Re: New Version Notification for draft-divilly-atompub-discovery-00

["Nikunj R. Mehta" <nikunj.mehta@xxxxxxxxxx>]

Actually, if feed and service are available, then one could use the  
app:collection mechanism to discover the feeds that are editable as  
well as its publishing metadata.

Actually, HTML manages its own registry of link rel values and uses a  
slightly different syntax for this attribute. Mark Nottingham's been  
working to get a unified Link registry that can be used with a new  
HTTP header [1]. As to where a link should be specified for  
collections and services, the HTML5 spec [2] identifies "link" as  
document metadata and hence expects such links to be placed in the  
head element. N.B. rel=service has not been included in the initial  
registry of HTML link types.

XHTML2 spec [3] identifies "link" as metainformation and expects links  
to be placed in the head document

Is there some other use you have in mind?

Nikunj Mehta
http://o-micron.blogspot.com

[1] http://tools.ietf.org/html/draft-nottingham-http-link-header-05
[2] http://dev.w3.org/html5/spec/Overview.html#the-link-element
[3] http://www.w3.org/TR/xhtml2/mod-document.html#edef_document_head

On May 20, 2009, at 10:46 AM, Hadrien Gardeur wrote:

> Nikunj,
>
> I agree: it makes sense to divide hierarchy and collection discovery  
> in two separate I-D.
>
> About discovery in HTML documents: you specify rel="service" for  
> AtomPub Service Documents and rel="feed" for an Atom feed, don't you  
> think that it might be useful to have a specific way to discover  
> writable feeds too (backed by a collection) ?
>
> Where should these links be specified ? rel="service" is usually  
> strictly used in the header, but some other rel values from the IANA  
> registry are also used in the body XHTML documents (rel values for  
> pagination for example).
>
> Hadrien