Jump to content

oukej

Former Developer
  • Content Count

    1063
  • Joined

  • Last visited

  • Medals

  • Medals

Everything posted by oukej

  1. And exactly that you can simulate when creating the mission using the setSkill command or Skill slider in the Insert Unit dialogue (which I previously mentioned as subject to improvement - I hope we can still find some time for it)
  2. oukej

    AI Discussion (dev branch)

    Yup, it's dependent on the configuration of weapon's fire modes. When aimed the AI just does what's set in there. But some randomization is created by the dynamic error introduced in Arma 3 (the AI looses its aim more often which creates irregular pauses in the shooting) Unless the situation is super-static I think the intervals are by far not so obvious as before. http://forums.bistudio.com/showthread.php?159710-AI-Discussion-(dev-branch)&p=2685753&viewfull=1#post2685753 Nothing interesting to share here, move along ;) In a long-run. So far it is not intended to override combat behavior, but solve the cases where it clearly "sucks". Either - pretty vaguely said - by new functionality or by improving some of the conditions.
  3. oukej

    AI Discussion (dev branch)

    Does it happen always with units local to the "headless" client? Or does it need to have specific circumstances? Could you please check the individual AI skills and AI level on the client? (final skills below 0.1 can eventually render some broken behavior - currently that can happen with AI set to lower bound skill in the editor + e.g. Novice AI level. We may want to readjust the curves to achieve better propagation and cut off the "broken extremes". (And possibly break everything by doing so ;)))
  4. oukej

    AI Discussion (dev branch)

    Headless, aren't they? Ok, seriously now - it probably caused by http://feedback.arma3.com/view.php?id=18365
  5. Hm, we've just rolled the patch out. Awaiting blizzard.
  6. Sure, should happen pretty soon ;)
  7. oukej

    Weapon tests

    This has sadly been around ever since, we are aware of it. There's also a FT ticket here - http://feedback.arma3.com/view.php?id=18367 Regarding the above-water to under-water detection, please follow ticket http://feedback.arma3.com/view.php?id=18986
  8. http://feedback.arma3.com/view.php?id=18328#c71582 The isolated representative issue has been fixed (fix is on the Dev branch). Let's hope other not-so-easily reproducible symptoms were caused by the same bug. For confirmation of that we'll leave it open for some time.
  9. Yup, it's always in the end down to global modification of each AIs subskills. SkillAI and precisionAI each modify a specific group of sub-skills. There are 4 AiLevelPresets - 3 of them have skillAI and precisionAI values hard-coded, while the last one, custom, leaves the customization of those two values up to you. ---------- Post added at 14:41 ---------- Previous post was at 14:38 ---------- I have created a quick poll, where I'd like to ask you for your thoughts about current presets. Take the survey here please Quick survey closed due to it's limits. Thanks for the feedback! Thank you!
  10. oukej

    AI Discussion (dev branch)

    Quickly tested (player, owner of Zeus module, creates an unit, takes control of it via RC, engages one enemy behind a corner and lets the RC unit get killed) and have to disclaim - the enemy AI had no clue about the player. But there still could be a specific case where this would happen
  11. Btw, those two cases should be improved ;)
  12. oukej

    AI Discussion (dev branch)

    Instant and temporary, as a result from the hit (player looses his/her aim due to the shock as well). Low health level, similarly to high fatigue, affects the weapon sway for both AI and players in a similar fashion already. I must admit though that AI is rather good at compensating it and would probably benefit if the range (of how good it is at it, proportionally to the skill) was extended. --- So far it seems the subordinate's disobedience (movement, target acquisition, in combat behavior) and driving are the most pressing issues, right? The driving is complicated in regard to different causes (and their combinations) of bugs - from vehicle configuration, map data, dynamic circumstances to AI behavior itself, precision and time. For both, even though the issues may seem obvious, we are gathering the data and isolated situations. With the help of our fresh reinforcement Iceman we've categorized the AI stuff in feedback tracker a bit more and there is also a path-finding primary report here http://feedback.arma3.com/view.php?id=18340 For exact spots on Altis and Stratis there's a ticket here: http://feedback.arma3.com/view.php?id=18324
  13. oukej

    AI Discussion (dev branch)

    E.g. If you run into some cover the AI should be less prone to head-shot you once you reappear ;) Part of http://forums.bistudio.com/showthread.php?159711-Development-Branch-Captain-s-AI-Log&p=2535506&viewfull=1#post2535506 (FT links there) The other one - wounds and reloading, that's probably clear but jic - if an AI is firing at you and runs out of ammo, needs to reload, or you manage to hit it, but not kill it, it will need some time to reacquire the aim after such action. Getting hit obviously causes much more disturbance. All 3 changes are also part of http://forums.bistudio.com/showthread.php?159711-Development-Branch-Captain-s-AI-Log&p=2598737&viewfull=1#post2598737
  14. oukej

    Development Branch Captain's AI Log

    Change log notes, v. 124216 All changes are also generally part of Dynamic Calculated Error for AI. :cool:
  15. oukej

    AI Discussion (dev branch)

    More precisely with "next .exe update" (which probably won't be today due to a different issue)
  16. oukej

    AI Discussion (dev branch)

    wut? TRLld :computer:
  17. http://forums.bistudio.com/showthread.php?159710-AI-Discussion-(dev-branch)&p=2679062&viewfull=1#post2679062
  18. oukej

    AI Discussion (dev branch)

    Despite the fact there's not a single use of word "AI" in http://www.bistudio.com/english/company/developers-blog/460-arma-3-roadmap-201415 there's no need to worry. The AI will have to face new situations (and our engine masters great challenges), but the mentioned overall ongoing effort is still on and we'll try as much as we can to give AI the needed love, fixes, improvements, tweaks and polish ;) What difficulty have you played it on?
  19. There's one important information in http://www.bistudio.com/english/company/developers-blog/460-arma-3-roadmap-201415 ;)
  20. oukej

    AI Discussion (dev branch)

    This was a nasty one. It appeared only on Steam Development branch, not in our internal build. Also the newly deployed Stable branch patch does not suffer this problem. The fix of the issue with move waypoints (affecting also majority of the playable content) will be on the Development branch with the next update. We are sorry again for the inconvenience.
  21. We are currently investigating it. Please follow in http://forums.bistudio.com/showthread.php?159710-AI-Discussion-%28dev-branch%29&p=2678645&viewfull=1#post2678645
  22. We are currently investigating it. Please follow in http://forums.bistudio.com/showthread.php?159710-AI-Discussion-%28dev-branch%29&p=2678645&viewfull=1#post2678645
  23. oukej

    AI Discussion (dev branch)

    We are on it with prio. We are sorry about any inconvenience. Any further info (video) will be helpful.
  24. oukej

    AI Discussion (dev branch)

    The disobedience of AI subordinates is no less important issue to us. We try to avoid responding to questions where the answer could inevitably impose false expectation - unless we are 95% sure that we can deliver the subject. Ain't it better than false promises and consequent disappointment? ;)
×