Arma 3 Engine - What would have been a better option and what can we learn?

[lightspeed_aust 681]

Another thread by someone who clearly has no concept of engines and the requirements for Arma. It seems so easy to just pick up and move to a new platform. Reality is soooo different.

[calin_banc 19]

... AI talk...

I was in the first showcase - infantry and then combined arms. With AI settings turned up, they could hit me and most of the time one shot killed me even when I was just looking a little bit above a rock. Seeing through vegetation was easy, shooting like a sniper or DMR a pieces of cake for them with an AR. It was just a matter of seconds, most of the time I couldn't get any of them before I was put out. If I stayed behind a little, the squad died or I remain alone before I could make it to a relative "safe point". They see you way too easy, shot way to good and at that point they are pretty much USELESS. That's the problem. It's somewhat "independent" compared to other games, but cheats more than a human player who intends to do just that.

I prefer a more scripted AI that at least gives the impression of a human being rather than this. Of course, is playable if you put them on lower settings, but you may come into that "dumber than a brick" situation. I don't know, I would prefer that the engine would collapse at 20 GOOD AI that could really stage a decent performance, not at 100 or 200 that will collapse your nerves faster. :)

But, AI aside, there is no excuse for the poor performance on MP when there is none on the server. Plus, I have no issues running the game with settings turned way up and 4k+ viewing distance. That's not the performance killer, it's the AI and poor MP code.

Edited September 30, 2013 by calin_banc

[Leon86 13]

What engine would be good for arma? all the "big" engines seem to be mainly about graphics.

In battlefield everything agianst infantry happens clientside, vehicles shoot slow projectiles that are done serverside, but it's dodgy, tv missile flies through helos or hits a tank on one side, takes out the reactive armor on the other side. Buildings can be destroyed, but it's always the same animation and result, AI is terrible and only works in waves. Push a box somewhere in multiplayer and it'll be in another place on another client, they dont exist on the server.

Cryengine has decent AI, although it gets it's stupid ass stuck on terrain a lot, not sure how it would perform if you put a significant number of them on a big map.

for both I wonder if they'll be able to have decent performance with lots of ai active in multiplayer, which is the main problem in arma. Cryengine and frostbyte could possibly perform a bit better when it comes to rendering lots of objects, which would be nice, but I dont think it'll solve the multiplayer performance problem.

Maybe the network bubble they're working on for dayz will eventually make it to A3, would be interesting.

[clydefrog 3]

Maybe the network bubble they're working on for dayz will eventually make it to A3, would be interesting.

Whatever a network bubble is, shouldn't it be the other way around? Is DayZ their priority now?

[instagoat 133]

Are you talking about Outerra? Lol it's still Work in Progress but they will eventually get to that point. Far better performance and the game world is Earth... Enter longitude and lattitude of any area in the world and it gets created in game.. What other game does that?

And in the end it'll turn out to be an engine that can do one thing great (dynamic world generation) and sucks at everything else.You can't have everything wrapped into one, and people generally seem to overestimate the potential that is in these engines. There is a reason why so much vintage stuff is still around: it is known, it is documented, you have an industry backing the technology up as far as hardware, and thus it is less risky to develop on it. Since the potentials are known, people know what they can do on an engine and what they can't. Very few companies today use their own in-house engines anymore, simply because they are a pain to work with when doing new game concepts.

Example: OF:Dragon Rising. A racing game engine with a 64 entity cap, streaming problems, memory leaks that they tried to build an Arma competitor on. The result was a failiure.

Changing from RV to another engine would be suicide, especially to a new engine like outerra that is undocumented, untested and unproven.

[clydefrog 3]

Changing from RV to another engine would be suicide, especially to a new engine like outerra that is undocumented, untested and unproven.

Sticking with RV for another Arma game - say 5 years from now- could be suicide as well. If it can't be altered to be 64bit or utilise multicore CPU's properly (or make use of any of the other advantages of contemporary hardware at the time) I'd say it's time could be up by then. They surely aren't going to try to base the whole life of their company on 1 engine that doesn't keep up to speed with advances in technology forever are they, that would be suicide. So if RV can't do that then BI will have to develop something else / have something developed for them that can.

Edited September 30, 2013 by clydefrog

[froggyluv 2130]

+1 InstaGoat.

Not only that but there is not one other engine that can do what RV does -huge terrain non-player centric shooter with real ballistics and the potential for 100's to if not thousands of AI infantry, seamen, airmen, tanks, helos, planes, subs, bicycles, fish..........

When I see an engine doing something remotely close to this, let alone beating RV ...I'll take the request more seriously.

If it can't be altered to be 64bit or utilise multicore CPU's properly

I have VBS2 64bit and it does not perform any better -fact is, it performs worse -way worse.

Edited September 30, 2013 by froggyluv

[das attorney 857]

Whatever a network bubble is, shouldn't it be the other way around? Is DayZ their priority now?

