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

[Tantek Ç elik]

On 4/25/06 8:02 PM, "Ross Singer" <ross.singer at library.gatech.edu> wrote:

> On 4/25/06, Tantek Çelik <tantek at cs.stanford.edu> wrote:
>> Yes, we get things done because we dismiss the 20% in favor of getting the
>> 80% working well.
>> 
> 
> Except, in this case, you've dismissed the entire problem that we're
> trying to solve here by merging into some only quasi-related "80%"
> case.

Ross, if the problem you're trying to solve doesn't involve common real
world publishing cases on the *Web*, then yes, it should be dismissed as far
as microformats are concerned.  This is the wrong place to solve problems
that don't fit those constraints.

> I would also like to note, that you are, in fact, the only dissenting
> voice on a lot of these.  You, in fact, seem like the 20%.

Voices on a mailing list don't determine the 80 vs. 20.  That's the point of
the process.  Documented real world publishing examples on the web determine
the 80 vs. the 20.


>>> And, quite frankly, that's a disturbing viewpoint.
>> 
>> To some perhaps.
>> 
>> To those that seek a perfect schema or format to cover all (or even most)
>> edge cases, microformats may be quite disturbing.
>> 
> 
> I think it's easy to cast anything as an "edge case" when it doesn't
> pertain to your business model or current occupation.

When it doesn't pertain to common real world publishing on the Web, then
yes, for our purposes, it is easily cast as an edge case.


> To cast URI
> (or, rather, the entire universe of "identification" outside of what
> currently has a URL) as "edge case" seems particularly short sighted.

You're right.  On the *Web*, URIs which are not URLs are an *edge case*.


>> We've deliberately, not lightly, taken this stance, and no amount of
>> theoretical problem raising will change it.
>> 
> 
> When you can prove, beyond a "theory" that "url" is the predominant
> "identifier" you can make that claim. Until then, it's just
> stubbornly sticking to a dogma.

Given that everything on the Web uses URLs, it is you who have the burden of
proof in this matter.  Until you can prove that URNs are nothing more than
edge case on the Web, then this is not worth discussing any further.

Tantek