Modular approach? Better vehicle simulation!

[scajolly 14]

I suggest BIS start working on Arma3. With it, a new engine. The current one has met many of its limitations (in the eyes of this scrub), most ghastly perhaps the physics handling and collision detection.

In short, a new engine is needed to do away with some of the most glaring descrepancies between real life and Arma - or better, between other simulations and Arma. Try turning a car, dropping a bomb or doing anything else in a vehicle - those things need new algorithms.

I think Arma3 could become not just an infantry simulator, but a military simulator. Right now, Arma2 excels as a simulator for individual soldiers. With a bit of scripting and some modifications, the community has polished the gem even further. The desire is there to keep expanding on this functionality. Unfortunately, I think BIS do not perceive they need a new engine, if they are to provide improved infantry simulation.

What is severely lacking is the will to simulate vehicles. Here is a good rationale, I feel, to promote the need for a new engine. The evidence for this lack of desire is apparent whenever you sit in the Warrior's driver seat, or occupy any number of other vehicles. Most subtly, perhaps, with the moon-like gravity for bombs from aircraft. Armoured units, part to do with the AI, will drive back and forth and turn around like impatient six-year olds.

I argue, however, that to simulate these different methods of warfare demands a new approach. These are not - I suspect - commensurable paradigms unless one runs a supercomputer. Once you're in a ground vehicle, you must be presented with another type of interface than when in an aircraft - and this is separate from the infantry simulation.

A modular approach may therefore be simpler - although I unfortunately don't know jack about making games, so I can't say.

Such an approach would have the game support different manners of handling within vehicles than outside. Can you imagine if we could start our jet's engines like in Allied Force: Falcon 4, use autopilots, adhere to 'real physics', drag calculations, speed profiles, stalling etc? Or similarly in Apaches (collective!) and in Abrams tanks, rev the engine - use GEARS? Or just have a more realistic interface than sighting and looking at... nothing. A helicopter can taxi by adjusting its main rotor; an Abrams does not have 1,650 hitpoints. It's time to leave the compromise of the vehicles behind, and take on a greater challenge. I think this endeavour could produce the highest yield to the developers.

If Arma3 could incorporate enough elements of realism from the worlds of armoured vehicles, helicopters and airplanes, can you imagine the reviews? That would be as groundbreaking as what was launched in 2001. Militaries should desire to use it for training, and Bohemia would gain the absolutely massive audience that is the flight sim/combat flight sim community.

If 'modularity' helps in this regard, I don't know. But it doesn't seem to me as if spit-polishing today's game will see Bohemia into the future.

(btw, this is not for "our Next Generation Game ArmA 2 & the standalone expansion Operation Arrowhead" - it's for a new game. I didn't see any such suggestion-forum, so you'll forgive me if I missed a more appropriate forum!)

Edited February 1, 2011 by SCAJolly

[sickboy 13]

Simple vehicle simulation is a design choice, not an 'engine problem'.

You don't need a new engine for better vehicle simulation, instead, they need to want better vehicle simulation first, and then implement it.

Switching engines or creating new engine, especially when it has to be as versatile as the arma series, is not a small challenge.

And considering real virtuality actually performs pretty good for the open world and all else that defines the arma series, I wouldn't know why they should switch.

If big changes would be required, backwards compatibility would perhaps need to be (partially) sacrified, but still no need for a new engine.

Past years plenty of people suggested e.g switching to CryEngine - yet they serve two very different purposes, aside from the fact that you don't just switch overnight.

[metalcraze 290]

The dude just thinks that Engine is something static and unchangeable, that appears out of the vacuum overnight and when you switch to a new Engine that also appears overnight everything suddenly changes in the way the dude wants.

Why does it always like that in every "omg I don't like X in your game, please use a new engine" thread?

[maddogx 13]

At one point I also fell into the trap of thinking that a potential sequel to Arma2 would need a whole new engine, but looking at what BIS have done since the games initial release, I think that conclusion is completely wrong. With the exception of a few features that would require major overhauls of core game code (for example loading and unloading of addons at runtime), many - perhaps most - of the feature/improvement requests put forward by the community would concievably be doable in a reasonable time frame, within the engine we have now. If the devs worked on one of those features to the exclusion of all else, that is.

