Two of the most interesting things here at Code4Lib are OSU’s LibraryFind and dchud’s talk on taking OpenURLs to the next step — making them understand where you (say, by IP address) are and using OCLC’s registry to write links that reference your resolver.

In short, it would finally let library finding tools work sanely for people from any location. It doesn’t address off-campus uses, but it doesn’t make things any worse, either — just proxy like you’re already doing. (OpenID could solve this very elegantly… if libraries decide to jump and be early adopters here. That’ll totally happen, too.)

LibraryFind is a combination multi-search tool and harvester — they think multi-search is inherently a flawed approach (hint: they’re right) and that harvesting is ultimately going to be a better solution.

But multi-search is what we have right now… and man, even in an early stage, it kicks Metalib squarely in the testicles. It provides hits from the catalog as well as article databases (why doesn’t everyone do this?). It caches results, which makes it fast while also hugely pleasing database vendors. It’s really very excellent.

Back to harvesting now — OSU is looking to harvest stuff. Lots of stuff. Unless I mis-heard, they’ve got 200,000 items now… and they’re investigating what it’ll take to scale up to two trillion items.

To reach that kind of scale, you need some very serious investment in infrastructure — something not everyone is gonna want to do. If they, say, implemented the Magic Resolving URLs, this tool would kick ass for everyone.

No tags