Jump to content

ToxaBes

Member
  • Content Count

    18
  • Joined

  • Last visited

  • Medals

Community Reputation

12 Good

About ToxaBes

  • Rank
    Private First Class

Recent Profile Visitors

215 profile views
  1. ToxaBes

    1.90 update - Troubleshooting

    BattlEye = 0 is enough to start mods working. You can keep verifySignatures = 2. Somebody can say what this is not an option. bla bla bla. BIS can't fix signature check issue during 3 releases <- this is not an option.
  2. ToxaBes

    1.90 update - Troubleshooting

    Also, for some reasons BattlEye block standard Malden map with the mods which perfectly works with Altis and Tanoa. Malden, Carl. This forum definitely need some facepalm emoticons...
  3. ToxaBes

    1.90 update - Troubleshooting

    This mean what different mods blocked on different servers but core of issue is BattlEye related. And it seems we don't have any chance to fix it on our (server) side! BIS devs - do you copy? psh psh
  4. ToxaBes

    1.90 update - Troubleshooting

    I think we also have some issue with BattleEye and terrain mods. Example: If I use -mod=@cba_a3;@sma;@cup_core;@cup_weapons;@cup_units;@cup_vehicles; (CBA A3, Specialist Military Arms and CUP mods without CUP Maps) all working fine But if I try to include some map mod like CUP Maps: -mod=@cba_a3;@sma;@cup_core;@cup_weapons;@cup_units;@cup_vehicles;@cup_maps; nobody can join to server. I tested about 10 custom terrain mods and all they have this issue. If I disable BattleEye - all this terrains works fine. Checked on Linux Dedicated Arma 3 server on stable Arma version and Dwarden's builds (v01,v05). This issue is strongly related to terrain mods (maps).
  5. ToxaBes

    1.90 update - Troubleshooting

    Nope, as far as I can see perf build use system user directory. This mean one user = one Arma server instance. I hope it will use -profiles option with next update.
  6. ToxaBes

    1.90 update - Troubleshooting

    Guys, let's try to do next: 1. Download and install v 1.90 perf build v01 (32 or 64 bit) from Dwarden post on previous page 2. Go to C:\Users\{Your name}\AppData\Local\Arma 3\ (Windows) /root/.local/share/Arma 3/ (Linux with Arma started by root user) /home/steam/.local/share/Arma 3/ (Linux with Arma started by steam user) 3. Check what you have "DataCache" folder. If not exist - create it manually 4. Start arma3server without any params (just double click on it or call from console) 5. You'll see cache lock file in DataCache folder. Wait while it gone and cache.ch file appears 6. Stop arma3 server and start it with any mods which you want to use 7. Cach lock file appears again, wait while it gone 8. Play Tested on Linux Dedicated server with CBA_A3, All CUP, SMA and a lot of map mods. Tested on Windows Dedicated server (x64) with the same mods.
  7. @Dwarden - Thank you, I will test it. Btw, Linux versions present on Dropbox but not uploaded on Google Drive. Update for Linux Dedicated Server with a lot of mods: it works if you create "DataCache" folder manually. How to check? Simply start ./arma3server and you'll see repeated message like: Waiting for '/root/.local/share/Arma 3/DataCache/cache_lock' to be available It seems Dwarden moved cache.ch from Arma folder to user profile but forgot to create cache folder? First run will be quite long (up to 5 mins) due to cache calculation for server and each headless client.
  8. ToxaBes

    Machine learning / AI projects using ArmA?

    About improvement of inside group tactic and individual unit behviour. We have FSM and this is good step forward, but we want more (we always want more). FSM consist of number of statements with different priorities. We can be in only in one statement in the same time. Each statement have conditions. If several statements conditions match current situation we use statement with highest priority. Priority is just a floating point number from range 0..100. A high value indicates higher priority. This is the basic logic of FSM. For example, "combait brain" in latest version of VCOM AI have about 10 statements. Let's think what statement is like some action: move to cover, supress fire, arm static weapon, etc. The more statements we have the more "tactical" ways we can achive. But here is one problem: priorities in statements are constants and placed by author of FSM mod during creation. Question is: can we use neural network for changing priorities of statements inside FSM? Is it possible? Usually, neural network answer is array of probabilities. So if answer of network is suprress : 0.15596 move to cover : 0.9256 ... call for mortar : 0.87352 we can use this numbers as priorities for FSM statements. On this case unit will try move to cover then call friendly mortar if it placed around (if condition match). In other words output layer of neural network have 100 neurons, each neuron related to one of statements. Imagine: FSM-mod with 100 statements. It have some initial priorities for each statements and ~90 statements have zero priority (disabled) and neural network start changing it reacting to changing situation. May be it will be more realistic than waiting for Headless Server|GPU computing for A3 or introducing it in A4. Make sense?
  9. ToxaBes

    Machine learning / AI projects using ArmA?