I think the network bubble is to help stem the tide of Memory Hacking, where the "hacker" will quit the game, read their RAM and then know all of the positions of things like vehicles and tents. That way BE can't do anything about it.

The network bubble limits the amount of information sent to a client so they only keep positions of items within a say 2km distance of their avatar. They can still read the RAM on exit of game, but with far less information.

I would have thought this would decrease network traffic, but place extra load on the server.

You can do something similar in Arma 3 by setting a script on the server to only enableSimulation for units and game objects within a set distance of each player. (This decreases bandwidth but in theory you could still hack the memory and learn the location of every object in the game). However, in Dayz, it's more critical as the gameplay is built around hiding things, whereas Arma tends to be more about general unit positions.

As for priority, Jay Crowe has bemoaned the lack of programmers at times, so you'll have to draw your own conclusions. I would imagine they will share any applicable code though (in time).

[calin_banc 19]

"Down to Earth" in Outera (jeep, all terain, sedan) -

.

Looks quite good

Anyway, with enough man power RV could head the same way plus Outera needs some love as well in order to be at the same level of detail as ArmA 3, on the ground at least.

[LoneCrow66 10]

somebody mentioned Outerra Alpha engine in this thread.. they have a tech demo available if anyone is interested.

http://www.outerra.com/

Holy crap that is awesome.

[doveman 7]

Lol, not like in Arma. Arma is FAR worse than DCS. I can play DCS on a laptop. Playing Arma on the same laptop, I get about 5-8 FPS. And thats on ALL Low.

Besides, those people complaining about stuttering are using SLI and Crossfire. And we know all games can have stuttering problems with that.

Well on my desktop (see sig) DCS still drops to a stuttery mess of 8 FPS sometimes, mostly in battle when there's a lot of projectiles flying around (seems to be an issue with large amounts of Objects), which makes it impossible to operate effectively. I get around 40 FPS away from the airfield but it drops to around 21-25 FPS when I'm approaching it or looking out my side windows on the ground at the start. I can only hope that things will improve for me with EDGE.

ArmA seems generally OK in SP but I've been on multiple servers where it's dropped to 10-15 FPS, which is unplayable for me.

[aoshi 1]

I only think they need a better shadown (that work with flash lights fires etc, and a more soft one)

the light shaft in dev is a good adition 've seen better but it's still good

[tacticalone 10]

I did not know about the game engine thing when i purchased the game back in march, if I would know this at that moment i would wait until release, and probably would not buy the final game.

However the game itself is not a total disaster, is actually a good thing, i really enjoy a lot of things on Arma3, the realistic thing overall.

1- Bohemia should stop adding new things/features on arma3 and focus from now to fix what we have.

2- Bohemia should think about porting this game to linux so they can avoid the windows overhead and be ready for steamOS.

3- complaining about game campaign is no sense.

4- Bohemia is not a small company.

5- the comunity leads should be the first ones complaining about this, and not be such bootlicker.

Regards

[dayglow 2]

Wow 3 whole posts and every single one throwing insults around. Don't expect anyone to take you seriously.

[tacticalone 10]

Wow 3 whole posts and every single one throwing insults around. Don't expect anyone to take you seriously.

I think I can live with that. Hope Bohemia can do real changes, like hire a better developer or something like that, maybe the project lead is the problem here.

[Opticalsnare 12]

Outerra is pretty nice.

http://www.youtube.com/watch?v=afFb2JU45gA

http://www.youtube.com/watch?v=Xfo3uGPNLd0

Edited October 1, 2013 by Opticalsnare

[big_t 1]

Well, Dean Hall said a game panel recently, that if battlefield would offer modding tolls, ArmA would be pretty much done. Not sure if that would actually happen, tough you never know.

There is the "BF3 Real Life Mod" /hack that takes out much of the flashy neon hud and other graphical nonsense that seems to make the game look a lot more immersive and playable. I think a fully moddable BF4 could be pretty decent. I don't really get the idea of blocking a game from becoming moddable, to me the worst that could happen is more people would be interested in your game. I'm sure EA has their reason though.

Edited October 1, 2013 by Big_T

[dayglow 2]

I think I can live with that. Hope Bohemia can do real changes, like hire a better developer or something like that, maybe the project lead is the problem here.

If you want to share your awesome skills then you could download the Dev Branch version and participate in that forum and make suggestions.

[tacticalone 10]

If you want to share your awesome skills then you could download the Dev Branch version and participate in that forum and make suggestions.

