Monday, June 27, 2005

Peter Brantley's new blog #

I only very recently discovered that Peter Brantley, Director of Technology at the California Digital Library, has started a public blog, shimenawa. Even though I already have semi-regular conversations with Peter, I'm happy to read his latest thoughts about libraries and whatever else he will choose to write about. I'm envious that Peter will be able to attend Where 2.0 and look forward to seeing how he reacts to the conference. I have been experimenting a lot recently with geospatial services, specifically combining Flickr and Google Maps. I've been working on a write-up of my experimentation (e.g., Google Maps Hacking); you can see some of the pictures that I have georeferenced centered around 2195 Hearst, the location of my office. Peter is really grooving on the technology. Today is his first podcast, involving Jane Lee and Colleen Whitney on the perceptions of libraries at SIMS.   Posted by Raymond Yee on 6/27/05; 5:47:41 PM from the Libraries dept. Discuss (1 response)

Should I start embedding URNs in my HTML? #

In response to What is a definitive URI for books?, Bruce D'Arcus pointed me to using URNs as identifiers. e.g., urn:isbn:0520237048, noting that both ISBNs and ISSNs are registered URNs. I didn't give URNs a serious look because I had actually never seen any "real support" for them. That is, I can write urn:isbn:0520237048, but what would that do? Still, Bruce's email prompted me to take a closer look, which led to Ben Meadowcroft's URNs, citations in web authoring: This article covers the techniques authors on the web can use to reference articles and publications in a robust manner. It was written to highlight some of the inadequeces of relying on non persistant URI references for academic and technical papers. It was partially born out of my considerations on holistic hypertext and how this could be integrated into both CMS and KMS. which I found via URN, some URI scheme - Anne’s Weblog about Markup & Style. I learned two interesting things: the cite attribute for blockquote and q in XHTML 1.0 that although there doesn't yet to be widespread support for URNs in browsers, there is an URN Support extension for Mozilla/Firefox, which worked for Ben Meadowcroft. I installed the extension and clicking on a urn:isbn led to the corresponding amazon.com page. Here I provide my example: Introduction to California Plant Life: <Q cite=urn:isbn:0520237048>Nearly one-fourth of the plants found in North America north of Mexico, and more than are found in any other state, grow in California.</Q> Clicking on the name of the book does not lead anywhere in most browsers because there is no native support for URNs, but any bot/search engine reading my wiki can now pick up on the ISBN, stated in a RDF-friendly way. One might find the default linking of the browser extension to amazon.com rather ironic. There's a lot of jumping through hoops to write URNs, which then gets directed by the browser extension to where people are already linking to in the first place. But note that the decoupling of the identifier from the locator does make the important statement that ISBNs are not synonymous with amazon.com and allows for other explicit aasociations to be made. For instance, the extension could be rewritten to make links to the Library of Congress instead. I'm still mulling how to make practical use of URNs....   Posted by Raymond Yee on 6/27/05; 5:01:30 PM from the Web Technology dept. Discuss

What is a definitive URI for books? #

I come back to problem of what URL to provide for a book so that others can most easily manipulate thet book information. I often discuss books on my weblogs and wiki and have long wanted a definitive and neutral way to identify a particular book, a "definitive URI for books", as Kellan Elliott-McCrea (a.k.a. Laughing Meme) puts it. It seems that the de facto standard for indentifying books on the Web is to link to its amazon.com page (providing that page exists). This state of affair is not surprising given how easy and attractive amazon.com has made linking to its pages. The book pages are typically attractive and informative, full of reviews. Amazon.com is easy and fast to search and provides an API to boot. When allconsuming.net was still primarily about books, it spidered weblogs for links to amazon.com (and other book sites). The new technoratic.com tracks new technorati.com tracks popular books, "ordered by new links to Amazon in the last 48 hours". I have certainly done my share of linking to amazon.com to reference books. However, there are good reasons why linking to amazon.com is less than ideal. As LaughingMeme writes: And there are reasons not to link to Amazon, primarily I don't want people buying their books from Amazon but from their local bookstores! Also you are: bombarded with ads for People's magazine, Epson printers, and clean underwear, tracked and indexed, and have handed over the keys to controlling conversations about books to a single corporation who aspire to be just like Walmart, not the what I look for in a Muse. To provide myself a place other than amazon.com to which to link for books, I have also experimented with my own little cgi script (e.g., http://raymondyee.net/projects/biblio/bookInfo.py?isbn=0520237048) that returns links to a variety of book-related sources (including amazon.com, Library of Congress, the Berkeley Public Library, melvyl) given an ISBN. I've made it easy for me to generate links to the script on my wiki through a macro. (e.g., [[BookWithISBN(0520237048,"Introduction to California Plant Life")]] generates Introduction to California Plant Life. Although placing such links to raymondyee.net reflects my sense of books as items available from many sources, the links will not be readily reconizable to search engines such as technorati.com as pointers to books. I've decided to improve my bookInfo page because it would be behooves us in the library/cultural heritage communities to find better ways to link to books. I also need a place to experiment with various ideas of mine. On the smaller scale, I'd like to: Allow customization of the links that come up. I link, for example, to the Berkeley Public Library -- but why should readers who have no interest in the Berkeley Public Library see that link? Ideally, I'd like a user to add any link that makes use of the ISBN. It would be great to find or create a directory of such services. For instance, I should make it easier to link to isbn.nu if one would like. Figure out how to link to the Open WorldCat system at OCLC, which is what the big players Google and Yahoo and other partners are doing. When can others play? Specifically, how I can programmatically generate the Open WorldCat link that corresponds to "Introduction to California Plant Life" whose ISBN is 0520237048. Now I can do a search on Google scholar and pull the URL for "Library Search" to get a working link....but I would guess that OCLC and/or google would not like others' doing so. (I may need to re-ask a previously posed question to Lorcan Dempsey.) Link to redlightgreen. I'm not sure whether the link to Orndufff's book is as simple as http://redlightgreen.com/ucwprod/servlet/ucw.servlets.UCWController?ACTION=search&SRCHBY=identifiers&SRCHTERM=0520237048 Explore the use of Latent OpenURLs in HTML in this context. We have book information. Let the user route that information in any way he or she chooses. See Main Articles: 'Opening up OpenURLs with Autodiscovery', Ariadne Issue 43 for more context. Figure how how to think of this problem in the context of the Semantic Web and RDF. Add a technorati link for book discussion in the blogging world. e.g., http://www.technorati.com/search/www.amazon.com/exec/obidos/ASIN/0520237048 links to discussion on the book with ISBN 0520237048 Decide on whether to retain the link to allconsuming.net, which the last time I checked, does not work. On a larger scale, I might want to: Do some more research on communal efforts to create a definitive Uniform Resource Identifier (URI) for books. Do many people even care about the issue? Are librarians content with people linking to amazon.com as a way of pointing to a book for the lack of other places to point to? If not, what would we rather have people link to? Turn bookInfo.py into a real service that I can rally others to use. Obviously, the cgi script is not the place for others in search of a definitive URI for books to use. What would it take to make the page the definitive site? I'd be glad to have Library of Congress or some big library or a meta-library system like OCLC take over this job. But until I can convince a major organization to do so, what can I do to rally some use around a system?. Get beyond ISBNs.to handle books that don't have ISBNs and for a service to identify related works.   Posted by Raymond Yee on 6/27/05; 11:39:05 AM from the Web Technology dept. Discuss