Web-CT and E-Campus

WebCT and E-Campus

In response to the Instructional Enhancement Initiative (I.E.I.) in the summer of 1997, Dr. Wayne Miller, the head of Academic Services of what was then called the UCLA Humanities Computing Facility, researched ways to host individual class web sites for several hundred UCLA undergraduate humanities classes, under a mandate that required all undergraduate Humanities classes to have a Web site with content by the start of the Fall Quarter. Dr. Miller liked Course in a Box from MadDuck Technologies (later purchased by Blackboard), but Course in a Box broke under the latest NT service pack. Lotus LearningSpace, based on Lotus Notes, was prohibitively expensive. The decision was made to use WebCT.

What Dr. Miller, and my predecessor as Instructional Technology Coordinator, Evan Nisonson, did was truly astonishing—even the Chronicle said so. They managed to create, propagate, and support hundreds of web sites, and their faculty and student users, on strikingly short notice.

WebCT and Me

After several years working off campus as a multimedia developer and technical consultant, I returned to teaching in the spring of 1998, largely because I was intrigued by the new support for teaching with technology. I was a T.A., so I had to use the faculty member of record’s log in code in order to add content to the site for the classes I was teaching. The login code was a long string of arbitrarily assigned numbers and letters, difficult by design for security reasons. The faculty member was largely uninterested in the use of the technology, finding it cumbersome, but enthusiastically asked me to help create instructional materials for the faculty member’s own classes as well after seeing the possibilities for his classes based on my site.

I was completely under whelmed by WebCT. I thought it was slow, clumsy, ugly, and navigationally silly. It was also downright hostile in terms of users with disabilities, to the point where you couldn’t use the UI with a screen reader. I used WebCT because that was what was provided, but I subverted much of the built Javascript, and relied on my own hand-built html. I wasn’t, in fact, much impressed with the way the I.E.I. was implemented either, as I wrote in a letter to my departmental chair, later printed in his review of the initiative for ADE.

When I was hired by HCF as the Instructional Technology Coordinator in January of 2000 we were running WebCT 1.3, and using ColdFusion and custom Perl scripts as middle ware. HCF had been looking for a Unix Administrator/Programmer for some time, and was relying on pre-built templates and class generation scripts that the last admin had thoughtfully written far in advance, through Summer 2000. In Spring of 2000 the I.E.I. third year review took place. Part of that review included a “software” review. Dr. Miller offered to do the contact-the-vendors and get a quote part, if I did a general study of what our users wanted and needed, and the products that met those needs.

I.E.I 2000 Review

Working with Dr. Miller, I looked at other course hosting solutions, including SSCNET’s homegrown ClassWeb. We consulted faculty, in person, via surveys, and in email, as well as undergraduate students and T.A.s, to determine what they felt was important. I also contacted my peers on campus, and lots of the technologists I’d met in my previous job.

Criteria? You mean features, right?

We established a list of criteria that we felt were crucial for any hosting system. I looked at what other schools were using, contacted people I knew personally, emailed others, and created sample class sites on a wide variety of products. You can see my bibliography.

You can read the notes I made for my final report, but given that E-Campus depended almost entirely Dr. Miller and on me, with end user support from the trained graduate student assistants, ITCs, I felt the choice really came down to WebCT versus BlackBoard, and favored WebCT. My recommendation was based primarily on user features, support from WebCT and local familiarty, and concerns about data access, though cost was also important. I didn’t like WebCT, but I didn’t find any of the alternatives any better for our users. In fact, no one I contacted really felt any true product loyalty&mdash. People generally felt their software was “adequate.” We decided to upgrade to WebCT 3.x, realizing we’d need to have a full year to implement all the changes, and to allow for unexpected problems, and the fact that much of the work would have to wait until summer.

Enter: The Unix Guy

But then, finally, in June of 2000, we were able to find, and hire our Unix Guy, more properly described as a UNIX Engineer. Bruce Dumes’ first task was to set up WebCT 3.0 on a new server, relegating the previous courses to an archive server where they could be restored and accessed if faculty needed older class sites. This was not a simple undertaking; aside from installing WebCT, Bruce also had to create new Perl scripts to massage data, and communicate between the data collection and WebCT, as well as generate accounts and create classes, though initially we still depended on Cold Fusion for much of the process. WebCT promised an API, and source, but neither ever really materialized. Had I known in March of 2000 that we would have a top notch systems engineer/programmer on board, I’d likely have not been so enthused about upgrading WebCT. Over and over again, Bruce Dumes proved to be able to code maintainable, reliable solutions with user-friendly GUIs.

New E-Campus

New E-Campus, as we called it informally, went live in Fall 2000, after extensive testing in the summer. We offered all users, for the first time, a single login page, and global permanent IDs. Bruce continued to refine and improve the admin side, creating custom utility scripts and admin pages at my request, and, by Fall 2001, had removed all but one very small ColdFusion routine in terms of E-Campus proper.

As Bruce and I thought about and refined My E-Campus, we began to wonder about the viability of slowly, gradually replacing WebCT completely, using a Perl back end for processing, and incorporating open source, local, and commercial tool modules.

To Upgrade or Not to Upgrade

In March of 2000 WebCT sent a representative to pitch the forthcoming WebCT 3.5 to us. As far as I could tell, WebCT proposed to raise our yearly licensing fee from $4,000.00 to $21,000.00, for the year, with thousand dollar price increases built in to renewal fees for the next two years, a promise of portal support, and, if we paid now, more features in  the all-Java 4.0 version of WebCT. Since they hadn’t fixed bugs in 3.1 that were there in 3.0 and 1.3, I wasn’t impressed. Nor did I see any point in “upgrading,” since we weren’t getting anything new, no new features, no bug fixes and no improvements in terms of UI and user experience. We invited WebCT to persuade us, and even participated in a conference call, but Bruce, Wayne, and I were not persuaded upgrading made sense. The decision was made to stay where we were.