"uid" microformats? (was Re: [uf-discuss] ISBN mark-up)

[Scott Reynen]

On Apr 26, 2006, at 11:44 AM, Tantek Çelik wrote:

>> A large portion of what is published on the web references things
>> that don't exist on the web, and thus don't have a canonical URL.
>
> Right, and to resolve whether it is a "large portion" or not, we  
> ask that
> such things are documented in examples pages, and the citation- 
> examples are
> a good example (so to speak) of this.

Ok, here are a hundred thousand references to "a cup of coffee":

http://technorati.com/search/%22a%20cup%20of%20coffee%22

None of these cups have canonical URLs that allow me to uniquely  
reference the cup of coffee on the web.

>> Having
>> a canonical URL on the web suitable for a UID is a bad constraint for
>> a microformat.
>
> This I am not sure about.  It certainly seems like a *good thing*  
> to provide
> incentive for more UIDs to become canonical URLs on the web.
>
> But I would agree that this shouldn't be a constraint/requirement  
> per se,
> but rather should be a bias (i.e. SHOULD) so that we provide  
> incentive or at
> least preference in the direction of more canonical URLs.

And everyone else seems to be arguing that this shouldn't be a MUST.   
MUST is a constraint.  SHOULD is a *good thing*.  Where is the  
disagreement here?  Let's not waste our time explaining why we  
disagree before making sure we really do.

>> Right.  I think saying "80% of people do X" without pointing to the
>> real world publishing examples that back up such a statement makes it
>> look like voices on this mailing list are determining the 80 vs. 20.
>
> Right.  Hopefully we only do so in "obvious" cases (e.g. things on  
> the Web
> have URLs :)

That's what I thought, but I apparently illustrated my own point with  
my "large portion" comment.

Peace,
Scott