No more handmade resource collections passing for professional Web publishing, please! I hereby declare the merciful death of the one-person "<fill in your favorite subject here> HomePage." The complexity of maintaining resource collections has eclipsed an individual's ability to keep one current. As someone who maintains a Web site for a large organizations (I work for the Maricopa County Community College District), I know firsthand that the hands-on approach to resource collections on the Web is just not efficient, especially as a handful of pages yearns to grow into a complex Web site.

When I started creating Web sites in the old days (October 1993), like many Web masters of the time, I organized collections of links, static HTML documents maintained by hand — in my case, my hand. That original site for the Maricopa Center for Learning and Instruction (MCLI) had links to every Web starting point I could find in subjects as diverse as art, computers, government, medicine and health, and what I called "Category-Defying Information." At that time it was easy to discover what was new by visiting any major Web site, the NCSA What's New Web page [formerly http://archive.ncsa.uiuc.edu/SDG/Software/Mosaic/Docs/whats-new.html] that every Mosaic browser pointed to. (That What's New site was the standard way to announce a new site, as we did in January 1994.)

Even at the beginning I had a problem categorizing links. If I had a great link for computer art, did it go on the computer page or the art page? If I linked from both places, then I had to edit two files when I updated. So in the fall of 1995 we stopped designing static HTML documents, and created a searchable front end to the content of our site [formerly http://www.mcli.dist.maricopa.edu/451f/], allowing users to search by keyword across all documents, restrict a keyword search to a particular subject heading, or view all items under a particular subject heading. In essence, a program running on our Web server created the new pages on the fly, drawing links from the database, and arranging them on the page as the user called for them. That way I needed only to put links into the database, and characterize them by keyword or subject; the computer program did the rest.

As part of that new system we built a functionality that allowed anyone who visited our Web site to add a new link to our collection. We adapted that approach to a searchable collection of examples of how teachers at all levels are using the Web for part or all of their instruction.

As an aside, it was interesting, odd, and somehow disturbing to see the keywords that some people entered as search terms on our education-related site. When you use a Web page for even something as simple as a search, behind the scenes you may be collecting information that can be stored somewhere, and perhaps used one day. That may sound like a security panic alarm, and perhaps it is. On the other hand, that information also provides an interesting sociological view into people's Web habits.

"The Web underscores the notion that no single person can grasp all of the content."

We learned fast how an unmoderated site gains inappropriate links. Some of the things just did not belong in the collection (like sites that were mainly advertisements, e.g., "Tom's Video Powershop") and the others that were for offensive material ("Tom's Raunchy Video Powershop"). And there were the ones sent with frustrating frequency by people who did not know how to use a Web form or write a correct URL. I had to delete those unwanted items from our database manually, a tedious task. We needed a gatekeeper, and some programs to support that gatekeeper, and then someone to check the program . .

Over time, we've built an entire system of checks and balances. Now, when someone adds a new item, the site automatically acknowledges the submission, and the item is stored in a "review" area. The site administrator (me, or one of my staff) reviews the link for suitability and accuracy (sometimes links are entered wrong) before adding it to the database. Our work doesn't stop there, however; periodically we run automated programs to check the entire list of links to make sure they are still functional, and remove those that no longer work.

We have a number of systems (programmed by a part-time student worker) that provide a suite of administrators' tools for maintaining resources sites. We can maintain them without touching any HTML code. Every day I can click a link from the administrative page and see how many new links are waiting to be added. I can follow a hypertext link to jump to the submitted site. Also, I can edit any part of the submitted information. And, with one more click I can choose to add it to the active database or to delete it. Either way, the system is set up so that it sends standard e-mail messages to the people who submitted the information so that they know what happened.

Our searchable, subject-organized resource collections are designed to fit the needs of both visitors who would prefer to browse by topic (e.g., the Yahoo approach to categorizing information) and visitors who want to do a keyword search of the entire collection (e.g., the Alta Vista approach to keyword searching for information).

From a site-management view, we are not the maintainers of the content, our visitors to the site are. For example, a searchable index of community college Web sites, created as a resource for anyone interested in locating particular community colleges, grew from the 150 or so I collected from 1994 to 1995 as static pages, to more than 740 today. The links are obtained from folks who visit our Web site, and know the URL of a community college that we do not yet have listed. They can expand our collection by sending us the information via a linked Web form.

From a management standpoint, this works well because it frees me of the responsibility of finding new sites. That is the exact reason I created those systems; my job is not to be a "Web master" but to be actively involved in many aspects of technology in education. Are the so-called shifts into the "Information Age" real? Where do we shift from the memorization of facts to the cultivation of knowledge skills? The Web underscores the notion that no single person can grasp all of the content (there really are no experts) and that the Internet design underlying the Web is truly a decentralized information system. And truly a "chaordic"[formerly http://www.funderstanding.com/mailing1.htm] one. ("Chaordic" was coined by Dee Hock, the organizational guru in the credit industry, to describe systems that are oxymoronic hybrids of chaos and order).

"There is still a place for handmade Web pages."

We've created another system just for folks to send us interesting URLs [formerly http://www.mcli.dist.maricopa.edu/eye/bag/], so every few weeks we can compile a collection to share from a Web page. Like the other systems I described earlier, that Web site has a form to add new content and a page to search the entire collection. The online administrative tool allows us to review the submitted information, and a click automatically generates a new Web page listing the latest list of sites and sends out a notification message to anyone subscribed to our e-mail list.

Do not get me wrong; there is still a place for handmade Web pages. Creating them is part of the wonderful learning process of how to use the Web. The ability to be amateur publishers is definitely part of infectious popularity of the Web, as one can see in the energy of online communities such as GeoCities. One of my favorite pastimes is tinkering with my own handmade home page, and it pleases me to no end when someone sends me e-mail about one of my obscure offerings.

The situation is different, however, when you aim to offer a comprehensive and/or dynamic Web site. At the Maricopa Center for Learning and Instruction, we haven't given up the function of list of links, but we have given up creating them the way they did in the "olden days" of Web site design.

No more Home Pages for me: I'm moving on to real Web site functionality.



Alan Levine is an instructional technologist (not a "Web master") for the Maricopa Center for Learning and Instruction (MCLI) at the Maricopa Community Colleges. The MCLI site has won its share of Web award badges for sites such as the widely used Writing HTML and How to be a WebHound tutorials. More information is available from his Web presentations including the inspiration for this article's title. Alan maintains (by hand) his own retro-1994-Mosaic home page.