The thing is: there are thousands of suggestions for improvements floating around, and BIS is a comparatively small company. So priorities need to be set. Suma and other devs have frequently stated that they decide which features and fixes to work on based partially on demand for it in the community at large, but mostly on the cost/benefit ratio.

So if they really felt the need for new vehicle or aircraft simulation physics, they would probably work to implement them. But that would require a large portion of the community to be dissatisfied with those things, not just a vocal minority who happen to have a special interest in perfect vehicle or aircraft handling.

Switching to a completely new engine isn't going to make things like that happen any faster. Quite the opposite.

Edited February 1, 2011 by MadDogX

[Zipper5 74]

I suggest BIS start working on Arma3.

In the broad scheme of things, don't you think that's rather soon? It was 5 years between OFP and Armed Assault. 3 years between Armed Assault and Arma 2. It's been much less than a year since Operation Arrowhead launched.

I'm happy sticking to Arma 2 for as long as BIS continue to patch it and release content for it.

[Primarch 10]

In the broad scheme of things, don't you think that's rather soon? It was 5 years between OFP and Armed Assault. 3 years between Armed Assault and Arma 2. It's been much less than a year since Operation Arrowhead launched.

I'm happy sticking to Arma 2 for as long as BIS continue to patch it and release content for it.

If you give me something that is as much of an improvement as arma 2 is from Armed Assault, I can wait for those 3 years. BIS didn't make the game in the release day... If they are going to do arma 3 with a new engine I'd suggest BIS to start working on it now so they won't have to patch it so much, so they can release DLC which would make BIS more money.

[wolfbite 8]

Yeah I'd rather they just averhaul each vehicle class at a time with dlc.... The Driving does need a major overhaul I feel. I can deal with planes andchoppers atm.... Although I want them to get a major overhaul....

But I never drive cars in this asdrivings absoloutely shite...

If the cars were like GTA handling I would literally shit a bat of happiness

[maddogx 13]

Yeah I'd rather they just averhaul each vehicle class at a time with dlc...

Actually, that's quite an interesting DLC idea. A succession of patches introducing improved handling and a few nifty features for a certain vehicle class, each coming with a DLC for corresponding content. I'm not a huge vehicle fanatic, but that does sound cool.

Not sure how high the community demand for better vehicle handling really is though. After all, not everyone plays this game for the flying, and the driving is... adequate, I would say. If BIS wanted to create a fighter plane DLC for example, I think they would first have to gauge how many people would actually buy it, just to be sure it would make any sense at all.

[wolfbite 8]

If BIS made an aircraft DLC that seriously overhauled Flying mechanics I would buy a ticket to Egypt kick the shit out of the government and the opposition singlehandedly become Presidente ( It has a good ring to it) and Rename it Bohemiatopia......

[Mr. Charles 22]

They worked on Game2 which has become ArmA2 for more than 3 years...

[-=Grunt=- 10]

A real physics engine would "solve" most of the vehicle problems in my opinion, like the tank and car handling, no more "missile bombs" (Like those lock-on Mk82s and FAB-250s) and maybe some kind of "crash simulation" for choppers, that it doesn't blow up as soon as it hits a pebble on the ground. Maybe better bullet penetration simulation on different surfaces etc? :)

Think of it... No more "invincible" glass windows on buildings.. .50 M2 piercing the enemies' cover in a building!

I guess we'll have to see if BIS is up for this for ArmA 3.

[scajolly 14]

I'm overjoyed I've been able to spark debate with only an uneducated opinion to show =)

Metalcraze (that dude?) will have my understanding - and concession - that it's no black and white issue. When someone without a programming background sees Arma(1) and calls it out on being OFP with a lot of bandaids and with bad stitches, he's not necessarily correct. I began playing OFP in 2001, and the leap that has taken place from OFP to Arma2 has me astonished and amazed. When I call for a new engine, it's not necessarily about starting from scratch. It's about revising the algorithms at its core that produces the laws by which all elements abide. I will condense my case below!

My premise is vehicle-centered first and foremost because I think it will open up a new market to BIS, most essentially the combat sim fans - avid, insane people that build communities and contribute with the years of experience that the open genre has provided.

