Aidan Delaney’s weblog

I have a few issues with Python. One of which is the unit-testing mantra. I’m a big fan of unit testing, however I don’t see how it can take the place of static type checking. It can supplement static type checking, but not replace it. For example:

File “/home/balor/tmp/revisor/revisor/base.py”, line 1410, in buildInstallationMedia
self.mode.BuildMedia.extend_task_list()
AttributeError: RevisorBase instance has no attribute ‘mode’

The above would never occur in a statically typed languages such as Java. The Java compiler would simply tell the developer at compile time that RevisorBase has no mode attribute. Unfortunately, this application takes about 20 minutes before it hits the offending line. So what would be a quick spin in a compiler in a staticly typed language is a 20 minute run in Python. It’s also a 20 minute delta between testing patches If I write a change it takes 20 minutes before I see if the change is effective.

I understand that one can get similar runtime bugs in statically typed language, however I’ve seen several instances of this type of issue in Python which would be solved by the simple application of a statically typed language.

Anyway, why should I, the user, see this issue.

As I’ve said before, one of the biggest lies about academia is that you have your summers off. You don’t. But I’ve finished all my teaching for this academic year. So I’ve got two months ahead where there should be no student interruptions except for setting repeat papers in August.

Recently I’ve had two papers accepted for publication (and there’s one in submission):

• Stapleton G., Delaney A. Towards Overcoming Deficiencies in Constraint Diagrams . Accepted for VLHCC2007, Visual Languages and Human-Centric Computing, Coeur d’Alene, Idaho, USA, September 2007.
• Delaney A., Stapleton G. On the Descriptional Complexity of a Diagrammatic Notation. Accepted for VLC07, Visual Languages and Computing, San Francisco, USA, September 2007.
• Another paper has been submitted to VLL, which is a workshop of VLHCC, with partial reference Delaney A., Stapleton G. Spider Diagrams of Order.

So the plan for the summer is to :

• Turn the papers into a thesis chapter.
• Finish another technical thesis chapter.
• Get my registration (finally) transferred to University of Brighton.
• Finish the integration of Eclipse and scratchbox-PPC for deployment in our labs.
• Write the technical content for our new Computer Games degree.
• Rewrite my Computer Graphics Algorithms course as I want to take it in a different direction.
• Turn a thesis chapter into a possible journal paper.
• Go to GUADEC and have a good time.

So when I said that you don’t get a summer off in academia, I’m not moaning. I could probably get away with not doing most of the above list but then I’d (a) not get my Ph.D. finished and (b) not be providing the best teaching I can to my students. Both of which are important to me.

Basically I have two months of heavily self-directed tasks ahead of me. Which is just as good as a summer off, because I work on what I want to.

I’ve been banging on about source and configuration management for years. I teach SCM to students and have even written a paper about how to use it in teaching software engineering. I sent a patch to Eric S. Raymond for the documentation to his comparator, and got back the following reply.

Applied — but necessarily by hand, because you sent an -e patch. You
should ship patches in -c or -u format.

D’oh. I’ve been told that before with respect to source code. I didn’t think it would be the same for documentation. Well, you live and learn.

I’ve just posted up an initial revision of my source and configuration management mini-book. I wrote it as a reaction to the questions I get from a lot of students. I aslo wrote it in four hours…so be kind. It’s rough, likely to contain mistakes and has no references currently. It does, however, take a different approach to most howto’s or guides. It goes into more depth on why you might want to use SCM and it give examples using two different SCM tools.

The web URL is http://foss.it.brighton.ac.uk/notes/SoftwareArchitectures/scm.html and the bazaar (bzr) repository is at http://foss.it.brighton.ac.uk/~balor/bzr/SourceAndConfigurationManagement/ patches and suggestions will be gladly accepted.

So summer is finally over. I know this because we had our first league game of the season last Saturday (and lost ) and there are quite a number of confused students walking around the building today. I remember my first day at University. I took a lift in with Donie who was going in to fix the lights in the student union bar (I worked for a sound and lighting company then. There was no room in the van, so I had to lie on a ladder in the back. But now the tables have turned…

And there’s a lot planned for this year. I’m rolling out the method of teaching software engineering that I presented at PPIG 2006. Hopefully I’ll be able to collect some data to support my hypothesis and present that next year. I’ve also got a few new courses such as Real-Time Systems and Client-Server Technologies. I’m planning to roll out the Real-Time Systems course using RT-Linux on ARM (using qemu), which sould be a lot of fun. I’ll be using the hardware/software plaform to underline the theoretical issues of Real Time Systems. This should also form a basis of an Advanced Linux course that we intend to introduce 2007-2008.

Plus I’ve got a lot of research to finish this year. It’s going to be quite an eventful one, though I don’t know if it’ll top getting married. Which reminds me that I should blog more about food. On Friday (our two-month anniversary [yes I know the word means annual]) I made L. a nice meal, starting with spinach and feta cheese in filo parcels, Thai fish curry with Japanese steamed rice for main followed by Tiramasu for desert. It all worked out very well, and it didn’t take too long to do. Once the Tiramasu is in the fridge about 5 hours in advance, everything else only takes about an hour to make. I’ll try to remember to post the recipies tomorrow.

Havn’t been blogging much lately as there’s been too much comment spam. Hopefully it has been fixed now. Though, dear reader, I need your help. Could you please tell me what you think I’m communicating via this diagram? Update: I’ve already modified the diagram, based on some feedback. More is appreciated.

I’m giving a similar (or modified) image to my second year students. I’d like to know if I’m actually communicating what I intend to communicate. Consider this a usability test(ish).

I’m getting married in ten days. Unsurprisingly I’ve little else on my mind. However I do still have to work.

I just submitted a camera-ready copy of my paper to PPIG2006 (change history is available at ‘svn co http://foss.it.brighton.ac.uk/svn/ppig2006/ppig2006-release’). I think it’s a good paper, but it needs to be developed over a longer term, which is something that I want to do.

I’ve also inherited next years real-time systems course. Which places me in a strange position. I’m going to blog about it but….. Some of the techniques I’ll be using to teach the course are experimental. The problem is with the “Show no fear” mantra, where lecturers are advised not to show that they are non-expert in an area or that they make mistakes. I may use a technique that turns out to be a mistake. If I don’t blog I can always “show no fear” and point-blank refuse to accept that I made a mistake, if I do blog I may be land in some interesting situations. I’m willing to trust that students want the best from me that I can provide, and I’m willing to trust that they understand that in order to do this I have to push the limits of my knowledge, technical skills and teaching skills. Basically, I want to turn this away from me teaching the students and more towards having a conversation with the students. In this way they can tell me how they learn best and I can accommodate them.

So at the moment the set-up for real-time systems reflects my wish that students actually get down and dirty with a real time system. In order to do this we’ll be using RTLinux on ARM emulated by Qemu. I’m getting two ARM development boards (thanks JD) and I have two robotic armatures. I want to “simulate” the control rods in a nuclear power plant using the armatures and a childrens’ toy (you know the one where kids get to put cylinders into cylindrical holes). So we’ll see how this all goes