Ok, Moesday (Monday+Tuesday) was a little rough on me. I felt too sick to come to the office on Monday to work with Dr. Nguyen on Connexions, instead I extended the Thread Checker to support XOR @Combine annotations, as suggested by Bill.

At the group meeting, Corky announced today that IT had agreed to add RedHad 5 virtualization techniques to at least some of the computers within our reach. That would enable us to automatically test both on Linux and then on a virtualized Windows. MacOS seems more difficult, unless you use an x86 Mac and Parallels, for example. I think this is a great idea. I wonder if it’s based on my comment a few weeks ago about Parallels and always running tests on all major operating systems. It feels like I can take some credit for this decision.

Connexions is a terrible user experience. To take some HTML website (and HTML kind of resembles XML), the best way we have found so far is to paste it into Microsoft Word, mark the different paragraphs, headings and code examples with the CNX template values, then import the Word file into CNX. Doing so wrecks everything else you have done so far, so if you have uploaded assignment files and used their URLs for hyperlinks, the files will be gone, and you need to-reupload them. Of course, then the URLs change, so you need do change your document again… Since everything is running on a boneheaded-thick server and think clients, clicking on one “Next” button can make you wait 20 seconds.

I think there’s great value in formatting teaching material in a coherent way. But saying “hey, we’re making teaching material coherent, let’s first invent our own language”, that’s completely wrong. The inventors of Connexions project said it would change Computer Science curricula: When Corky, Stephen and Zung asked how it’s going to change CS curricula, they received no answer. After making a few attempts at translating my “Programming for Change” lecture and assignment for COMP 201 and the SIGCSE 2006 workshop to Connexions, I know exactly how it’s going to change CS curricula. CNX is going to mutate entire CS departments into packs of monkeys that take existing material, then use the mouse to select text passages and mark them with CNXML styles…

I really think Connexions would make a wonderful study of what went wrong in human-computer interaction. What they really need are good converters from plain text, HTML, DocBook, and perhaps TeX to CNXML so you don’t have to go through Microsoft Word, which is incredibly boring and tedious. The translations doesn’t have to be perfect, but work the most common cases. Then, Connexions needs a local editing system: There either needs to be a CNX system that can be installed on localhost so that all the work can first be done locally, with very low latency, or an offline-editing application.

I’m also usually a very late packer, but I already had all my things packed by 9 PM on Monday, because we had planned to spend the night at Diana’s mom’s place, so I wouldn’t have to be dropped off and picked up again. I was ready to spend an additional day away from my apartment. Then I found out I was a persona non grata in that household. Very encouraging, but I’m not going to let that spoil anything, at least not on my side.

Diana and I went to our first ballroom dancing lesson today, and I wasn’t surprised when we started with Foxtrot. It was taught a bit differently than what I had learned in Germany (during the four lessons I received; I don’t know much); for example, the men’s leading foot is the left in the US, but it was the right foot in Germany. I think Both Diana and I had fun, and we’ll definitely do it again. Sadly, we miss the Waltz lesson next Tuesday, so maybe we’ll just have to Waltz on our own ;)

The instructor was less forceful than the instructors I had at the “wedding dance crash course”; it was a more enjoyable experience, even though I think the biggest part was that I was dancing with Diana, my girlfriend (or stepping on the toes). I’m sure we’ll do it again.

After that, though, I was so tired, I pretty much fell asleep right away. When I woke up at midnight and 2 AM, I was surprised to find Diana still here. I thought she’d go home to finish packing. But we’ll make it, no doubt. I still have to pack my computer. We were able to change our seats to an exit row and sit next to each other, by the way. And I packed extra batteries so Diana can play her video game… ;)

For those who somehow do not know, Diana and I will be going to California for a week. I don’t know how accessible I will be, but I will let you know.

I’m looking forward to this trip. It’s going to be completely different. Of course, I’ve still taken all my code and two versions of my thesis with me…


About Mathias

Software development engineer. Principal developer of DrJava. Recent Ph.D. graduate from the Department of Computer Science at Rice University.
This entry was posted in Concurrent Unit Testing, DrJava, Ramblings. Bookmark the permalink.

Leave a Reply