W3C home > Mailing lists > Public > public-html@w3.org > February 2008

Re: rel="contact"

From: Ryan King <ryan@theryanking.com>
Date: Mon, 11 Feb 2008 22:11:44 -0800
Cc: public-html@w3.org
Message-Id: <DEA6ADEF-9CB4-47C8-BEF9-0E6F406078F4@theryanking.com>
To: Matthew Raymond <mattraymond@earthlink.net>

On Feb 11, 2008, at 7:49 PM, Matthew Raymond wrote:
> Ryan King wrote:
>> html5 defines rel="contact"[1] in a way that clashes with XFN[2],   
>> which has wide deployment on the web today.
>> I suggest that rel="contact" be dropped from HTML5, since you can   
>> accomplish the same things with the existing combination of XFN 
>> +hCard.
>> [...]
>> 1. [http://www.whatwg.org/specs/web-apps/current-work/multipage/section-links.html#link-type3 
>> ]
>> 2. http://gmpg.org/xfn/
>
>   While I understand the obvious conflict with existing microformat  
> content, I can't say that XFN's definition of "contact" is at all  
> intuitive. If I were to see |rel="contact"| in an HTML document  
> without knowing anything about XFN, my first thought would be that  
> the hyperlink points to contact information as opposed to being a  
> link for the web page of someone who is a contact. In fact, I don't  
> know anyone I'd refer to as a "contact", and even if I did, I'm not  
> sure how that helps a third party, except so that they can bother me  
> to pass on some sort of message to the "contact" in question.
>
>   (Similarly, I would interpret "parent" to be the same as "up" in a  
> hyperlink context. By contrast, "friend" or "crush" are far more  
> obvious.)

Whether or not XFN has a useful definition of these terms, there are  
already millions of instances on the web of people using these terms  
in this fashion, and there are tools to index and extract the data[1].

-ryan

1. http://code.google.com/apis/socialgraph/
Received on Tuesday, 12 February 2008 06:11:59 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 9 May 2012 00:16:12 GMT