While I do get what you are meaning, I have to disagree on the 'it's just big', in comparison to RCT (which I love BTW, my daughters make some wicked coasters just trying to make me sick, ) Your examples have a lot of 'just's in them (implied simplicity). 'Just make a graphics engine, tie the spacial aspects of that into the occlusion mapping for the sound, and AI logics for nodeless cover and in-building positioning, and ballistics, and distance for the sound.. AND make that optimized for 1000's of troops.. AND make it optimized (cpu loading auto adjust) so I can see everything 10k out. And the rest is easy. Just make the AI work alone, or as a team, or as an army, as close to lifelike response as possible, in an open, all possibilities exist, non scripted environment, whether under my command or that of an AI commander (initiative) to my satisfaction. And just make your sound system capable of handling all these units, and said spacial anomolies, scale... I'm tired, there is simply too many subtleties when you play, to ignore that the scale alone, and allowance for it's effect on EVERYTHING else is simply staggering. And no, just no. It's not something that simply works again and again like in BF/COD on a 500x500m map. The fact they programmed past the 32bit memory limitation alone speaks of complexity (see the BIS blog).

The 'one success means all success engine' argument doesn't hold water. As a fact to that point, you are about to find out how hard it is when you see the neutered outcome of OFPR. They followed your idea, the simple 'pop it out of the can and put it in the pot' method, using a successful game engine- Ego, for something it's just not meant to do in scale. May be bug free out of the box, but will not -at all- be capable of what Arma II does.

IF/when you get OFPR, please place even 600-700 units in a pitched battle and fight it out in MP with 40-50 of your mates, THEN we can chat about simple or complex, and why it's bugged and being repaired. So, for what it delivers, I can deal with the birthing pains, hope you can manage it too. Hope to see you in the field.

Have a good one.

Well, if you don't care for bugs...
Imho even a very few unfitting game elements have a severe impact on my game experience.


The "complexity" you are describing isn't "complex". It's just big.
But big doesn't mean a lot more bugs, especially it doesn't require more testing. Wether the GFX system renders 10m correctly or 10km - I just need to look at the rendering and I see, wether there is a bug or not. As well with the streaming: Make a map with the maximum number of handles and test, wether it works - the process ist the same, no matter how this number actually is.
So while a game of this style is definetly a pain in the ass to optimize or, more correctly, to develop an engine, that makes it playable at all, it shouldn't be so much harder to debug. Especially as any program running so many times to handle all these objects should be a quite short program.
You might even raise the point, that a sandbox game is a whole lot easier to debug, as there is no linear sequence of actions expected anyway. (again: It can be a lot harder to develop, for example every script has to be bound to an area and this area must suit it from every angle, while in a limited corridor you can place a single switch at "the" entrance - but once you have it working, it works you don't have to care for people finding a way of skipping your "entrance". The only way I could imagine for this to result in an increased number of bugs, is when you don't fix them on purpose -because of complex code?-, but deliver a program with known bugs. But this certainly doesn't attrack my sympathy...)



BTW: RCT has to handle AI for well over 1000 independently operating visitors, basic physics of dozens of attractions and an I-suppose-extreme-streaming-load, as every single animation step, movement or change in viewing angle is not a simple command to the GPU, but requieres the display of a different, prerenderd sprite.
Admittedly there is no 10km view distance and destructability, but still some high numbers giving the impression of complexity. (and certainly a lot more of optimization, as all this runs fine on a couple of 100MHz)

You speak of two different things, enjoying the product (which depends almost entirely on the subjective experience and tastes of the user), and bug free. Not the same thing. With Arma2, I'm having a absolute blast (subjective), with a reasonably bugged product.

And Sooo.. These games you speak of had the complexity of:

1- Entire core processes to handle streaming terrain with over a million 'destruction interactive' objects on it?
2- Supports 1000's of unique AI (700-800 visible at any moment, much more and you'll melt an average GPU) that use raycasting to find cover, and can engage on the ground, fly, and drive (def. needs patching, but you get the idea)
3- Optimized GFX system that can render all this to 10km, and is quite future proof. You get more bang for your buck over time.
4- Ballistics for ALL those units over ALL that space (you can cover the entire island with artillery)
5- Burdended by that, you still can reasonably have 50-60 people playing tactically with or against AI/players using the existing netcode.

Of course, there's more but I'll leave it at that.

I think not. You are compairing a neatly packaged shoebox for one person, with an open sandbox for the whole damn neighborhood. Honestly there is nothing that you CAN compare this to. OFPR isn't going to come close to this level of burden, or possibility. This really is the only game of it's type with the scale and flexibility it shows.

My opinion of course,
Cheers

System Specs:
ATI 3870
Phenom quad core
6gb memory
1.2TB RAID array

yea...me too. Until 19th of this month, the game won´t be released here. It is going to be a long waiting...

With STALKER Call of Pripyat, one of my most fps expected games for this 2009.

thanks.. i will be waiting for that..

You don't have to look back this far. Famous American McGee's Alice was already quite sophisticated and "needed" exactly one patch - to fix a few graphic-glitches on exotic matrox-videocards. And for small companies: RollerCoaster Tycoon is still consiederd a very good game and was developed entirely by practically one person. Oh and as far as I see, theres only one patch and that this deals with a minor bug, that can appear in a few scenarios in an add-on and only unter Win2000.
So while its certainly not easy, to make a complex, bug-free software - it's not impossible to have practically everyone enjoy your product.
(if you don't decide to safe a few bucks in testing and instead shut down the developerstudio, as happened with Troika Games of the still very good, very ambitous, but only half-ready and bug-infested Vampire Bloodlines)

I'm not talking about small companies with limited resources. I'm talking about companies with virtually unlimited resources that releasing game that is almost broken or half made. Companies like EA or Sega with their recent Empire Total war (it could be Creative Assembly's fault but isn't publishers are there to make sure the game works and plays fine?).

Good thing we have, like you said, the good small companies who make great effort (sometimes, until they become big) or companies like Blizzard.

Easy to say in the era of Duke Nukem and Doom, where the games have the depth and complexity of a mudpuddle on Route 66 in July. A different matter entirely when a courageous small firm, like BIS, is constantly riding the edge of liquidity and still pushing boundaries of game development with monsters like this. Declaring an 'open beta' would have been a wiser choice in my opinion, but what's done is done. This is already a great game, and with support commitment that BIS has given previously, I'm fairly sure it will only continue to get better.

Its really sad when its already imprinted in ppl minds that patches are quite necessary after game release. Where are the good old days...

Unfortunately I don't know the exact settings that have been used, but I guess that they are quite high. I will deliver more information as soon as possible.