Tremulous Forum
Community => Mod Ideas and Desires => Topic started by: floodbud on January 24, 2007, 06:41:26 pm
-
Hello everyone, these mods are nice additions and everything, but seriously, why are ther no AI mods??? :evil: If anyone is working on one, keep going. Even a basic AI mod (aliens move to human and click rapidly, humans aim at aliens, move random, shoot) would be wildly popular. Or even just some still aliens for target practice...personally I don't know enough about Tremulous source to create a mod myself but I will help any way I can.
-----
Floodbud, owner of the seven-foot tyrant jump
-
Making an alien AI is way to hard to be worth the effort.
Humans could be done, but there is no point with out alien.
Just like it says in the FAQ.
-
well, for all of us who suck reasonably at aliens there would be a huge point, a human AI that we could train on, or for a clan training match it could be clan = aliens, ai = humans
-
With current popularity there is no player who is capabel to build proper ai.
-
NO
-
It's possible but difficult. Here is what you need:
1.) Quake BSP pathing -- see Quake/ioq3 source
2.) Equipment and combat strategy -- attack pattern scripts, simple "walking turret", run-at-and-bite (see Counterstrike)
3.) Generalized building strategy -- priority scripts, zones of influence (see Starcraft etc)
Need I mention good bots exist for Natural Selection?
-
Would maps need to be edited to provide bot support or could they just be programed to do certain things around certain map features?
EX:
Long Hallway = Sit and shoot as human, avoid as alien
Door = possible good building spot for either team, good ambush point for aliens
I think you get the idea.
Possibly the hardest thing for human bots would be making them NOT be aimbots.
-
Possibly the hardest thing for human bots would be making them NOT be aimbots.
i think that would possibly be the easiest thing :/
maybe u can make a camper bot, sits next to the armory, upgrading whenever possible, and occasionally blowing up the base wht grenades.
-
Would maps need to be edited to provide bot support or could they just be programed to do certain things around certain map features?
This was done in Quake 2. Map authors can make the bots smarter by hinting out paths etc, maybe makes the server a bit more efficient by precompiling a few things. It's a bonus but not necessary.
-
maybe u can make a camper bot, sits next to the armory, upgrading whenever possible, and occasionally blowing up the base wht grenades.
That is an art form. Bots will never be as good at it as people.
-
How can you path the wallwalking? That would be impossible to program bots for.
-
How can you path the wallwalking? That would be impossible to program bots for.
No, just fiendishly difficult.
-
It's possible but difficult. Here is what you need:
1.) Quake BSP pathing -- see Quake/ioq3 source
2.) Equipment and combat strategy -- attack pattern scripts, simple "walking turret", run-at-and-bite (see Counterstrike)
3.) Generalized building strategy -- priority scripts, zones of influence (see Starcraft etc)
Need I mention good bots exist for Natural Selection?
To generalize it further from an AI perspective, there would be 2 types of AI for Tremulous, combat bot and building bot.
A killbot would ideally do several things, from easiest to hardest:
1. Recognize and attack enemy players and structures.
2. Navigate maps without getting lost.
3. Attempt to satisfy the objective of the game, i.e. find and kill the enemy base.
4. Attempt teamwork with fellow bots and players alike.
A basebot would ideally do several things, from easiest to hardest:
1. Navigate inside friendly base.
2. Understand Generalized building strategy -- zones of influence ( :reactor: :overmind: ), choke-points ( :trapper: :turret: :tesla: :barricade: :hive: :acidtube: ), power needs ( :repeater: :egg: ), convenience vs. safety ( :telenode: :medstat: :armoury: :defcomp: :egg: :booster: ), etc.
3. Prioritize which buildings are more important than others.
-
Also, a builder bot would have to know how many spawns is sane for the team it is on. Imagine a bot deconning everything because it thinks the team needs more nodes, eventually using up all of the build points. It might be better to just make them build pre-defined bases, until someone can figure out how to make them think up their own designs.
-
I did say ideally, and those instances would be covered in prioritizing the BP.
-
To generalize it further from an AI perspective, there would be 2 types of AI for Tremulous, combat bot and building bot.
The nice thing is, while these two bots share much of the same plumbing and in a real game would switch between roles, you can create a combat-only or a builder-only bot for a start.
1. Recognize and attack enemy players and structures.
This is probably easiest for humans. The trick is often knowing when you can attack and how to do it without getting killed.
2. Navigate maps without getting lost.
More like navigate the map without getting glitched. :-?
4. Attempt teamwork with fellow bots and players alike.
The NS WhichBot actually forms attack "packs" with other WhichBots. A simpler alternative would be to just follow a human player. :)
2. Understand Generalized building strategy -- zones of influence ( :reactor: :overmind: ), choke-points ( :trapper: :turret: :tesla: :barricade: :hive: :acidtube: ), power needs ( :repeater: :egg: ), convenience vs. safety ( :telenode: :medstat: :armoury: :defcomp: :egg: :booster: ), etc.
Bots would need to know the map layout and the starting positions of enemy buildings otherwise they would put all the turrets facing a closet or something similarly stupid.
3. Prioritize which buildings are more important than others.
A bit of enlightened scripting would do the trick here. :)
-
spawn places in trem are dynamic
and you cannot path everywhere on the map
so you all fail
unless you managed to complete complex path finding algorithms in 1 ms
-
spawn places in trem are dynamic
... but still visible to server-side bots. If it comes to it, making the bot cheat would be pretty easy. :)
and you cannot path everywhere on the map
so you all fail
unless you managed to complete complex path finding algorithms in 1 ms
You don't have to do it every ms. There are typically 50 ms per frame for one thing and you don't need to build the map nodes every frame or even in-game for another. You can build path graphs at map start for maps that are not pathed. You would have different paths for each class presumably.
-
3d path finding algorithms that evaluate in <1000 ms is not yet widely available
I suggest brain reading
-
3d path finding algorithms that evaluate in <1000 ms is not yet widely available
I suggest brain reading
This is wrong, wallwalking is not true 3D pathfinding because you are still constrained on a plane :) Thus, a simple graph/node based navigation should do the trick. It won't support advanced tactics like jumping from one wall to another though.
-
3d path finding algorithms that evaluate in <1000 ms is not yet widely available
I suggest brain reading
This is wrong, wallwalking is not true 3D pathfinding because you are still constrained on a plane :) Thus, a simple graph/node based navigation should do the trick. It won't support advanced tactics like jumping from one wall to another though.
yes if you say so I think you know path finding better than me like everything else
oh btw did I mention 2d pathfinding is used on 2d images and if you cant project something on a plane it wont be 2d. But you are so leet you can do it easily
-
I think making an AI learn its things, is gonna be harder then u think. However i would agree with the idea of making the bots... as 0din said, fiendishly hard... Because you need to make a superior bot, that learns all the moves of a player, and expects where it will be next... That, and you would have to tone its aim down, so it doesnt get too smart for the player, and the bot would be unkillable... lol
-
mappers could use key for buildable to indicate stage to build each object, so the bases would be designed by a human but built by a bot
-
If there are human players on the map, build a map-node-grid from them.
