I'm fairly well involved with Open Source. Nothing really big though, no kernel hacking (yet).
Big projects
WorldForge
Homepage: www.worldforge.org
Worldforge's aim is to create a framework for Massively Multi-player Online Role Playing Games (MMORPG). They are currently working on their second bootstrap game "Mason". I have been involved in a few areas since September 1999.
I have some wf-related junk here.
Worlds development
I currently am the (Acting) Coordinator for Worlds Development. I keep world design moving. We are currently working toward the second release of the Duralsaur, built from the documents in the CVS.
Muse
I starting working on this in early 2002. It was going to be the media retrieval and caching library (in C++) for WF clients. Stopped due to lack of progress with getting requirements and standards.
Stage
I wrote Galileo and Pegasus, both of which have been rewritten since I stopped actively developing Stage. This was my first real experience in C++, I think I could do better now (see TP below).
Others
I have been following Dime development and help out early on with the build/configure system. I have also written a metaserver client in Java, which is used in the servet. I am also the lead (and only) developer of Archipelago. I am thinking of changing it a lot from its original description.
Thousand Parsec
Homepage: www.thousandparsec.net - local mirror (gone now)
This is sort of a side project of WorldForge. It is a Massively Multi-player Space Strategy Game. I am writing the server in C++ for it. Progress is rapid and good. About June there should be something to look at and move around. Should be really good when finished.
PVMPOV
Homepage: pvmpov.sf.net
PVMPOV is a patch to POVRay, a popular Ray Tracer, to run on PVM (Parallel Virtual Machine) clusters. I did the initial patch for POVRay 3.5b, since undone by other developers in 3.5c. I would probably be more active on this project if I had my cluster.
OpenRISC
Homepage: www.opencores.org/projects/ork1/
Open Source Hardware is fun. The OpenRISC 1000 is a 32/64 bit computer ISA. They are/were in need of a port of Linux (or32), but I've been busy and didn't completely understand the or1k. Someone has offer to do a port, as have several other people. It will be interesting to see when one comes (and whether I write it or not).
Small projects (AKA, my projects)
TrapBit
TrapBit is a passive Gnutella client. Run it on a firewall, point our Gnutella client at it as if it was a pinhole in the firewall and try to download off people. If the other person is behind a firewall (common and a pain), they will connect to TrapBit and TrapBit will try to download off them. A very dumb piece of software I would rewrite now if I still used Gnutella.
Dart
Dart ray tracer now has it's own page.
Stellar
Stellar was designed to be a distributed program that would run various projects, or maybe a hierarchical PVM-style environment. Didn't get very far. BONIC seems to do the former, and I don't have a cluster to do that latter.
LeeOS
A friend of mine started writing an operating system in late 2001. I also started one as well, called LeeOS. I only got the bootloader going. This led to my involvement with the OpenRISC. BTW: there is at least two other LeeOS around, trying googling for the other ones.
Processors 1 & 2
This project just doesn't go away. I think I'm going to have to focus some more on a processor design and build it. Currently, working on a 32-bit RISC with 64 registers, fp/vector/big-int instructions. Also thinking about doing a much larger processor, either 128 or 256 bits. Operating system will either be Linux or something like LeeOS.
Ankia
Perl ircBot. "The only way to learn Perl is to write an IRC bot". She has attitude, just like the real Ankia (Moa, NZ Singer). Might look at again some day.
Prime finder
Another project that doesn't go away. Also related to the next project. Finding primes is sort of easy (easy up to about 32 bits). Raw sieving should work fairly well in parallel. I need to learn how do to multi-word multiplication and division. I also need my cluster.
RSA factorisation contest
I found out about this not long ago. Nice prizes from US$10,000 up to US$200000, but almost impossible to do. Have some ideas, also works nicely with prime finding above.
LLV
Lee's Lossless Video. What a flop. Might try again in future. I think 100MB per 10 seconds isn't good enough. Encoding method was too simplistic (and not complete, only implemented part of what I designed). I also think I won't continue with lossless, but will try to keep to high quality.
Now, nearly 2 years later, that I have done a data/text compression course, I might stand a chance of doing better. Might look at it sometime.
Others
I have done quite a few other, even smaller projects.