Second

I start to get into right mood only when i'm in somewhat bad situation... Harder to get the right mood in games, these days, don't know why. Otherwise i just play with "left hand"-style. Shooting like this: and usually from standing position.

And i usually just start to shoot sharper and more efficently, but other wise i doesn't effect me either. Maybe some times suppression has effect when i feel to roleplay the character, but usually this is not the case. player can't feel fear of death and that should be simulated somehow... Blurring or black outs are quite efficent way to reduce player's will to do something, atleast this stands for me. Other way would be that suppression forces charater to do things and take control to charater from player for a short while... But that wouldn't be fun or even good thing. Forcing enemy to seek for cover and not to advance and reduce it's skills and eventually to make it to hold it's fire and flee is what i used with AI. Combat mission uses quite good suppression system, which has 8 steps, from mild cautious behaviour (nothing more that men watch bit more their doings and use cover bit more) to sever broken (units will to fight is gone, not fit to fight anymore). But i quess that ArmA won't need that much, three might be just fine. Individual changes would be also one thing, which ArmA should represent too. Someones are unaffected by it, many WW2 veterans have said that those individuals become even more efficent.

Is that direct or indirect fire? I mean... there isn't direct fire distances up to 4 km in sahrani... Or is there?

Some Ssimple Artillery script which attaches it's targetarea to forexample to gamelogic and then move gamelogic via waypoint and there you have artillerybarrage, which will advance infront of you just like IRL-style (having RL friendlyfire change too ) For n00bs it could be made so that it follow player moves (when player advances barrage advances) or "hardcore" style with timed adjustements to barrage (player needs to keep in timeline, if player can't keep up enemy has better change of recovery, or if player advances too fast he ends up in barrage). Nothing beats the sight of several artillerybatallions bounding their targetareas or AI FO guided fire of several batterys on attacking enemy's neck.

Right now i'm more intrested about that 'movetarget' scripting command... Altough it can more be related to FSM and combatroutines, and not affect to position of target or engage square. Those setskill-things only affect to AI's skill in certain aspect, altought who knows what everything they have impact on, once BIS gets them working. Maybe it affects also to pinpointing.

Your right. ArmA's AI is guite good in the end. With 1.08 patch their urban warfare abilities have gone up, it seems to be much tougher opponents than before. There are things that are lacking... Or is there? -In dynamic-field that Funnyguy1 is asking (yes i understand now what you are after, mostly i agree) all or most they are already present in ArmA: AI has wide variety of aspects, mission editor can adjust AI's aggressiveness (will to engage), cowardy etc... In that sector what could be asked from BIS is that they could make those easier to select (no need to use scripting language) so that mission designer would be actually using them more. -Then there are standart military routines missing (that's what i'm asking)... But is that problem with waypoint system more than with AI? AI can be made to do most of the things... But after all it requires scripting or lots of work in mission editing phase... Which usually reduces dynamic structure of mission. But i don't see this relating in "balance vs realism" as Ofpforum hinted. AI isn't balanced from n00bs and it isn't made to be stupid, but there are things that affects negatively in taking advantage of AI's or engines felxibility.

And ArmA works like same... Only that their hearing gives them rather good hint about location of threat. ArmA does the same quessing, looking at engage square shows that it isn't always there where opponent is. Movement has big impact on this. 1. I see your moving there 2. Now i can't see you so (you went out of my sight)... So i think about your moving speed and direction and calculate your path. 3. I'm going to check it or 3. Damn calculation were wrong! (i can't see anything in the spot that square is pointing) So i need do quess some spot in between. NOTE: About this routine: i'm not sure how is it determed, but this is my quess 4. You weren't there... So i quess i need to go back to my position in formation This is how AI works when opponent is moving and AI sees that, but them they can't see opponent.Only issue is that hearing is too good, well not too good, but they can pinpoint of locations with hearing very well... Which then again can cover some of their flaws (like they have no actual logic that where enemy would be hiding, under that bush or in that house)