A human will automatically build the bots map-node-grid as they wander around.
It might be a pretty lousy map if the player had never seen the map before.
and the wall walking problem is overstated.
Only dretches really wall walk S1, and even then all other aliens (except baslilisk, adv. granger) are ground based.
Considering the number of 'cheap' tactics a Alien-bot could employ, even if they can barely pathfind they will be VERY difficult adversaries.
Cheap tactics like:
Smart-Straife: Move left/Right when human is aiming gun at you.
'Perfect' Pounces (could be tricky?)
Flawless judge of range and aim.
0 Fear.
Sees you in the dark.
Auto-Sneak, when a Bot is behind a human and the human has NOT seen the bot.
-
If there are human players on the map, build a map-node-grid from them.
Thats interesting, though obviously with problems. The bot basically follows the other players around.
and the wall walking problem is overstated.
It's not necessary for the bots to wallwalk, they are playable without ... if a little dull. :)
Cheap tactics like:
Smart-Straife: Move left/Right when human is aiming gun at you.
'Perfect' Pounces (could be tricky?)
Flawless judge of range and aim.
zomg h4x! not to mention clairvoyance for server-side bots?
0 Fear.
Actually you do want fear. If you are dying consistently at certain areas by, say, turrets or MDers you don't want to keep running back and feeding. Bots would become temporarily "afraid" of those areas ... the effect wearing off with time.
-
Ahh, a fight/run logic structure.
Run from bigger aliens, unless you know what you are doing.
Basilisks run far away from flamethowers and Lucifers.
Tyrants run away from a couple chain-suits.
-
GOOD ai would be impossible, BAD ai would be possible.
-
And how hard would bad AI be to make?
And it wouldn't be that bad, compared to all the Unnamed Players.
In fact, in some areas they would be better. (Aim, for example)
-
And how hard would bad AI be to make?
And it wouldn't be that bad, compared to all the Unnamed Players.
In fact, in some areas they would be better. (Aim, for example)
and they wouldent decon the reactor right before you reload you chaingun during a tyrant rush. they wouldnt luci or nade your base either
-
@vcxzet
3d environments can be represented as graphs just like 2d ones. the only drawback for pathing algos like A* is that 2d heuristics are more efficient. however q3 pathfinding doesn't even use heuristics.
the amout of wall and ceiling areas is ~ 500 - 2000 per cluster. the botlib pathfinding sure can handle this, team arena and world of padman maps have much worse areacounts.
so 1000ms per route is as unrealistic as 1ms. even if it took more than 40ms, the algo could still be time sliced.
if anyone decides to go about bot development, i'd like to help with the aas part.
-
if anyone decides to go about bot development, i'd like to help with the aas part.
Timbo, in his infinite wisdom, removed the bot code from Tremulous in SVN. Before you do anything you need to either add it back or find a way to do without, either way talk with Timbo and decide on how to do it beforehand.
-
i know. the 1.1 source has at least the botlib and the server part of botlib trapcalls. all it took to init the lib and load the aas map was to add some traps to g_syscalls.c
if the botlib was removed from svn within one reversion step (and nothing else was done), then it will be easy to restore. thats what svn is good for :)
one important thing i forgot about the pathfinding issue: all classes have different bounding boxes, so they require seperate aas files, so cached pathes can be used only by the same class... means a lot extra memory and extra cpu time for the server :/
-
i know. the 1.1 source has at least the botlib and the server part of botlib trapcalls. all it took to init the lib and load the aas map was to add some traps to g_syscalls.c
if the botlib was removed from svn within one reversion step (and nothing else was done), then it will be easy to restore. thats what svn is good for :)
one important thing i forgot about the pathfinding issue: all classes have different bounding boxes, so they require seperate aas files, so cached pathes can be used only by the same class... means a lot extra memory and extra cpu time for the server :/
plus you dont need ai for a multiplayer game
( I know even the most crappy ai implementation would be smarter than some human players ; sad )
-
As I see it the builder could be set out like an rts ai the dretches,goons,mara's,basi and tyrants like standard q3 bots but then you must take the 2ndary attacks and wall walking
-
Maybe we aren't understanding this. Any Tremulous AI would be largely ignored, because Tremulous plays best as a multi-player game. The point would be to create some bots and customizable scenarios so that the noobs could work themselves up to pro without being kicked from 83 servers first. I remember how difficult it was to get to pro when I started.
But if anyone is a super-genius and wants to take it up to the level of campaign-style bots that act like real players, go ahead.
For now, the best way to do it is to split the labor. Since no one has any better suggestions, here is how we will split the AI programming, for all those interested. By the way, alien AI will be much harder, which is why it takes up the bulk of the programming.
(Divided by month of birth):
January: Dretch AI
February: Human Lucifer Cannon AI
March-April: Basilisk AI; great programmers, add Adv. Bassy AI
May-June: Goon AI and Human close-range weapon AI
July-August: Adv. Goon AI
September-October: Maurauder AI; just add a little lightning for Adv.
November: Tyrant
December: Human long-range AI
:roll:
Good luck, everyone!
____________
Floodbud, owner of the 7.65 foot Tyrant jump
-
And how hard would bad AI be to make?
And it wouldn't be that bad, compared to all the Unnamed Players.
In fact, in some areas they would be better. (Aim, for example)
and names. Bot 1 is better than UnnamedPlayer
-
:D By the way, I'm pretty busy most of the time, anyone want to head up the AI project? We would all send in our code to you. I'll approve you if you post a good reason on this thread. Not like I'm the master of the world or anything, but someone has to choose a leader.
Hey, good luck, anyone interested.
__________________
Floodbud, owner of the 5.87 foot Tyrant jump
-
Yeah sure, a guy with 16 posts on the forum with visibly no idea how AI work and who is 2 months late before even planning anything (may i recall you January was 1 month ago ?) is definitly the guy i would choose as a team leader. :roll:
floodbud, McArtman, coding an AI is NOT easy. It s a whole area in computer science. And Tremulous has a lot of additionnal difficulties. You wont get AI for Trem, it s not worth it (as you said it wouldn t be used much). Get over it and stop making useless posts.
-
:evil: Ksempac, we don't need your pessimistic tongue-waggling. The poll says that most people agree, good Tremulous AI is possible. If you don't want to do it, you don't have to try. The rest of us, who can see how it would work, will continue.
Your discouraging remarks will never do any good, and are just a waste of space.
And as for January, you must have no idea how the forums work or something, because you forgot to read the part where I clearly stated that the workload is split based on month of birth. For example, if you were born in January, than you would do the January section. If you need more examples, show this to your three-year-old sister and ask her. Seriously, get your head in the game.
