ArmA 3 and backwards compatibility

[Preacher1974 0]

Assuming there is an ArmA 3, what do people think about the question of backwards compatibility vs new features. BI have built ArmA 1 and 2 on top of OFP, enabling relatively easy porting of scripts and units/vehicles between games. The downside is that ArmA 2 inherited a lot of the limitations and clunkiness of OFP amd ArmA 1. I swear the muzzleflash is still from OFP, for example.

Personally, I would rather BIS just start with a clean slate and sacrifice backwards compatibility for a new engine that eliminates a lot of the limitations of the ArmA 2 engine and introduces new features such as ragdoll physics, possibly a new scripting language (LUA?), new weapon interfaces, better damage models, etc. I know that Arrowhead will go partly this way, but I reckon that BI should go further and just start planning the feature set from scratch without worrying about backwards compatibility.

As a point of clarification, I'm not necessarily advocating coding a new engine from scratch but modifying features of the current engine enough to essentially break old addons and code. An alternative approach to recoding might be for BIS to license another mature engine and customise it for ArmA.

Please vote and discuss :)

Edited November 27, 2009 by [ZSU]Preacher

[NeMeSiS 11]

ArmA3?

[Jockson 10]

There is a long-ish thread on this topic in the suggestions forum. IMO there is very little chance of BIS rewriting the whole thing from scratch IF there is an ARMA III at all which I wouldn't really count on.

[walker 0]

Hi all

Maintaining the same software and improving it is the most efficient and effective method of development. That is software engineering 101.

Only a fool abondons legacy code.

Not one major developer of any software ever does as you suggest [ZSU]Preacher.

Why do you think Microsoft are so powerful?

It is because the keep what they have, refine and improve it, adding new features and improving the efficency of individual modules.

Kind Regards walker

[Preacher1974 0]

OK. I think some of you guys missed my point. I'm talking about starting from scratch from a features perspective and not being constrained by backwards compatibility. The end result would probably contain a lot of existing code (as yes, it's not commercially feasible to start coding from scratch). I'm sure BIS could completely break backwards compatibility with only relatively minor code changes - THAT'S what I'm referring to.

[De_little_Bubi 1]

well i missunderstood the question. I thought "new engine features" so old engine with some new features which will destroy backward compatibility.

the author obviously means a completly new engine.

a complete new engine is out of the question i think but new features in a new game iteration should imho have higher priority then total backward compatibility.

edit: damn it. should watch tv while typing posts. meanwhile everything got said...

[columdrum 11]

i think that backward compatibility its always good, but for example use a normal scripting language( OMG, lua, for example, i really hate the actual array management) would be a good idea :), or focusing a little more on optimization will be better.

[Leopardi 0]

Hi all

Maintaining the same software and improving it is the most efficient and effective method of development. That is software engineering 101.

Only a fool abondons legacy code.

Not one major developer of any software ever does as you suggest [ZSU]Preacher.

Why do you think Microsoft are so powerful?

It is because the keep what they have, refine and improve it, adding new features and improving the efficency of individual modules.

Kind Regards walker

But Real Virtuality is like the Cell processor, a technological dead end.

[Flash Thunder 10]

Engine needs more support of new Operating Systems.

[ActionMan 10]

Maintaining the same software and improving it is the most efficient and effective method of development. That is software engineering 101.

Only a fool abondons legacy code.

Uh, no.

"legacy code" is often a euphemism for "that horrible code that I wish we didn't have to keep around". At some point, the cost of maintaining that horrible code becomes larger than replacing it. At that point, you start fresh.

Look at ID Software, they start fresh every generation...

[Windexglow 10]

Maintaining the same software and improving it is the most efficient and effective method of development. That is software engineering 101.

Helloooo code creep.

[Zipper5 74]

(This reply is all in context if there is to really be an ArmA III.)

If you remember back to ArmA's release, there was almost zero backwards compatibility with OFP addons. What did this cause? An outrage to say the least. It took a very long time for OFP addons to be ported to the game. And this was all because of the changes from RV1 to RV2.

