Insanatrix

My mistake. Rereading my post though I noticed that what you described was kind of how I made it sound though.

Well that's kind of what I meant. Whatever you want to call it, it would be a simulation of gravity, gravity recenter's your weapon. I'm not saying there should be something that constantly pulls your aim down while shooting, but rather after your shot or shots have been fired. What I meant by larger deviation and loss of aimpoint was that if you full auto'd an entire clip, your aim would not be 1:1 to where you started after you stopped firing. Not like an active system reducing recoil per se. Watch that and you kind of see during the second half speed clip that even on full auto, there is some resettling of the weapon due to muscle tension and gravity. The current system feels like it's lacking that, hopefully the new system they introduce will factor it in.

Maybe they could introduce a system that gradually draws your weapon back to it's original aimpoint over time, say like a half second or so? For every consecutive shot the deviation becomes greater and the loss of the original aimpoint becomes greater?

No, You're a hypocrite because you apparently feel entitled to tell others how feeling entitled is wrong. If you can't see the logic, you're apparently also stupid.

TL;DR I feel entitled to tell you how you feel entitled.... Do you always feel the need to gloat after banning someone or is this a special occurrence? I think most of us could do without the power trip conjecture.

In a run of the mill combat scenario, you might get 60fps (average/constant) at 1080p with say like a GTX Titan or a 680/7970. I get anywhere from 18-55fps with a GTX 480 at 1080p. I would say my average fps in a combat scenario is around 28fps or so. A lot is going to depend on your detail settings and how well BI can get the game engine to run moving towards retail.

So in other words, even say at like 1.0 skill level, they should be really really damn good but not headshot magnets like they basically are now? If so, AWESOME! heh.

I'm interested in this as well as if we can tweak the ragdoll variables or not.

It was very hit and miss. Come to think of it, since I got this new mobo and cpu I haven't actually experienced that problem. It could have been an issue with my old mobo and cpu, which I know the cpu was bad ( the IHS wasn't on the chip properly and caused it to run at insane temperature's, who knows what other problems it caused. ).

I've had issue's with MSI Afterburner and EVGA Precision with ArmA 2 and 3. Stuff like clocks being limited to low power or UVD clocks that require a restart to fix. I thought it was driver issue's at first but it's persisted with both an AMD 4870, a GTX 460 and a GTX 480. Of course I can't really find an alternative to them that I like other than Overdrive, but I don't have an AMD GPU anymore.

I'm sure there would be a way to work around that.

I agree with you on the actual magnified optics PurePassion. I think that is the type of reload they are trying to mimic with the new animations, a more tactical reload where the weapon orientation remains in the same vector as the threat. I agree that you don't need to keep the sights up, the weapon needs to turn more so that the mag port is in peripheral view. It's more about keeping the weapon pointed on target or towards the threat rather than having to reorient the weapon off target then back on target. I do agree though about the point of view following the optics or sight as being rather unrealistic.

I take that statement to mean that in A3's current state of Alpha, it's hard to say whether or not it's optimized. Meaning that for an Alpha, they may very well consider it optimized since optimization isn't a priority, for a release product however, they may not consider it optimized. I agree that if they intend for the ArmA series to be ran on something like an i7 3980X at 4.8ghz and a GTX Titan for a playable game and frame rate, they need to specify that in the requirements. I also agree that they could be more up front about the issue's and have a little more transparency about what's going on behind the scene's to fix the problems. The only difference is in how you word your posts, how you portray yourself and your attitude. You come off as more interested in trashing them as developers than expressing concern in seeing that the issue's are fixed.

I agree with Maverick. Not to come off as hostile, I have been very adamant and outspoken about this issue, but your posts tend to come off as nitpicking of words and semantics and very hostile overall. I'm very pessimistic of what they can do with the current engine. I know there are aspects of it that need to be rewritten to take advantage of multi-threading and parallel processing, things like scripting that make up an extremely large portion of the game and how it runs from the audio down to the AI behavior routines. So any little bit that they can give us performance wise, I'm content with. I'm fairly confident that they don't want the performance to be a dog, but I'm also fairly confident that there is only so much that they can do given their time frame as well as engine limitations. I'm all for pushing for better, but I'm also not trying to demean them as developers.

I hope it's intentional, though I hope they do a better animation for the underbarrel GL's. Having my gun block my entire view gets quite annoying in a firefight.

I would venture to say that it is interpreted rather than native/compiled. Bytecode is still interpreted, it's just designed more efficiently. As far as what everyone is talking about, Imagine that your script is written in French(SQF) and the game is compiled in English(Machine Code) and your machine only reads English. You would need something to interpret/translate the French to English so it can be processed by the Machine.

I think DnA said they were working with Nvidia and other partners even now. That was how they got some quick fixes and optimizations for PhysX. We probably won't see any big improvements though until release though.

Some of it is rage, some of it is just distrust. Can you honestly blame people for not trusting them when they give the same line for the past 4-6 years or so that it's "been known"? I got all the respect for them for producing games in a niche market, but that doesn't mean that I like buying something that doesn't run on a system that's above or on par with the recommended specs.

I'm just talking about the scripts running concurrently in one thread. I'm pretty sure the scripting is only single threaded.

A set benchmark with key factors that are segmented and isolated so that you can see how each one affects performance and usage would probably help the most. For instance AI with all their different behaviors running separately instead of concurrently like they normally do so you could see if it's any one specific behavior script or if it's a combination of them all together. @Dwarden It's about the fact that 4 years later we are getting the same blanket answer. If you came to me and said "Man your cars performance is terrible" and I said "Yeah I know about it", that's not really an indication that I'm going to fix it, especially if I gave you that same answer 4 years after you initially reported that to me and I gave you the same answer again. What people want is to know it really is actually going to be addressed and fixed. Seeing it persist for 4 years or more doesn't really instill much faith.

I get lowerish utilization on my GPU even without AI present too. It's usually around 40-50% sometimes upwards of like 60%. That's actually really not that bad though. I would expect it to peak around 75-80% just rendering normal scenes.

I'm sure it's a priority, I just don't think there's much they can do without a larger rewrite of the engine. Doing some more testing trying to find out exactly what is causing massive fps drops, I found that the AI behavioral functions are a large cause of low fps with even just a few AI. I did a test with 3 blufor squads and 3 opfor squads on opposite side's of Agia Marina and gave them each 1 Seek and Destroy WP with open fire engage, full speed and Aware, Logged the FPS and then simply changed the Aware behavior to Careless. There was a massive fps difference between Aware and Careless, and you could see a large jump after mission load when the behavior switched. Looking at the FSM's when their behavior is set to careless, it skips a couple of the larger behavioral FSM's. The FSM's themselves are not overly complicated or complex, they simply call upon pre-compiled functions for each state. From what I read, the scripting is done in one thread, correct me if I'm wrong, I'm guessing the behavior functions are single threaded because of this. This could explain the slowdown with AI as a separate function would have to run for every AI unit for every state in one single thread. If the rendering thread is synchronized with the scripting thread, this would probably be why you see a large drop in GPU usage as more entities are placed. I think if that is the major cause of the AI reducing performance, the only way to alleviate it would be to find a new way to simulate the AI, or make the scripting multi-threaded, Both of which would probably require a large rewrite of the code base.

I found .60-.62 to be a good approximation of recoil. I think the bigger problem is that it's not the recoil as much as it's how the recoil settles.

Yeah that's probably the biggest thing.

Good to hear. Sometimes with large downloads file corruption happens. I downloaded it as well after the sale and it installed first time no problems. Glad to hear you got it fixed.