I am not running for team leader! :x I already explained that! I am only looking for someone to head it up. EASY job, team leader, just look at code scripts sent to you, choose the best from each category, compatibilize them (as far as variables, etc.), and throw them together as a mod. I'm still looking for candidates.
I'm aware that Tremulous AI would be very difficult to do. I am not looking for perfect, smarter-than-human AI, only something sketchy for target practice and strategy development. It doesn't need to be perfect AI!.
All right, everyone interested keep up the good work. It can definitely be done. Go for it!
And everyone else, shut up.
_______________
Floodbud, owner of the 18 foot tyrant jump
-
I think it would be best to start with the human A.I., because that's possible to program. I have 0 programming skills (yet) and can't help you with this though i suggest that starting with one side first will make things easier. I just feel srry for the dretches who walk upon an A.I. controlled md user...
-
I agree with you floodbud. I see great potential in this mod. It would help beginners with training(cause it was supposed to a singleplayer option, true?) I only fear that people will hack it and use it in a server as a aimbot. If I had knowledge programming I would volunteer to help making this mod, but tobad I don't have this skills. (also add different difficulty levels like Easy, Normal, Hard)
Please, make this possible! :D
-
The poll says that most people agree, good Tremulous AI is possible.
Yeah sure. If the poll filled by random users of this forum (with unknown programming skills) say that programming a good AI is possible, that must be true. What i see is that people who knows what they are talking about (vcxzet, Risujin,...) are all saying that this is way too hard to be worth it.
And if you want me to show you one more proof that you dont know what you re talking about :
Dividing work by saying "each coder will do one class" is DUMB.
AI has several generic functions : Path-Finding, Aiming, etc... So many AI can share a lot of generic code.
For example : A close-range weapon AI would have the same path-finding as the long-range one or a lucy one (btw why did you separate that one ?). A Tyrant would have quite the same , you would only needs to remove use of ladders.
On the contrary, you re saying "everyone will make a path-finding algorithm for each class, because doing the same thing 11 times when there are only 3 or 4 different algorithms is the best way to do things"
A smarter way to do it would have been :
"i will do the generic path-finding algorithm for humans, you will do the special path-finding for wallwalkers, etc..." Although i suspect we could even have only one algorithm by implementing waypoints of differents values "walkable" "wallwalkable" "ladder" etc...
Understand me well : I would LOVE to have AI in Tremulous so that i could show it to people without Internet access or play in LAN with some friends without needing 16 computers in the same room.
But, as a programmer i see far too much difficulties for that.
-
Afaik the trem devs have said THEY won't be coding an AI system for trem but that if somebody is willing to code it they will have a look at it and maybe add it into trem. Unless one of the people here asking for it steps up and codes it and submits it to the devs it won't get implemented.
-
Sorry I got mad at you ksempac. I see what you're saying now, except I still firmly believe that good tremulous AI is possible. If you can't understand the programming part, that's your problem, not mine. Personally, I know several programming languages well, and I may tackle some AI work myself over the summer. And you still don't seem to understand that the AI doesn't need to be perfect. It could be as drunken, imperfect, and retarded as possible and still be a perfect way for noobs to work themselves up to pro.
I also understand the worry about players using a hack like this to really mess up internet servers. This problem is solved with imperfection as well, because the AI will be very stupid, and if anything, just feed.
Again, it just needs to be adequate enough to provide skills like goon-dancing and targeting. :)
Time is not an issue here. If it takes years to do this, (1) it will be worth it, and (2) it will just be for fun and something to do. We don't need halo-standard AI or anything.
I still encourage anyone with appropriate skills to work out algorithms such as the ones ksempac mentioned, and someone to volunteer to head up this project.
P.S. The AI should probably be worked out initially to work for ATCS, bells and whistles like ladders and wallwalking can be done later.
------------------
Floodbud, owner of the 2 foot tyrant jump
-
I still encourage anyone with appropriate skills to work out algorithms such as the ones ksempac mentioned, and someone to volunteer to head up this project.
there are very few people with the appropriate skills, and even those who have those skills would have trouble writing an AI for tremulous because of the large time investment involved, you aren't going to find someone who can do this by posting a forum thread, because if a person had the time, the knowledge, and the interest in undertaking such a project, they would already be working on it.
-
just get some slaves and make them play trem with you everyday
-
The thing about AI is that they are going to be very very very stupid if we try to to put it on a game like this. There are going to be two modes: Suicide run, which is really no better than the usual feeders. OR we have ones that camp a base for the whole game, and we don't need more of that. Human's movement will most likely walk in the middle of the hallway, and though doors towards the alien base no matter what besides firing on targets that they come across accidentally., even if there is no alien base. Aliens same thing, but they wont be wall climbing or doing fancy wall maneuvers.
-
too much talk no work
-
too much talk no work
How many people on this forum actually do work?
-
I understand that very few people are the kind of people that would work on this. Believe you me, I know enough programming languages to know the difficulties of AI programming.
Finally someone gets it. Thank you Nagato!!!!!!!!!!!!!!!!!!!!!!!!!!! :D
The Tremulous AI would be very stupid. That's the point. It would be for TARGET PRACTICE as I have said many, many, many times. It wouldn't have to be intelligent at all. I just want a bit of training for n00bs to come about. It wouldn't be aimbots, and ideally only the server could institute it. That way there would be no cheating on well-meaning servers.
Plus, some algorithms (most for the humans) could be cut-paste from machinegun turrets. The way they work is ideal: They swivel/aim slowly and shoot fast. If humans used this, it would be totally fair. As for the aliens, basically aimbot code could be used except the aliens would move very slowly to make it fair. See what I mean?
As for Risujin's comment, I appreciate the pointless pessimism. I will try to tackle some of this myself over the summer but I will suck at it alone.
___________________
Floodbud. Owner.
-
hey, why they say that tremulous is a bias to humans? the developer "maybe" thought of that already... i think no game is bias to 1.. or watever.. the point in tremulous is, 1 human, 3 aliens i think that was there point of this game.
In AI perspectives:
if there would be AI aliens, this creatures likely knows best of the MAP..
[3 warrior aliens + 1 builder] x human, this [3+1] would be better to go out rooming around as a group, protecting the alien builder, AI making is very hard, it is very hard, not very very hard, but just very hard, get my point, having intelligent bots is not impossible at all.
The MOD they say about Troopers/Star ship troopers they say? maybe it would be cool, if that would be, 1 vs 1 would do.
C'MON NUGGITS, HIT THAT ALIENS IN THE HEAD!!
-
:-? I don't quite get your response Ifs_id. And please keep the reply to the thread topic.
In AI, there would be no builders:
FOR THE LAST F***IN TIME, Tremulous AI would be TARGET PRACTICE, NOT SUPREME AMAZING GENIUS AI!!!!!!!!!!!
