ok. to make things clearer, i'll describe my case at hand. when visiting the pages of one of our journals (<a href="http://www.livingreviews.org/">http://www.livingreviews.org/</a>), zotero is not able to pick up much of the metadata. this is because the currently supported mechanisms to get data - embedded rdf or dc xml - won't work for pages delivered in html
4.01. on the other hand, we do have an oai pmh interface, thus a perfect mechanism to harvest metadata. the only thing lacking is a standard way to tell zotero where to find this interface, and which identifiers to use for querying it.
<br>you have a point in saying that unAPI may be better suited anyway. and since we also do have a unAPI interface, we're all set as soon as zotero supports this.<br>i was just wondering, whether auto-discovery (of the baseurl only) wasn't worth including into oai pmh. and i think auto-discovery links are still rather unobtrusive, even if there's 20 of them in the head element of a page.