2006/12/1, Mark Baker:
Urgh, sorry for my tardiness; I'm falling behind on my reading.
On 11/30/06, Thomas Broyer wrote:
> I'd prefer basing autodiscovery on the media types and not at all on
> the relationships.
All a media type tells you (non-authoritatively too) is the spec you
need to interpret the document at the other end of the link. That has
very little to do with the reasons that you might want to follow the
link, subscribe to it, etc.. Which is why you need a mechanism
independent from the media type. Like link types.
See the mail I just sent in response to Ian.
Consider hAtom. If you went by media types alone, you'd be confronted with;
<link type="text/html" href="hatom.html" />
Not particularly useful for subscription (or anything else for that
matter) is it?
How does hatom.html relates to the current page? Is it an alternate?
is it a "container" (rel="up", rel="index")? why would I subscribe to
such a thing if I don't know what it is about?
(also, note that rel="" is required for <link> elements).
This would be better;
<link rel="feed" type="text/html" href="hatom.html" />
It still doesn't tell me what it has to do with the page I'm looking at.
I do agree there is a "problem" in these cases, and that's why I
originally proposed keeping a rel="feed", but with a clear definition
as a relationship (opposed to a "kind of resource I'm linking to").
Autodiscovery should ideally be based primarily on link types, and
only secondarily - as an optimization - on media types. Even this
should work;
<link rel="feed" href="hatom.html" />
As long as hatom.html is a feed where the current page is (or has
been) linked to as an "item".
If you are already looking at hatom.html, your hAtom-aware browser
should already provide you with a "subscribe to this page"
link/button/etc.
If you can't describe the relationship between the current page and
hatom.html, there is little chance that this is a resource of interest
and that the person reading the page will subscribe to it (at least
without "visiting" it).
With rel="feed" as a real relationship (à la rel="index"),
autodiscovery can be (as it should have already been) based on media
types (am I able to subscribe to such a thing?) *or* rel="feed", with
an equal "priority".
If it appears than my proposed rel="feed" really is identical to
rel="index", then a new mean should be found (e.g. a new attribute
<link rel="index" href="hatom.html" type="text/html" subscribable>)
Saying "this is something you can subscribe to (it's a feed)" is not
talking about relationships. On the contrary, saying "this is an
'index' and it incidentally is something you can subscribe to (it's a
feed; either by using the 'type' attribute an hypothetical
'subscribable' attribute)" is.
--
Thomas Broyer
