[COMS W4995] Topics in Computer Science: VoIP Security

I agree with the previous reviewer. Stay way clear of this class unless they change the layout of it severly.

The lectures were just student presentations of different more or less interesting subjects. It was supposed to be seminar style but since no one had a good VoIP background the presentations (including mine) were naturally shallow and buzz word centric.

The labs (30%) were enormously time-consuming and not very rewarding. My group had hands on experience with setting up VoIP servers before, otherwise I would had been completely screwed. The presentations were alright but harshly graded. The projects were rather vague and too wide in scope and in my opinion most of the groups did not produce anything of substantial value.

Even though this was the class that caused me the most stress and took almost as much time as OS I still got a rather lousy grade compare to other classes. This class gave me almost nothing. I still don't know how VoIP works :-)

Steer clear of this course. It is supposed to be a research course, but except for the project (40%), everything else is a lot of manual (I'd prefer the word 'menial') labor that comes to nothing. Schulzrinne is probably a brilliant professor, but this course is the worst I have ever taken in my whole life, and this course is also the worst mistake I made in Columbia.

As it seemed to me, there is nothing worthwhile taught in the class, the TA was not very useful if you went up to him with problems, the labs (3, 30%) were long and monotonous that didn't teach you anything about how VoIP really worked, 2 presentations (30%) that is supposed to make you research the topics and give a talk on (but everybody just makes short work of it).

The project can be really meaningful if you choose a good topic, but make sure you either do it alone or get a partner who is sincere about the work and is not looking for a free ride. The TA was not very helpful in resolving either technical or interpersonal problems. You might get a mentor in one of the professor's PhD students, but remember that you might not get a lot of help unless you put it yourself. Everyone I went to talked in arcane terms that they expected you to know, but you wouldn't know anything about them unless you have prior background in that field.

And compared to the fact that it has a 40% weightage, the professor does expect a little more than that. If you do take this course, brace yourself for some boring moments in the semester, and do not take up a project that would require you to write a lot of code from scratch. No one seemed to care that writing original, running source code for a nascent idea wasn't trivial (that was what my project was about, and unfortunately I had a partner who thought his job was only to supervise).

I wouldn't recommend this to anyone, but to each his own.