While practicing algorithms, I suddenly felt urged to learn more about IEEE 754 double representation. I also made a little tool in C++ to display doubles as binary. Check it out :)

I just spent the whole day implementing a pathfinding algorithm: A*. The A* algorithm is based on Dijkstra, with the addition of a heuristic (estimation) that allows to favor one path rather than another.

The A* algorithm also has an interesting property: you can easily alter the estimates to make it more accurate or faster (not both, of course ;). That’s why the path depicted above is not actually the shortest one. That’s also why I spent the whole day searching why my implementation of A* was slower than the one I found on the web…

This is all properly explained on Amit’s A* pages. You can also find an example implementation and training material in Java on the Memoization blog. Thanks to these two people for their excellent content :)

Have you heard of it? The problem was to find a path through the city that would cross each bridge once and only once. Well it is probably the origin of graph theory :)

So yes, today’s theory: graphs. And I’m starting with these slides from the Princeton University. Pretty interesting :)

I wanted to finish my day with a little review of the quicksort algorithm. Instead of boring theory on random websites, I actually stumbled upon this very nice video-explanation. Also visit author’s website.

And by the way… If you happen to have a planning poker (or any card deck actually!), that’s exactly what you need to train yourself to sorting :)

Nothing’s better than a little game to learn! As today’s theory is TCP/IP networking, how about playing around with you own packets? Scapy allows you to do so. Built on top of the Python interpreter, it offers tools to forge your own network packets. If you’re interested in networking, it’s definitely something you want to try :)

And now… Let’s explore my network! See you soon.