As this points out, Jeff Young's ERRoLs are a form of cool URLs, much as PURLs are, another system my group here in OCLC Research maintains. PURLs basically add the 'another level of indirection' that is one of the basic ways of solving problems in computer science. You give someone a PURL and when they use it the PURL system resolves it to another URL, allowing a central place to manage the resolution of your URLs.
ERRoLs give you a bit more, however, in that the ERRoL system will do protocol transformations too.
ERRoLs are built on top OAI-PMH registries. In practice, we build our OAI registries in turn on top of SRU/SRW servers, so you can already see some protocol transformations happening. Since OAI repositories have date stamps, that makes it fairly easy to turn them into RSS servers.
The ERRoL home page lists a number of repositories it knows about (via the Experimental OAI Registry at Illinois). Following the links will give you a set of screens that help you see the different transformations offered.
What I'm really leading up to, though, is Jeff's latest twist on this whole thing: the MetaWiki which combines the ERRoL approach to metadata with Wiki editing and display. More on that in a later post, as well as what Jeff's been working on the last few days--to extend ERRoLs to the latest OpenURL syntax(available now as ANSI/NISO Z39.88-2004 here).
It's been interesting to watch this whole area develop. We've been working with OAI almost since its inception, and there's always been a few more things that we (mainly Jeff) has been able to see that keeps us working with it. A D-Lib® paper we wrote with Herbert Van de Sompel lays out some of these.
--Th
Comments