We love databases!

The key to virtual libraries is database technology. When it comes to virtual libraries of Internet resources, there are many options. Depending on the size of your collection, consider the use of your existing OPAC software, ROADS, or a microcomputer-based databased application like FileMaker Pro to manage the resources. This column describes approaches to using each of these database technologies to create virtual libraries of Internet resources.

The bread and butter of our profession has traditionally been lists; lists of books, lists of journals, lists of articles, etc. Each of these lists contain many items, and each item contains similar characteristics; authors, titles, subjects, dates, publishers, paginations, and possibly URLs. In terms of computer technologies, these lists are very easily manifested as electronic databases.

Why we, as a profession, have not become database construction experts is beyond me. Yes, we know how to search databases, but in general, we do not know how to create them. To make matters worse, we, in general, do not have ready access to desktop database applications. Hopefully this will change with the advent of easier-to-use relational database programs like Microsoft Access or Claris FileMaker Pro.

Until we become database construction experts, we will be restricted to using the monolithic databases applications hosting our OPACS. While there are quite a number of good reasons for using your OPAC software to create and maintain collections of Internet resources, there are other applications that pick up where our OPACs leave off.

OPAC software

I am proud to say that one of the very first collections of Internet resources hosted by OPAC software was located at the North Carolina State University Libraries in 1995. Namely, Alcuin - A Database of Internet Resources. [1] Granted, the database has not been maintained in forever, but its purpose was to demonstrate the feasibility of using subfield u of the 856 field of MARC records. The Alcuin database was one of the very first, if not the first databases to take advantage of this technology.

Since then, quite a number of World Wide Web OPAC interfaces have made themselves available. A list of many of these interfaces is maintained by myself. [2] ("Yes, the list is maintained through a very simple, if not rudimentary, Hypercard-based database application.") The list is not intended to be comprehensive. Rather, the list is intended to demonstrate how different OPACs can be implemented using World Wide Web interfaces. A much more comprehensive list of Internet-accessible OPACs is Peter Scott and Doug Macdonald's webCATS: Library Catalogues on the World Wide Web. [3]

Cataloging Internet resources has a number of advantages. First, the record structure, the MARC record, will be familiar to you and your fellow staff. Data entry should not be too difficult. Second, assuming your OPAC comes with a World Wide Web interface, mixing Internet resources in with your physical collection records provides a sort of "one stop shopping" experience for you clientele. Your patrons will only have to look in one database to locate items of interest. Third, there are a growing number of MARC records of Internet resources that you can acquire from OCLC and their InterCAT project. [4]

The downside of using your OPAC to serve a virtual library of Internet resources is the dynamic nature of the Web. URLs change and our OPAC software has no way of checking for broken links. Consequently, keeping up with illusive URLs can be a problem. Second, while placing all your resources (Internet and not) in one database makes a lot of logistical sense, it has been my experience that our clientele would like one database for electronic-only resources and another database for traditional materials. Thus, even when Internet resources are placed in our OPACs we still end up creating HTML pages containing those same links. Silly.


ROADS (Resource Organisation And Discovery in Subject-based services) is a suite of software that attempts to overcome many of the limitations of OPAC database software. As its home page states:

The project is producing software which can be used to set up subject services or gateways, i.e. indexes of selected Internet resources which can be both searched and browsed. The software is highly configurable, from a maintenance and end-user interface point of view. It is developed with a "toolkit perspective", so that individuals and organisations can choose what parts of the software they want to use, and which parts they may wish to use with other/existing gateway software. [5]

ROADS is a Unix-only solution. It's maintenance interface is almost 100% Web-based. Almost no Unix knowledge is required to set it up. Configuration is done through the Web interface and simple ASCII files that can be FTP'ed to their proper locations. It's free (or as free as a free kitten). It is well documented. It works. It works well, and it's getting better.

The system is based on rudimentary ASCII files called templates and a little known Internet protocol called WHOIS++. These text files are much like MARC records, only easier to create. Once the software is installed, adding and deleting records from the ROADS database is a matter of using HTML forms.

After learning how the system was intended to be set up, I converted the HTML interface of my Alex Catalogue of Electronic Texts into ROADS templates (with the help of FileMaker) and consequently reincarnated Alex as a ROADS service. [6] The whole process initially took two to three weeks.

Unlike OPAC interfaces, no knowledge of the underlining database record structure is necessary for maintenance. Unlike OPAC interfaces, the ROADS software does link checking verifying the integrity of your URLs. Unlike OPAC interfaces, ROADS services can search other ROADS services across the globe much the way Z39.50 is intended to do. Ideally, if many people were to create ROADS services of subject-specific disciplines, then there could literally be a World Wide Web of classified Internet resources field searchable through one interface. Like OPAC databases, it would be possible to share ROADS database records.

The ROADS system holds and immense amount of promise and potential. I advocate its use whole heartedly!

Claris Software's FileMaker Pro

Another approach to the creation of virtual libraries of Internet resources is the utilization of microcomputer databases. Database de jour is FileMaker Pro version 4.0 from Claris. [7] This cross-platform database is well integrated with both Windows and Macintosh operating systems. It is almost 100% graphical thereby removing the need to know any programing. What programing there exists is done with a simple script editor making it virtually impossible to produce syntactically incorrect scripts.

FileMaker is a relational database program. This means it can operate in more than the two dimensions of rows and columns of spreadsheets. Instead, its database can have "depth." For example, you can create a listing of Internet resources complete with titles and URLs, as well as any number of authors and subject specifications. Unlike "flat file" database structures where you must specify the maximum number of occurances for any one record, relational databases can specify unlimited numbers of field occurances. This is ideal for much of a librarian's work.

Perhaps one of the most advanced features of FileMaker is its ability to be a HTTP (hypertext transfer protocol) server. This means you can both serve and manage your FileMaker databases thorough a Web browser. This is all done without the need of any CGI (common gateway interface) scripts. At the very least, all you have to do is create your database and "turn on" Web serving. Really! I couldn't believe it myself when I first tried it, but it works. If you want to customize the Web interface, then you must use Claris's CDML (custom document markup language), but CDML is a lot like HTML and the creation of CDML pages is a matter of cutting, pasting, and tweaking ready-made templates.

We love databases!

Databases are the technology to understand and master if you are going to create virtual libraries of any sort. This is true since much of what we do as librarians is manage organized lists of information. Organized lists of information put into electronic form are computer databases. Therefore, computer databases are at least a part of virtual libraries. Learn to use the databases you have at hand. They may be your OPAC. They may be something you find on the Internet (ROADS). They may be the something you purchase and play with locally (FileMaker). Whatever you choose, databases play a central role in the creation of library services of the present as well as the future.


  1. http://library.ncsu.edu/drabin/alcuin/
  2. http://www.lib.ncsu.edu/staff/morgan/alcuin/wwwed-catalogs.html
  3. http://library.usask.ca/hywebcat/
  4. http://www.oclc.org/oclc/man/catproj/catcall.htm
  5. http://www.ukoln.ac.uk/roads/
  6. http://sunsite.berkeley.edu/~emorgan/alex/
  7. http://www.filemaker.com/

Creator: Eric Lease Morgan <eric_morgan@infomotions.com>
Source: This is a pre-edited edited copy for Eric Lease Morgan, "We Love Databases" Computers In Libraries. 18(2):38-39, Februrary 1998.
Date created: 1997-12-12
Date updated: 2004-11-14
Subject(s): Filemaker; virtual libraries; ROADS (Resource Organisation And Discovery in Subject-based services); OPAC (Online Public Access Catalogs);
URL: http://infomotions.com/musings/virtual-libraries/