I started my latest project this last week. Essentially it's the final year project for my CS course and counts for 17.5% of the degree final mark. I also cannot pass unless I at least pass my project. So nothing important then. On to what it is.

Essentially I've got to build a system to render terrain heightmaps in 3D (hence my extremely bad blog title). That in itself wouldn't be too bad apart from the fact it needs to be delivered over a network (where naturally I implement both server and client), in real time, using a distributed delivery model.

Further more it needs to be a multi-resolution system. I need to be able to deliver the map by resolution level (so that I fill the low resolution stuff first then build up). It also should deal with T-vertices (where sections at differing resolutions leave gaps due to a lack of detail) and scene management to decide what parts of the terrain need full detail, what parts can be less detailed and what parts not to render at all.

None of this is ground breaking in any way but I don't know of any project that simultaneously considered this particular combination of issues (of course there is no good real life reason to deliver a heightmap in real time but I'll ignore that). It should be interesting to read my excuses as to why it didn't work come the April, 23rd deadline.

Naturally I'm using C++ for this. Along with OpenGL for rendering and SDL to hold everything together and handle events. I'm going to try and get it open sourced when I finish so if anyone is interested it will probably be available here next August or so. It will probably work on all three major platforms (I'm using Linux but I will probably be testing against Windows throughout. Hopefully the Mac will live up to it's 'just works' claim because I'm not hunting down a Mac user to test it).

Other than escaping with my sanity I hope to learn a lot about proper scene management data structures and algorithms and get a decent exposure to network programming.

I will hopefully post more details here when there is something slightly more interesting to say.