Ok, assuming that we've come to agreement on this, we need a pace
written up. Specifically:
a. require that APP implementations, at the very least, serve the
introspection document up using an APP specific mime type
b. defining link[rel="introspection"] as a thing that can point to the
introspection document from (x)html
c. specify that link[rel="introspection"] SHOULD point to the
introspection document that is being served up using the APP specific
mime type but that link[rel="introspection"] but may be used to point to
introspection documents served up using other mime types, e.g.:
<link rel="introspection" type="application/atompub+xml"
href="/introspection" />
<link rel="introspection" type="application/xhtml+xml"
href="/introspection?f=xhtml" />
d. define what it means (if anything) if an (x)html document has
multiple <link rel="introspection" />, e.g.:
<link rel="introspection" type="application/atompub+xml"
href="/introspection1" />
<link rel="introspection" type="application/atompub+xml"
href="/introspection2" />
<link rel="introspection" type="application/xhtml+xml"
href="/introspection1?f=xhtml" />
<link rel="introspection" type="application/xhtml+xml"
href="/introspection2?f=xhtml" />
Luke Arno wrote:
On 10/29/05, Eric Scheid <[EMAIL PROTECTED]> wrote:
On 30/10/05 1:05 AM, "Luke Arno" <[EMAIL PROTECTED]> wrote:
Yes. rel="atom-introspection" or something like that.
@rel="introspection" and a mime type in @type would be sufficient,
I should think.
Good point.
Agreed. I don't think we should legislate anymore than we
have to. Give them an optimally useful format and a mime
type to kick off the apps and leave the door open for
creative implementations.
which might include some non-Atom interfaces, so a more general @rel might
be better.
That too.
- Luke
