- From: Julian Reschke <julian.reschke@gmx.de>
- Date: Sun, 03 Oct 2010 21:14:36 +0200
- To: Willy Tarreau <w@1wt.eu>
- CC: HTTP Working Group <ietf-http-wg@w3.org>
On 03.10.2010 19:21, Willy Tarreau wrote: > On Sun, Oct 03, 2010 at 07:06:22PM +0200, Julian Reschke wrote: >>> In case nobody would object, how could we move on ? Doing so involves a >>> new header ("content-status"), so that should probably require proper >> >> Do we really need Content-Status? > > I don't know. Maybe we can deduce it from the presence of the Link > headers. But when reading the Link header spec below : > http://tools.ietf.org/id/draft-nottingham-http-link-header-10.txt > > I did not get the impression that the header is supposed to cause UAs > to request specific user action. But maybe we could imagine that some > specific values of the header are automatically detected by UAs and > make them ask the user for chosing the most suitable version. I'm not Well, rel=stylesheet already "makes" UA do something. > certain this practice is good for the long term (eg: should browser > present a popup when encountering unknown types). Probably that Mark No, the browser certainly should *not* display a popup for unknown types :-) > has some insights on the best way to use the header. > >>> registration. Also, I've not seen any registry for all relation-types, >>> so we might need to define a few. Also, does a draft need to be written >>> to advance ? >> >> The Link Relations registry is defined in RFC5988-to-be >> (<http://greenbytes.de/tech/webdav/#draft-nottingham-http-link-header>). > > So if I understand it correctly, we refer to the types enumerated in > RFC4287 ? Then we have 5 available types : "alternate", "related", "self", > "enclosure", and "via". I thought that "alternate" would fit but it's > indicated that no more than one link may be present, which limits the > usefulness. No no, I just was pointing out that this is the new definition of the registry. I would suggest that we either pick one of those listed in http://www.iana.org/assignments/link-relations/link-relations.txt ... - "latest-version" might make sense, or define a new one. Best regards, Julian
Received on Sunday, 3 October 2010 19:21:51 UTC