Well, Z-Day has come and gone and it’s time to tidy up the wrapping paper. My first impression of Zotero is positive: the integration of citation capture into the browser is looking good.

There are a range of generic options (create a new citation from scratch, capture current web page, capture link, create snapshot). More importantly, there are specific options when Zotero recognizes a page it knows how to scrape from: you see an icon in the location bar, like the RSS icon when Firefox discovers a feed. Click it to capture the citation. It did a decent job of grabbing book citations from our OPAC and from Amazon, and quite a nice job of grabbing article citations from JSTOR–with one important caveat, discussed below. By default it takes a snapshot when it creates a record, but this seems a little slow and I turned it off. Beyond individual books and articles, you see a folder icon when you’re on the search results screen in JSTOR: you can save the whole set of 25 records with one operation (though you need to check them off one by one: it needs a “check all” option). It takes a while, but it’s very cool. But turn off the snapshots: not only were they excruciatingly slow, they ended up as empty PDFs.

Once you’ve captured some stuff, you can use the “Locate” button to find them using an OpenURL pointed at your institutional resolver. Here a problem in the capture of JSTOR citations shows up: Zotero captures the date in an unstructured format (“Dec., 1931”), and then omits it from the OpenURL, presumably because it didn’t have a clean year and month. Without dates, SFX (for one) does a lousy job of managing thresholds, and so many of the OpenURLs failed to resolve back to JSTOR, although they would if they included the year. If I edit the date down to just the year “1931”, it gets included in the OpenURL and all is well. The problems with date handling pointed out by Bruce D’Arcus therefore have very practical consequences.

I’ve only played with the capture mechanisms so far; there are also functions for exporting and managing citations. Zotero’s beta 1.0 release has come so far, and has such strong backing, and such attractive plans, that I’m confident we can look forward to a really useful tool.