Zulu Hour

it’s been a while since my last post, but that certainly doesn’t mean i haven’t been working on the game. i know it’s a bogus reason, but it’s not as motivating to write if you know nobody will read it (i never kept a journal). however, i just found 2 comments that weren’t spam -that’s right, live PEOPLE! so, no lame excuse not to write now. i’ll try to do better.

so, what have i been doing? menus and windows! while i wouldn’t want to do it for months on end, it’s been pretty fun. karl has been supplying me with fantastic art, so zulu hour is looking more and more like a “real” game. of course, in the grand scheme, it doesn’t help the game play at all, but at least it doesn’t give you a bad taste in your mouth before you even start playing it. plus, polish is one rating factor in DBP, so it can only help us, right?

closely related to the GUI is the overall framework of the game, meaning starting the game, selecting a level, beating a level, then continuing to the next one. i’m getting close to havig “campaign mode” ready, and once that’s done, it’s off to playtest for some public feeedback!

oh yeah, EMP weapons have been added! man, i’ve got to post more often.

i love it when people get on the XNA forums and ask, “how do i make a game?” that is one of the rare occassions where the bastardly answer, “if you don’t know, i can’t tell you” might actually be appropriate. games are designed to be fun to play, and it’s easy to look at a beautiful let-down like far cry 2 and think you coud do better. but my god, video games are not trivial things to make. it’s not like golf…pretty simple to create a game like golf, but as they say, it takes a lifetime to master playing. i would say video games are quite the opposite.

so what brought this mini-rant on? i spent my weekend optimizing. i try to follow the rule of optimization, “don’t do it”, but when you’re running at 20 fps, you don’t have much choice. my newest xbox-crippling feature was about 1200 trees. they’re all the same, so hardware instancing was the obvious solution. now, i could have cut+pasted the xna instancing sample, but where’s the fun in that? actually, getting quick results can be extremely fun, but i’m doing this project partly to learn something so i dissected every LOC in the sample and recreated the whole damn thing. i can’t say it was the funnest thing i’ve done, but the technology is cool, i learned a lot, and it works (what more can a programmer ask for?)!

results? could be better, but on the xbox i’m bottoming out at 45 fps. still needs work (better culling), but if you think about it, if every tree is 100 polys…

100polys * 1200 trees = 120,000
120,000 * 2 (reflection in water) =240,000 polys

plus the terrain…not too shabby.

and what was the point of this post? i’m not sure myself. for one, the game now has trees! second, for those who want to know how to make games, multiply this kinda stuff by…oh i don’t know…the xbox’s fillrate or some other really big number.

hey all, ‘nother video has been posted!