LITA: Information Technology and Libraries (ITAL)
ITAL Editorial: The Open Source Movement and Libraries
I've been thinking about open source software 1 lately for a couple of reasons. One reason is that the March 2002 issue of ITAL will be a theme issue devoted to open source applications in libraries, guest edited by Jeremy Frumkin of the University of Arizona.
The other reason is a bit closer to home. I hired Eric Lease Morgan to be head of a new department at the Notre Dame Libraries called the Digital Access and Information Architecture Department (DAIAD, pronounced "die-ad," frontward or backward). Eric is something of a force in the open source movement and was the primary developer of the MyLibrary@NCState ( http:// my.lib.ncsu.edu) open source software created at North Carolina State University. (He is contributing an article to Jeremy's issue, incidentally.)
One of Eric's first responsibilities here at Notre Dame is to convert our library Web site to a database-driven technology and he wants to do it on an open source platform. My job is to decide whether to go along with that or to dictate a more conventional approach using something like Oracle and ColdFusion or Active Server Pages (ASP). Hence my immediate reason for devoting my time to the subject.
So what's the problem? Why not just do it? One reason is that despite a few well-publicized examples, there is not yet a whole lot of open source software available, and the movement is still fairly young and unproven. Another is that if we go this direction, we will have to commit some substantial human resources to do the required programming.
Does that mean we shouldn't do it? Why put my eggs in an unproven basket? Why commit valuable human resources to create more of this unproven software?
Maybe it isn't all that unproven. There are, as I mentioned, some fairly well known and widespread open source applications, such as the Apache Web server and the Linux operating system. My sense is that Linux is finding its way into more and more libraries these days, and that Apache is the leading Web server in colleges and universities. So to call those two unproven is really subject to debate. Likewise, open source products such as MySQL and Perl are being used in many applications.
While it's true that open source applications require programmers, the fact is that ColdFusion and ASP also require programming, and I would surely not want to run a critical library application based on Oracle without someone with extensive Oracle knowledge and experience on hand. Thus the human resources argument is also less compelling, especially when you realize that not only do you have to invest dollars in people to make those commercial products works properly, you also have a substantial cost in licenses for them that you don't have with open source software.
Another thing about open source software, in theory at least, is that if somebody creates something that is truly good and makes it freely available, a community of users will emerge and collectively support, maintain, and improve upon it. That's an intriguing, almost '60s concept that is also unproven. Certainly for Apache and Linux it seems to be working. On the other hand, a brilliant piece of work like jake (the Jointly Administered Knowledge Environment) is, ultimately, a failure, because even though the software is great, the user community did not keep current the data upon which jake depends.
One last point to make regarding open source applications is that there does seem to be a potential for something truly beneficial to libraries. MyLibrary@NCState is a good example. Many libraries do not have the resources to create software from scratch, but because there is no ongoing license fee for open source software, perhaps they can provide the resources needed to implement it at their library. There are other library-related open source projects underway, such as OpenBook, a library automation system developed by Technology Resource Foundation ( www.trfoundation.org) for smaller public and school libraries.
On a different note, I call your attention to the lead article in this issue by Peter Murray, the first-ever winner of the LITA/Endeavor Student Writing Award. I was on the judging panel and can attest that we had many good papers submitted, so I'm pleased to know there are some burgeoning young writers out there. This is a new annual award-see www.lita.org/a&s/awards.htm#writing for more details.
1. I assume most readers of this journal know what the phrase "open source" means, but just in case (and greatly simplified) it refers to software that is freely given away, along with source code, according to certain explicit distribution terms that allow it to be just as freely redistributed (although not necessarily for free). The concept actually goes back to the GNU Project ( www.gnu.org/gnu/thegnuproject.html) that began in the mid-1980s. Back then the preferred term was "free software," meaning that users of the software were free to modify it and redistribute it. Most open source software is distributed under the terms of the GNU General Public License (GPL), ensuring that it retains this "free" status.
Dan Marmion ( firstname.lastname@example.org) is Associate Director for Information Systems and Access at the University of Notre Dame Libraries, Notre Dame, Indiana.