Some of my readers might be following Dan Chudnov and company's work on the "appropriate resolver" problem. The problem under consideration is how to pass OpenURLs around inter-institutionally. OpenURLs are made up of two parts: 1) the address of the OpenURL resolver (which depends on who the user is) and 2) the referent, which contains the user-independent bibliographic information. The solution that Dan and others have been exploring (as a proof of concept) is to pass only the OpenURL referent and then programmatically prepend the appropriate resolver. Specifically, :

1) have information providers generate href-less anchors with the OpenURL referent. For example (Note: I've broken the referent in half to facilitate formatting):

<a name='spage=222&isbn=1581138105&id=doi%3a10%2e1145%2f1031607%2e1031643&date=2004&epage=231&auinit=BA&aufirst=Bonnie&aulast=Nardi
&atitle=Blogging+as+social+activity%2c+or%2c+would+you+let+900+million+people+read+your+diary%3f&genre=article' rel='alternate' title='OpenURL'></a>

Note the a tag with the rel='attribute' and title='OpenURL' attributes and the absent href.

2) having users run a bookmarklet customized with the appropriate resolver to rewrite the a tag to stick in a href with the resolver andreferent. Dan has tried to make it easier for folks to get the correct bookmarklet by creating a little demonstration bookmarklet directory. (which I would surmise is dynamically generated from a table of OpenURL resolvers..

Why should we care, since, admittedly, this demonstration is a hack. Nonetheless, it is certainly a way to get at sharing bibliographic information across instititutions. More on this subject later.

Tom Schirmer and I can now participate in that hack with the Scholar's Box. We've made OpenURLs from Melvyl and MetaLib search results and write them out to a web page. If you go to an entry on my blog, you won't see any OpenURLs. If you "view source", you will see embedded OpenURL referents, which you can then tie to a resolver if you apply one of the bookmarklets.