lozz08

A SSD will only ever show an average frame rate increase because of less stutters, because a benchmark will see a stutter as 0 fps for a short time and therefore reduce the average frame rate for the session. The idea that your storage medium being faster can increase actual frame rate and not average frame rate over a benchmark in this game is bogus. If anything a SSD could show a minutely lower frame rate because its file transfers during streaming happen much faster, and so utilize the CPU much more, but for a shorter time. Ie. With a normal HDD a texture load may take 3 frames to complete, whereas with a SSD that operation only takes one, but it takes that frame a slightly longer time to be rendered because the cpu has a tiny bit more work to do for that frame.

@Kbourne In the video he did say "all the buildings are enterable, and the AI can use them quite well" So yeah, thats got me a little hyped

Get the cheapest phenom II X4 BE you can (i.e. lowest number BE) because while higher numbers may have higher preset clocks, the lower numbers are definitely capable of reaching the same max overclock as the higher numbers (obviously luck comes into it- no two cpus are made identical) What I'm trying to say is that if you can overclock at all get the cheapest black edition X4 you can. And the price difference between the 560ti and 560 is pretty much relative to the performance difference. (~15%) apparently 560 (non-ti)=6870 The 560ti is really beaten by AMD's 6950 when it comes to value though.

It does not surprise me because the maps that came with Operation Arrowhead (Zargabad, Takistan) contained no "block" non-enterable buildings, or if they did there weren't many because I didn't spot them. Personally I would ditch enterable buildings for more convincing towns i.e. better footpaths, environmental effects. I mean, the poor inhabitants of Chenaurus didn't have one supermarket or grocery store between them. No wonder they ended up killing and eating each other

Me and my brother like to play ARMA 2. He lives in a different house but it is only a few kilometers away. When we try to connect in ARMA 2 OA, It takes a very long time for the game to show up if either of us try to host. When it finally does show up, it says "connection failed." What is strange is that we can connect perfectly fine in ARMA 2 non-OA.

Pretty much every country is expensive compared to USA these days. Those egypt prices are only a little over Aus prices at the moment. Keep in mind he may be buying from a retailer and not online (more expensive in most countries)

In OA I can hip fire with an M60 and hit a target 200m away with more accuracy than aiming down the sights. I know that the hip fire reticule can be disabled, but the weapons' fire is still hitting perfectly at the point on the screen where the reticule would be. Will there be something which reduces the accuracy of hip fire in AA3. Because you should need to aim down sight to hit targets at range.

The reload animation seems to be exactly the same and of the same length for all weapons at the moment... Lame.

try this: Go into editor, place a tank, player as driver. In the init field, type: this setvelocity [s * sin(getdir this),s * cos(getdir this), 0] where s is your desired speed. You can setup some pretty cool jumps, stunts and collisions with this. Collisions between tanks and lighter vehicles are fun.

Because in many of today's games, there is an option to enable physx. For instance Mafia 2. Now this game slows to a crawl on systems without an nvidia GPU, because nvidia GPUs are able to offload physx processing, and ATI cards cannot. However, I understand that most of that load is caused by things such as cloth simulation and destruction, which ARMA 3 does not seem it will implement. The RV implementation of physx is going to be "light" you might say. Including features that require relatively little parallel processing (the thing that GPUs are made for)

Changing the mobo is simpler and easier than you may think... And from what I've seen good AM3+ motherboards are running at similar prices to intel's z68 boards. From what I've seen, the bulldozer 8-core only just nudges out the 2600k. (possibly the leaked benchmarks are bogus, but there have been a few released) Seeing as Arma 2 can only use 4 cores, it would seem that Sandy bridge will still be the best performing platform for this game. I hope Arma 3 can make use of 8+ cores. I am waiting till post release to upgrade. Hopefully reasonably priced Intel 8 core processors will be out by then.

That makes me sad that you can't enjoy this great game. I'll wish upon a star for some miracle silicon to come your way.

snoops_213: Get the 955 and a vertex 2. You will not notice the difference between a vertex 2 and a 3 with your system. IMO 60gb is not worth buying; get a 120gb vertex 2. The 1090T will be slower than the 955 for arma2. The new mobo will not help at all; neither the new RAM. The above upgrade will give you a massive boost in performance. You probably won't even feel like upgrading until ARMA 3. And in my opinion buying hardware for Arma 3 would be silly at this time anyway.

Better close range combat abilities is all that is missing for me. Long range firefights are currently the best thing about ARMA IMO. It is true that mods like ZEUS AI have improved AI quite a lot. I wonder if mods like that have ever affected the development of any of the RV games...

I disagree, because I think that delay was because of a fundamental shift from single threaded to multi threaded programming. I would expect the process of utilizing more cores in programming to accelerate from here on.

I'll be interested to see how the creation of physics assets for the physx simulation goes. BIS are talking about an overhaul of the animation system. Integration with physics would mean a lot of screwing around to get things looking right in game.