Thank you.
And there is definitely no bias in Tremulous :o . The better player wins, regardless of team.
__________________
Floodbud aka opwn3nt
-
Regardless if we can or can not make AI, its is completely true that NO one is going to make a "smart" AI like DOOM 3, or the news and best FPS game to date. Go play some StarCraft/ Brood War, and go play with the computer, you will find that there will happen here if we try...maybe even worse.
-
Yes no one will make a F@#<!NG good AI because no one puts $ to it. Make an Amazing AI for what? enjoyment?happiness?famous? somebody will spend time for nothing. no!!!! cut the crap just play tremulous w/ or w/o AI its still the game that matters.
You know guys, what ever war you do in the forum, GM/Game Author of this game will just laugh at you.. just play the game instead, be happy for what it have and dont have. the thing iv said above was just my "VIEW" some idea.
-
as I under stand it the human ai would be simple to code for the non building part, the selection of starting weapons/alien could be like in the Quake III Mod Urban Terror or the Soldier of Fortune II bot mod,they a bit like CS have a weapons selection system no buying,that could be implemented for class or weapon selection, going to the armoury would be hard (I think),then you would have to implement the ai for attacking/feeding, which might be hard,but the build system would be very very hard.
-
You don't really need a build bot. If you're going to practice, use layout and give both of your a base or so.
I'd like some target practice. :D
-
Idea: make training server and make the bot human. all people who join the server must join aliens and try to eat the bot. but make hummie huge and give him 5000 health or somthing.
It would be fun to have a End boss!!!!!
Make him move slow. How more people connect how powerful it will become :D
All on Saturday night END BOSS FIGHT!!!! :D
28 aliens attacking huge human LOL :D
-
Finally some positive input. Thank you.
Ifs_id, of course there is no pay for this (*although some sites try to sell Tremulous). We work on the same principle as Linux: To put in what we can for the common good. If we fail, we lose nothing. If we succeed, then we gain some amount of limited fame, many happy players, and some degree of self-satisfaction. If nothing else, we do it because it's there.
Baxter, finally you see my dream. There would not need to be any builders; they don't provide very exciting targets. Remember, this would be solely for very bored people without an internet connection, or noobs working themselves up to pro, or for modders who want to test their mod and can't do it alone (i.e. need other players to be able to stage up). I know raytray25 said the same thing (& t/y) but I just wanted to confirm it.
E-Mxp, sounds like fun, but not quite what we're focusing on. Maybe you can do that with the template the rest of us provide.
_______________________
Floodbud, Owner.
-
E-Mxp, sounds like fun, but not quite what we're focusing on. Maybe you can do that with the template the rest of us provide.
:( Sorry, I don't know how to do this kind of stuff. Just hoping somebody els will. :D
//Idea: Less hard to do(??) Give Human bot 1hp!! :D :granger:s would try to spit at it :D Battle[/size]Granger[/size]Galore!!![/size]
-
Apology accepted 8) .
Like I said, I'm just going to focus on general AI, and strictly that, without going deeper into Trem code.
______________
Floodbud
-
If we were to have AI, they better not count towards the kill count. Their ignorance can not be forgiven. :evil: LOL
-
Apology accepted 8) .
Like I said, I'm just going to focus on general AI, and strictly that, without going deeper into Trem code.
If you wont go deeper into Trem code, don't dream of coding an AI for it. If you don't understand at least 50% (estimate) of Trem code, the probabilty of you succeeding is ~0. I do support the notion of creating a dumb bot feature though (running in perfect circles and eights).
-
Exactly. Just what I want. If you can do more than talk about supporting it, let me know...it could be as much as six months before I am able to work on implementing that.
_______________
Floodbud
Bubba sez:
Eat lots of donuts.
-
Could the q3 bot code simply be copied and pasted in with minor "dont shoot teammates/things" additions, at least at first? A rifle with no armor, but being able to know where everything is with more accuracy than a helmet and being able to shoot more accurately than your average md, would be a difficult target.
EDIT: Another use for bots would be to give SOME competition on an otherwise empty server.
-
I have managed to put back in the botlib from SVN 777 against my copy of SVN 901. It compiled and ran fine. But there are no syscall traps hooking into the botlib at all as of now. That will be in my next plan. Don't expect too much :)
-
I skipped most of the thread so sorry if this has already been mentioned (might also have come up in the tutorial mod thread)
But what about simplified bots, that say only do one class, and could be played on simplified maps? The point of this would be to train the player. For example one 'level' could be the player shooting a dretch as a naked human. Another level could have say 2 dretches and a goon, with a squad of humans backing him up, etc.
-
All botlib syscalls are back in and working fine. Bots can be added and join alien/human team. The behavior isn't implemented yet, so they don't spawn just yet. I'll report here when I have the first bot to spawn.
-
All botlib syscalls are back in and working fine. Bots can be added and join alien/human team. The behavior isn't implemented yet, so they don't spawn just yet. I'll report here when I have the first bot to spawn.
Post a patch when you've got something.
I think, actually, a different approach might be easiest. Here is my idea: Create a new class for the AI (yes, give it its own model too!) on humans and aliens. Because the AI is very likely to be retarded, give this new class advantages to balance out the deficencies. If all AIs have the same size, you only need to run a pathfinding algorithm on each map once (though botlib does this, no?).
For the human team you could probably create, without too much hassle, a kind of walking turret. It'd be capable of using a few weapons (hitscans probably), and would run some kind of simple strategy. Give it lots of HP and make it fast and it might even be useful. ;)
-
Give hummies drones (See David Lynch's Dune)
-
Give hummies drones (See David Lynch's Dune)
doesn't really have to do with Human AI now do we?
But yea, that'd be cool.
But didn't people already talked about that in mod?
-
Sounds good everyone. Sleekslacker, let us know when you have something. Fourth quarter in school will end in about 8 weeks, so I could do a little work then if necessary.
______________
Floodbud (Opwn3nt in the game)
-
Give hummies drones (See David Lynch's Dune)
doesn't really have to do with Human AI now do we?
But yea, that'd be cool.
But didn't people already talked about that in mod?
he meant AS the human AI...someone suggested giving the ai different models, the human's AI could take the form of a defence/offence/medical drone.
-
if you wont do anything, please stfu
-
we are just discussing. and the OP doesn't make mods too. He was just asking why.
-
if you wont do anything, please stfu
Though I kind of like this attitude, I think it is important to suggest and discuss things before undertaking them...that's what separates man from the apes. Don't be an ape please; if you have a constructive comment, post it.
___________
Floodbud
-
if you wont do anything, please stfu
Though I kind of like this attitude, I think it is important to suggest and discuss things before undertaking them...that's what separates man from the apes. Don't be an ape please; if you have a constructive comment, post it.
