Couldn't wait for tomorrow to make this entry.

Got a major work-related project done today. Took about a week, with most of the time going into a tiny fraction of the overall project. Why is it always that way? The work goes along as well as could be expected, then there's a giant blockage of some kind.

Once in a while it's something unanticipated, but more often it's like this one: I coded my stuff to be just like an example in a manual. And presto hey, it didn't work.

Looking at it for several days in a row, posting to help boards, trying variations, etc., etc., ad nauseum. Even turned to Kenny for help ;-) who in his methodology said "what's it look like in view source mode?" which led me spot on to the problem.

The server was bouncing my submission back and redirecting to a secure port, but that redirection didn't include the data that had been sent originally. Modifying my code to send directly to the secure port resulted in an instant gratification moment as the whole project hinged around that working. Que buena!

So what's this great project do? It is my raison de etre at CSUN right now....

Right now whenever an instructor wants a new course made in our on-line learning system called WebCT, either Kate or I have to go in to administrative mode and manually create the class, then manually connect to the central student information system to download a list of enrolled students. We then have to format that list in a special way, and run it through another script that puts the students into WebCT.

We have to do that for each course.

Every semester.

Can you imaging how boring and tedious that is?

I can.... but that's why I learned to program. (The tediousness of D&D NPC character sheets was the beginning on a TRS-80 in 1982 actually.)

So now, he says with great joy, that boring stuff is no longer necessary.

Now the instructor fills out a little web form as before, and gets back a list of all the courses he/she is teaching during the specified term. The script checks to see which of these do not yet exist in WebCT and provides little checkboxes for the instructor to choose which to have made.

On the click of a button, a script creates the WebCT course (named according to our standard scheme) and fills it up with students, then adds the course to a list of them that will be updated every day to ensure student course add/drops are reflected in our system for every class.

This all means that neither Kate nor I will ever have to create and manually populate classes in our teaching system again! Yay!

More technically, my perl scripts use the WebCT IMS API to communicate with WebCT directly and via XML files, plus they fetch data from PeopleSoft servers in Utah, selecting about a dozen fields from 40,000 in about four tables of some 12,000.

I will be having my bosses and colleagues test this next week, but all the hard work is done. They're going to love it.

Harder work still to come: making a similar system to deal with our other courseware system called Blackboard. That one runs in Windows and uses Java, neither of which I have interest in dealing with at this point. Well, maybe a little Java, but not Windows.

---

Sorry if all t his technical stuff bores you. I feel that way a lot of the time too, but it does cause good feelings to accomplish a worthwhile project like this.

I'm still working on the Panoscan business. Am currently trying to re-contact Hornet to see what's up with them. Will probably start parallel contacts going with other places within the week, so that SOMETHING starts happening somewhere.

Must get out of CSUN. Must not become complacent.