Openly Linking

Miles Poindexter
Openly Informatics, LinkBaton Conductor Program: http://my.linkbaton.com/conductors
miles@openly.com

Presented at the OpCit technical seminar, 13 July 2001
This is a partial transcript of Miles' talk, had the Web connection been working at the time!

LinkBaton

First, a short demonstration linking using LinkBaton.

Here is a Web page where a book has been reviewed. The book sounds interesting, so I decide to click on the link to buy it. http://www.self-propelled-city.com/cgi-bin/news.pl?task=article&artfile=e-tgtaic.txt

The link I click on does not go to Amazon or any other book vendor, but to a link transformation engine, which will use cookies in the user's browser to customize the linking experience for each person who clicks on that link
http://my.linkbaton.com/isbn/0879518847

This same scenario can be repeated with a link to a journal resource. Here's an example link to Nuclear Instruments and Methods in Physics Research Journal, volume 170, issue 1-2, page 199:
http://my.linkbaton.com/jvip/0168-583X/170/1-2/199/

Openly Informatics has been developing and implementing intelligent linking technology for almost two years now. At the heart of the LinkBaton system is a link server which accepts various forms of search information from the link, as well as user preferences from cookies, and then redirects the user to an appropriate resource. There are substantial areas of functional overlap between SFX and LinkBaton. While SFX concentrated on integrating library resources in a local link server, LinkBaton focuses on providing appropriate links to the internet at large. SFX focuses on the presentation of links, while LinkBaton focuses on the transport of links.

Link.Openly

Another Openly project, "Link.Openly" provides an XML framework that can be used to define the linking syntaxes. For example, here is a template for the OpenURL linking syntax:
<?xml version="1.0"?>
<!DOCTYPE slinks SYSTEM "slinks.dtd">
<slinks ID="slinks:openURL/article">  <URL>&baseURL;issn=&ISSN;&amp;volume=&volume;&amp;<option>issue=&issue;&amp;</option>spage=&startPage;&amp;genre=article</URL> </slinks>

We've put link.openly on a server system: http://link.openly.com/

Openly supports the OpenURL syntax and is serving on the NISO standardization committee. But we think libraries will need low-cost linking systems to deploy OpenURL. We are currently packaging Link.Openly with systems like Jake, which tracks the location of journals, in order to create a full service for libraries to manage their online resources. Openly has created OpenURL linkserver wrappers for Jake using both Java servlets (http://jake.openly.com/), and Javascript (ECMAScript) (http://www.openly.com/jake/instant.html)

Our ultimate goal is to allow a researcher to click once on a resource and immediately go to it without any intervening login pages, search forms or anything else. Thus we've named our library link server project 1Cate, which stands for "One Click Access to Everything."

So that's what Openly Informatics is up to now. I took a few days out of my European vacation in order to learn as much as I could about OAI and the people working to make this effort happen. I've been very impressed with the work going on but I've been more affected by the monolithic powers arrayed against OAI. Thankfully, the Web is changing the rules for scholarly publishing, just as they changed them for music publishing. It's just going a little slower on the scholarly front.

Even though Openly operates in the business sector, and OAI in the academic sector, we have similar goals to improve access to scholarly publishing for everyone. And I look forward to the opportunity for future collaboration.