Fairly recently on the Web4Lib mailing list, a thread started by Jim Campbell (of UVA) and David Walker (of Cal State San Marcos) (and others) prompted me to ask what the role of OPAC is in the modern library.

Outside of "Inventory Control System", I don't feel like I got a very good or meaningful response.

I have been thinking a lot about something that Karen Schneider had written a while ago about the need for search interfaces to be search/browse. By this, I mean you begin your session by typing some words in a box and your interface adapts itself contextually to the results and what you should be looking at, so your "browse" options would be logical based on the context of your results.

If your terms were to bring back government documents, say, you would also have the ability to browse our GovDocs research guide or email our GovDocs librarian. If your search brought back a database (for example, ABI/Inform), then the page should also link to the subject guide that includes ABI/Inform (in this case, the Business guide).

This, of course, requires that the "library website" be in a format that makes it potentially servable in this manner. For our site, I have proposed that our content be broken down into small sections (rather than pages) that can be classified and served as necessary.

If you are an undergrad and one of your results happens to be one of your reserves items (which I'll get to in a minute), there's not much need to see the faculty policies for placing something on reserve. There is a use, however, in seeing the circulation policies regarding said reserve as appropriate to an undergrad.

If your search results in a journal that we get through an aggregator that sucks (meaning Lexis-Nexis or Factiva or their ilk), present tutorials on getting to the journal through that aggregator (or just a tutorial, in general).

Searches should be weighted contextually, as well. Objects that appear in your reserves lists or subject disciplines should have more relevance than other things. Circulation/clickthroughs should boost relevance (although I realize that non-circulating items present a problem here).

The important thing I want to see is the relationship between objects and content. My search brings back a journal. Besides the obvious information I want to know about thing (esp. things not included, like, what is it about?), tell me what databases index this thing; what other journals are similar; what is the current ToC (if available via RSS); are there preprints from this journal in our institutional repository/ETDs; etc. If there is any library created content related to a particular object, I want that, too.

I want to break down the silos between our resources and content and different collections.

And, yes, I think articles and other database content should be included in that as well (if you have the credentials to view them - if not, an indication of what you'd see if you were logged in).

If we don't include the entirety of our collections, I am not entirely sure what the purpose of the catalog is.