Bohemia Interactive's future ARMA 3!

[dmarkwick 261]

They need to make that giant leap forward to 64-bit only. If they aren't doing the next ArmA/VBS in 64-bit they might as well pack it in. They will be seen as just rehashing the same thing over and over.

64-bit is here. Almost all of upcoming Microsoft products are 64-bit only. In the next three years you will not be able to purchase a 32-bit product from Microsoft. That is their mandate. Windows7 is the last 32-bit home OS.

I wish they would stand-up and be that first company out there to embrace and demand 64-bit only. So what that you will lose some customers up front, but in the long haul your 64-bit engine and game will be light years ahead of the competition like COD:MW2/Battlefield/Crysis/etc... Make it 64-bit / 6 GB of RAM minimum. Once the community sees what you could do with that combination they will flok to you.

Be bold BIS. Get off the 32-bit engine rehash.

I'm going to go out on a limb and guess that you are running a 64-bit system :)

---------- Post added at 12:04 AM ---------- Previous post was Yesterday at 11:58 PM ----------

One thing that would improve ArmA2 beyond anything a mere engine rewrite could do is to implement deferred lighting.

This will fix the "8 lights only" hardware limitations (or whatever the number is, in any case it's quite small) you'll note that the link specifies that lighting performance is independent from scene complexity. That means that, given your PC can handle the scene complexity, the lighting complexity can be UNLIMITED. An UNLIMITED number of lights and lighting effects (negative lights etc).

That single implementation has GOT to be worth serious study IMO.

[Flash Thunder 10]

But sometimes opinions are stupid when you look at the facts.

64-bit is still hardly used these days and not needed for ArmA 2.

2 light sources? Looking at all the towns lit up at night I don't think that's right...

Edit: Oh you mean shadows.

The lack of a feature doesn't mean they can't do it. ArmA 2 is already very demanding. Many of your suggestions would mean more resources are needed to run it smoothly. Or the game world and scale of missions is drastically reduced.

But I'm sure we'll see all that stuff in future games.

This is an Arma 3 thread you know not Arma 2, Arma 3 probably wont even be out for another 10 years. Jeez you people need to relax...............

[Minoza 11]

10 years, seriously? It wasn't that long between Arma 1 and Arma 2...

[Flash Thunder 10]

10 years, seriously? It wasn't that long between Arma 1 and Arma 2...

Seeing how BIS is a small company and that Carrier command is their next game and the upcoming arma expansions etc, its reasonable, they might wait for better game technology.

CPU's in 2020 would be pretty damn powerful, we could see physics?? Actually intelligent AI

10 years maxed but if BIS doesnt plan alot of expansions for A2 then I would say 5 years for Arma 3.

[Minoza 11]

Flash, I see it sooner if it's just what you've mentioned. OpenCL brings that doesn't it? Besides physics, it brings AI calculating too right? What I'm trying to say is that I wouldn't guess what 10 years brings... It's funny because 6 months ago DX11 was ''far'' and 4870 were kings and look at them now... Technology is going so fast man, I think we'll have all that by the end of the next year - heck - 3 years most!

[-Shifty- 10]

The game needs a new engine for sure. Whether or not BIS is up to it I wouldn't know but I somehow think we won't be getting an ArmA 3, unfortunately, all good things come to an end.

The main problem with ArmA is the AI, as it's always been. They need to be autonomous, and take cover and return fire in a suppressive manner when fired at.

[PuFu 4600]

The game needs a new engine for sure. Whether or not BIS is up to it I wouldn't know but I somehow think we won't be getting an ArmA 3, unfortunately, all good things come to an end.

The main problem with ArmA is the AI, as it's always been. They need to be autonomous, and take cover and return fire in a suppressive manner when fired at.

Did you even bother trying 1.05 before posting that??

Do you even know how the suppression system works?

Why does a very improbable A3 need a new engine? Can you link me any game engine available or in the works being able to withstand the size of A2, its AI and the overall complexity? In a SDK format?

BTW: regarding a possible A3.

How long did it take to have A2, after OFP release? 2001 - 2007 (6 years).

What makes you think there will be a new sequel to that in a relative near future (if any), especially knowing that BIS is a small game developer, and they have already started work on another game: Carrier Command? (and that is very improbable they have the manpower to fuel 2 games in the same time)

[echo1 0]

lol

I love that people still insist on calling Real Virtuality "outdated"

Yeah, it's that time of year again...

[Minoza 11]

Well it's not hard to do that, it's scripting part... Problem is mission makers don't usually use good scripts but rather use Arma's engine to do the job which is at times just fine but it can also be horrible...

For example, we still don't have AI in god damn houses or buildings... There's like 2 year old script that changes that but no one uses it... And there's plenty of other examples where people make great scripts to make AI act more natural/real but you got to talk to specific mission maker if you want it implemented in his mission...

