Open Source Programs for the Reference Librarian: When Your Budget is More Limited Than Your Vision
LITA Open Source Systems Interest Group
Sunday June 25 8:30-10
Speakers (in order):
Ranti Junus, Michigan State
Teria Curry, Johns Hopkins
Kirsten Allen, American University
Mary Evangeline, Univ. of Arizona
George Harmon, Florida State
(note: editorial parentheticals are by the scribe. Otherwise this is a loose paraphrase)
(About 100 chairs. Two-thirds full at start of session).
(A festive mood on the panel; much good-natured laughter. Remarkable for 8:30am).
Intro by Gwendolyn Reece, chair of OSSIG
Definition of open source software: freely available computer programs. Usually monetarily free also, but not necessarily. The freedom is the freedom to make copies and to make changes to improve the program or make it do exactly what you want. Open source programs are collaboratively developed and tested in a process very much like peer review for academics. However, it can require an investment in staff time and learning to get it working.
Invitation to OSSIG meeting, Monday at 1:30 at the Loews in the Beauregard room.
iVia, an open source software package for building virtual libraries. Developed by the University of California’s INFOMINE project and available for download at http://ivia.ucr.edu
The result as it appears to be in use at INFOMINE and at Michigan State University, is an almost infinitely configurable database of resources of all types. It could be an e-journal portal, but it could also integrate librarian-generated electronic materials like course/subject guides with content automatically generated by a web crawler which is part of the package.
System requirements: Linux/Unix (best with Debian, but also SUSE and Redhat). Apache and MySQL.
Installation scripts make the install process relatively painless.
Records can be manually created, or automatically generated by crawler. The crawler can also add metadata as it goes. Possible to have an expert point the crawler at a web resource and crawl that specifically.
Search and browse features very customizable. Allows for a great deal of precision. Appears to allow a wide variety of content (maps, grey lit, datasets, journals, other)
Can handle OAI and MARC content.
Extensive documentation at http://ivia.ucr.edu/manuals/. This is required reading!
LibX Firefox extension
Created by Virginia Tech libraries. Available at http://www.libx.org.
(An extension is an add-on to the Firefox web browser, which is also free and open source and available at http://www.mozilla.com)
The hosting necessary to make the extension work can be done locally, or can be hosted on the LibX server–which has the advantage of automatically handling updates and such.
The extension itself is a program which students will have to download and install on their computers. (In response to a question it appears that many libraries offer Firefox on their public PCs already; the extension could be pre-installed there)
What it does: creates a toolbar which can include direct searching of the library catalog, and a means to connect to the library’s OpenURL resolver. With a look and feel customized for your library.
Google Scholar (and presumably Microsoft Academic Live) can be included pretty readily.
What this does is to push the library’s services out to where students are already, in their web browser. A downside is that that web browser has to be Firefox. (though it would probably also work in SeaMonkey and Camino, which are also browsers using Mozilla’s code)
It adds a right-click menu which allows a one-click catalog search for whatever is highlighted on whatever web page the student is using (the example was a book title on Amazon).
An “embedded cue” feature puts your library’s icon next to, for example, the title of a book the student is looking at which your library owns. That is, the program is constantly checking to see if it can match what students are looking at with library content.
Advantages: great for digital natives–makes library and “The Internet” the same place. The toolbar is infinitely customizable to suit your particular patrons.
Disadvantages: may be confusing for non digital natives (i.e., old people). Only usable with Firefox at this time, which limits patron choice.
Jabber IM and Gaim
Jabber (htttp://www.jabber.org) is an actual chat protocol which uses streaming XML and other Internet communications standards. Can be run on Windows, Mac, or Linux/Unix.
Used by Google Talk, but can communicate with other commercial services like Yahoo and AIM.
You can set up your own server–talk to your IT folks first, as this is not necessarily easy and has some security ramifications! A simpler option is to sign up with a remote provider, of which there are many who provide free accounts.
There are many clients (that is, software you can install on your desktop), of which Exodus is probably the best for Windows.
Possible to have multiple chats going on multiple networks at once.
Gaim is similar to the commercial product Trillian. Both provide one client program which can communicate with all the major IM formats. You can then access all of them at once from one clean, uncluttered, non-commercial interface.
Easy to install. Available from Sourceforge. (there is also a portable version which can be installed on a USB key…or on a desktop where you don’t have permission to install software. http://portableapps.com/apps/internet/chat/portable_gaim)
There are plugins which provide almost all of the features of the major IM clients.
There are some bugs, but patches for open source programs often happen almost as quickly as they are discovered.
The wiki software used by Wikipedia. Whatever one’s feelings about Wikipedia, the tools are powerful and useful.
Used at UA to provide training for the complex environment at UA’s Information Commons–24/7, 250 computers, 50+ software programs, 400 databases, 20 students, 15 staff–and constant training needs. The current solution is a static webpage which only one person can work update. This limits how interactive and current it can be.
Wiki software allows updates to be done anywhere or anywhen there’s a web browser. It can be secured to prevent loss of important data by careless editing. As social software it has the potential of allowing students to post their photos and thoughts to share with each other, and to associate the library with that. It could also be used to deliver constantly updated information on events and student life.
Cons: Requires some forethought to set up the right categories, as new pages are harder to create. There is also a real possiblity of information overload if you’re trying to do a wiki and a blog and podcasting (and the rest of your job)…so decide what this can REPLACE.
JabRef is an open source citation manager, similar in concept but not in polish to Refworks–not a replacement yet. But in a year it might be.
Requires Java, which has pros and cons. Can be slow, can be tricky to use. (but ARTstor needs it, too).
Documentation is at Sourceforge, and designed to be used on the web rather than printed.
Uses BibTeX–pronounced “Bib-tech”–a venerable format familiar to folks in the sciences, but also used as an export format by Google Scholar and Microsoft Academic.
No restrictions on how many desktops you can install it on (the beauty of open source is never having to say you’re sorry for making copies)
Recognizes several database export formats: ISI, CSA, INSPEC, JSTOR, MEDLINE, and others.
Recognizes Endnote and RIS formats, so you can move data around.
Doesn’t recognize MARC. Limited Z39.50 support. So far.
Doesn’t print references, but will export them to your clipboard so you can copy and paste into your favorite word processor as RTF or CSV. Works with OpenOffice.org, a free/open source suite designed to replicate Microsoft Office (available at htttp://www.openoffice.org in Windows, Linux, and Mac format)–it will natively export to an OOo database.
Q: What kinds of things are you adding manually to iVia? A: Little. Primarily use crawler, with human direction. Software can be used to integrate, say, subject guides with subscription electronic content.
Q: Does iVia handle Dublin Core metadata? A: Probably.
The presentations will be posted either by LITA’s Online Software Systems Interest Group, or hosted by American University. There may be a brief delay….
Sources on the handout
Programs mentioned in sessions