[microformats-discuss] Referencing hCards
ryan at technorati.com
Sat Jul 16 18:32:39 PDT 2005
On Jul 16, 2005, at 3:47 PM, brian suda wrote:
> Ryan King wrote:
>> On Jul 16, 2005, at 1:32 PM, brian suda wrote:
>>> One idea for referencing remote hCards was to use the rel="hcard"
>> There's a couple of difficulties with this.
>> First of all, rel="" defines the relationship between the two
>> resources, not the format of the target.
> You are absolutely right. There is the HTML 'type' attribute for
> content-type of the resulting link, but like 'hreflang' there is no
> to control someone else's content. So i guess i ment:
> <a href="../hcard.html" type="text/hCard">Another hCard</a>
>> Second, page A can't assert anything about the format of page B in
>> any reliable manner.
> This is absolutely true.
So don't even try. Even your above example of using the type
attribute falls into the same trap.
>>> This could also lead to the following link on
>>> <a href="http://example.com/hcard.html#brian_suda" rel="me hcard">my
>>> This gives a link to an HTML page, but it is an anchor directly
>>> to the
>>> hCard. The idea being, that X2V could then also search for any
>>> rel="hcard" attributes and spider those links as well. Then you
>>> build a disconnected network of hCard nodes.
>> If we're building a network, why not just use XFN?
> XFN just shows relationships between two people ('websites'), but the
> hCard would also build a distributed address book.
>> Yes, it makes sense, but I don't think the rel="hcard" is necessary
>> in any of it. First of all, its unreliable (which was the point I was
>> trying to make above) and secondly, a spider could just hit all of
>> your friends links to find hcards.
> The unreliablity is certainly a factor. I would perfer NOT to
> spider all
Right. So, just spider XFN'ed links. So:
1. I link to you.
2. You have an hCard on your homepage OR
3. link to your hCard with rel="me"
voila! distributed addressbook with XFN + hCard.
> otherwise i might be traversing the internet. It could be limited
> to a depth of 1, but that is still un-needed work or bandwidth.
More information about the microformats-discuss