I just posted the first demo release of the 3d engine I’ve been working on, any feedback is greatly appreciated.
You can find it here: http://www.geocities.com/kyberteknik
Would like to complain about the size of the world, could have been slightly larger with some more scenery. 
Joke aside. Works nice, look s nice. Nothing to say.
Win2k, 2x733 MHz (useless to say, since it was only using one processor 
) and a good ol’ TnT (I really like those cards, they are awesome, it’s the programmers that don’t know how to use the card to it’s true potential. GeForce-like cards are just too fast for todays consumer-level computers)
Just out of curiosity, Bob, did you get the scrolling clouds (both layers?) I had a problem with a Riva 128 a while back and I havent been able to test it again on a TnT. I should have put a really good description There is supposed to be 2 layers of scrolling clouds and a big star in the sky as well as a purple gradient coming up from behind the mountains.
Ok, I tried it out on my TNT using the 5.22 drivers since the newer ones are slower for me. I noticed a few quirks.
- The model’s shadow did not appear until I raised the viewpoint up a bit from the starting position. Any lower and z fighting hid it partially or altogether.
- While the fog was on, the shadow was bright red, regardless of distance viewpoint was from model.
- Was the model supposed to be walking in slow motion? It was smooth and real slow.
- Bringing down the unfinished console background dropped the FPS to a paltry 2 FPS.
(Otherwise it topped out at 30 FPS) - I was able to get rain to go up and down at the same time!
Nice
Well, DFrey, spounds like it works perfectly (except for the shadows . Yes, the model was animated in a heavy space suit to look like she was walking in an antigravity situation. The console is unfinished (and hence is why I didnt mention it kudos for finding it though And yeah, the snow and rain can go up and down I didnt add the code to fix this because it just wasnt important to me. About the fog, does anyone know why thats happening? I use the polygon offset (and I check if the extension is available and use that if 1.1 isnt supported) to get rid of the Z fighting, it works on my card explicitly it seems (and the shadows are black on my card too). I use the function from ShadowyLeapinLizards demo to produce my shadow matrix, why would that cause the shadows to be fogged entirely? The non-homogenous transformation? I guess I could turn off the fog while rendering the shadows… I’m quite happy its running on other computers though, and for the speed, I starting to work on view frustum culling to help that out. Any other suggestions?
Cound’t see any strange thing with the sky. Could see the clouds flowing, and the starts behind the clouds. So I assume everything is OK on that part
I haven’t been able to make it work, i get a strange error message.
GLError #1282 in P:\Zeus\Engine\zFont.cpp(72)[invalid operation]
and then a failed assertion in line 72 of same file that says Expression(0).
Maybe it’s because I run Win2K? Also i don’t have a drive P: so it kind of baffles me to have an error like that .
Originally posted by DarkAngel:
[b]I haven’t been able to make it work, i get a strange error message.
GLError #1282 in P:\Zeus\Engine\zFont.cpp(72)[invalid operation]
and then a failed assertion in line 72 of same file that says Expression(0).
Maybe it’s because I run Win2K? Also i don’t have a drive P: so it kind of baffles me to have an error like that
What kind of hardware do you have? I know that the demo works with Win2K, I’ve tried it. The reference to drive P: is completely a debugging reference. If you noticed you dont have a zFont.cpp file either… it’s only relevent to my computer… Line (72) is the the line that calls glBegin() within a glNewList(). I’m not sure why that would be crashing. Aren’t compiled display lists available in all GL versions? And do the forbidden commands apply within glBegin/glEnd when compiling a display list?
I bet thats what the problem is can anyone confirm? I’ll take out the glGetError call and upload a new version late tonight (unless someone can tell me this isnt the problem of course