The question is, will a vehicle handle realistically within the current confines? (eg taxiing choppers, vehicle handling)

The premise here is that if the answer is yes, BIS need 'only' work 'a little' on that. No need for major revision.

Because a vehicle handles today as it did in OFP, I'm lead to believe it is impossible or very difficult. The counter-argument is of course that some vehicles HAVE improved handling: Harrier, JSF, Osprey and Venom all represent different parts of the spectrum of aircraft handling.

If the answer is no, it follows that BIS needs to revise the current constraints under which they operate. It becomes a question of maximum return for each man-hour: How much will they gain by adding xyz features and/or solving problems in the infantry simulation - will they gain more by polishing the vehicle simulation?

What I am saying is that the infantry simulation is excellent, and that where BIS has not produced features, the community has. This is important. It means that BIS can set the benchmark somewhere within reasonable time limits, and then leave it to the community to create mods like ACE.

BIS can -polish- their game, but not add features; we are effectively in front of BIS with our patchwork of scripts and modifications. Therefore I am saying that BIS can create a game with perhaps no more advances in infantry simulation than today's Arma2, but spend time on making it about military simulation.

Whether or not it takes a new engine is entirely tangential to the majority of us, since we (I!) are laymen of no education. What we can promote are ideas we think make sense to the business model, and state the perceived constraints that limit Arma.

Unfortunately the vast majority of feedback on the game comes in the form of "X is broken!", or even "Y is missing!" - it's taken for granted that features are polished and perfected. But when you want to add a feature, I think it makes more sense to promote it as a cost-benefit decision, than just polishing a product. It just happens that with Arma, as I perceive it (and as ten years of little improvement leads me to believe), a great upheaval is needed to get to that benefit.

When judging by the volume of feedback on particular issues, BIS will see that people are concerned with their infantry, with their coding and the mission editing. There's no reason we should complain daily about aircraft or tanks - they've been the way they are for about ten years. It's a natural compromise that the community is complacent with. What I'm stating is that there's an audience out there that was turned down in 2001, and there's an audience out there that has yet to be invited.

BIS can perfect their MP and even approach the CoD-audience to a greater extent (cost-benefit argument for you there), or they can choose to make the Arma series about seamlessly going from one branch of military to another, finding that each rigour of warfare is modelled as closely as their infantry component is today. It may well be outside Bohemia's scope to ever do that much work, but I see two huge arguments against that: The decade from OFP to Arma2:OA, and the Arma2 community.

[Big Dawg KS 6]

A real physics engine would "solve" most of the vehicle problems in my opinion, like the tank and car handling, no more "missile bombs" (Like those lock-on Mk82s and FAB-250s) and maybe some kind of "crash simulation" for choppers, that it doesn't blow up as soon as it hits a pebble on the ground. Maybe better bullet penetration simulation on different surfaces etc? :)

How is this any more "real" than RV's simulation of such physics?

Honestly, comments like this are nothing but ignorant. Simply dropping physics code X into game engine Y does not solve problem set Z.

RV does have a real physics engine. It's real because it exists (yes, real physics calculations are computed in RV, otherwise the game would be an unrealistic piece of crap). By design, it lacks the amount of detail in some places that other common game engines have, but I assure you it's there and it's very real. And what's more important is that it works for what the game needs it to do. Besidse that, there are many instances where ultra-realistic/utra-detailed physics simulations would do more harm to the game than good.

Furthermore, we (community members who are not devs) know nothing about RV's modularity. For all we know, it could be extremely modular by design (ok, most signs point to this being unlikely), or certain parts such as the physics simulation could be integrated into the most core componenets, making it impossible to change. These are the two extremes of course; based on feedback from BI and the evolution we've seen of RV it's certainly maintainable enough for them to dramatically enhance parts of it, but it is certainly not a trivial process. Based on this, I would imagine the devs are content with continuing to use the current iteration of RV as a base for (while developing further iterations) future projects. Also try to remember that a game engine is not necessarily a concrete, immutable object; BI could completely redesign and replace 80% of the core engine code and still call it RV.

