Re: In support of draft-nottingham-http-link-header-01

... and it's in the editing phase still because I haven't yet received  
much in the way of critical review, even though it proposes some non- 
trivial changes.

I'll solicit reviews one more time, and if that goes well, will issue  
a revision and ask for publication; that may draw more useful feedback  
out of the woodwork.


On 22/04/2008, at 9:25 AM, Lisa Dusseault wrote:

> Hi Phil,
> Thanks for the input and keeping us in the loop on the POWDER WG and  
> plans for DRs and PICS, that's good to know.
> FYI, the link-header draft is still in editing phase, and Mark has  
> not requested publication yet -- so the IESG has not yet been  
> officially called on to take any action on this document, let alone  
> approve as an RFC.  Still, early input can be useful so I don't  
> intend this information to deter such input.
> You might also provide this input to the HTTP WG, because getting WG  
> consensus around such a document is always better for getting it  
> approved as a standards track RFC, implemented and deployed.  I tend  
> to agree that the Link header work will be quite useful for HTTP  
> extensions and connections to non-HTTP functionality as well, which  
> makes it all the more important to get this done right and quickly.
> Regards,
> Lisa
> On Apr 22, 2008, at 4:32 AM, Phil Archer wrote:
>> Dear IESG members,
>> I'm writing on behalf of the POWDER Working Group at W3C [1] to  
>> support the draft submitted by Mark Nottingham, dated 14 March 2008  
>> [2]. The WG would like to see this become an RFC.
>> The use case we have for the HTTP Link header is set out in an e- 
>> mail sent to the W3C TAG mailing list [3] which quotes from and  
>> builds directly on our use cases document [4]. In essence, the  
>> Protocol for Web Description Resources (POWDER) is designed to  
>> facilitate greater personalisation of Internet content through the  
>> provision of metadata that can be created separately from the  
>> multiple resources it describes and that can be authenticated.
>> A typical use case would be whether or not to include links to  
>> other resources on a page delivered to a mobile device, whether to  
>> recommend certain resources for school study and so on. In each  
>> case, the ability to find the metadata without having to parse the  
>> relevant resource offers a substantial optimisation in processing.
>> HTTP Link, as set our by Mark Nottingham, achieves this. A HEAD  
>> request to a given resource would be sufficient to identify the  
>> location of any Description Resources that may be available.  
>> Moreover, for some resource types, it offers the only practical way  
>> to provide the link to the Description Resource.
>> The POWDER WG intends to submit a proposal for at least one  
>> relationship type to be used in HTTP Link Headers.
>> The relevant section of our Recommendation Track documentation is  
>> at [5].
>> Incidentally, if we are successful in reaching full Recommendation  
>> status, it is likely that PICS [6] will be withdrawn.  It would be  
>> appropriate in that case to withdraw the PICS HTTP Header too.
>> Yours faithfully
>> Phil Archer
>> POWDER WG Chair.
>> [1]
>> [2]
>> [3]
>> [4]
>> [5]
>> [6]
>> -- 
>> Phil Archer
>> Chief Technical Officer,
>> Family Online Safety Institute
>> w.

Mark Nottingham

Received on Tuesday, 22 April 2008 17:00:32 UTC