I tried to suggest these and similar things to some mission makers out there but with no result... :(

So, with right script you don't even need new engine... well... at least not for AI part...

[Dogtags 0]

Well it's not hard to do that, it's scripting part... Problem is mission makers don't usually use good scripts but rather use Arma's engine to do the job which is at times just fine but it can also be horrible...

For example, we still don't have AI in god damn houses or buildings... There's like 2 year old script that changes that but no one uses it... And there's plenty of other examples where people make great scripts to make AI act more natural/real but you got to talk to specific mission maker if you want it implemented in his mission...

I tried to suggest these and similar things to some mission makers out there but with no result... :(

So, with right script you don't even need new engine... well... at least not for AI part...

The problem for me and many others and I mean MANY others is that it is a pain to constantly add scripts, mods, addons etc to be able to play. Many people love ARMA as I do but (I am over 40 BTW) I don't have the time or by the time I get home from working all day I can't be bothered to muck around setting things up, all I want to do is click load and play online or solo. The addons just get out of hand. I wanted to play some new SP mission so I went to the forums and found one that sounded great but to run it I had to download another dozen addons!

Anyway I play ARMA2 often but only online or SP missions that require no addons, I have to live with the shortcomings. It is a real shame that BIS can't integrate some of these scripts into the core ARMA2 AI code.

Anyway ARMA3 needs to be a whole new engine or a massive upgrade to the current one. As mentioned lighting, AI, popping textures are the main issues, for me anyway.

[Flash Thunder 10]

Flash, I see it sooner if it's just what you've mentioned. OpenCL brings that doesn't it? Besides physics, it brings AI calculating too right? What I'm trying to say is that I wouldn't guess what 10 years brings... It's funny because 6 months ago DX11 was ''far'' and 4870 were kings and look at them now... Technology is going so fast man, I think we'll have all that by the end of the next year - heck - 3 years most!

You dont get the fact that BIS is A SMALL COMPANY like 30 devs or so, and they would have to rewrite the entire engine from scratch to implement something like what you mentioned and then the programming team would have to relearn alot of stuff and get experience coding it.

Plus making these kind of games takes time, to fix bugs, test every feature in the game which would be alot ARMA 3 would probably be twice the size of Arma 2 in content, and thats a Gigantic workload for BIS.

Technology advances fast nowadays, but Developers still need alot of time and programming a game like Arma 2 cannot be done in 2 years flat, unless you want a complete beta of the game that is, I don't.

[Yoma 0]

no need to rely just on Havok

better to utilize P.A.L. (Physics Abstraction Layer) while use as base physical engine e.g. Bullet Physics Library

this way You get easy access to cost effective (no license fees) OpenCL accelerated physics

You can also then interact Bullet with e.g. DMM and accelerate both via OpenCL

while Euphoria is nice

yet any proper realtime (semi/full)procedural animation blending would be enough ...

also there are similar middlewares around with better price tag ...

e.g. Locomotion for walking

the LUA part raise the question is if it's worth it

or why not look for something what may challenge LUA, one of candidates may be AngelScript

some may fight for good old Python

in the end just evolve SQF + allow direct C++ API linking to the application (hint VBS2Fusion)

C++ code build with Intel TBB/IPP/etc or Parallel Studio

combined with DX11 to squeeze max of todays PC platforms

the above is just example that You should always consider other ways

You don't realise the seed of hope you planted sir.

However, there are so much issues (MP client sync, blabla) that must hamper really getting there.

---------- Post added at 02:19 AM ---------- Previous post was at 02:16 AM ----------

You dont get the fact that BIS is A SMALL COMPANY like 30 devs or so, and they would have to rewrite the entire engine from scratch to implement something like what you mentioned and then the programming team would have to relearn alot of stuff and get experience coding it.

Allthough they are small compared to bigger studio's, don't underestimate the brains they have.

[Flash Thunder 10]

Allthough they are small compared to bigger studio's, don't underestimate the brains they have.

I hope they prove me wrong. Arma games are excellent products, but I sure hope the game wont launch with huge issues and bugs left and right.

[Yoma 0]

I hope they prove me wrong. Arma games are excellent products, but I sure hope the game wont launch with huge issues and bugs left and right.

ATM there is no Arma3 announced as far as I know...

I'll be happy to look into their take on Carrier Command, but I personally hope they do make a sequal to OA / make an Arma3.

[Dwarden 1125]

You don't realise the seed of hope you planted sir.

However, there are so much issues (MP client sync, blabla) that must hamper really getting there.

---------- Post added at 02:19 AM ---------- Previous post was at 02:16 AM ----------

Allthough they are small compared to bigger studio's, don't underestimate the brains they have.

don't loose hope :)

some of these were already captured and now being examined in our underground facilities ...

[froggyluv 2136]

don't loose hope :)

some of these were already captured and now being examined in our underground facilities ...

