- From: Julian Reschke <julian.reschke@gmx.de>
- Date: Wed, 03 Dec 2008 12:29:04 +0100
- To: Mark Nottingham <mnot@mnot.net>
- CC: Eran Hammer-Lahav <eran@hueniverse.com>, "ietf-http-wg@w3.org Group" <ietf-http-wg@w3.org>
Mark Nottingham wrote: > ... > I think the issue here is more around trust than just semantics; i.e., > you need to know the source of the statement to evaulate it. The thing > is, that's really defined by the context of use; i.e., if you're working > from within some wonderful semantic framework, you might be able to > trust inbound links. > > So, if this were defined by the application, would you be more happy? > The effect on *this* spec, I think, would be that the link *header* > section would say that rev links aren't necessarily authoritative. > > Make sense?s > ... Not sure. What would that achieve? Why would you trust rel=author more then rev=made ? In general, the level of trust for a relation exposed with a resource probably depends *both* on the resource and the link relation being used, not the fact whether it's exposed as "rel" or "rev". Best regards, Julian
Received on Wednesday, 3 December 2008 11:29:51 UTC