Re
>> Engelbeekstraat 8
>> B-3300 Tienen
>> tel: (+32) 486 425803   e-mail: [email protected]
>>

--=20
Werner Donn=E9  --  Re
Engelbeekstraat 8
B-3300 Tienen
tel: (+32) 486 425803	e-mail: [email protected]