___________
Floodbud
well discuss ... but you wont get anything by discussing. try and see.
moreover if someone was about to program ai. he wont even look at the this thread. because this thread does not help programming ai. But hey just entertain yourselves with lies
-
well discuss ... but you wont get anything by discussing. try and see. this thread does not help programming ai. But hey just entertain yourselves with lies
Party pooper.
___________________
Floodbud
-
Target practice!!!
And AI will probably end up with smarter team-play once they implement vchat.
-
if you wont do anything, please stfu
Though I kind of like this attitude, I think it is important to suggest and discuss things before undertaking them...that's what separates man from the apes. Don't be an ape please; if you have a constructive comment, post it.
___________
Floodbud
well discuss ... but you wont get anything by discussing. try and see.
moreover if someone was about to program ai. he wont even look at the this thread. because this thread does not help programming ai. But hey just entertain yourselves with lies
Actually, i agree with the troll. I'm posting here since you guys started talking about AI and stuffs. So be informed that I'm working on one but I can't guarantee anything useful will be produced out of it. This kind of discussion is pretty much useless unless you can tell me where in the old quake3 code the spawning commence and stuffs. In other words, look into the code and maybe start by documenting stuffs if you want people to join you as a group.
-
Despite all of Floodbuds and powerover's bickering, lets get back to the topic, I think that could be quite simple. Human, sit in corner, aim shoot. Alien, look at head run aimlessly.
-
Maybe we aren't understanding this. Any Tremulous AI would be largely ignored, because Tremulous plays best as a multi-player game. The point would be to create some bots and customizable scenarios so that the noobs could work themselves up to pro without being kicked from 83 servers first. I remember how difficult it was to get to pro when I started.
But if anyone is a super-genius and wants to take it up to the level of campaign-style bots that act like real players, go ahead.
For now, the best way to do it is to split the labor. Since no one has any better suggestions, here is how we will split the AI programming, for all those interested. By the way, alien AI will be much harder, which is why it takes up the bulk of the programming.
(Divided by month of birth):
January: Dretch AI
February: Human Lucifer Cannon AI
March-April: Basilisk AI; great programmers, add Adv. Bassy AI
May-June: Goon AI and Human close-range weapon AI
July-August: Adv. Goon AI
September-October: Maurauder AI; just add a little lightning for Adv.
November: Tyrant
December: Human long-range AI
:roll:
Good luck, everyone!
____________
Floodbud, owner of the 7.65 foot Tyrant jump
You are a complete idiot. You are nowhere near a pro tremulous player. And what makes you think because you post you want an AI and you give the schedule to work on it that everyone will just be like ok and get started. You ignorance is beyond that of which I have ever seen. Go get a life.
-
All botlib syscalls are back in and working fine. Bots can be added and join alien/human team. The behavior isn't implemented yet, so they don't spawn just yet. I'll report here when I have the first bot to spawn.
Post a patch when you've got something.
I think, actually, a different approach might be easiest. Here is my idea: Create a new class for the AI (yes, give it its own model too!) on humans and aliens. Because the AI is very likely to be retarded, give this new class advantages to balance out the deficencies. If all AIs have the same size, you only need to run a pathfinding algorithm on each map once (though botlib does this, no?).
For the human team you could probably create, without too much hassle, a kind of walking turret. It'd be capable of using a few weapons (hitscans probably), and would run some kind of simple strategy. Give it lots of HP and make it fast and it might even be useful. ;)
Say, Ris, sorry to butt in since I'm not going to actually code anything, but:
This is a really hip idea for a n00b botmatch thing. Somebody else mentioned medical drones. So, thinking along the lines of the Deus Ex drones, you could have various classes. And, Quake3arena already has the proper bot behavoir (Giant Eyeball, except it would be looking for the rocket launcher and the flag). :P
Thematically, you could play as either Human or Alien against it, because Humans would use them to train, and Aliens would use captured ones to train also, or just meet bots in forward areas where people haven't been sent yet, or in areas that have already been "secured", so the Humans left the bots in charge again. It's a trainer: Humans v. Aliens would only be happening at "The Front"; that is, in contested areas where active Alien presence requires direct Human intervention.
Since it's mostly about attacking and using weapons against a fairly dumb bot, you could have various classes, no buildings except spawn points (Nodes or Eggs) and Armouries (Aliens would just evo up), and medbots instead of Medistations. You would only be able to join one contingent in any particular match (Humans vs. Bots or Aliens vs. Bots). Bots could be about as smart as they were in Deus Ex (two kinds: lumbering but strong; quick and deadly). Medbots as usual would be like a dog that won't heel. I know this doesn't help much, because the other places to get Open Source models are the wrong engines (Deus Ex is Unreal and it's not open source, afaik; Stargiege is open source, but different engine). Perhaps some Quake mod somewhere has an appropriate open-source model? Didn't vcxzet/optear already import the Giant Eyeball and Lara Croft model into Trem maps? Giant Eyeball would be the Attacker and Lucy/Lara would be the Defender.
I know. It ain't Alien v. Human. And I can't really help much.
Bot classes:
Medbot - roving medistation
practically same as Deus Ex, you gotta find the health
Defender - slower, powerful (defends "enemy" OM/reactor)
in DX: the big drones; in Q3a: Lucy
Attacker - quicker, deadlier (attacks players, deathmatch-style)
in DX: those little mean buggers; in Q3a: the giant eyeball
Medbot wanders aimlessly heals only Humans (Aliens regenerate)
[Possibly damages Aliens with tesla/reactor attack]
{Yes, much like an Adv Mara, a teammate, I know}
Defender uses more Stage 3 Human weapons (chainguns or lucis)
[Yes I know chaingun is S1, but few can afford to die with one until S3]
{What I mean is slow weapons you move lumberingly in a battlesuit with}
Attacker uses what Ris said, the hitscan stuff (iirc lasgun, rifle, etc.)
[Generally, all of the Light Weapons or Fast Weapons]
{What I mean are all the other stuff except what the Heavies use}
Paint 'em all Human Blue (or make 'em Alien Red in Human v. Bot)
Call 'em Spot (Medi), Heavy (Defbot), Meanie (Perpetrator)
OK, since I can't do any of it, I'll stfu & gtfo. :P
Thanks. And have a lovely Trem day. :D
Oops, sorry, edit: Make them all them same size, just like Ris says (yes, I missed it, and I'm too lazy to edit the entire thing). Just armored and powerful for defenders and quick and dangerous for attackers. And hard to find for Medibots. :roll:
Player vs. Bots
not Human vs. Alien
but as Alien you have to destroy a Reactor
and as Human you have to destroy an Overmind
which some bots defend
while others attack you (but maybe your structures, too?)
no builder classes?
teambotmatch/assault/mission/coop-style play only
like a coop SP level
or assault MP map
evade attackers
destroy defenders
accomplish SP/coop mission
default maps & Aegis & the ATCS & UTCS variations (the training maps)
so sorry if that's not what you meant/want/thought/were trying to say
thx agn
k c-ya
edited for clarity
yes, i've got nothing better to do right now
-
You are a complete idiot.
