Yes, it took a while for me to get back to blogging, but i was waiting for something interesting to write, but then again...

The group-forming process of this semester has been my best so far - it took about 10 seconds. We were 4 guys who had already decided on a forming a group, we are 28 guys on the game programming line, and 4 groups of 7 had to be formed. 3 guys sat between us, and asked if we should go for it - and so we did! Two other groups were already formed, and the last 7 was then a group - done! Took longer to explain it in this blog.

Well, our project for this semester should be using some form of agent behavior. We have decideded on a turn-based strategy game, looking at Warlords (mainly the second) and Heroes of Might & Magic for inspiration. The graphics have been but in the hands of MadsG and one of the new guys Anders 'Duck', whick will probably result in "The Drunken Vikings vs. Dragon Ninja" graphics - well the prime target of a game should always be to entertain!

We have decided on using Java, because some of the group members have had some rather bad pointer experiences with C (and some haven't touched the language yet), and if they have to both learn a new language and code a game at the same time - it just seems to hard. So Java it is - our preliminary code seems to prove that the graphics performance of Java (using the awt toolkit) is good enough for our purposes, eventhough the Linux implementation leaves much to be desired (faster blits, more display modes, etc.). I'll post some screenshots when there is more to see.

One the AnyBus front, i made some good progress today, i actually understood how to do network programming, guess i should have read Beej Guide http://jcatki.no-ip.org/beej/ instead of pretending to know everyting. I think i might even try to take up SDL_net (for the third time) to see if i didn't just interpret the API wrong. Actually i hope to make the library support both Netxx and SDL_net, and just make a preference towards Netxx. Transfering a 512 million bytes file over AnyBus (locally) yeilded about 3.12 MB/sec, which seems alright (altough initially, when the library only worked sometimes, i had rates higher than what Apache could perform).

I can now go on to design the real API for AnyBus - a more high-level Bus class and a BusServer class. This should make AnyBus much easier to use for programmers. More to come soon (i hope) - when i get time off from the university and don't waste my spare-time drinking.