Train hard, win easy

In order to keep up with IT my next challenge for myself was to become Sun Certified Java Programmer (SCJP, CX-310-065). I took the exam yesterday and scored a nice 95%. I'll share some observations about the whole process.

Preparation
I first took a two day exam training (utilizing some special HBO-fund ). Looking back this wasn't really necessary, but didn't hurt me either.
I already had an impressive exam book by Rasmussen but the book I got at the training was even better. The Rasmussen book is very comprehensive but the (for me) difficult topics of generics and inner classes were too dry to get through. The Sierra/Bates book was just more readable.
At the end of the training we took a mock exam and I scored 60%. The current pass norm is 58% (used to be 65%). So I decided to study my weak topics and all should be ok. After studying my weak topics I took the 2nd mock exam: 60%! Hmmm, shouldn't I be scoring higher?
After about 30 hours of going through the entire book during the christmas holidays I took the 3rd mock exam. 61% WTF!? I never had so little return for so much study.

I was starting to get pretty annoyed with the whole thing. The Sierra/Bates questions were often about nitty gritty details, or contained tricks within tricks within tricks. I'd study a whole chapter and the first test question was about the exception to the exception in some small corner.

The last day of study I toyed around with live code, instead of just reading the book. This also helped and I could have done more of that.

Exam
Then the exam came. It turned out to be way more easy than the Sierra/Bates mock exams. One big difference is that Sierra/Bates phrased most questions as "choose all that can apply". On the real exam most often the number of options to choose was specified ("choose the 3 correct answers"). That made a huge difference. The other big difference was just that the questions were less tricky. There were some, but most questions were straight about testing your understanding on the topic at hand.

Another weird thing was that I identified 3 (out of 60) questions to have clear reproducible errors in them. And these were unintentionally since the answer "will not compile" was no option. One was a declaration "private name;" (missing String), the other was "private void() methodName{}" (parentheses misplaced). The other was "choose the two correct options" but then providing a radiobutton . I find this pretty astounding for an official exam taken by thousands of people.

Afterthoughts
Then came my score: 95%! At first I was happy, but then felt a bit disappointed. If it is this easy and the norm is 58% then everyone can to this. The certification isn't very exclusive and I could have saved a lot of studytime. Oh well, I'll just make sure to mention my score to everyone ;-)

So does this whole experience make me a better programmer (/lecturer)? I'd say that 50% of the effort was useful in seeing things about the Java language in which I didn't have much experience. The other 50% is more a testimony of determination, concentration, focus and study capabilities. But I could also have shown that by learning Spanish or how to play the piano (oh, if only a day had 48 hours... :-)

Next stop is to educate myself more on research but I'll be back for the developer exam (SCJD) in the future.

The best way to fail for a student

Today we will finish our project for Datema with demo's, presentation and prizegiving. We had five groups working on a web app, Android app and iPad app. In general the results are satisfactory.

Pete* however had not been up to the task. He never got his hands on the keyboard and learning objective C was too difficult for him. I had noticed that and the peerassessment confirmed it. I invited Pete over for an individual examination which he probably wouldn't pass.

When Pete entered the room I asked him if we should go through with it, or rather discuss his issues. We ended up talking about his problems, possible solution, his future ambitions and his choices for the rest of his studies. This was much more valuable to him than doing the individual test. When he left he had a new vision about what he wanted to do and how to get there. If you fail for a subject, make sure you get something out of it.

Twitter

Blogging about Twitter, ehhh... right. I've looked into Twitter for the Datema-project (www.wikipilot.org) and think I have the hang of it now. But the big question is: will it be really useful? I'll give it some time to find out.
I see a great use case for Twitter in the chess world. If every player or captain gave some tweets after the game, you'd have a really nice way of following all the matches. But I'm afraid I'm probably the only member of our club who has a clue about Twitter.

Research

Last week I went to a seminar on research. Boring? Dry? No! The seminar was titled "Research in professional education, there's more possible than you think" and was run by Bas Haring and Maarten Lamers. Bas Haring is best known of the two, but they make a great couple and complement eachother seamlessly.

The angle of the seminar was that the universities of applied science (which in the Netherlands are just getting started with research) shouldn't try to emulate what the research universities are doing for centuries. Instead we should make use of our own strengths to find new subjects and research them in creative new ways.

For me the biggest takeaway however was an enhanced vision of what research is, separated from how it's done. Another big point we kind of discovered as a group is the confusion between "uitzoeken" and "onderzoeken". (Trying to translate these in English doesn't resolve the confusion.) Probably a lot of activities which are called 'onderzoek' in reality are 'uitzoeken'.
Finally I also have some ideas about how to integrate research in our curriculum. That might be useful because the rest of the organisation is kinda struggling with that. :-)

A new project for Agile Software Development

For years we used the Cab Dispatcher as the standard project in our Agile Software Development course. It still is a great (fictional) case study but it was time for a change. Last year we experimented with the case study on monitoring mentally handicapped people. Doing a new case poses a number of challenges to the lecturers, but it was closer to the real world.
This year we're taking it a step further by doing a real-life project with a client in the business of shippingnavigation. I think it will be very interesting. Let's see what the students think about it.

Summertime & mobile devices

It has been a long summer which I have used to study a lot of stuff, both professional and private. One of the things we did was buy a Nintendo for the kids and find out how to (...).

What amazes me is the amount of game- and userexperience comes from such a small device, i.e. in a nice game like Picross3D. Compare this to the new HTC smartphone (Windows mobile) I got from my office a while ago. It looks cool but has the size of a refridgerator, the powerconsumption of a SUV and I'm not enjoying the usability.

When is Nintendo developing a mobile phone?

Lean architecture

Just before the summer holiday started I went to a NLJUG university session by Xebia. The evening was about their concept of "lean architecture". Lean architecture means applying lean principles like "eliminate waste" to the architecting process. And there is quite some waste to eliminate at most companies :)

After presenting their principles (http://www.slideshare.net/xebia/lean-architecture-university-session-for-nljug-at-xebia-on-july-8th-2010 ) we quickly got to work on a case study. And maybe this was the most instructive part for me. Getting a fuzzy assignment, too little time and a bunch of people. "Go and do it." As an outsider I would say that there were way too much degrees of freedom in the workshop (I counted 8). While you were working on 1 or 2, you would be interrupted about one of the others. Never mind, my take away point is to keep this in mind when I set my students to work.

At the questions at the end they put on their "process" slide. I felt a bit uneasy with it. Agile was al about keeping things simple. But now the scrum process was extended with a productbacklog kanban and an architecture kanban and relationships between the three. Oh well, when you need to choose between changing people or processes, maybe changing processes is easier?