W3C home > Mailing lists > Public > www-tag@w3.org > February 2009

Re: google supports on rel="canonical" ISSUE-27 rel-ownership

From: Phil Archer <phil@philarcher.org>
Date: Wed, 18 Feb 2009 15:39:27 +0000
Message-ID: <499C2BAF.1010005@philarcher.org>
To: Dan Connolly <connolly@w3.org>
CC: www-tag@w3.org
Dan,

This is interesting on a number of levels.

Wearing my new W3C Mobile Web Initiative hat, this appears (falsely as 
it turns out) to be a contribution to the One Web debate. MWI Best 
Practice implores developers to use a single URI even it it dereferences 
into mobile and desktop representations [1] depending on the user agent 
  - and using a 302 to redirect to a particular version is a valid 
method. The problem then is that one might bookmark the representation 
rather than the original URI. OK, so far so well known by the TAG! So 
might rel="canonical" be relevant here?

No... because of this (from the blog post):

"Is it okay if the canonical is not an exact duplicate of the content?
We allow slight differences, e.g., in the sort order of a table of 
products. We also recognize that we may crawl the canonical and the 
duplicate pages at different points in time, so we may occasionally see 
different versions of your content. All of that is okay with us."

A mobile-friendly representation should be thematically consistent (the 
mantra of MWBP) but it might be very different in other ways.

So no, rel="canonical" isn't relevant here. The @rel value bookmark is 
probably a better fit but, if memory serves, implementation of that is a 
little sketchy.

Removing my MWI hat and donning my POWDER/general irritant where 
metadata is concerned hat... here's another @rel value that has popped 
into existence without going through any kind of registration.

Going back to last November's open session of the HTML 5 WG at TPAC, I 
took away from that an unofficial action to try registering a new @rel 
value through IANA. This I have done, eventually, without any 
difficulty. However... that word 'eventually' needs explaining.

The main cause of delay was not the IANA system - it was me and my 
inexperience in this area. Firstly, the actual registry of @rel values 
maintained by IANA relates exclusively to ATOM [2] so if you want to get 
a new entry in there, it's best to talk about ATOM and not, as I 
mistakenly did at first, things like XHTML and HTTP Link. When I 
eventually got my end of the bargain sorted out and submitted a properly 
formatted request [3], I got a (positive) reply back less than 2 weeks 
later (yesterday in fact [4] so I'm hoping to see 'describedby' added to 
the registry in the coming days).

Now... if it really is /that easy/ to register a new @rel value with 
IANA, then IMO, things like rel=canonical really should be registered too.

If, as I dearly hope, Mark Nottingham's HTTP Link: draft [5] (and the 
related work by him and Eran Hammer-Lahav on site-meta and discovery) 
becomes an RFC then the ATOM link registry will become much more generic 
  and the addition of new terms really should become the norm.

Cheers

Phil.



[1] http://www.w3.org/TR/mobile-bp/#OneWeb
[2] http://www.iana.org/assignments/link-relations/link-relations.xhtml
[3] http://lists.w3.org/Archives/Public/public-powderwg/2009Feb/0007.html
[4] http://lists.w3.org/Archives/Public/public-powderwg/2009Feb/0014.html
[5] http://tools.ietf.org/html/draft-nottingham-http-link-header-03

Dan Connolly wrote:
> I just discovered this:
> 
> [[
> Now, you can simply add this <link> tag to specify your preferred
> version: 
> 
> <link rel="canonical"
> href="http://www.example.com/product.php?item=swedish-fish" />
> ]]
> --
> http://googlewebmastercentral.blogspot.com/2009/02/specify-your-canonical.html
> 
> 
> Perhaps worth updating the generic resources finding?
>   ISSUE-53 genericResources-53
>   http://www.w3.org/2001/tag/group/track/issues/53
> 
> It sorta goes against our advice on follow-your-nose and
> self-describing web, but I guess it's not new information,
> since we considered rel="nofollow" before.
> 
> ISSUE-51 selfDescribingWeb-51 well known formats and URI based
> extensibility
> http://www.w3.org/2001/tag/group/track/issues/51
> 
> 
> 
> See also an HTML WG issue:
>   ISSUE-27 rel-ownership @rel value ownership, registry consideration
>   http://www.w3.org/html/wg/tracker/issues/27
> 

-- 

Phil Archer
http://philarcher.org/

i-sieve technologies                |      W3C Mobile Web Initiative
Making Sense of the Buzz            |      www.w3.org/Mobile
Received on Wednesday, 18 February 2009 15:55:09 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Thursday, 26 April 2012 12:48:12 GMT