BIBLIOTHEKE

So the ILL-L interlibrary loan listserv was recently ablaze with discussion about Infotrieve’s mess with Ariel. Ariel is an electronic document delivery (EDD) P2P system for libraries. With it, libraries can scan, send and receive documents and post them online for patrons. It was originally developed by RLG, but was bought by Infotrieve about two years ago.

At the same time, there are two other important EDD applications. The first is OCLC Illiad, the popular ILL management system. Illiad has a lot of great features, including EDD functionality integrated relatively seamlessly into the overall system. It is also the source of the open Odyssey protocol (PDF). The other is a little program called Prospero originally designed for converting Ariel-transmitted TIFFs to PDFs and posting them to a web server for patrons back when Ariel didn’t have patron delivery. Not only did it do this job very well, but it was released under the GPL.

At Northwestern University, we have used the Ariel and Prospero combination for a long time. During the past four years that I’ve used EDD software, Ariel gained a Prospero-like web posting function and Prospero gained, in version 2, the ability to scan and send documents. However, the Ariel post to web feature has some serious workflow problems, bugs and a just plain silly/obnoxious design, while the Prospero scanning feature was tremendously buggy. So we stuck with the traditional combination of the two.

With the slow pace of Ariel development we hoped to eventually shift to using only Prospero. When it came time to put together a system for EDD from a closed storage site directly to patrons, the Library focused on Prospero, taking advantage of its open source to link version 2 into our LDAP directory and create a unified authentication system, a vital feature for us. EDD tech was progressing smoothly, albeit slowly.

Then came a licensing bombshell. Soon after Ariel was purchased by Infotrieve, version 2 of Prospero was pulled due to an unspecified licensing issue. Not too long after this, Prospero 3 was announced as only using the open Odyssey protocol. The sequence of events led many to speculate that Infotrieve might somehow be involved, possibly trying to push a free and open source software (FOSS) competitor out of the game. Users don’t know one way or the other, however, since those involved with Prospero are keeping quiet about it.

Today, everyone is still waiting for Prospero 3. The listserv is basically dead. The site seemed to go down for some time, then came back. No one really knows what’s happening with it and there is nothing on its SourceForge project page.

And that brings us to Ariel 4. Ariel 4 isn’t compatible with any version before it aside from version 3 systems that have been updated to Ariel 3.4 and a compatibility patch. This patched Ariel 3.4 is also not compatible with any version below it or unpatched 3.4 installations. Needless to say, many users are upset. Infotrieve claims they let all the libraries know, but they didn’t actually find out that the ILL-L listserv even existed until a month ago. Ariel 4 adds no new features that are, in my opinion, significant or terribly useful for many libraries (according to Infotrieve, patron delivery has not received its much needed overhaul) and costs a ton of money to buy. Finally, Infotrieve has released version 4 without testing it with Illiad or Prospero, also according to Infotrieve, and can’t guarantee that it will work with them.

Why all of this is important.

Essentially, this situation with Ariel demonstrates pretty clearly a number of mistakes that have been made on the part of the libraries. Most importantly, there was never an effort on the part of the ILL community to support any FOSS project for the sake of FOSS. This had a number of effects. First, no one had any problem with RLG creating an application with closed source and a proprietary method of sending and receiving documents. As a result, libraries have become dependent on proprietary software over which they have little control. Sure, there is an advisory board and Infotrieve has some communication with its users and a support listserv, but it’s obviously not doing any good since it’s progressing at a snail’s pace. The set of features in Ariel 4 is, to put it bluntly, pathetic for 2005. According to members of the Ariel team during a recent phone conversation, little has changed on the patron delivery side in version 4 even though version 3 hasn’t caught up with Prospero 1’s ease of use and is just bloated with useless features.

[edit 1.10.4: Previously, I railed on Infotrieve for lack of documentation. Infotrieve does indeed have documentation that is helpful for basic set-up, so I feel my criticism was overly harsh and inaccurate. However, most of their documentation is focused on patron delivery and, in my experience, unhelpful and dense. For example: huh?. Comparing Ariel documentation to the Odyssey documentation (PDF) and makes my actual point a little clearer.] On top of this, Infotrieve doesn’t seem to know much at all about what its customers really need. As noted before, they only very recently learned about the ILL-L listserv. Now they’ve proposed an upgrade based on geography, apparently oblivious to the fact that many libraries pay little or no attention to geography when conducting ILL. The incompatibility of Ariel 4 with previous versions will undoubtably be a tremendous inconvenience for library staff and patrons everywhere.

But the most important problem is that libraries have relinquished their independence to a commercial company. If Infotrieve at some point figures out a way to commercialize ILL transactions, we’ve lost a lot of leverage. Strapping on some OCR to gather data is something publishers would love. That’s perfectly fine, but libraries no longer have the ability to determine exactly how that development would proceed.

Last, if staff or librarians at any library wanted to take an established EDD system and help it evolve into an advanced networked web application for sharing information, that dream is basically shot for the time being. We don’t have access to the code of the dominant application and even if someone does create something from the ground up, the dominant application clearly isn’t interested in participating.

Now, apparently, many staff and librarians in ILL are upset with Infotrieve for some of these problems. I, for one, am looking at what will likely be a sizable chunk of time testing configurations and meeting with coworkers about whether to use Prospero, if we can, or move patron delivery to the more awkward and less customizable Ariel. It greatly affects our workflow and prevents us from efficiently serving our patrons.

Conclusion

Clearly the best step is to support Odyssey and the development of applications based on it. If Prospero is going to be continued as an open source project, then more should be put on SourceForge to make development more transparent and encourage participation. If there is a reason that it can’t be, that reason should be communicated with users. There also needs to be a strong recognition that, to patrons, EDD is just another web application. To staff, it should be designed as a P2P web application with staff administrative and scanning tools. In other words, the focus should be on the web/patron end, and the staff end needs to be designed around actual ILL workflow. OCLC seems to have figured this out, both with Illiad and WorldCat.

I honestly don’t think libraries or even interlibrary loan departments take ILL as seriously as they should. ILL EDD is a hidden backbone of online resources. Google is scanning books? We’ve been doing it on demand for years. We go beyond providing our patrons with information that’s unavailable online; we give them the information they can’t even get from their own libraries.

If you think I may have said something in error, by all means, please correct it in comments.

This entry was posted by William J Melody @ 9:43 am and is filed under L2L. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site. Note that there may be a delay when posting your comment because continued comment spam has forced me to turn on partial moderation.