Listen powerover, stop flaming me and only me.
Has everyone noticed how every single thing I've posted, powerover has attacked me on it?
Obviously it's something personal, so I suggest you either laugh at his n00bish nature or ignore him.
He actually isn't that bad once you get to know him...he's worse.
_________________
Floodbud
-
Okay, powerover and floodbud should be ban, I know I am being a little hypocritical when I am saying this. Tremulous is a game, games are made to have fun in. We are on a forum for a game, forums are to learn new ways to have fun playing a game. Simple, so both of you since all you do is spend your time insulting each other, just leave and go sit in your self pitty.
Back to the original AI, I am not total genius in any language, but instead no some of most. Trem is based off of Q3, Q3 has bots, this has been mentioned before, and I agree is a good idea to use that code with some minor updates. Risijun had a good idea, and could be implemented since Trem 1.2 could be out soon.
Despite what everyone is saying, I think that AI is posible, infact even plauseible and has enough support and knowledge if we all to give alittle bit of effort.
-
All botlib syscalls are back in and working fine. Bots can be added and join alien/human team. The behavior isn't implemented yet, so they don't spawn just yet. I'll report here when I have the first bot to spawn.
Post a patch when you've got something.
I think, actually, a different approach might be easiest. Here is my idea: Create a new class for the AI (yes, give it its own model too!) on humans and aliens. Because the AI is very likely to be retarded, give this new class advantages to balance out the deficencies. If all AIs have the same size, you only need to run a pathfinding algorithm on each map once (though botlib does this, no?).
For the human team you could probably create, without too much hassle, a kind of walking turret. It'd be capable of using a few weapons (hitscans probably), and would run some kind of simple strategy. Give it lots of HP and make it fast and it might even be useful. ;)
You wouldnt even need to give it lots of health--presumably it would effectively have radar.
-
Okay, powerover and floodbud should be ban, I know I am being a little hypocritical when I am saying this. Tremulous is a game, games are made to have fun in. We are on a forum for a game, forums are to learn new ways to have fun playing a game. Simple, so both of you since all you do is spend your time insulting each other, just leave and go sit in your self pitty.
Well, he started it!
No more, I promise. Except in self-defense...I'm not going to sit, wounded, on the ground with a fully charged rifle and not counterattack.
_____________
Floodbud
-
Okay. If you two do know each other than fight it out off the forums.
As for the real code, each and every one of us knows that AI itself is possible, well atleast for games. We all know that other games have AI, from simple games like a scrolling shooter, or to something very complex like Splinter Cell or FEAR. The more complex one is written my an educated proffesional, the scroller, any one who takes the effort and knows something about computers.
Now, Trem is an open source game, probably the best, but that means that there are people who understand this code very well, and not just the makers of Trem. What we need are more people like that, and more people who would be willing to make the effort. But I really am a huge fan of Ris's idea, add a new class, now what would be nice to do is make it so that tey could side with one team, in example if you are playing agianst a newer team or clan, the host just puts them on that team, also they could have different difficulty settings example:
Easy - Low health, medium damage, medium accuracy.
Normal - Health is higher than a real players, damage is higher, Accuracy is the same.
Hard - Health is even higher than medium, Damage is even higher accuracy is about 95%.
Maybe its alittle overpowered, but tweaking could be done, that way AI script would be less elaborate, also something that would help would be for map editors to put in zones, like OPEN they would avoid, DEFENSE they would play defensive there, and ATTACK these would be like high spots and places they would like to attack from.
-
Demo
http://mercenariesguild.net/files/233755
Flash Video
http://Mercenariesguild.net/files/234436
He doesn't respawn after dying yet though.
-
I wasn't expecting anything to happen yet. Not after everyone turning against me. Thanks for not being a slacker.
____________
Floodbud
-
haha aimbot!
-
Demo
http://mercenariesguild.net/files/233755
Flash Video
http://Mercenariesguild.net/files/234436
He doesn't respawn after dying yet though.
i want his connection speed.
-
If there is one thing we have proved, it is possible, now the last thing is that respawm, I find it very simple actually there are two ways:
1. Change the code so its like Q3, you just spawn, no clicking.
2. By default the game has selected Rifleman or Dretch and the mouse is on has something selcted, cant remeber what right now but but if its OKAY then just have code like
if(dead == true)
keypress enter;
or something like that, for all of you hardcore C++ or C# fans sorry for me only knowing Java.
So after we are all friends, we can get something done, even if it starts with something as simple as aimbot.
-
Demo
http://mercenariesguild.net/files/233755
Flash Video
http://Mercenariesguild.net/files/234436
He doesn't respawn after dying yet though.
He's good. I want him in my clan.
-
Shut him up (The AI, not anyone posting in this topis :P). He gives away his position by screming like that. Make him taunt every time he makes a kill though =P. Oh, and how do we know you didn't just get a friendd to join an empty server, join humans, and then jump up and taunt a few times? This is not a violent accusation, just a little skepticism.
-
Shut him up. He gives away his position by screming like that. Make him taunt every time he makes a kill though =P. Oh, and how do we know you didn't just get a friendd to join an empty server, join humans, and then jump up and taunt a few times? This is not a violent accusation, just a little skepticism.
I think the insane connection speed takes care of any skeptics that say it is a human player.
-
stop spamming me with PMs asking for the mod. I wont release an alpha or beta.
-
Why?
-
stop spamming me with PMs asking for the mod. I wont release an alpha or beta.
Plz Relse it, it looks really cool!
people will love it,(besides that its god hehehehehe)
plz finish it and release it or tell me the server u tested it in i want 2 try it XD
-
stop spamming me with PMs asking for the mod. I wont release an alpha or beta.
Plz Relse it, it looks really cool!
people will love it,(besides that its god hehehehehe)
plz finish it and release it or tell me the server u tested it in i want 2 try it XD
[rant]MAKE YOUR OWN IF YOU WANT IT THAT MUCH!! [/rant]
@ sleekslacker:
Which revision was the last one with the bot calls in it? (Yes, I am too lazy to go to the repository and search) I dont want your code, but I would like to try for an alien bot, among other things.
EDIT:
Just a thought: Would it be possible to make it save a 'config' file server side about what works and what doesnt? Meaning: It would try and find out to never go naked with a flamer or luci, for example.
-
sleek's just being greedy by not releasing the code. There's no good reason I can think of to keep it all for himself.
Unless, of course, he's faking it.
-
Its real, (that or he's doing an amazing job...)
And there is good reason to keep it.
If you read what he said, its not that advanced yet.
You would all download it and bitch about how shit it is, or worse servers would start running it.
