Ben Bartlett
WebDB
v. 2.0
    WebDB has a client-server architecture. WebDB itself is the server,
and the clients are the users who log into WebDB to gather
information. WebDB also acts as a client, however, in that it accesses a
database in order to retrieve the information which the user asks for. So,
to be exact, WebDB is a web-based interface between a database and a user.
    The major components of WebDB include the five Perl modules, which
hold many of the procedures which WebDB needs in order to run, as well as
index.cgi, login.cgi, course_search.cgi, person_search.cgi,
registrar_tool.cgi, and timesheet.cgi. Index.cgi gives access to the
various functions of WebDB. Login.cgi is what allows a user to log in as
his or her user name, and thus gain access to certain options unavailable
to the general public. Course_search.cgi allows a user to search through
courses by department, professor, time, term, and year, among other
criteria. Person_search.cgi allows a user to find a list of people, their
mailboxes, and their phone extensions by searching for their first name,
last name, major, or minor. Registrar_tool.cgi allows the registrar to set
up class schedules and the like. Finally, timesheet.cgi allows people who
work to sign in and off on an electronic timesheet.
    Many of the links in WebDB are actually invocations of cgi scripts
with certain arguments passed to them. For instance, when you click on the
"My File" link, you are actually invoking a script called person.cgi,
telling it to take the action myfile, telling it the pidm of the user (the
pidm is a number which is assigned to each user of WebDB; it is attached
to a particular session number when the person logs in through login.cgi),
and a bunch of preferences which tell the program how the user wishes to
view the My File window.

    Another example of a cgi script is the course search, which runs under
course_search.cgi. Once again, the course search has several possible
actions which can be passed to it, such as registration tool or info. Much
more interesting, however, is the course search itself. The script reads
the input the user has posted to the display page, and then connects to an
Oracle database in order to retrieve information which fits the criteria
the user entered. It then displays this information to the user.
    What I find most interesting is the idea of having three "areas", the
navigation area, the display area, and the work area. The navigation area
is created using nav.cgi; the other two areas vary. The display area shows
everything from the login screen to course searches to the person
search. The work area only appears from time to time, such as when the
user retrieves a course schedule. I think that the way in which the screen
is divided up is very useful.
    I don't use WebDB enough to really want any major changes in it in
terms of its appearance. I do think the code could use quite a bit of
cleaning up. It is rather difficult to read, at best. Some organization,
better documentation and commenting, and some rewriting of code would
really improve things.