GameDev is suprisingly unpopular in general... more to the point, ASM gamedev is relatively unheard of these days, mostly due to the public perception that development in ASM takes too long compared to other newer languages, and that the resources just are not there for ASM gamecoders. The reality is that ASM has grown up, as the people around here are well aware, it's possible for us to write oopasm etc..

To be honest, there's simply very little interest in asm gamedev... Scronty is a perfect example, his DirectX (asm) site had only something like 600 (unique) hits in a year.. compare that to the average C/C++ gamedev tutorial site which gets 1000 hits a week, and thats not a heavily trafficked site.

I am not worried in the least :)
As far as I am concerned, if the rest of the world insists on writing code in those languages, I must cater to them by restricting my general coding activities to writing superfast and ultrasleek code modules callable by those languages. To me, it means that the rest of the coding world is a potential customer.

It still won't prevent me from working on larger projects in asm either, it just means that I must as I said adopt a modular approach to my coding.
Posted on 2004-12-09 01:22:42 by Homer
what about emulate the spine "AI" we have?
check for what poly will be under the foot next time and adjust foot underside to turn so it becomes parallel to this poly and collisiondetect this and stop motion of that foot and start with shifting your weight to this leg, then lift next leg

human walking is done that way to consume minimum energy, by slowly "falling" forward by tilting your body slighlty forward and lift a leg forward to take care of impact on ground,shifting weight between legs
your sight+balance is main input here+underside foot skin feels if you detect any collision, in uneven terrain you must look down on ground to see how to carefully put down leg and not to fall over the dead orc you just killed, or just a rock or dead tree
Posted on 2004-12-09 03:46:04 by daydreamer
I agree with your posting - human walking is a "controlled fall" - but how to apply it to a skeletal model? We could use an IK solver by generating our animation based on constraints applied to joint rotations, but would it look smooth? It sounds like overkill - the method I proposed attempts to converge a walk animation with surface collisions - it still sounds like the most realistic solution so far.. keep the ideas coming :)
Posted on 2004-12-09 04:23:31 by Homer
Hello guys

Yes, I'm also very interested in this topic but, very busy at work.

Maybe some interesting sites ?

Real-time inverse kinematics techniques for anthropomorphic limbs:
http://hms.upenn.edu/software/ik/ikan_gm.pdf

Here's an archive and link site for some related resources:
http://www.imonk.com/baboon/bones

and also have a look at "Game Developer Magazine" there are a lot
articles on how to do character animations:
http://www.gdmag.com
Posted on 2004-12-09 07:08:05 by Siekmanski