W3C home > Mailing lists > Public > public-grddl-wg@w3.org > February 2007

Re: GRDDL and IETF: Need "HTTP Header Linking"

From: Graham Klyne <GK@ninebynine.org>
Date: Thu, 08 Feb 2007 12:47:39 +0000
Message-ID: <45CB1BEB.6060205@ninebynine.org>
To: Harry Halpin <hhalpin@ibiblio.org>
CC: Dan Connolly <connolly@w3.org>, w3c-semweb-cg@w3.org, public-grddl-wg <public-grddl-wg@w3.org>

Formally, entry in the permanent header field registry requires a
standards-track specification or equivalent.  Entry in the provisional registry
simply requires a specification, and could be done now based on Mark's draft,
but doesn't guarantee that a different Link: specification can't also be
registered (but that would be discouraged).

I think that what could be done now is to request a permanent registry entry
based on the RFC 2068 specification (based on that, Link headers can
legitimately appear in the wild, even though its not present in later
specifications), with a commentary noting that it has been dropped from later
HTTP specifications and is being developed separately with a reference to Mark's


Harry Halpin wrote:
> (cc'ing Graham Klyne and Mark Nottingham)
> GRDDL [1], a W3C Rec-track spec, now has a dependency on the success of
> Mark Nottingham's IETF draft renewal of the "link" header and possibly
> the "profile" header as given in his "HTTP Header Linking" IETF draft. [2]
> My goal is not have waiting for the IETF process prevent GRDDL from
> going into CR stage when our current charter [3], which states that we
> go to CR the first quarter of 2007. I am not as familiar with IETF
> process as W3C Process, but is this possible?
> If not, what are the current issues preventing [2] from going through
> IETF process to become a RFC? Is help needed? Ian Davis, a member of the
> W3C GRDDL WG, seems interested in helping [4].
> GRDDL does *not* define a new header per se, but simply a using the
> currently deprecated header name "link" that Mark [2] is drafting for
> re-inclusion, and giving it a new field value, since "link-param = ( (
> "rel" "=" relationship )" and "Relationship values are case-insensitive
> and MAY be extended within the constraints of the sgml-name syntax." The
> exact text we want to use is here [6].
>   Note that there is not a field value repository for HTTP, for while
> there appears  appears to be a relatively straightforward IETF process
> for headers [5], so this should not be a problem for [2].
> "Neither repository tracks the syntax, semantics or type of
> field-values. Only the field-names, applicable protocols and status are
> registered; all other details are specified in the defining document are
> referenced by repository entries."
> Since we are depending the "link" header name (given in part 2 of Ian's
> e-mail [4] to the WG is acceped at our next telecon and also the
> "profile" header name [6],  all we really need is for Mark's draft [2]
> to become RFC. Ian claimed that the "link" header was dropped from RFC
> 2616 and now needs to be put back in RFC 2068 [4,7]
>  Can we reasonably get this by Q1 2007?
> [1] http://www.w3.org/2001/sw/grddl-wg/
> [2] http://www.mnot.net/drafts/draft-nottingham-http-link-header-00.txt
> [3] http://www.w3.org/2006/07/grddl-charter.html
> [4] http://lists.w3.org/Archives/Public/public-grddl-wg/2007Jan/0087.html
> [5] http://tools.ietf.org/html/rfc3864#ref-24
> [6] http://lists.w3.org/Archives/Public/public-grddl-wg/2007Feb/0020.html
> [7] http://www.ietf.org/rfc/rfc2616.txt
> If you tell me the process for getting a HTTP Profile header registered,
> Ill
> Dan Connolly wrote:
>> On Thu, 2007-02-01 at 23:07 -0500, Dan Connolly wrote:
>>> On Jan 30, 2007, at 11:20 PM, Harry Halpin wrote:
>>>> We've been working hard in the GRDDL WG, but we're still falling a bit 
>>>> behind our rather ambitious schedule.
>> We closed the last open issue today. Yay!
>> http://www.w3.org/2007/02/07-grddl-wg-minutes.html
>> Meanwhile, we picked up a dependency on
>> getting an HTTP Profile header registered.
>> That probably means we're going to hang out
>> in CR for a while, rather than going straight
>> to PR. Hmm.

Graham Klyne
For email:
Received on Thursday, 8 February 2007 12:48:59 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:39:10 UTC