So if BIS were to do it again, you might all say "sure, remove backwards compatibility" but come the release of the game, you'll all be up in arms for them doing such a "ridiculous" thing. If BIS want to play it safe, I'd say that they should make it backwards compatible. Otherwise that will essentially make peoples' hard work in ArmA II go unplayed as the community moves on to the "newer and better" game.

[reconteam 19]

By the time (2011?) of ArmA III (or should it get a new title) I think it would be good to start with a clean slate. I've heard the engine is pretty much based off the same one from OFP, is that true?

[NeMeSiS 11]

By the time (2011?) of ArmA III (or should it get a new title) I think it would be good to start with a clean slate. I've heard the engine is pretty much based off the same one from OFP, is that true?

True in the same way that MW2 is based of the quake3 engine. :p

[Pyronick 21]

Hi all

Maintaining the same software and improving it is the most efficient and effective method of development. That is software engineering 101.

Only a fool abondons legacy code.

Not one major developer of any software ever does as you suggest [ZSU]Preacher.

Why do you think Microsoft are so powerful?

It is because the keep what they have, refine and improve it, adding new features and improving the efficency of individual modules.

Kind Regards walker

Sounds like "640 kilobyte ought to be enough for everyone" to me.

Microsoft is powerful due to it's monopoly position and abuse (vendor lock-in). I.e. DirectX, Internet Explorer, GDI, .NET, etc.

The only reason why Windows is so succesful is because it was one of the very first OSses which had backwards compatibility with DOS. Which was stripped out at a certain time aswell.

If this wasn't the case we'd probably use *nix-based OSses, OS/2, Amiga OS, BeOS or even Solaris/Java. Windows is the worst of all those.

Windows Vista/7 isn't actually backwards compatible with previous NT-based kernels (Windows XP, Windows 2000), instead it's all done on an in-kernel virtual machine.

Microsoft is just mindfucking us all. Don't get fooled by Apple either.

Edited November 28, 2009 by SgtH3nry3

[Bobby2 10]

Preacher;1498724']Assuming there is an ArmA 3' date=' what do people think about the question of backwards compatibility vs new features. BI have built ArmA 1 and 2 on top of OFP, enabling relatively easy porting of scripts and units/vehicles between games. The downside is that ArmA 2 inherited a lot of the limitations and clunkiness of OFP amd ArmA 1. I swear the muzzleflash is still from OFP, for example.

Personally, I would rather BIS just start with a clean slate and sacrifice backwards compatibility for a new engine that eliminates a lot of the limitations of the ArmA 2 engine and introduces new features such as ragdoll physics, possibly a new scripting language (LUA?), new weapon interfaces, better damage models, etc. I know that Arrowhead will go partly this way, but I reckon that BI should go further and just start planning the feature set from scratch without worrying about backwards compatibility.

As a point of clarification, I'm not necessarily advocating coding a new engine from scratch but modifying features of the current engine enough to essentially break old addons and code. An alternative approach to recoding might be for BIS to license another mature engine and customise it for ArmA.

Please vote and discuss :)[/quote']

I agree; New engine with ragdoll. If BI could make the game like it is now but with features Physics and animations like "Red Dead Redemption". It would be a five star game(simulator). Also a new engine might be better for computer hardware.

---------- Post added at 10:51 PM ---------- Previous post was at 10:46 PM ----------

(This reply is all in context if there is to really be an ArmA III.)

If you remember back to ArmA's release, there was almost zero backwards compatibility with OFP addons. What did this cause? An outrage to say the least. It took a very long time for OFP addons to be ported to the game. And this was all because of the changes from RV1 to RV2.

So if BIS were to do it again, you might all say "sure, remove backwards compatibility" but come the release of the game, you'll all be up in arms for them doing such a "ridiculous" thing. If BIS want to play it safe, I'd say that they should make it backwards compatible. Otherwise that will essentially make peoples' hard work in ArmA II go unplayed as the community moves on to the "newer and better" game.

