Hi Eric,

* Eric Scheid <[EMAIL PROTECTED]> [2006-01-19 21:10]:
>Here is a link to a resource:
>
>    <link type="application/atom+xml" href="..." />
>
>Please explain how a tool can decide whether that is a link to a
><atom:feed> document, or is a link to an <atom:entry> document?

this is an excellent point. How does either Pace address it?

>> In fact, semantically, we should be encouraging people to move
>> things out of their `<link>`s and into `<a>`s in the page.
>
>Sounds like PaceAnchorSupport. How do you propose we do this
>encouragement, if not by codifying it into a spec?

PaceAnchorSupport merely promotes `<a>` to parity with `<link>`,
for which the I-D says a `rel` attribute consisting of at least
`alternate` is required.

But I don’t see how “this is a feed I want to link to” implies a
different relationship with the linking page than “this is some
resource I want to link to.” The mere fact that the linked
resource is a feed (or entry document) does not constitute a
relationship with the linking page in and of itself; and that
makes `rel` the wrong place to shoehorn this information into.

So no, encouraging people to add advisory `type` information to
their links is not equivalent to PaceAnchorSupport.

In any case, the important part is that when someone clicks the
link in a browser, the right things happen. For that, the `type`
is not even necessary, so long as the feed is served with the
correct MIME type and contains a self-link. The type makes some
harvesting easier, but ultimately, it’s not *necessary*.

We’re trying to shoehorn this functionality in with autodiscovery
right now because as of yet, the right things do not happen, even
though they *could*, whereas something closer to the right things
does happen when autodiscovery links are added willy-nilly.

>The existing behaviour is based on the various incarnations of
>RSS where the only document type involved are feeds. RFC 4287
>introduces a new document type, the Atom Entry Document, which
>autodiscovery-01 fails to take into consideration. That doesn't
>meet my definition of "well-written".

I don’t know how that is relevant. I am trying to think of a
scenario where I’d want to autodiscover an entry document (as
opposed to simply linking to it) and the inability to distinguish
between feed and entry documents is causing a problem, but I
can’t come up with anything. Can you provide an example?

Regards,
-- 
Aristotle Pagaltzis // <http://plasmasturm.org/>

Reply via email to