- From: Martin Thomson <martin.thomson@gmail.com>
- Date: Fri, 10 Jan 2014 10:40:33 -0800
- To: Julian Reschke <julian.reschke@gmx.de>
- Cc: HTTP Working Group <ietf-http-wg@w3.org>
On 10 January 2014 10:12, Julian Reschke <julian.reschke@gmx.de> wrote: > Not sure about that. Maybe there are use cases I'm not aware of. Sometimes negotiation is just formats, in which case it's not a bad idea exactly. It is, after all, just representations we're talking about. For instance, I've seen numerous cases where PUT of XML is perfectly able to cause the JSON representation to update correctly. For natural language, I'm sure that automatic translation is possible to a certain extent too, if perhaps inadvisable... I'm not sure what to say here. If it's clear that the URI for a negotiated resource is effectively a label for multiple things that are logically the one thing, and the other URIs that might appear in Content-Location are identifiers for the individual facets of that uber-thing, then that should be enough to allow people to reach their own conclusions. Maybe: A PUT to a negotiated resource is expected to set the state of the resource for all alternative representations. This is easily possible if the translation between representations is purely mechanical, though less mechanical translations, such as natural language translation might cause this sort of operation to be difficult and therefore inadvisable.
Received on Friday, 10 January 2014 18:41:00 UTC