Key words... newer and better. So why would you go back to older and worse?? A new game that plays better out of the box and with the ability to mod and make addons that are newer and better.

[nodunit 397]

I agree; New engine with ragdoll. If BI could make the game like it is now but with features Physics and animations like "Red Dead Redemption". It would be a five star game(simulator). Also a new engine might be better for computer hardware.

What is with you people and thinking people can just build or move to an engine within the snap of a finger, it takes years to develop engines, cry enegine, quake, havoc, hell any engine used in 'big' games to date has been in development for years and improved upon over time.

Simulation isnt about ragdoll (vehicle physics I will give though), it's about how things work and what they can do, what does ragdoll simulate? A more fancy, resource hungry death? No thanks, I'd rather put it towards weapon ballistics and systems.

What makes you think a new engine out of the box is going to be better, or even moddable for that matter? As it stands we have SDK's, a new engine would require resource time in creating EVERYTHING again, and another SDK for the public, top on that you now also have even more to learn.

This reminds me of a conversation I had with a friend about the apache, it is a very costly vehicle to maintain, so why not just build something less advanced and costly? Because it has the same engine as many of its brothers makes it far less costly to begin with, because it is made by the manufacturer of another helicopter the cost in parts is substantially less plus you would need years to do something else..

Newer is not always better, what do we lose as opposed to what we gain and already have.

Edited September 29, 2010 by NodUnit

[ffur2007slx2_5 11]

i feel quite confused when i see the result after voting! Why so many people wanna a new engine rather than improving it? I think BIS can make them done and make the game perfect through consistantly fixing.

I know that maybe most players do not feel satisfied with ArmA2's physical performance. But it does not mean BIS should make a total new engine, you know that's time consuming as well as money consuming! I hope ArmA2 can make their patch chargable, like VBS2 functions for ArmA2 DLC, physical performance enhancement DLC or AI enhancement DLC. lol.

[Iroquois Pliskin 0]

Zipper, do you want the same people to be playing the game, with a few newcomers that don't seem to stay here for very long?

The industry has changed and so have BI priorities, I would believe. Arma needs a proper physics engine if they are to succeed with the sequels.

[-)rStrangelove 0]

If we can do amazing things with new stuff sure go ahead. We learned it back then, we can learn it all again. :)

[celery 8]

In all truthfullness, despite the huge moddability and flexible scripting, the engine of Arma 2 is horrendously limited when it comes to things that are taken for granted in many fps games. The animation system, while giving 1:1 output in and outside your character, is ultimately a hindrance to gameplay and the AI has no soul whatsoever.

Getting a better game without the limitations of the past should be a priority.

[wolfbite 8]

I dont think they need to make a new engine. But it needs some significant upgrades on physics and damage.... And if it means we have to sacrifice the old addons and learn new ways of making them so be it... Its a bit sad to see some of those Ofp addons popping up however many years later in ArmA2

[dmarkwick 261]

I'd be happy if there was mo backward computability, if the reasons behind it made it worth the loss. We had that with OFP/ArmA and things turned out OK :)

DAC would be the only thing I would really miss.

[MrXToTheN 10]

I think the most important part to improve is the enemy AI. It seems to get criticized the most. It's good on a bigger scale, but at fighting man-to-man it seems to be really bad.

Vehicle physics need to be improved too.

If they concentrated on that instead of just implementing more and more additional features and content their games would improve greatly.

[wipman 1]

Hi, i've voted Yea, because keep the actual way of work of the code and models unables you by force to use outdated and caped code & models; that's why we can't fire from the cargo possitions of a vehicle already and things like that you'll think that should be there on "the ultimate military simulation" but they aren't there, or the way that the units walk... that unables you to hit on target at close to medium range aiming down the sights as the weapon sway is entirely wrong, unrealistic and useless. Let's C ya