Sure i will think about it, thanks for the suggestion :j:

[fabio_chavez 103]

I don't really get the idea of blocking a game from becoming moddable... I'm sure EA has their reason though.

this one is really easy to answer, the purpose of the no-mod policy is to systematically eliminate homebrewn competition to your sales of overpriced DLC of inferior volume and quality and to decrease the average lifecycle of your product so that you can shove the next marginally different major title up the numbed throat of your cheering customership, within 24 month. I would add, that according to this, EAs "reason" must be described as "voraciousness" but the truth is, EA is not a person, its a pretty huge and powerfull corporation which motivations may not be described best within the frame of human reasoning but in the frame of soulless moneymakingmachines submitted to shareholder interessts and dedicated only to suck as much money out of the pockets of teenagers as possible.

you are welcome.

[smokedog3para 365]

@opticalsnare been watching that engine for the last couple of years defo the way to go in the future for big open worlds with good fps can't wait for it to be fully functional.

[fumanwho 10]

this one is really easy to answer, the purpose of the no-mod policy is to systematically eliminate homebrewn competition to your sales of overpriced DLC of inferior volume and quality and to decrease the average lifecycle of your product so that you can shove the next marginally different major title up the numbed throat of your cheering customership, within 24 month. I would add, that according to this, EAs "reason" must be described as "voraciousness" but the truth is, EA is not a person, its a pretty huge and powerfull corporation which motivations may not be described best within the frame of human reasoning but in the frame of soulless moneymakingmachines submitted to shareholder interessts and dedicated only to suck as much money out of the pockets of teenagers as possible.

you are welcome.

+1

I think i love you

[txalin 2]

Well, Dean Hall said a game panel recently, that if battlefield would offer modding tolls, ArmA would be pretty much done. Not sure if that would actually happen, tough you never know.

Like on bf1942/bf2 times? Arma is still here.

Looking far away from mods, graphics and stuff like that i remember a post i typed when ArmA? can't remember, comes out, and i think is still valid, so let me retype it again, everything down here is my personal opinion, based on being "working" on gaming industry and being computer engineer for more than 10 years (damn, i'm too old!)

BIS is an amazing company, but has an amazing problem. They focus on making the game they want, not the game that market demands, and on this point i'm not talking about having a Arma: COD series. The key problem is that, as BIS is making almost the same type of game since 2001, looks like that they can't change anything on game creation process. Since OFP, BIS releases ALWAYS had the same problems:

- Bugs: Sometime more, sometime less, but we always had a big issue on release: On Arma 3 is game performance and game not using all cores available.

Solution: Hire a good test team, deploy new test procedures, white/black boxes, etc... The impact on BIS sales of this point is critical. Gamers all arround the world always think that BIS games on release are always bugged, so they won't buy it until patch 3 or 4. BIS, you are loosing tons of money with this problem, instead of updating graphics you need to focus on having a product as bugfree as possible, example: On my first MP game i saw the no blades on chopper bug. ¿No one on BIS saw it in all this years of development, seriously?

- GUI: For god shake, we have the same gui since OFP, seriously, can't be updated? take a look to Battlefield and his radio communication!!

Solution: Hire a good GUI creator. Let me show you a good example, take a look:

http://www.gamesajare.com/2.0/wp-content/uploads/2010/07/excel_teclas.png

Those are all the possible keys combination used for making all actions in arma 2. Can you imagine the face of that newbie guy that wants to join to the arma multiverse when he saw that screenshot? Making things easier for new users is not related to lower the simulation quality of the game. BIS need to revamp the whole GUI and reduce the amount of keystrokes needed for playing the game. A good example on this point is Battlefield 3. You can do this actions without using a keystroke: http://battlefield.wikia.com/wiki/Commo_Rose . Ok, i know that we can do more thing in arma 3 than un BF, but the key point is that BIS need to make the interface more user-friendly, we cannot have a 2001 gui on a 2013 game..

From my point of view improvements on those points should be mandatory for BIS, but, as i said, it's only my point of view

Edited October 1, 2013 by txalin

[froggyluv 2130]

^^^ Have to agree on the GUI. Although I'm fully accustomed to it, was playing with an old Warband friend who is older and intelligent gamer-but had no clue how to control our AI squad. I was shocked and thought less of him but then realized it is pretty arduous to figure out and it really isn't a 'fun' command system.

By 'fun' I mean clear, concise, snappy to look at and use.

[Heebiegeebie 10]

"What would have been a better option?"

I really don't know.

"What can we learn?"

That a smallish company trying to balance financial constraints and make as much money as possible at the same time (i.e. appeal to the widest possible demographic) is never going to make the MilSim we all want to play.

Look at how many major requests by the community were ignored. I'm sure these were seen as necessary compromises given their monetary, time, personnel (and god knows what else)... etc constraints.

The answer?

Crowdfunding. Look at games in development like Star Citizen, the ambition is huge and can only be realized given the funding model they've gone for. It's the community that makes Arma worth playing. They put in a s**t ton of work for the love of it because they share a passion to play the game we all want to play. Given a decent engine, funds going to the real Arma developers and the Suits out of the way, it could happen. Arma with no compromises... think about it.

Cheers,

Heebiegeebie.