    If we talking about tactic inside group (finding covers, moving, supressing fire, controlling sectors, garrisoning, mining etc.) I fully agree with you. Situation changed when we have 50-60 groups with armored vehicles/bots/helicopters up to 200-300 units in total placed in zone ~3.5 square km without human commander. In this case maximum what you can have here with FSM - share found targets between groups and move some groups to this target. For effective fight we need to answer on next questions: which group types should be send? we have tanks, infantry, helis etc. should we use infantry from roadblocks/bunkers inside zone? should use bots placed in buildings for scouting? how many groups to send? we should send effective minimum which will enough to win, so we need to store and use some data about previous fights for each group we should decide, in what position we need to move it (in front of enemy, in side etc) which group should fire first? what to do if we already fight with enemy group and it still alive and moving forward? what to do if we have several enemey groups infiltrated in our zone from different sides? what to do if we need reinforcements? call it or retreat? what to do if players attack from one side several times? (ambush) what to do after fight? just patrol this area or back to previous position or something else? Each of this question have a lot of sub-questions. With FSM you can, for example, call for friendly mortar but not answer on all this questions because FSM don't store information about previous fights. Good commander (player) can answer on this questions, but this is PvE - we have humans only on one side. That's why I use trained network. Again, Tactic inside group -> FSM based mods Tactic outside groups (HQ) -> trained neural network As far as I can see from your videos, SmartAI is really good. But this is in-group tactical solution.
  10. ToxaBes

    Machine learning / AI projects using ArmA?

    I know, we all could play in Arma by own way and this really great thing. Solo MP session will be an option as well as MP with teammates. But Solo MP make sense only with disabled self learning, it already trained to play against teams (5 or more human players) so learning on solo-enemy lead it to degradation. Even if you create second group of skilled bots it will lead to degradation (I checked it). Also playing against newbies/casual players in period will leed to degradation (also checked twice, and than restored AI files from backup). So, Singleplayer/Solo MP modes are not interesting for me because AI can't learn.
  11. ToxaBes

    Machine learning / AI projects using ArmA?

    You are correct. I mean "Lone wolf" tactic is not a tactic at all. Even if you have some amount of friendly bots. Using Neural network in such variant is overhead. VCOM AI is all what you need. Also, extDB 3 required for AI will not work in singleplayer mode (for me it works only in multiplayer).
  12. ToxaBes

    Machine learning / AI projects using ArmA?

    I'm afraid for one player in singleplayer mode it just don't make any sense.
  13. ToxaBes

    Machine learning / AI projects using ArmA?

    Yep. VCOM AI provide good improvement of individual and in-group tactic, but it don't work with high command. For neural network group is smallest operation unit. For VCOM AI group is highest operation unit. So they work on different operation levels and supplement each other. Neural network trained for manage limited number of forces on limited territory (defend zone with R~1km). It understand type of terrain/height gradient, types of units in groups, vehicles etc etc. in both own and enemy forces.
  14. ToxaBes

    Machine learning / AI projects using ArmA?

    I'm in the middle of refactoring mission code and logic so let me finish it first. After that I will be glad to see all of you on server, beaten by my Red Queen, he he. I will post here or add new thread about server with details when training mode will be ready for public. But here is one problem. Mission is not translated to English. Let me clarify, why. I try to make server for team training where results of good fights will be used for additional neural network learning. Mission logic divided into two parts: training and qualification. Training mode is similar to main AO in Invade&Annex mission. Qualification mode is like linked side missions from Invade&Annex but generated by my second neural network (it's like quest engine and it's in hard WIP). We can't train quest AI in 2 languages in the same time and our team is Russian-speaking (not all guys know english on good level) so we use russian language. So training mode will be good for all no matter of language but for playing in qualification mode you will need somebody who will translate task details generated by AI and other things.
  15. ToxaBes

    Machine learning / AI projects using ArmA?

    Yep, I use trained (over 2000 fights in training set) neural network based on Fast Artificial Neural Network Library as High Command structure for bots on my PvE server. Using it on low level don't make any sense because we have Final State Machines for that (good example is VCOM AI). Every X mins script get information about known enemies for each group, prepare data and send to Mysql DB via extDB3 as command request (+ get response for previous command requests). PHP script called by cron every minute check DB for new command requests, use Neural network for build command response and save it back into DB. Quite simple. Main things I got after more than year of using this: 1. 90% of players are not tactical geniuses, so if you will use all played battles as training sets for additional learning you faced in degradation of neural network (why complex things needed if simple attack from side is effective) 2. 90% of players don't like pain and humiliation from AI so your server could become empty quite fast 3. 90% of cases are covered by VCOM AI (thanks to Genesis for this) and if you don't think so - you can add needed parts inside FSM logic (I added some things like: if players have defend position in house, bots with RPG 7/RPG42/PCML will fire rockets in this house). So if you planning to use some neural network in Arma, just use VCOM AI first and expand it as needed. It dramatically save your time (and population of players on server).
×