When it gets to beta stage, I'm sure he will release it.
-
first make a command
/spawn dretch
that adds a local client, makes it join aliens and spawn as a dretch
then we can proceed with the rest
-
Its real, (that or he's doing an amazing job...)
And there is good reason to keep it.
If you read what he said, its not that advanced yet.
You would all download it and bitch about how shit it is, or worse servers would start running it.
When it gets to beta stage, I'm sure he will release it.
I guess those are good points.
There still is the possibility that that is one of his friends, they're on a lan in the same room, and he told his friend when to spawn while using an echo bind to make it appear as if he just joined.
Whatever, I'll back off now.
-
It's a bot. FYI dante is a bot of many from the spiterbot set (can't remember where I got it from).
Sorry if this disappoints you, but the work has stalled for some time now. I just had my last paper for this term yesterday, and there is another paper I'm sitting at the end of May (summer term they call it, don't ask me). I might start again on it later, but as I said don't hope too much.
I won't release the code until I think it's ready (or if I find someone who might actually be able to continue developing it).
-
It's a bot. FYI dante is a bot of many from the spiterbot set (can't remember where I got it from).
Sorry if this disappoints you, but the work has stalled for some time now. I just had my last paper for this term yesterday, and there is another paper I'm sitting at the end of May (summer term they call it, don't ask me). I might start again on it later, but as I said don't hope too much.
I won't release the code until I think it's ready (or if I find someone who might actually be able to continue developing it).
ok that makes more sense
ps. any 1 who voted u cant make a bot HA he did muhahahaha
-
first make a command
/spawn dretch
that adds a local client, makes it join aliens and spawn as a dretch
then we can proceed with the rest
stage 1 clear
http://www.mercenariesguild.net/patches/?do=details&task_id=58
-
Haven't coded in re-spawning yet?
Does it know if it is alive or dead?
If it does know, just tell it to '/class dretch' or '/class rifle'.
-
If you wanted to run this on a server, just make a little mod that fixes our problem! In the code where it identifies Rifle and Dretch, just replace that word with say basic, then the only thing you have to do is put in /respawn Basic... But then what about evolving, and buying guns, getting more ammo... We could write two seperate, very complicated scripts, but then that leaves my earlier idea useless, so another road block.
WHERE IS SOMEONE WHO KNOWS ADVANCED GAME AI SCRIPPINTG?
sorry for the caps....
and the dots.....
...
-
*Fatherly sigh*
My little thread has gone a long way. :)
Maybe AI could be ported from Quake 3 or Alien Arena? It wouldn't protect its own base and you would need to teach it to avoid turrets/acid...would work for target practice though.
I don't need 8 people to tell me how stupid I am, so if you disagree with me, PM me or something.
___________________
Floodbud
-
botlib is bloated period
-
I think building bots would be easier to make and better than regular bots. I mean, make them so they know acid tubes go next to doors, and 1 spawn per three people, never decon reactor, etc. Then again, I'm not a coder, so basically I'm just spouting shit about something I have no prior knowledge about.
-
Maybe AI could be ported from Quake 3 or Alien Arena? It wouldn't protect its own base and you would need to teach it to avoid turrets/acid...would work for target practice though.
Quake 3 has a CTF mode with a flag to defend, so maybe modify that. Just for fun, I am trying to make my own bots, and this is the path I am taking.
-
I agree, we are PC gamers, we can just take the AI from another game and mod it, its what we do best.
-
Maybe AI could be ported from Quake 3 or Alien Arena? It wouldn't protect its own base and you would need to teach it to avoid turrets/acid...would work for target practice though.
Quake 3 has a CTF mode with a flag to defend, so maybe modify that. Just for fun, I am trying to make my own bots, and this is the path I am taking.
Alien Arena has even better CTF...Google it. Plus Alien Arena is more modern and has some similar weapons to trem.
Not that I know anything.
______________________
Floodbud
-
Alien Arena has even better CTF...Google it. Plus Alien Arena is more modern and has some similar weapons to trem.
Not that I know anything.
______________________
Floodbud
Is it quake3 based?
-
No, I hate that game.
-
Voted yes, but then thought NO.. They cannot actually beat real players but.. In the other hand, there is so many newbies which they can actually beat so.. This is really a matter of the perspective (read 'player') .
NAAB: yes
PRO: no
-
Voted yes, but then thought NO.. They cannot actually beat real players but.. In the other hand, there is so many newbies which they can actually beat so.. This is really a matter of the perspective (read 'player') .
The simplest bot would probably be little more than a walking turret with no rotation speed limit = similar to trying to dretch a tesla with maybe 2-3x the range.
-
Not that I know anything.
______________________
Floodbud
Okay from what I understand, while we may not be able to make an intelligent strategic bot with the aviable knoweldge we can just make the bot stronger, thus meaning that the player would have to out smart the bot rather than brute it.
-
I can deal with stupid.
ET even has some relatively smart client-side bots that will prioritize objectives versus players.
Here are some tasks, in order of priority:
1. Runs away from nearby stronger enemy.
2. Runs away from nearby enemy if low hit-points and/or human ammo.
3. Attacks nearest enemy or building.
4. Buys ammo/heals if nearby appropriate building and not at full (Human Only).
5. Upgrades current equipment (From a selection of equipment sets (i.e., what we can code)).
5. Follows random nearby teammate.
6. Wanders map randomly.
Nearby means in same vis area.
Mind you, we need definitions of 'manageable' and 'low hp/ammo', and probably have to hard code those definitions those ourselves.
-
I can deal with stupid.
ET even has some relatively smart client-side bots that will prioritize objectives versus players.
Here are some tasks, in order of priority:
1. Runs away from nearby stronger enemy.
2. Runs away from nearby enemy if low hit-points and/or human ammo.
3. Attacks nearest enemy or building.
4. Buys ammo/heals if nearby appropriate building and not at full (Human Only).
5. Upgrades current equipment (From a selection of equipment sets (i.e., what we can code)).
5. Follows random nearby teammate.
6. Wanders map randomly.
Nearby means in same vis area.
Mind you, we need definitions of 'manageable' and 'low hp/ammo', and probably have to hard code those definitions those ourselves.
How about start with:
1) Compiles (possibly hardest part)
2) Still works fine WITHOUT bots (nothing broken)
3) Bots spawn and get off spawn
From there, they should probably:
1) Attack any target in sight
2) Defend base
3) Wander randomly without running into walls
4) Deal with low health/ammo (return to base)
5) Obtain better items/classes
6) Attack enemy base
-
I meant it as a logic process to give the bot instructions.
Not only would wander randomly would probably make him step off the spawn, I added wander to my list as a "if you got nothing better to do" option.
It should not take priority over health, ammo, attacking the Base, or getting better gear, as it can do it forever if unimpeded.
Ignoring the enemy base to wander randomly is a fairly bad idea.
