[uf-discuss] class="url"?

[Scott Reynen]

On Nov 5, 2006, at 11:06 AM, Siegfried Gipp wrote:

> That's not true. hCard data within a vCard record is well related  
> to that
> hCard and thus absolutely not random data.

Not random data, but not necessarily a URL of the contact within the  
hCard, not the "homepage" as you've been calling it.  There are other  
links within hCards and it's not practical to demand publishers  
remove these links.  The very first example in the wild from the wiki  
demonstrates this:

http://www.finds.org.uk/

Here's the markup:
	
<div id="footer2" class="vcard">
   <address><span class="fn org">The British Museum</span>, <span  
class="street-address">Great Russell Street</span>, <span  
class="locality">London</span> <span class="postal-code">WC1B 3DG</ 
span>  | <a href="http://www.finds.org.uk/legal/disclaimer.php"  
title="Legal disclaimer">Disclaimer</a> | <a href="http:// 
www.finds.org.uk/legal/privacy.php" title="Our privacy  
policy">Privacy Policy</a> | &copy; 2006 <br>
   E: <a href="mailto:info at finds.org.uk" title="Contact the Scheme"  
class="email">info at finds.org.uk</a> <span class="tel">T: +44 (0)20  
7323 8611</span></address> </div>

They have a link to their privacy policy and legal disclaimer within  
the vCard.  Any suggested change to hCard needs to allow for this,  
because this is what publishers do on the web today.

>> It may seem redundant, but it is the only way to disambiguate random
>> links with links associated with the vcard/vevent
> That is a point already mentioned. And this may indeed be true.  
> Although still
> i think _inside_ a vCard or vEvent record there sould not be random  
> links,
> but only links related to that hCard or vEvent.

We're not telling publishers what they should be publishing.  We're  
just helping them publish it with more descriptive markup.

>> It may seem redundant, but it is the only way to disambiguate random
>> links with links associated with the vcard/vevent
> So what now?

That depends: what practical problem are we trying to solve here?

> Do we define that any link within a hCard/vEvent record is by
> definition related to that record?

No.  That doesn't accurately describe the intended meaning of  
publishers using links with a block of contact information.

> Or,
> does a hCard/vEvent record contain random urls, of which only one  
> is related
> to the record?

Yes.

> Then you may consider to classify this one and only link with
> class="url".

That's what hCard does.

> But then the meaning of this url is not "any general url", since
> it is not any url to any random data, but it is _the one specific_  
> url. This
> is different from the w3c definitions.

Yes, it is different.  The more specific meaning comes from the more  
specific context.  Why is this a problem?

Peace,
Scott