[uf-discuss] format for identifiers?

David Osolkowski qidydl at gmail.com
Mon Nov 28 13:59:15 PST 2005


On 11/24/05, Ryan King <ryan at technorati.com> wrote:
>
> On Nov 23, 2005, at 1:09 PM, David Osolkowski wrote:
>
> I don't believe the question was misread. Tantek's point was that
> there's no need for a microformat, we already have it covered with URIs.


Well, the OP agreed with my interpretation, but I was also roughly
suggesting that we already have URIs to handle the concept of an
identifier.  Basically, I'm agreeing with Tantek.

So, you're saying URIs are only useful when they're URIs?


The alternative would be a URI that isn't a URI... what would that be?

Given a page containing the string "qwerty123", the OP would like a way to
mark up "qwerty123" as being an identifier.

I'll refactor my point for clarity: An identifier isn't useful if you can't
determine what is being identified.  I suggest that creating arbitrary
identifier schemes/namespaces is not a great idea, and one should consider
using a URL or existing URI scheme/namespace whenever possible.

Second, I would suggest that given an identifier whose format is documented
(and thus, an identifier where one can determine what is being identified),
a good way to mark it up would be using a href="".

Also, C. Hudley suggested that class="url" in hCard is redundant; I'm
guessing that class is needed so that authors can provide links for other
elements in the hCard that wouldn't necessarily be the URL for the person
identified in the card.  For example, you could put a tel:// link around
their phone number.  And I would guess that a good ID for a flickr photo
would be its URL.

- David
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://microformats.org/discuss/mail/microformats-discuss/attachments/20051128/5efb9b4d/attachment.html


More information about the microformats-discuss mailing list