Re: Schema.org External Enumerations mechanism

On 22 April 2012 03:26, Guha <guha@google.com> wrote:
> Good to see a nice heated discussion :)

:)

> There is nothing stopping anyone today from using wikipedia urls as
> identifiers for countries or whatever else, just as there is nothing
> stopping anyone from using any vocabulary they choose.

I think it's reasonable to assume we are talking within the context of
schema.org.

> The goal of schema.org (as Dulitz points out) is to provide a single place
> (for the kind of webmaster who doesn't read this group for fun) to get as
> much of the vocabulary they need as schema.org can possibly provide.

Hmm, wasn't the idea to be able to get (most of) the common pieces of
vocabulary, rather than as much as possible..? Did I misunderstand, or
is that a little scope drift?

Which
> is why we want to 'bless' certain external vocabularies as being recognized
> by us.

Right, I do think blessing established vocabularies/dictionaries etc.
is a good idea.

> We can do this by saying on say the http://schema.org/Country page that
> http://en.wikipedia.org/Country lists the canonical list of countries and
> let the webmaster deal with the rest or have schema.org urls for each of the
> countries and have the schema.org pages point to the corresponding wikipedia
> pages. This is no more an attack on the decentralized nature of the web than
> are url shorteners.

But URL shorteners *are* an attack on the decentralized nature of the
web ('attack' seems a little strong, but I'll go with it). If the use
of shortened URLs from a given service is encouraged in preference to
the original URLs, then that centralized bottleneck is added to the
system. But a little compromise of the system is acceptable, as there
is tangible utility - URL shorteners allow people to communicate URLs
in systems that can't cope very well with long URLs.

But where is the tangible utility in using schema.org URLs? As noted
earlier, they actually add friction to the system.

> And of course, people are absolutely free to use wikipedia or whatever other
> urls directly.

Yes, but isn't it plausible that implementors will tend to scope their
tools to schema.org terms over anyURL? (In a similar fashion to the
way certain implementors have hardcoded against namespace prefixes).
The effect there could be into the range of 'attack'. A hypothetical
risk, sure, but that's balanced against no tangible utility :)

Cheers,
Danny.

-- 
http://dannyayers.com

http://webbeep.it  - text to tones and back again

Received on Sunday, 22 April 2012 13:41:38 UTC