What are you working on?

[jitspoe]

Still working on AI stuff. It would go a lot faster if I could focus...

I've made progress on dynamic waypoint placement. Basically, it observes player movement and places waypoints while they're on the ground up until a hardcoded limit. Once the waypoint limit is reached, it "randomly" deletes and reconnects existing waypoints, favoring removal of waypoints that are clustered together (no sense keeping a bunch of waypoints that are right next to each other).

It actually... works a little better than I expected:







3 shots of the same area as I've walked around it a bunch with a waypoint cap of 100.

[goldenboy]

Very good, jitspoe. That looks useful.

Being an obsessive-compulsive level designer, I'd still place mine by hand, but this is very impressive.

[toneddu2000]

great work, jitspoe. Do you plan to release the code open source? I could really use it a dynamic waypoint system

[jitspoe]

Very good, jitspoe. That looks useful.

Being an obsessive-compulsive level designer, I'd still place mine by hand, but this is very impressive.

That's all well and good when you're working on a new level, but if you're just a player who wants to fire up one of a thousand preexisting levels and play with bots, something a bit more dynamic is in order, I think.

great work, jitspoe. Do you plan to release the code open source? I could really use it a dynamic waypoint system

I actually have the source up on the sf.net cvs here:

http://paintball2.cvs.sourceforge.net/v ... all2/bots/

Be warned that it's still very early WIP/prototyping code, so it's super messy, and half the functions don't get used or don't work correctly. Feel free to try to salvage some of it, though.

Most of what you'd need for dynamic waypoints is in bot_waypoint.c

[toneddu2000]

wow, thanks for the link,jitspoe.

That's all well and good when you're working on a new level, but if you're just a player who wants to fire up one of a thousand preexisting levels and play with bots, something a bit more dynamic is in order, I think.

This isn't the only case. Imagine you want to design a game where levels are randomized every time you restart them (there could be a dungeon -like but even a city environment like game, it's irrelevant): in this case a solid dynamic waypoint system is required to give credibility to AI.
thanks again

[frag.machine]

Impressive work, jitspoe. I haven't seen the code yet, but I imagine that would be feasible to adapt the standard AI QC builtins to use this waypoint network, so already existing mods could take advantage of the better navigation, right ?

[jitspoe]

toneddu2000: With a randomly made world, I think you'd be better off with pre-placed waypoints on each playable prefab area. The method I'm doing requires that the player actually traverse the space, so AI would not be able to navigate until the player walked there. Of course, you could still use my logic for auto-connecting nearby waypoints and just have waypoints procedurally scattered around the map.

frag.machine: This is being developed with Quake2 as a base, so I'm not sure how difficult it would be to support Quake1.

[ceriux]

Working on monsters for an android app called minimon he's 176 tri's

[toneddu2000]

Of course, you could still use my logic for auto-connecting nearby waypoints and just have waypoints procedurally scattered around the map

Yeah, that's what I meant.

[jim]

jitspoe: Seems like a good/fast way to insert the waypoints.

ceriux: You could scale the 2nd edge loop above the eyes bigger, so it would appear rounder..

--

Some HUD graphics.. (the mid part wasn't from ingame screenshot). Will improve placement of stuff later with CSQC, but otherwise have it look something like this:


Banana Container.. going to improve this one a bit, as the other one got more detailed. Was thinking it would get some dark grey things in it like the other one.


Tomato Juice Blender.. this is the other one:

[Spirit]

Those look awesome!

[qbism]

Hud and models look low-res/pixely, so I have to ask if that unfiltered appearance will remain in game?

[jim]

Thanks

I'll leave the no/filtering up to the user. I'll probably try to have default setting for it "gl_nearest_mipmap_linear".

[leileilol]

for OA I made a cvar that slows down the game with a huge for loop that is amplified by on-screen rendered indices.

Pointless for gaming maybe, but there is a use for it in mappers for map optimization to find choking points in areas since r_speeds is not illustrative enough. and boy does OA have a lot of places that slow down

[hogsy]

More work towards the Linux port of OpenKatana.



Other than some obvious bugs, everything's working now