[OAI-general] Getting to grips with OAI

[Douglas Campbell]

I just stumbled (as you do) across the OAI site as I am currently looking into metadata harvesting.  However I am mainly looking at HTML meta data. I'm just trying to get clear in my head what OAI intends to do...

It appears to me that metadata is available for harvesting on any HTML page which contains it.  But the intention of OAI is to make it possible to get metadata for records [locked] within databases whose content isn't available directly via individual HTML pages.

One way would be to create an HTML page for each record in the database so the information is available for harvesting - not a particularly nice solution.  So OAI defines an interface protocol where you can query a database directly for metadata using programs over the Web.  I think I saw that you can ask for ALL the records, or specific subsets - this may be where its main difference to Z39.50 is [apart from how non-trivial it is to bolt a Z39.50 server to the front of your database!].

Mind you, once you have the metadata, how would you then provide access to the described resource, say if it came up in a search result from your harvested metadata, because it is locked up in a database behind a website?

One other thing I was wondering - was SOAP (http://www.w3.org/TR/SOAP/) looked at for the HTTP request/response formats?

Any enlightenment would be appreciated.

Thanx,
Douglas Campbell
National Library of New Zealand