Finally, I'd like to address the idea that organizational/resource management issues are more likely to blame for any hinderence on engine improvements than technical limitations with RV. Technical limitations can usually be overcome if you devote enough resources to it, but even easy changes could be pushed off because of a lack of resources (time, money, people, what-have-you) or because it's not worth the investment. Changing to a new technology base probably isn't going to improve that much. I honestly don't believe that BI's clinging to RV is in any way severely impeding the potential of the ArmA series. We've also seen that reven recently BI has been continually growing, so things that may of seemed unfeasible to do before might suddenly become a real possibility.

I realize of course that the majority of my rant was off-topic, but I hope that it will save the thread from countless "New engine" related posts.

[Zipper5 74]

Furthermore, we (community members who are not devs) know nothing about RV's modularity. For all we know, it could be extremely modular by design (ok, most signs point to this being unlikely), or certain parts such as the physics simulation could be integrated into the most core componenets, making it impossible to change.

Ever seen the many new features brought to VBS2 regularly since its release? It seems to be more of a "you provide the money, we'll get you the feature" scenario. Unfortunately there's a lot more money catering to armies than there is with catering to military simulation aficionados. :p

[Big Dawg KS 6]

And now for on-topic discussion.

I will admit that I have pictured such a modular simulation being pretty cool, but I don't really agree with a lot of your points.

ArmA (and all it's incarnations) is about Combined Arms; at least that's where it excels and that's where (IMO) the devs should focus. Enhancing the different simulations would be welcomed, but in the end they have to integrate well with eachother and have the same level of attention. I mean it really isn't worth anything to me if tank combat is extremely detailed and ultra-realistic if as an infantryman I can't appreciate it. I would agree that this approach might be very useful for military training or other serious markets, but for the entertainment market it's just not worth the investment.

To be honest, I find ArmA 2 to have enough gameplay/game mechanic features for now. I'd rather see BI shift focus to polishing what they have and improve the flexibility of the game (of course it is already miles ahead of most games in this department). Polishing the game will enhance the experience, and providing more flexibility to modders/addon makers will extend the scope of that experience in other contexts. That isn't to say BI can also take advantage of the flexibility and release some nice quality DLCs to expand said experience.

And I don't really agree with your point about ACE, or letting the community sort of "pick up the pieces". Most modders (and myself) will tell you that they'd prefer to use what BI has already provided than to throw together a bunch of hacks (ACE has many "hacks", and as talented as their members may be they are still just using what limited tools they have to cover up limitations imposed by BI), provided that BI takes advantage of their ability to make said things more flexible. For example; who would want to spend hours trying to code a complex simulation of armor penetration when BI can more easily provide an even better solution that's more flexible? Of course when I say flexible here I mean it in a sort of software engineering sense, but of course extended to include support for modders too.

---------- Post added at 02:00 PM ---------- Previous post was at 01:55 PM ----------

Ever seen the many new features brought to VBS2 regularly since its release? It seems to be more of a "you provide the money, we'll get you the feature" scenario. Unfortunately there's a lot more money catering to armies than there is with catering to military simulation aficionados. :p

I try to avoid mentioning VBS in a discussion about RV, mainly because I'm not very knowledgable (and I know others on this forum are much more so). But I would argue that this is part of the organizational aspect: BIA/Sim (or whoever's involved these days) have the resources and the desire to incorporate a more modular design; especially since they are in fact catering to specific contracts/requirements.

[scajolly 14]

An aside, Dawg: When you label the comment ignorant, I hope you realize we're all engaged in making only educated guesses. Not only can RV be interpreted not to be a "real" engine because it models the gravity of the moon, but many will find the simplification an easy way of discussing unknowns without making fifty reservations.

The discrepancies I mentioned earlier are the ones that stand out the most. Until Arma chews as much as it has bitten over (simulates equally well all its different methods of joint warfare), it can make a lot of sense to disregard RV's as a wholesome engine, especially so long as some 'basic' physical behaviour like turning a car is lacking.

More or less every point that denies the need for a "new engine", 1) directly draws on the assumption that the current iteration can be modified or used to create the requested feature, and 2) INdirectly assumes that it is sustainable in the long run to continue this policy.

Obviously we can't produce a good argument based on these, unless we have some intimate knowledge of how Bohemia and RV function. I see it as a question of time, whether Arma3 can be postponed indefinitely.