You are speaking the right words. What AI right now is lacking are basic ways how AI should do things. -Forming defence is one part of it's lacks, -Urbanwarfare in whole scale in missing. -There's no way to operate crew served weapons. -It's takecover routine is busted (in FSM interupting with it?) -what else from infantry's point of view? There are more, just can't remember what all. But then again who wants that kind realism, that he has to spent hours waiting that he's units can advance again. Shooting tens of minutes at bushes and other shady objects as enemy in nowhere to be seen, but it is still shooting back. Or that he can have just short climpse of already dead enemy and rest enemies are killed by artillery and other units. Player can see some enemies taken as POWs, but he can't shoot them as it's against militarylaw. Rest of enemy have run away... That can be bit unsaticfying for player... Even to "hardcore" oh btw. I've made suppression to ArmA Nothing about actual suppressive fire yet to ArmA, as without suppression it's pointless to have that (they would just get themselves killed). Search for suppression script. It's not perfect, but i'd say that it reaches some amount of realism in modelling fear of death to AI, still i'm improving it but other pojects are in a away. But most likely most of community wont give a damn about it, which is understandable (it's heavy and non-elegant). I don't expect it to become solid part of becoming mods.

Are you sure that it works I honestly can't verify that. Sometimes seems to work, but when i test it in "shooting range" it doesn't seem to work. Forexample: I mean i haven't noticed that aiming would shake more if i tweak that attribute (aiming shake was it?) to direction or another. I tried to look it with TroopMon, but aiming shake didn't seem to chage if that attribute was set to 1 or 0. But in some situations i could swear that it (and other attributes) affects... So i don't know

Please tell me, what are these realistic and yet upredictable things that AI needs to be scripted to make it perform better? In military sense (as we are discussing about realism), mission being capable of being supprising after replay is obsolent (there are on replays in real life), so doesn't have nothing to do with realism. So AI being unpredictable and still military realistic isn't the wisest word. Dynamic isn't the word also when talking about military operations in small scale. Platoon leader is expeted to do what CO tells him to do and so on. This is basic mind-set. There are very few things that require dynamic reaction (in sense that you are speaking, ). Basically everything has been ordered, so that superiour knows that what everyone is doing. Ofcourse dynamic comes in play in small unit level too, but ArmA's AI already does that (flees, engages etc...). Basically it can't do anything more (like sending it's own squad to counterattack, when squad should hold area) as it has it's own orders (waypoint) which it needs to fulfill. I've been thinking this ALOT, as i mess with AI and military ways (from squad to company level) mostly when "modding" ArmA, and dynamic and realism just mostly won't fit together. Dynamic starts from the top and lower it get less it can affect... What we would need would be chain of command between all existing units on mission and horrifing amount of tactical solutions for AI(leader), to make missions "script free". Right now scripts and triggers does replace chain of command, "tactical awareness" and orders and most of all mission designer. So i ask again (as i presume that you already forgot what i was asking you in beginning of my long post ): Please tell me, what are these realistic and yet upredictable things that AI needs to be scripted to make it perform better? You mentioned few, but i don't think that they have nothing to do with military ways (aka realism), just making missions more replayable. Or if you think that they have, then tell me what i might be missiong a point here. EDIT: I've been flashing dynamic things to ArmA too. But our point of view seems to be different. I'm intrested of implenting chain of command (units' organization in better words) and some basic methods to that organization (from squad to company/batallion level) so that units in organization can have guide lines of what it can do and what are it's tasks etc, it wouldn't be so dependant of mission designers insight and editing skills as now... I would say that it's impossible to do in other than very limited scale... So basically better to script and trigger it.

Look at your skill settings in difficulty menu. There's answer If you can't cope with thing that enemy is as good as you or your friendles just set enemy AI's skill to 0 and friendly AI's skill to 1.

Noone hasn't explained it complitely, but i think (as most do) that it just sets all enemy AIs to expert skilllevel or maybe bit beyond that if it's possible (which i doupt). Don't know about CS hack, but that super AI-setting usually breaks almost every mission, where friendly AIs are supposed to acomplish something and their skill level isn't at maximum. If they are experts too, then it shouldn't be that big problem as they are almost (or complitely) at same level than super-AI.

