Charles, Have you read the new Annotea draft protocol [1]? This is not clear from your message. It actually documents what you're proposing and Appendix B [2] explains why we were using POST. In a previous message we asked developers for feedback wrt making PUT the prefered practice and dropping altogether POST for updating annotations. -jose On Sat, Jan 18, 2003 at 09:05:53PM -0500, Charles McCathieNevile wrote: > > Hi, > > I prefer to have the HTTP semantics less ambiguous. If a PUT always means > updating an existing thing, and a POST always means creating a new one, then > it is possible to create simpler servers - if they have to check the HTTP > protocol used and the message on the wire it introduces complexity and I > don't see any value in it. > > Backwards compatibility with systems that don't do PUT seems a backwards step > given that Annotea is a new system anyway, and I would prefer to remove such > complications. [1] http://www.w3.org/2002/12/AnnoteaProtocol-20021219 [2] http://www.w3.org/2002/12/AnnoteaProtocol-20021219#PostUpdateReceived on Monday, 20 January 2003 05:40:46 GMT
This archive was generated by hypermail 2.2.0 + w3c-0.30 : Friday, 25 March 2005 11:19:19 GMT