That would be the freakin Muthaship :eek:

[Minoza 11]

I hope they prove me wrong. Arma games are excellent products, but I sure hope the game wont launch with huge issues and bugs left and right.

I honestly don't even see this as a problem since we got this with every Arma out of the box so far... (maybe wrong term - I mean, without patches...)

So as far as they plan to patch their ''beta'' products I'm not concerned about what I get at start since this game is simply too awesome!

[fantsu 10]

I hope that they make ArmA3, but with downgraded graphichs.. maybe original OPF engine? Just make everything that this game has and more, just so that it would work.

[Minoza 11]

I hope that they make ArmA3, but with downgraded graphichs.. maybe original OPF engine? Just make everything that this game has and more, just so that it would work.

Actually this isn't bad idea at all! Really!

Maybe not literally downgraded, but to work more on optimization so when it comes out that we get good and optimized game (that would be a first time for Arma to come out optimized lol...). That would make a lot people happy. I mean sometimes I find really weird that on my (and even stronger CRAZY systems) rig Arma 2 can act like I have crappy low-end PC... At times it runs great, and then for some reason, at times, it runs like crap...

[NeMeSiS 11]

Actually this isn't bad idea at all! Really!

Maybe not literally downgraded, but to work more on optimization so when it comes out that we get good and optimized game (that would be a first time for Arma to come out optimized lol...). That would make a lot people happy. I mean sometimes I find really weird that on my (and even stronger CRAZY systems) rig Arma 2 can act like I have crappy low-end PC... At times it runs great, and then for some reason, at times, it runs like crap...

You could make it look like pong but you will still most likely be limited by the CPU in most cases, giving you a crappy looking game with 0 fps increase. :p

[Minoza 11]

Yeah I understand. Arma tends to torture your CPU... But that's where DirectCompute comes in play, right? With OpenCL capabilities AI calculations could be done on both CPU and GPU if I'm correct. Also, alowing a game to actually use all of your cores, I mean I'm not sure that it uses them now... If so, damn, then it's one hell of a demanding game... Running like that on 2 and 4 cores CPUs...

[th3flyboy 0]

I wouldn't mind seeing most of these, however the losing remaining ammo is a no-go. A better option, and one which can not be currently scripted in ArmA 2, is to be able to set it so that you can reload, and the bullet in chamber remains in chamber, and is removed from the old magazine's count, but is added to the new magazine, so unloading a fresh mag, and loading in a new fresh mag would result in this:

Mag 1: original count 30 rounds

Mag 1: count after reload 29 rounds

mag 2: original count: 30 rounds

Mag 2: count when loaded in: 31 rounds, base mag+ bullet in chamber

Oh, and bullet physics may be a better option than havok.

[Fiya 10]

MulleDK pretty much forfeit his oppinion when he said this.

Ya, I know... I'm dreaming too much.

But seriously. I don't care if I have to buy Roadrunner or pay double the price to get all this.

Perhaps everyone else does care. Performance is poor at best, its been improving and I am sure it will in the future. Yet right now, it would cut even more people out of the community. DMM alone could be a performance hog, depending on how detailed it was.

Second of all, the bugz. I could not imagine what introducing completely new systems into the mix would do. We still have some everlasting flaws since OFP. Man I cannot imagine the bug squashing that would have to happen to make Real Virtuality play nice with the others.

Finally, are you going to pay for the licenses? Havok itself is a pretty penny and I do not know how much a license for Euphoria would be.

To be honest your being pretty ridiculous considering every creator in the world would introduce these systems into there games if it was free.

I don't mean to be a total douchebag, but it seems that is what your aiming at with your initial post. I support BIS, and I can completely understand why they wouldn't use third party software with the next release, IF there is another release of Arma.

Again, my personal opinion. I'm not saying it wouldn't be kickass, (Because it would.) but I'm looking at it from another perspective.

Personally I'm looking for, REAL 64bit support instead of what we currently have. Better multi-core support/usage. Better bug squashing, and another great game.

[LeonAdi 10]

Arma 3 should be a mmo fps military simulation, with 1milion vs 1milion on the same map with real powers USA, Russia, China, European Union, India, Brazil, Japan

Use military organisation, with divisions, brigade, headquarter who will make decision of strategy of war, and the objectives for every division, then for every platoon, squad, team.

For example Chernarussia could simulate a limited American-Russian war with 50000vs50000 battle

[Simon C 0]

Arma 3 should be a mmo fps military simulation, with 1milion vs 1milion on the same map with real powers USA, Russia, China, European Union, India, Brazil, Japan

Use military organisation, with divisions, brigade, headquarter who will make decision of strategy of war, and the objectives for every division, then for every platoon, squad, team.

For example Chernarussia could simulate a limited American-Russian war with 50000vs50000 battle

1 Million vs 1 Million? Well, if you want to melt your CPU...