Good old isometric view?

[BLANKBLANK 4,414]

I've seen quite a few of the screenshots for this game, and I've noticed various people complaining about the textures and the cartoonish look of cities and buildings, plus there are ugly rumblings about small city sizes. I think all of these problems could be solved if SC5 just stuck with the isometric view, like that of SC4. Four different angles, several zooms and that's it.

That alone would cut down on the graphics requirements but also allow for more detailed textures on buildings, props, terrain, roads and so forth and give this game the 'SC4' look. Also it would allow for bigger city tiles because again there's less demand on the graphics engine.

I just think 3D city simulators always get the colours, tones, hues and lighting wrong and oversimplify on building models because they must seen from a 3D perspective. I would be perfectly happy to buy SC5 (or SC 2013 as it known) with just an isometric view because it would have curved roads, a new stimulation, those building add-ons (which are really clever and cool!) and more, plus you would get detailed textures because everything is not being rendered in 3D.

[T Wrecks 10,209]

According to statements from EA/Maxis the limiting factor is NOT graphics, but the GlassBox engine, which is used for the data simulation and visualisation. If this is true, then it wouldn't matter at all whether this engine is at the background of a full 3D game or of an isometric game.

As for cartoonish look being tied to 3D vs. isometric, I beg to differ. While it is true that models must be simplified somewhat, this doesn't hold true for the textures. Even a simple cube would look cartoonish if it was textured like most buildings presented in the screenshots: shiny, smooth, possibly oversaturated. Turning down texture saturation doesn't influence graphics performance at all, and neither does giving the walls and roofs some actual texture, i.e. not just a single, solid colour, but e.g. an illusion of bricks (with slight colour variations) and mortar, plus possibly some grime, wear & tear.

I basically agree with you in that I would prefer detailed graphics and the possibility to build at least diagonally over 3D just for the sake of 3D, but I don't think this is the actual reason of the city tile size we are seeing or the cartoonish overall look. The former is said to be connected to the GlassBox engine (as mentioned above), the latter seems like a conscious design decision to me.

[Dewm 177]

^this

This holy grail "glassbox" is the biggest waist of time IMO.

[Skyrunner27 23]

^this

This holy grail "glassbox" is the biggest waist of time IMO.

You could say that but what difference would there be between this Simcity and Simcity 4. Just wondering but how would you have updated the game?

[Cob92 360]

The limitation on size is due to the simulation, not the graphics. I'm glad to see the isometric view go; I like being able to view my cities in 3D. IMO the real test of city aesthetics is what the streetscape looks like.

[Dewm 177]

^this

This holy grail "glassbox" is the biggest waist of time IMO.

You could say that but what difference would there be between this Simcity and Simcity 4. Just wondering but how would you have updated the game?

this is the firat time Ive used the forums app on my cell... so patience with my bad spelling please

things that i would have changed?

I would have looked into a better 3D engine, i know isometric are old hat and wouldnt have gone over to well.. but when you start cutting core game features because your new engine cant handle it... its time to start looking into a different engine.

You dont see Halo cutting out multiplayer because they built it on a engine that can only handle 3 people on the screen at once..

Second. i would look into more features.. not less. curved roads, more roads, more train stations..

Third. i would design this game with the modding community in mind. we have kept CS4 alive years past its experation date.

Third.. at would look into multiplayer. where you have regions that act as a server off of your computer. you can play localy by yourself, play with friends on your LAN or invite people from all around the world. Or even log into your server remotley (minecraft does just this...and its a blast)

...just a few ideas..

[Indiana Joe 1,750]

This is why they didn't name it SimCity 5. It's not supposed to be a sequel; they're letting us know they're starting over. God knows why.

[Cob92 360]

This is why they didn't name it SimCity 5. It's not supposed to be a sequel; they're letting us know they're starting over. God knows why.

I guess to build another generation of SimCity based on Glassbox? Honestly I'm not missing anything besides subways and big maps (which weren't even that big). I don't mind not having an ingame terraformer as long as there is a way to make new maps for the game. Personally I find terraforming exceedingly boring and would rather focus on building the city.

[Bobvila 8]

I think isometric looks very stale in this day and age. I'm glad it's gone.

I won't miss subways either. But I will miss freeways and big maps.

[Xenocity 343]

This is why they didn't name it SimCity 5. It's not supposed to be a sequel; they're letting us know they're starting over. God knows why.