In all honesty, I had hoped to inspire another one of those "new engine" discussions. Not because I in all honesty believe a new engine is needed, but because presenting the thought can yield useful answers: "NO, you idiot." - "Well then how?" - "THIS is how you do it -----", for instance with regards to technical details about modularity; entering 'new' simulators when crossing from a Rifleman to an Apache.

When, inevitably, the answers are "No, what we have is good enough", I can only ask if they will be good enough forever. Turn it into a question about finance, the best argument there is towards a business. I choose the niche of flight sim enthusiasts, to encourage people that fly Lock On and other games to join an atmosphere riddled with cooperation, customization and open-ended game features, and that has a caring developer.

From my POV it's a decent argument, and it isn't thoroughly thwarted when people say it'll be an expensive venture into the unknown. To keep status quo + patches, I imagine does not demand an entire Bohemia office, but the revenue can't be said to be phenomenal. If anything, it keeps them alive, but how long? At some point we/they need an astonishing ArmaX, but with another ArmaI disappointment in between?

To me it therefore seems the most logical to support Arma2 technically, but to focus the vigour on creating something that allows them to take a bigger bite out of the market share. With Arma2 they brought a more beautiful game with massive islands, and a system that was MP friendly and handled addons very well. While eg. the ACE team's efforts could (and should!) be honoured in a new game, I imagine Arma3 needs to appeal to more people than those of us already loyal.

The underlying premise is of course that Bohemia want to continue their Arma (infantry combat sim) franchise. I hope they will because they are the best at it, but I can't measure their 'tiredness' from over here.

[Flash Thunder 10]

I would like to see at least one more expansion before arma 3, the hardware is still not good enough/ not a big percentage of BIS customers have i7's and the lastest GPUs.

OP, place a sign down in the editor, shoot it and see what it does.

Its affected by physics in a very realistic way, BIS could add more physics simulations at anytime if they wished.

[dmarkwick 261]

Some kind of well-documented modular design would be welcome I think. What we would be talking about it the transferrence of certain datas between engines, not some magical insertion-of-game-A-into-game-B thing. So for example if we choose something appropriate like the excellent DCS Black Shark sim, we might assume that the similar-looking game engines have similar properties that could be traded, especially if they're designed to.

For a start, some sort of map trading standard needs to be in place, so that DCS uses it's equivalent of Chernarus, or perhaps even Chernarus itself if the DCS sim could use the BIS engine as a skybox-type affair and composit it into the sim, KA-50 cockpit rendered first with ArmA2 world displayed outside, this would mean two engines working at the same time (although the DCS world engine could be disabled).

For the DCS' part, it'd need to import all ingame units and their movements, as well as their ordnance, and export it's own position and movement, and also export any ordnance it sends out.

In the ArmA2 game world, the standard ingame KA-50 would represent the DCS one, updated just as any KA-50 is that's used by a client. In many ways the ArmA2's side of things is the simplest :) it needs to send & recieve mostly the same info it already does, DCS-specific requirements notwithstanding.

One of the big problems though is of the subsequent fidelity imbalance, the KA-50 as simulated by DCS would be up against much simpler units, although heck we got to start somewhere right? :D

Edited February 2, 2011 by DMarkwick

[Macadam Cow 1]

I trully believe this idea should seriously be considered by the devs.

it's the only way I see to properly increase their respective community.

For the moment each sim dev is caring about his own tiny niche.

IMHO (as stated in another post) even if ArmA had the best graphics, the best physic, no bug, thousands of units picturing every single country in the world, it wouldn't be much more popular.

Simply because it's a simulation and there's an huge gap between simmers and arcade players.

Someone whos playing DCS is more likely to enjoy ArmA then someone whos playing an arcade FPS.

We, simmers, all share the same taste for effort, learning.

We don't care if we have to wait for several week before achieving something decent in our favorite game, we actually love that.

So I only see 2 ways for increasing the number of player in ArmA :

1. Turn it into an Arcade game

2. Make it attrative for others simmers.

And I really hope they'll choose the second option.

Just imagine BIS associated with ED and another studio in charge of the tank/vehicles part.

Potentialy every ArmA players could (and probably would) buy :

ArmA + DCS + the Tank mod

Every DCS player would buy

DCS + ArmA + Tank mod

And so on.

