On Tue, Nov 25, 2008 at 20:40, Jonathan Rochkind <[log in to unmask]> wrote:
> With the giant bibliographic universe, your system often has an identifier
> that identifies something you don't have a local record for, you need to go
> fetch it. In general, in our bibliographic universe, we are increasingly
> having to deal with a data set spread accross multiple database instances.
> I write software that takes my bib record, and tries to find a match in
> Amazon, Google, Internet Archive, and maybe the Library of Congress. Maybe
> it tries to fetch authority records from LC or OCLC.
> So, yeah, the identifier is not purely inward-facing, I need to know what it
> IS, so I can know how to match it with external systems that may use a
> different set of identifiers.
You didn't read the info I linked to, did you? :)
> Our conceptual 'database' will be distributed, one way or another, it won't
> be entirely inside a server in my library. And it really already isn't, and
> never has been, with library data.
PSIs are designed for distributed, non-normative, adaptable systems,
and leave the resolving part to those who need or want it. There's a
guarantee to the identities seriousness through it being *published*
(as in, there's both a human and a computer resolving involved). If
you want to work in your own silo, go right ahead. If you need all
identities external, you're welcome. Mix and match? Help yourself.
Since it is published, distributed collaboration and extension is
welcomed.
Regards,
Alex
--
---------------------------------------------------------------------------
Project Wrangler, SOA, Information Alchemist, UX, RESTafarian, Topic Maps
------------------------------------------ http://shelter.nu/blog/ --------
|