CS50 looks more "fast paced" as it teaches you several paradigms, but I'd rather learn more C than Python.
I can't make up my mind. I've programmed a bit, in Python, C and now a bit of C++ (Also a bit of Haskell) but I never took any kind of classes nor studied things like Algorithms or Data Structures in deep.
What would you choose ?
It's a very popular course, and the majority of students are taking it as an elective. It doesn't go into particular depth with regards to algorithms or data structures, so if you're looking to learn more than just the basics (binary search, trees, etc) it probably isn't for you. It has a good reputation on-campus, and is one of the most popular classes. It's also already got an established distance education component, which is one reason why it's being offered as one of the first edX courses.
I have only taken CS50, so I wouldn't want to comment on MIT's offering, but hopefully that will help you out! If anyone had any questions about the course I can try and answer them - the course hasn't changed substantially since I took it.
CS50x looks less theoretical and more of a survey of different programming languages and patterns. It's a good "get up and running in several different ways" class.
To answer parent's original question: I'm going with Harvard's. I started watching the open courseware lectures from MIT's a while ago, and though Eric Grimson was intelligent and clear, he just couldn't hold my attention.
(My opinion based on having done the MITx Circuits & Electronics, and Game Theory and Cryptography from Coursera.)
Hopefully, the various platforms (Coursera, edX, Udacity) play nice together, i.e., it's only a small amount of work to setup the same course content on a new platform.
edit: woah, Dan!
I like that, as with the recent Caltech machine learning course, these courses look like they will really push people.
I don't really get the difference between these two: https://www.edx.org/courses/MITx/6.00x/2012_Fall/about https://www.edx.org/courses/HarvardX/CS50x/2012/about
Any ideas?
The online version only has half the material. However, this makes sense; the class is naturally divided into two parts. The first half was focused more on AI and search problems and the second part was more about machine learning. I suspect the only reason it is one class is because it has to fit in the semester system. Online, naturally, it's liberated from this constraint.
Anyhow, I definitely recommend this for anyone interested in AI. And even if you're not interested in the field as much, it's a fun course and only lasts three months.