What is this "manual"? Mission briefing or game's own manual. That campaign mission tells you bulls*it: Just use your radio 0-0-*, they have pr-defined spots where sniper moves and artillery fires (if it fires at all? i don't remeber hearing or seeing explossions of arty barrage). Frankly campaign doesn't have even one single artillery barrage, even when there's lots of mentioning about artillery support. Those are direct firing howitzers that player comes across.

Muzzleflash Altough it's has been problem way back from OFP: player is better at clear day and AI in nights and fog. In ArmA this seems to be even badder, but i don't know reason for it... Maybe it has to do with AI's better ears and spotting ability. With FFUR in OFP i think i noticed the same, AI becomes very dangerous opponent at low visibility conditions. But happens also during daytime. I shot their team leader in same conditions (had SVD and enemy was blufor's standart squad, which skill level was 1) at mid day ... They didn't see me instantly and i tried to stay low in quite crappy spot (only 1/2 of me was covered with bush, in top of the hill)... After about 5-10 seconds they finally spotted me (binoculars or sniper) and nailed me.

In 100% it got almost weird, expacely with some addon weapon's sights. Weapon handling very quite flexible and in small spaces it was good if body wasn't able to turn. Maximizing benefit from cover was also great with 100% as body stayed still even when weapon was turned alot Shame that it didn't affect to recoil and accuracy negatively. I reduced floationg zone to about 20-30% because of that... Felt like cheating.

DAMN! I bought GRAW to see how the AI works in there and what it's like when comparing to ArmA... Is there anyone who has been thinking this urban warfare aspect of AI and how to improve it? My ideas: -When unit/group halts it has to search cover for itself behind solid objects. -give AI ability to either lean, move or raise it's stance to be able to shoot and then move to cover when in firefight. -Some thing other, but what? I had that expiment to add alley checking ability to AI. It can be used to check solid objects and do it fast (calculation don't seem to put big strain to CPU), but is there others who would be intrested to give ideas or write code. Task feels pretty overwhelming to acomplish alone, there's too much for me to improve in ways of scripting etc and not much time. <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">#LOStest _dir = round (getdir leader _grp) ?((_dir < _dir2) and (_dir > _dir2)): goto "LOStest_start" _pos = getpos leader _grp #LOStest1 _spot = "LOSbeam1" CreateVehicle [0,0,0]; _spot setpos [(_pos select 0) + 5 * sin(_dir),(_pos select 1) + 5 * cos(_dir), 1.4] _spot setvelocity [10000 * sin(_dir + _n), 10000 * cos(_dir + _n), 0] ~0.01 ?((nearestObject [(_spot), "AllVehicles"]) distance _spot) <= 1): goto "lostest2" ?(_spot distance _pos > ((_directionarray select (_n + 90)) * 2) and (_spot distance _pos >= 5) and _dist2 <= 30): _changearray set [count _changearray, 1]; ?(count _changearray <= 0):_dist2 = getpos _spot distance _pos _dist = getpos _spot distance _pos; _directionarray set [(_n + 90), _dist] #lostest2 ?(count _changearray > 0): goto "dir_result" _n = _n - 5 deletevehicle _spot ?(_n <= - 90): _n = 90; goto "LOStest1" ?(_n <= 60 and _n >= 1): _n = -60;; _dir2 = round (getdir leader _grp); _dir = round (getdir leader _grp); goto "sensor" goto "LOStest" This is small part of my code. This part checks for drastic changes (possible alleys after solid wall) in vicinity of unit. I'm putting this piece of code so that everyone can see that it can be made and it worked. Is there someone other with better ideas or desire to try boost AI

I'm just curious: what type of ammunition was that? Is it from a D-10 gun of the T54/55? Most likely it was T-54, could have been T-55. But i don't remeber (over 6 years now)that it would have had that muzzlebreak in barrel's nose, so that indicates to T-54.

