W3C home > Mailing lists > Public > ietf-http-wg@w3.org > July to September 2009

Re: Input on request for link relation [ w/ proposed modifications to link draft ]

From: Sam Johnston <samj@samj.net>
Date: Mon, 21 Sep 2009 01:54:47 +0200
Message-ID: <21606dcf0909201654x7714981cu4af3def35d358bb8@mail.gmail.com>
To: Kris Zyp <kris@sitepen.com>
Cc: Robert Sayre <sayrer@gmail.com>, Mark Nottingham <mnot@mnot.net>, Brett Slatkin <brett@haxor.com>, Lisa Dusseault <lisa.dusseault@gmail.com>, Atom-Syntax Syntax <atom-syntax@imc.org>, ietf-http-wg@w3.org
On Mon, Sep 21, 2009 at 1:41 AM, Kris Zyp <kris@sitepen.com> wrote:

>
> I am not necessarily opposed to the addition of a
> simple/non-namespaced relation, but it still seems like it should be
> necessary to have some the relationship clearly articulated enough to
> delineate between "hub" (or whatever relation name that is used) and
> the existing "monitor" relation. A developer (like me) should be able
> to make an informed selection between the two for other notification
> formats.
>

At the very least, where there are overlapping relations (e.g. prev vs
previous, hub vs monitor, etc.) we should indicate which one is preferred,
perhaps by deprecating the other(s).

I appreciate Robert's concerns however and you can be quite sure that we'll
see more rather than less specs developed behind closed doors before being
unleashed on the world (Google Wave comes to mind). In that case I would
suggest that they supply a URI relation in addition to the proposed short
relation - reference implementations can send the URI and clients can accept
both in the interim.

Whatever we decide the Web Linking draft needs to reflect it (and probably
needs to go to some lengths to explain that URI relations are not evil).

Sam
Received on Sunday, 20 September 2009 23:55:24 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:10 GMT