-
The problem with the priority system is that we would have bots that would just run rather than fight later in the game, this would be bad if there was a rant attack the base and all he had was a rifle.
-
Bots will atleast need to have a 'follow' ability. They could then attack in groups, and a group of humans with good accuracy would do much better than one or two.
-
If all you have is a rifle, then running is suggested.
Maybe disable run if a friendly is closer to the alien than you are.
-
I also understand the worry about players using a hack like this to really mess up internet servers. This problem is solved with imperfection as well, because the AI will be very stupid, and if anything, just feed.
If you are able to hack it and be able to make an aim bot from it then your able to hack it and make it be able to make it good.
I am guessing that aimming is done with a random number generator? the higher the number the more chance of hitting target, and harder the bot the more time its a higher number. (if thats so) then they can just edit it to always have the highest random number, just switch the random variable with a high number and a perfect aimbot.
-
Make it an aimbot with an adaptive maximum turning speed?
The farther it has to turn, the faster it turns.
As it gets closer to its target, it turns slower, until lined up on the target.
From there, it tracks the target and shoots.
-
Hello. :)
A.
1. Runs away from nearby stronger enemy.
2. Runs away from nearby enemy if low hit-points and/or human ammo.
3. Attacks nearest enemy or building.
4. Buys ammo/heals if nearby appropriate building and not at full (Human Only).
5. Upgrades current equipment (From a selection of equipment sets (i.e., what we can code)).
5. Follows random nearby teammate.
6. Wanders map randomly.
B)
1) Compiles (possibly hardest part)
2) Still works fine WITHOUT bots (nothing broken)
3) Bots spawn and get off spawn
From there, they should probably:
1) Attack any target in sight
2) Defend base
3) Wander randomly without running into walls
4) Deal with low health/ammo (return to base)
5) Obtain better items/classes
6) Attack enemy base
- C
- 1 : Accompany the player screaming "Come on !"
- 2 : Add other commands : "Defend base", "attack base"...
Thanks
-
Further development on AI beyond simple tasks would certainly be easier in 1.2.0 with mini-maps, vchat, and such.
-
so when are people going to stop talking about bots and actually create them? Just wondering!
-
Exams are coming up; also, I'm sure that sleekslacker has a life other then flaming, and f0rqu3 has a life other then being (one of) the village idiots.
-
spam bringer ... try harder please. you hit like a little girl ( and also type like a little girl )
just spamming another thread? ( at least use your brain and spam something good )
-
I was offering an explination.
-
I was offering an explination.
yes sure. now leave us alone
-
Aww I was just about to get out of my room and get some 'life' and you guys have to hold me back. This is getting fun. LOL.
-
Wow, my spalling is impurfect. Big fucking whoop.
(Btw, I made those typos on purpose.)
aww dont cry. you sure made them on purpose
(but you realized it now.)
Dude, f0rqu3, do you have no life?! If the only thing you can do is bag on for being human, you have some serious issues.
Now if it is possible, can we please get back to the actual concept of AI. I agree it would be easier if we had Trem 1.2, but we don't. To make up for that, we have brains, exams are coming up and some are already here, that just means our mids are fresh! Even if all we need to do is install Aimbot on a player running off of the server and to write a program for movement it would be a start, the only problem is I don't know C, and not enough people do.
-
write it yourself ... I stop the development of my bot
but dont worry some guy will start working on tre bot
-
write it yourself ... I stop the development of my bot missing period
but dont worry some guy will start working on tre bot
Learn to spell and use grammar properly before you critisize me.
-
Is it possible for mods to move posts from one topic to another?
If you can, stick this flame war into a new topic in the "Off Topic" section and lock it.
-
Even if all we need to do is install Aimbot on a player running off of the server and to write a program for movement it would be a start, the only problem is I don't know C, and not enough people do.
It is called 'copy and paste'. Take the quake3 bot code for ctf, remove any dependencies on map file based paths, and get to compile.
-
first make a command
/spawn dretch
that adds a local client, makes it join aliens and spawn as a dretch
then we can proceed with the rest
stage 1 clear
http://www.mercenariesguild.net/patches/?do=details&task_id=58
quite cool, and motivating. here's my attempt: i'm trying to add those parts of the q3 botlib that can be usefull for trem (aas, movement) and the accordant framework in the game module. so far i have aas and botlib loaded, clients spawning, move-to-goal done. patch will follow as soon as i get one networking bug fixed.
move-to-goal in action( dm_70, needs currents svn version):
http://soclose.de/misc/trem/aas_atcs.zip
-
yeah.............YES!!!!!!!!!!
-
so how do we use the dm_70 file?
-
quite cool, and motivating. here's my attempt: i'm trying to add those parts of the q3 botlib that can be usefull for trem (aas, movement) and the accordant framework in the game module. so far i have aas and botlib loaded, clients spawning, move-to-goal done. patch will follow as soon as i get one networking bug fixed.
move-to-goal in action( dm_70, needs currents svn version):http://soclose.de/misc/trem/aas_atcs.zip
-
here is the patch:
http://www.mercenariesguild.net/patches/?do=details&task_id=72
apply, compile, add the aas file, start map and /addbot foo humans
enabling bot_developer will give you some console and hud messages.
meanwhile bots can create movement goals from buildings, buy and sell stuff and attack buildings/enemies. it's still very alpha though. if someone is interested in using this for further bot development, i'll provide documentation for the framework features.
btw: i just accidentally deleted that demo, sorry.
-
I got a mod idea that could use this, and stupid AI is better than nothing.
-
someone can compile this and give qvm ?
because i have error while patching and compiling ;-(
-
what kind of error did you get? here is the qvm:
www.soclose.de/misc/trem/vm.pk3
you'll also need the botlib'ed engine and the aas file. i posted an aas for atcs on the patchtracker page.
-
so, still working anybody on it?
-
i am, occasionally.
since the patch-release i got the bolib to handle multiple aas files and figured out how to pass a bounding box size and physics values as parameters to bspc. so every class gets its tailored aas.
-
How's it coming along cyrri? I think if you could get it up to a state where it is a simple mod everybody can install and use this would be great for training offline (especially A)...
-
i haven't touched the mod for one month. it's summer!
anyway, i'm not working on a full mod but on basic aas stuff. as noone seams interested in developing on top of it, there won't any offline training mod (this year).
-
Im no genius but this is hard to make and you get better at Tremulous naturally , would be nice for a AI but nope not necessary.
-
we have hundreds of people playing every seccond.
We could monitor the server for build patterns and attack vectors :)
and then make a bot that use the results of this data mining. lots of fun!
-
I am just hoping for something that doesnt move, and shoots you if it sees you lol
-
I am just hoping for something that doesnt move, and shoots you if it sees you lol
That's called a turret.
-
Its also is an aimbotter with no user at the keyboard. :roll: