Ed Summers tweaked the structured blogging template for books to include microformat ("microfauxmat", really, since it's not yet clear how microformats for citations will shape up) semantic html markup for books. Afaict, this removes nothing from the previous template and only adds the microfauxmat span class/value pairs, which are derived from the forma OpenURL standard's book key/value profile.
I've done the same thing for the review-article template (which has already been enhanced by Alf Eaton to support dynamic Pubmed lookups!). You can see an example entry here. Like what Ed did for books, this only changes the structured blogging template's display block to include a microfauxmat for article citations where the span class/value pairs are derived exactly from the OpenURL standard's journal key/value profile.
Here's the excerpted section with the microfauxmat markup:
<div> <h3 class='item fn'><a class='url' href='http://dx.doi.org/10.1016/S0140-6736(05)67894-4'>The first confirmed human case of avian influenza A (H5N1) in Mainland China</a></h3> <p><b>Authors</b>: Hongjie Yu et al.</p> <p><b>Journal</b>: <span class='jtitle'>The Lancet</span></p> <p><b>Year</b>: <span class='date'>2006</span></p> <p><b>Volume</b>: <span class='volume'>367</span></p> <p><b>Issue</b>: <span class='issue'>9504</span></p> <p><b>Pages</b>: <span class='pages'>84</span></p> <p><b>ISSN</b>: <span class='Z3988 issn' title='ctx_ver=Z39.88-2004& rft_val_fmt=info:ofi/fmt:kev:mtx:journal& rft.issn=0099-5355&rft.volume=367& rft.issue=9504&rft.pages=84&rft.date=2006& rft.id='>0099-5355</span></p> </div>
The COinS (which hasn't otherwise changed) CO is broken up into multiple lines for easier reading.
Note that the microfauxmat class/value pairs are exactly the same as those in the context object's rft attribute names, just without the dressing.
There are several issues still to consider with this: how to mark up the URI; how to surround the marked-up block to make it all "of one object"; are there opportunities for microformat class reuse, etc. In the meantime, though, as does Ed's example for books, this seems simple, and is based on a heavily used standard.
If you want to know about why we think reusing the book and journal profiles standardized in OpenURL profiles is a good idea, and are on the structured blogging list, see my writeup here (not sure why they made that private). It highlights that the Dublin Core's own inititative to define a citation format ended up saying, basically, "use OpenURL ContextObjects for machine-parseable citations" (read down through section 2.3). That's exactly what we're doing here. With the microformat approach, though, it's even better: this kind of markup is designed for humans first, and machines also. :)
The altered review-article.xml SB template is attached (derived from the version released with 1.0pre13).
| Attachment | Size |
|---|---|
| review-article.xml.txt | 3.17 KB |
DRY Principle?
I'm glad to see people pushing towards using microformats for structured data on the web. Your "microfauxmat" looks interesting, but seems to have at least one signifcant flaw- it violates the DRY principle.
As you note, the data all gets repeated in the OpenURL CO- needlessly so, it seems.
By following the DRY principle, we can make it much easier on publishers/writers.
Yes, but it's just an example :)
You're right, no argument there.
But, the COinS bit is already working, and already integrates with browser extensions and 1000+ resolvers, and is already in the structured blogging templates for articles and books. So the microfauxmat pieces are just additive, for now.
I hadn't really thought of it this way until I read your comment, but, from today's perspective, OpenURL is partly a solution to a problem we had to solve before there were microformats. We started doing stuff that looked like OpenURL in 1998, and it was immediately effective, and today resolvers all over the world can happily interoperate with COinS (however crappy the usability of today's resolver interfaces might be).
That doesn't mean we wouldn't abandon OpenURL and COinS altogether if a microformat took off that did solve all of this problem.
Fyi, the other part of the OpenURL equation is about what happens when you actuate the link at the resolver or remote site (the end point of a requested service offered by a resolver after you follow an OpenURL link). It doesn't seem likely that a proper microformat will address that piece. So, even if the book/journal microformats go forward successfully, there's still more work to do, and it might even still make sense to keep repeating ourselves. Or at least make sure that code we write can effectively generate the COinS piece from the microformat. :)
Thanks for taking a closer look!
Post new comment