vrijdag 28 november 2008

Landelijk Architectuur Congres 2008

In case I didn't mention it, I'm currently working with the RDW for 2d/wk. The ultimate way to keep up is to return to the industry and do some actual work. The focus is on 'architecture' which is a really broad term. We're looking at introducing the subject in our curriculum. I'll blog about this some other time.

One of the first activities was to attend the Landelijke Architectuur Congres (crappy site). The focus of this congres is not so much on "software" architecture as it is on "enterprise/IT/information" architecture. I was a bit worried about grey haired pompous speakers giving abstract talks. This didn't turn out to be the case. It was a nice way to get introduced into the subjects, get to know the names, see the vendors and what they are vending. Although the crowd was different from a technical conference, it was diverse.

Some talks were too academic, some talks were a bit too obvious, some were badly disguised sales pitches and some were just right. I won't describe all the talks since there were a lot, but mention some highlights.

Daan Rijsenbrij voiced his opinions on architects and what they should be doing. There still isn't consensus on that. But after a few days at the RDW I can already support the statement that it's 80% talking, explaining, influencing, guiding, changing. A lot of "peoplework". He had hired a professional cartoonist to create some visualisations for him, something well worth considering if you have to make an important pitch.

Michael Widjaja: nothing really interesting about architecture, but we might want to pay some more attention to web 2.0 in our webprofile? Did you already know http://www.midomi.com/ ?

Harmen vd Berg: commercial demo of the Archimate language (in the tool Architect). Internationally there are a number of competing products for creating architecture models. But in the Netherlands Archimate is winning the battle. And since we got a Surf license I'll give a look and see if it might be interesting for my (technically inclined) students.

Frank Baldinger from the NL Architectuur Forum spoke about education and presented a book "Competences of IT Architects" which I'll receive in the mail. This might prove useful when thinking about architecture in our curriculum.

Tobias Kuipers: is it an idea to make architects responsible for the end quality of a project? It would at least result in more effort of the architect to get his architecture understood and used. A famous quote: "The street remains long after the houses have gone", or rephrased: "Concentrate on the lines, not on the boxes".

Olaf Zimmerman gave a nice demonstration of applied research. He has collected a large number of design decisions made in SOA projects and put them in a database. He then created a decision supporting wiki which can search the database. This created a 'mentoring' tool which proved more effective in supporting developers than an architecting document full of "must and shalls".

donderdag 13 november 2008

JFall 2008

Today I went to the Jfall. First a big thanks to the NLJUG for allowing me to take 35 of my students for free. I suspect a number of them skipped it due to illness, oversleeping and whatever reasons. They don’t know what they missed.

I started off with a talk by Marco Mulder from Xebia. He talked about a large project they have done for Prorail, using agile software development (Scrum) and a distributed development team (Netherlands and India). This of course involved some challenges which they solved nicely. I also learned a bit more about Scrum to complement what I know about eXtreme Programming.

Then there was a talk about Constraint Satisfaction Problem solving with Java by Nicolas van Hanxleden (Iprofs). Constraint Satisfaction Problems (CSP) are all about scheduling and optimisation and are typically unsolvable by a brute force approach. Since I studied these kinds of problems when I was in university, I was interested to hear if there is anything new. The first 75% of the presentation was largely familiar, although he did a good job explaining it. Nicolas then introduced the open source tool Cream which can be used to model and solve CSP’s. He gave a good demo in which he applied Cream to solving the N-queen problem.

Lunchtime! The handing out was a bit annoying just like at the JSpring. After getting up pretty early you don’t want to stand in a long line which is not moving. I’d like to challenge the NLJUG to improve the logistics of the lunch. That shouldn’t be too difficult. :-)

The keynote after lunch by Reginald Hutcherson (Sun) was amusing but not really informative. (Dude, simplify your slides). Two other Sun guys showed some fun demo’s of Java FX but nothing really challenging. The thing I might remember most is the throwing of the T-shirts into the crowd. But ok, maybe having some fun was all they wanted to achieve.

Immediately after that there was a talk about Lean Software Development by Marc Evers and Rob Westgeest. This also fit in nicely in my objective to complete my knowledge about agile software development methodologies. But I must note that the speakers emphasized that Lean is not a methodology but a set of principles, a certain mindset. What was new for me was the notion of measuring the worktime for a specific development cycle and comparing it to the total time before the product available to the business. In the example given they calculated the “efficiency” as 6%, all the other time going wasted to errors and (mainly) waiting. One of the key principles of Lean is to eliminate this “waste”. To do this they use other techniques from the production industry like applying “pull” instead of “push”. A lot of these ideas stem from Toyota car manufacturing. This talk was ok.

This time I decided to skip the last two talks and try my luck at a hands-on workshop on unittesting. This didn’t really work out however. After finally getting my laptop from the wardrobe and some installation trouble I was lagging too much behind on the assignment which I would have found quite challenging even if I started right from the beginning. The other participants were clearly having a good time, so no problem.

See you next year at the Jspring!

woensdag 5 november 2008

What would you do with €5000 ?

Ok, this question is a little bit different from "What would you do with a million euro's?" The point is I've been nominated for a prize called "de innoveer" (this doesn't really translate, meaning innovation and inno-feather in Dutch).

Our dean Henk Zwetsloot proposed me based on the work done on the software engineering curriculum and the way I keep up with it. There where 12 applications and I'm among the last 3 nominees. On the one hand I'm honoured, on the other hand I feel like I've "just done my job".

The prize is a piece of art and 5000 euro which can be used for ones own development. So I can't help but think about what I would do if I won. Here are some ideas:
* Attending a nice big congres like the QCon last year would be nice.
* A studytrip to the silicon valley or something similar to what the HBO-I did in the past.
* A studytrip to Australia to visit some universities and discuss the possibilities of exchange of lecturers.
* Getting a masters degree in computer science (I have one in math) is always on the back of my mind. But I'm afraid 5000 euro still wouldn't solve problems of sleep, time, family and energy. Maybe later...
Update: it wasn't meant to be, the competition was pretty good. But still, being among the last 3 was pretty cool.

Assembla update

I previously mentioned the site http://www.assembla.com as a free site which offers a complete software development tooling suite (Subversion, Trac, Wiki, ...). The main benefit for us is that the students can form their own projectgroups and administer their own version control. This relieves me and our system administration of a lot of grunt work.

In the mean time the site has gone to a (very reasonable) commercial model but they still support students for free. Recommended!