- From: Ian Hickson <ian@hixie.ch>
- Date: Tue, 6 Nov 2007 15:31:14 +0000 (UTC)
On Mon, 5 Nov 2007, Martin Atkins wrote: > > James M Snell wrote: > > This is just off the top of my head so I'm certain that there are > > probably reasons why this wouldn't work, but could we not do something like, > > > > <link rel="edit put delete patch" href="http://example.org/foo" /> > > > > Edit indicates the purpose of the link, put, delete and patch indicate > > methods (in addition to GET). > > > > That doesn't tell you what sort of stuff you are allowed to PUT to that > URL, though. I'm not sure that HTML is really the place to tell you that > it expects an Atom entry, but without it knowing that it supports PUT is > not that useful either. > > However, I guess one could argue that there should be something that > acts as the opposite of the "type" attribute: rather than the type that > should result when you GET, instead indicate the type that you should > PUT. Probably too tricky to be worth it, though: it'd probably need all > of the capabilities of the HTTP "Accept" header to be useful, and it's > questionable whether that much detail should be defined externally of > the resource itself. It's an interesting idea, but as you say, I'm not sure we should do it immediately. It may be worth study for a future version though. -- Ian Hickson U+1047E )\._.,--....,'``. fL http://ln.hixie.ch/ U+263A /, _.. \ _\ ;`._ ,. Things that are impossible just take longer. `._.-(,_..'--(,_..'`-.;.'
Received on Tuesday, 6 November 2007 07:31:14 UTC