I have a nice trick. The soldiers who are shooting the civvies- make a looping function that does this: soldier addrating 10000 say, every second. This will make the civvies not see them as a threat, if you can get them to stop lying down.

There must be something wrong with the way you are calling the script. It should be called with nul = this execvm "script.sqf" Possibly you are putting this in brackets like nul = [this] execvm? which would mean the passed vehicle would be _this select 0 and not _this I'm clutching at straws would you like to show us your mission in its entirety so we can troubleshoot? like upload the mission

- Utilizing new zombie scripts Fight off the undead for as long as you can. What round can you reach? - single player mission Required addons: Undead mod v0.84 Fallujah v1.0 Download: http://www.mediafire.com/?zrh8d5vrsjwrs1r Armaholic: http://www.armaholic.com/page.php?id=13500 I would like to thank Charon for his excellent work on the undead mod. It was my inspiration for this project and without it I could not have created the mission. Most of the credit should go to him. Be sure to check out freesound.org- a fantastic site where you can find almost any sound you need that you can imagine. Any and all comments would be appreciated!

I would like to continue to push for separating some AI routines from the 1 cpu core and spreading the load to others. Surely it is possible to move some threads to other cores? The potential performance increase could be game-changing!

Hello all, I've just finished my Volition: Undead mission but I am having real difficulty converting it to multiplayer. The primary reason is that lots of things are happening on the server. For instance, the zombies have a random chance to say something every x seconds. The server is running this looping script. How can I broadcast the command to have the zombie say a certain sound, to all clients? Keep in mind that this zombie only has a local name to the script, _zom, and I read you can't pass local objects with publicvariable. I have the same problem with all the animations using switchmove, all the particle effects, all the sounds.

The co-op is coming along pretty well. I wanted to hold off on releasing this mission until I had something a little more... Spectacular, but I just wanted to get something out there or I was gonna go mad. This mission is just a taste of what is to come.

That is precisely what I was looking for neokitta. So if I want to broadcast any of the commands in that long list I just put r in front of it capitalised? excellent. Will putting any other letters in front do any other cool things? edit now, I do have another mp problem that I should't open a new thread for. It has to do with the scoring system. When a zom spawns, it has this done to it: _zom addeventhandler ["hit", {nul = _this spawn hit;}]; _zom addmpeventhandler ["mpkilled", {nul = _this spawn killed;}]; hit.sqf: if (isserver) then { _zom = _this select 0; _hitter = _this select 1; _score = _hitter getvariable "score"; _hitter setvariable ["score", _score + 10]; _totalscore = _hitter getvariable "totalscore"; _hitter setvariable ["totalscore", _totalscore + 10]; _zom setvariable ["lasthit", _hitter]; _sound = zombiehitarray select (round(random 2)); _random = round (random 9); _zom say _sound; }; killed.sqf: if (isserver) then { sleep 0.5; _zom = _this select 0; _hitter = _zom getvariable "lasthit"; sleep 0.1; _score = _hitter getvariable "score"; _hitter setvariable ["score", _score + 100]; _totalkills = _hitter getvariable "totalkills"; _hitter setvariable ["totalkills", _totalkills + 1]; _totalscore = _hitter getvariable "totalscore"; _hitter setvariable ["totalscore", _totalscore + 100]; _sound = zombiekilledarray select (round(random 5)); "logic" createunit [getpos _zom, "source=this"]; source say _sound; sleep 2; deletevehicle source; }; The reason I am adding the score for a kill in such a wierd way is because it wasn't giving score reliably before. Anyway, clients can't get any score. Only a host-player can. is setvariable local or something? The variables for the player have been defined on mission init, so that's not the problem. editedit: neokita, the code you posted works very well, however, I am unsure why. [nil_or_caller, nil_or_target_object,"loc", script_to_execute, par0, par1...] call RE; [nil, _zom, rSAY, _sound] call RE; why isn't _zom par0, and _sound par1? when I used rFADEMUSIC, this worked: [nil, nil, rFADEMUSIC, 2, 1] call RE; this caused 2 fademusic 1 to be executed on all clients. I don't get why _zom is being put in target_object and not a param.

class CfgSounds { sounds[] = {}; class scream { name = "scream"; sound[] = {"\sound\scream.ogg", 1, 1}; titles[] = {}; }; class crowd { name = "crowd"; sound[] = {"\sound\crowd.ogg", 1, 1}; titles[] = {}; }; }; class thank { name = "thank"; sound[] = {"\sound\thank.ogg", 1, 1}; titles[] = {}; }; }; That extra bracket there has told the script to end at that point. Careful of those. I really recommend this program: http://forums.bistudio.com/showthread.php?t=105860&highlight=squint It will automatically tell you about those errors.

the 4870 requires 2 6 pin as well. If you have a molex connector free then you can use a molex-to-pci power converter, they do exist. How many watts is your ps