Okay. HE round in very powerful. I've seen pattern of fragments caused by 100mm on white snow... Effectivity compared to ArmA's 120mm and 125mm "HE"-rounds is in other planet. ArmA's 120mm and 125mm HE could be compared to something like pipemine with 80-120 grams of TNT, which lethal radius is in 5 meters and wouding radius still at 10 meters. Not sure, but i would think that ArmA's 120mm and 125mm HE causes douple to that, 10 and 20 meters. RL 120mm HE-round would be bit differrent case. But in RL there's other things that affect to HE's effectivity, such as impact on ground (soft or hard) or in air (tree or timed/delayed fuze). What i saw happening with 100mm HE was hit in tree at 3-4 meters above ground. Fragments scattered in triangular shape to area in which could have been half-platoon in defenceformation 10-12 men with aprox. 7 meters spaces = Everyone (filled plasticbags) got hit from several fragments, even when they were lying down but not dug-in. Teaching was: "Choose your squads positions either from deep inside forest or from open space... BUT NOT FROM BORDER OF OPEN SPACE AND FOREST" That same HE-round hitting soft marshsoil and most likely hit should have been direct to man to cause even one casuality. ArmA doesn't have type of soil modelled, everything is hard as concrete. Back to ArmA: Comparing ArmA's 105mm or 122mm howitzer rounds to ArmA's MBT's HE-rounds, atleast to my eye, show that howitzer rounds are much more effective against infantry (longer effective radius with "fragments"). So i would say that in ArmA those are HEAT(-MPs or other) that MBTs are using. Not sure about "traditional" HEAT: they aren't complitelty safe to infantry near impact point (debris from shell etc), but not 100% killers too like in ArmA. But i don't have facts of lethality of traditional HEAT-rounds against infantry. Here's quess-list: Debris from shell, possible hot fumes from jet when it starts to scatter, debris from impactpoint (rocks, fragments?).

I would uninstall OFP if OFP CDs wouldn't be so scratchy (my kid got his hands on them). Last time i spent 6 hours in installing OFP. Maybe i just keep vanilla OFP without mods in my HD, as it doesn't take that much space. And i have to admit: too much memories with OFP... It doesn't deserve to be uninstalled.

Neartargets? which gives array of all vehicles and men inside given radius. Not much different than using suitable trigger... So i would go with trigger. But neartargets seems to have some intresting values that it returns, like that cost value that it returns.

Yeah. Problems be presented when that unit has been issued a new order, like engage. So script etc has to be writen so that those special cases or possible bugs can be countered. Using trigger is just too limited and comblicated way to sort out those undesired outcomes. <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">waitUntil{ currentCommand CAR == "" }; combined with <table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">getpos car distance position < 10 Medical tent or engaging enemy (instead of being there where they should be), caused few problems to me

Okay now i see it. (how in the earth did i miss that) Seems that we ended up in quite same solution

I'm growing weary of this. I didn't say anything that my way would be better (btw you haven't shown yours). I just said that i would happy to see what you have found out, as moveto doesn't work with vehicles. btw. I edited previous post. Is it what you found out? Or do you have some other way?

I don't see solution mentioned by you. Only that you have found workaround, that's all and nothing more about that workaround. Which i would be glad to hear as moveto doesn't work with vehicles. EDIT: Oh yes here's one more. Well it kinda crappy alternative, but here one: like this vehicle domove position; vehiclemoves = true make trigger with this condition: unitready vehicle and vehiclemoves and to triggers OnActivation: dostop vehicle and also (to give audio) commandstop vehicle. Which strangely works sometimes and sometimes doesn't. Don't know what is wrong with it. But in my tests sometimes my vehicle (which was 3rd) started to follow 2nd vehicle (which had that domove order). Those times we both returned happily to leader's formation (ignoring dostop and/or commandstop orders). But sometimes my vehicle didn't follow the 2nd vehicle (but remained in lea and those times it did work. I don't understand reasons for it, it just seems to happen and not happen. Best solution seemed to be to leave timeout for 1 second. Maybe dostop-order was overriden by hardcoded "return to formtion"-order after move-orders was fulfilled. Might be that my test was bad because that domove was given as soon as mission started. trying it out bit later seems to be much more "stabile"

But from what range that penetration value for Russian SABOT is given? 2 km would be my guess. I've heard that generally from nose of the barrel SABOT has about 30% more penetration that from 2 km. Other thing is that MBT's HEAT round doesn't have that good change of penetration than SABOT, as every MBT's armor resists HEAT's penetration method better. MBT's HEAT isn't obsolent, but not as good as SABOT in most cases.