Yesterday, four of my classmates and myself gave an hour long lecture on Erlang/OTP. That lecture was followed by a two hours long lab. It’s been a while I wanted to share about Erlang, I read a lot about it but unfortunately, I never found some project I was interrested enough to dive in. I now have an idea but the usual problem is showing up : not enough time.
So the presentation covered the Erlang language, with its roots (the Ericsson lab, industrial use), its paradigms (functionnal and concurrency oriented) and syntax (that reminded Prolog and OCaml to the attendance). I mentionned the capabilities offered by the OTP framework (generic behaviours, supersivion trees, hot code swapping, …) linking all that to reliability and availability concepts as we got a fault tolerance lecture few weeks ago.
The lab was based on a simple messenger application from the Erlang getting started tutorial. Given the amount of time, I structured it as a “fill in the blanks” exercise. All went OK and people seemed happy with their working system, at the end. That was really a great feeling to explain processus interaction and deal with students mistakes while helping them. It’s also great to open their eyes on an unknown technology, as they now could choose to pick it up to solve some industrial challenges they could run into during their career.
On my side, I’m now gonna find time to start that Erlang project I think about. I really need to get skills on it as it’s definitively part of my fields of interrest. That could help to find a related job later.
Back from holidays, we had a project week at school. This week (the one ending today) was supposed to be filled with project conception & development. I’m part of a eight guys team trying to improve a network simulator named simmobil. That project is already 2 years old and several things have been (more or less dirty) achieved on it. Simmobil is more exactly a wireless network simulator developped using the Java programming language. One of the improvements we’d like to bring to this software is the simulation of sensor networks. We plan to implement several routage and sleeping algorithm as well. Things that I’m really involved in are the repairing of the graphical user interface which sucks and the creation of a new way to use simmobil through a XML based configuration file. This week, we had to deliver the conception report. Since I’ve wrote the GUI part which is really not the most complicated, I had almost nothing to do (well, nothing to do twice because of a misunderstanding of a complicated algorithm). So next we started to code and that’s kinda boring : simmobil is like a dummy project, it will never be used. Our superviser is a researcher whose domain is networks but that tool is really not usable. How can you find motivation in such a context? I’d like my department to choose to attribute tasks on free software to students. That would be useful for the community. So I haven’t really achieved much work on simmobil, and my week was kinda… empty.