Would be a big win for each dev.

I know this would require a full cooperation between them, sharing their "secrets". But where's the problem ?

None of those are in competition (well in fact BIS is potentitally the only one who could represent a decent competitor for others devs)

They're all PC only developpers, and as everyone knows the PC is on the way down and it will probably never be as popular as it was.

My guess is in a few years (maybe 10, maybe less, maybe more) only titles we will see released on computer will be simulation.

This will make a huge difference, consoles will become more and more polyvalent, people will be able to go on internet, share their pics and videos,...

Nowadays teenager who are playing on consoles might never get a computer.

Don't forget the video games industry is the biggest industry of entertainment in the world, above music or movies, and this is mainly due to the console market.

Devlopers have to adapt, they have to catch this opportunity while they have enough ressources.

Edited February 2, 2011 by Macadam Cow

[-=Grunt=- 10]

Well.. If a "real" physics engine is in RV at the moment, most of the annoying things should be fixed to have a successful modular approach or vehicle simulation in my opinion. Such as a speeding M1 Abrams hitting a UAZ wreck and the wreck either flies 100 feet into the air or stops the 70+ ton tank in a split second, the ATV (Don't really have to explain this one hehe), bombs "slow falling", and more.

Better physics "tweaks" by BIS in the future (ArmA 3/next expansion) would be really good. The vehicle driving is alright at the moment, but it could be improved a lot by removing annoyances such as the ones above. I'm not asking for ragdolls, destructible environment ala BFBC2 and fancy things like that, just one that is implemented nicely. (By "nicely" I mean without most of the annoyances that break immersion)

[Pyronick 21]

The collision detection part is very awkward. But other physics attributes seem decent or playable.

Vehicle dynamics are horrible though. Currently it seems acceleration/deceleration is based on some sort of linear speed/time increase. At least, it very much looks like that.

To make the game more realistic it should at least have dynamic torque curves and various gear ratios to mimick transmissions.

Suspensions don't seem to react on lateral forces. There are no tyre dynamics.

At the moment it just seems that the height of the speed linearly increase turning radius.

An easy fix for the collision detection would probably be a higher frequency of collision checks.

[khugan 24]

Hello OP hope you don't mind me using your thread to post about a feature request ticket at CIT concerning vehicles.

I am requesting a crawl speed that can be reached by a key to essentially shift gears into low so that tactical vehicles could then creep along at infantry walk speed. I realize you can tap "q" but try that in a Warrior or Bradley... then try it down hill. It becomes an exercise in frustration trying to keep pace with human players when moving as a mechanized infantry team. I think the normal "q" speed/turbo and just normal forward speed should remain, but give us (shift 4) to shift into low, like engaging a PTO. There should be an initial jerk as it kicks in, then a smooth slow (maybe high RPM) movement and turns.

RL

You couldn't pay me to walk in front of a vehicle in Arma that way. :D He never gains on the guy walking.

Anyway, here is the LINK to the ticket. If you agree, go vote it up.

[John1000 10]

Actually, that's quite an interesting DLC idea. A succession of patches introducing improved handling and a few nifty features for a certain vehicle class, each coming with a DLC for corresponding content. I'm not a huge vehicle fanatic, but that does sound cool.

Not sure how high the community demand for better vehicle handling really is though. After all, not everyone plays this game for the flying, and the driving is... adequate, I would say. If BIS wanted to create a fighter plane DLC for example, I think they would first have to gauge how many people would actually buy it, just to be sure it would make any sense at all.

Here is a nifty solution tha Will Solv ALL Problems for you :Make the Dashboard instruments as open Plug-ins that can modifiable on the fly by anyone and in a format that corresponds with this:https:/https://picasaweb.google.com/112520604661993750465/SenastOverforda04#5741189764768935394/picasaweb.google.com/112520604661993750465/SenastOverforda04#5741189764768935394

If you make it as a Flight simulator add on called Fs Panel Studio and check out the new "X Plane v. 10.00

currently availeble you find all sorts of ways to make fantastic looking Dashbord as well as GPS,fully functioning

missiles systems and a lot of solutions on how to implement things in a simulator envirionment.

:cool: