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

Re: Link: grandfathered format-specific registered relations

From: Mark Nottingham <mnot@mnot.net>
Date: Sat, 22 Aug 2009 11:44:07 +1000
Cc: ietf-http-wg@w3.org
Message-Id: <9FB54EEF-2826-4687-8ADD-6BFE27398302@mnot.net>
To: Nicolas Alvarez <nicolas.alvarez@gmail.com>
I'm not sure what this has to do with my message.

While we could try to optimistically define the semantics of each of  
these in a generic fashion, it's likely to be a shot in the dark --  
i.e., there is as of yet no proven value to more than one application  
in these, just speculation that it may be later.

To reiterate - registered relations are supposed to be generic, while  
extension relations have no such constraint. However, the registry  
currently has some specific relations in, because they're already there.

As such, the proposal is to recognise this in the registry.


On 22/08/2009, at 10:45 AM, Nicolas Alvarez wrote:

> Mark Nottingham wrote:
>> I'm thinking that the following registered relations are in fact
>> specific to one format or application, and this should be noted in
>> their registrations.
>>
>> current
>> edit
>> edit-media
>> enclosure
>> next-archive
>> prev-archive
>> replies
>> service
>>
>> Comments?
>
> Wondering... As Sam says, relations should be generic and types  
> specific.
> That makes sense for GETtable URis, for example, rel=replies is  
> generic and
> type would say whether it's a RSS feed or an HTML page with replies.
>
> However, what about POSTable URIs? The type attribute, as I  
> understand it,
> says what data you'll get back, not what data to send. How can you  
> say that
> a rel=edit link is a URI where you can POST a specific data type,  
> maybe
> getting text/html in the response?
>
>
>


--
Mark Nottingham     http://www.mnot.net/
Received on Saturday, 22 August 2009 01:44:46 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Friday, 27 April 2012 06:51:08 GMT