PvP in the zombie apocalypse. Limited resources. Endless zombies. Last team alive wins.
Deadlock Repathing Demonstration
Post a comment
Description
This video discusses our plans for dynamic multiplayer environments.
Concept by Ryan Timoney and Eric Bellis
Video, Map, and Voice Over by Andrew Gordon
Audio Mixing by Tony Oetzmann
Graphics by Matthew Kibbler
Music by Vincent Girès
Full Resolution Version can be found here: Moddb.com
PROCESSING!!
I plan on uploading this in a higher quality ASAP
Okay, this is awesome!
Nicely done.
I'm watching this! This look very promising.
Very similar to what was done in cs_havana.
And Day of Defeat too I believe? Or maybe it was just the objectives which changed.
L4D maps are dynamic, theres no need for real-time repathing.
Whereas this demonstration displays maps such as CS:S maps, where players will encounter the same area multiple times in one session.
It seems that making multiple maps would make more sense, the distance of these interchanging paths wouldn't make too much of a difference when the path matches the rest of the map...
Although, this idea would be excellent in a singleplayer game. The thought that the game will be different every time would encourage multiple replays.
I apologize for not commending this creative idea, but the variable pathing has been used in (as stated above) CS:S, and I suggest you look at this idea rather then real-time.
Multiple triggers at round-load would make each round different. It just makes more sense.
Good luck with this mod.
While L4D maps may be slightly dynamic, they in no way approach the size, scale, or utility of this system. This is being built for a game in which you are either defending yourself from zombies, hunting down the enemy team, or searching out resources. Players will have no choice but to wander down alleys and search out buildings if they want to survive. The real horror aspect of the game will come from these lonely missions out into unknown territories. Couple that with ambient, lone zombies (that are stronger than the weak L4D ones) and randomized pickup placements and you've got yourself a fun, replayable multiplayer map.
We realize we aren't the first people doing this, but I think repathing will have a much larger role in our gameplay than in previous titles.
You understand the expanse that the map would require for this? Most of the time, a good map will have an open view at some point, whether it be windows, higher altitudes, et cetera. How will these changes be hidden from the player?
Again, I like the idea, I'm just trying to provoke some contrasting thoughts.
Also, is this video an in-game demonstration using the Source engine?
Hey, fresh eyes are good. As noted in the video, we're primarily looking at a round based repathing system. Deadlock is basically a team death match. When one team is wiped out, the round resets, everyone respawns, and the map will repath.
Deadlock maps will not be linear, and will have more of a large arena/open world-ish feel. As for the scale of the map, we won't know what sizes work best until we've at least got a playable alpha. I suspect the maps will be complicated from a production point of view, but I think repathing will allow us to get more mileage out of the space we do use.
And yes, the video is shot in Source. :)
AWESOME!
You have a lot of work to be done on this, but this is really cool. If only the original L4D had the same time as you to get this. This is a really nice system to be adding into this, now just to get the AI to work.
What about the paths that the zombies go? Don't you have to draw the spaces that are walkable? You do that in consol, in game, What if you change the layout of the map you wont have zombies able to walk there.
Please explain how you do that! Unless I have no idea what im talking about... You have a real problem!
We aren't using the L4D nav system. We will be using a modified version of the original HL2 pathing system, which is more basic but allows for more flexibility.
Extremely great job, i was seriously amazed.
It's a brilliant idea, hope you guys get it up and running!
Can you do the same for L4D? I mean, the pathing wouldnt matter since the infected can climb all over everything.
The survivor bots need a working nav_mesh, infected just need ANY nav_mesh.
I'm curious if you pulled this basic idea from any of my posts to the L4D Steam Forums, or possibly my write-up on the Valve wiki.
While I had pondered in-level changes, I hadn't done anything to this degree yet. Nice work. :)
Heh. Nah. I've been fond of procedural content since I saw it in Diablo. The Tribes guys figured out how to do procedural terrain... it's been in Hellgate London and RoboBlitz... At any rate, it's a feature that hasn't matured yet in the FPS and we'd like to tackle it. Definitely a next step for the medium.
One interesting way to make use of this type of map would be a crescendo event where the players have to perform multiple actions at different locations while under attack. Say they enter a warehouse area and, after wandering around it for a while, several fire alarms go off, calling hordes. The hordes continue to attack in endless waves until the players find all of the fire alarms and shut them off or destroy them. The fire alarms may always be in the same place, but the path to reach them would be continuously changing.
"Repaths" has been used in alot of valve mods I have made maps my self with your so called "repaths" all it is logic_case with a input of pic random and selects a outputs at random to a point_template to force spawn objects but you also have a delay on that. It also has a output from the logic_case to a entitie (probably a logic_timer)to delete that entite. Also from the logic_case you have it highlight the wall that was selected. Those triggers you speak of could be any number of triggers but they have outputs to the logic_Case which starts it all. Really you can use this method to spawn any entities in the map randomly based on events but that requires mapping skill and it can cause lag if you are missing any part of the half life 2 source files because it creates error and valve releases their source code with missing code.
Forgot to mention 1 more thing....
This is a good method but like I said it has some problems with some mods because some mods are missing those lines of code that are for entities and registering them. Specially with the orange box engine source files so if you can do not use the orange box engine code that valve released and if you are try to find out what is missing and try to fix it other wise there will be massive bugs.
VALVe will probably hire you because of you creating this system, as they did with the guys from Nabacular Drop ---> P0rtal because of the portal system :)