They already stated this new Simcity is a reboot/new line of games.

Simcity 3000 and Simcity 4 both suffered with simulation issues that prevented both games from reaching their full potential.

Maxis has been building the new engine to handle all the simulations their simulators need.

Simcity is a simulation first and foremost, hence what the Sim stands for in Simcity.

I hope this new Simcity runs smoothly, I do not want a repeat of Simcity 4 issues which nearly broke the Simcity 4.

Isometric view had a nice run, though 3D should add many benefits to the game.

[BLANKBLANK 4,414]

I hope this new Simcity runs smoothly, I do not want a repeat of Simcity 4 issues which nearly broke the Simcity 4.

Isometric view had a nice run, though 3D should add many benefits to the game.

Simcity 4 was just hard full stop, even with the patches the simulator still had a steep learning curve, and it seemed to encourage you to build neighbouring cities to stoke demand in your first city. I just hope the Glassbox simulator is as complicated otherwise SC 2013 will merely be an update of SC 3. The complexity of the simulator is what keeps me coming back to SC4, and if SC 2013's one is too simple. Well, it won't be very rewarding game...

[croxis 150]

Simcity 4 was just hard full stop, even with the patches the simulator still had a steep learning curve, and it seemed to encourage you to build neighbouring cities to stoke demand in your first city. I just hope the Glassbox simulator is as complicated otherwise SC 2013 will merely be an update of SC 3. The complexity of the simulator is what keeps me coming back to SC4, and if SC 2013's one is too simple. Well, it won't be very rewarding game...

On the other hand I know people who didn't like SC4 for that reason and prefer 3K

[P*Funk 12]

I think the engine limitations are interesting. People seem to gloss over it despite it being supposedly a step forward. I don't see how a brand spanking new engine that has more limitations on scale than something from 10 years ago has as being progress, just a bad decision.

Certainly a new engine that does new things is an advantage but engine limitations aren't something to be overlooked. The Old Republic MMO uses a new engine and it all looks very nice and all that but they're in trouble. Moving forward its basically been said by the developers that they're not able to do open world PvP because the engine can't handle it.

Imagine that, an MMO designed around an engine that can't actually handle massive numbers of online players in the same instance. Bad planning. Bad design, and its limiting the potential of the game going forward. Some people think MMOs without good open world PvP aren't worth their time, and they're right for them at least, but the problem here is that all that optimism people might espouse about the future of that game... well none of it applies to the rather substantial core of open world PvP fans. They're out of luck it seems.

Who knows what this engine's limitations might do to the rosey and as yet unexplored future of SimCity 2013. They could miraculously fix the engine to handle bigger maps and more stuff, but... my suspicion is that they won't bother wasting lots of money on that. I believe that these engine fixes may come from brilliant community modders, should the tools be given to them, in time just like modders have managed to create 128 player servers for Battlefield 2, and even then without an SDK.

The thing that strikes me though is that engine limitations are an awfully big elephant to be ignoring. Its not something you can brush aside. It poses serious implications for the future of this particular title, at least in terms of those who want to see things get bigger and at least return to SC4 size.

I'm curious to see how this hashes out, and whether the developers will be honest about it or BS their way around it like the SWTOR devs have been doing (in terms of firmly saying absolutely no).

[Evgik 1]

I am extremely disappointed to 5. 3D graphics leave a negative impression. Any transition to 3D involves downscaling. You look at these sidewalks? $%&^!.

Sprites are nicer and better aesthetically than crooked and ugly 3D

[Dewm 177]

I think the engine limitations are interesting. People seem to gloss over it despite it being supposedly a step forward. I don't see how a brand spanking new engine that has more limitations on scale than something from 10 years ago has as being progress, just a bad decision.

Certainly a new engine that does new things is an advantage but engine limitations aren't something to be overlooked. The Old Republic MMO uses a new engine and it all looks very nice and all that but they're in trouble. Moving forward its basically been said by the developers that they're not able to do open world PvP because the engine can't handle it.

Imagine that, an MMO designed around an engine that can't actually handle massive numbers of online players in the same instance. Bad planning. Bad design, and its limiting the potential of the game going forward. Some people think MMOs without good open world PvP aren't worth their time, and they're right for them at least, but the problem here is that all that optimism people might espouse about the future of that game... well none of it applies to the rather substantial core of open world PvP fans. They're out of luck it seems.

Who knows what this engine's limitations might do to the rosey and as yet unexplored future of SimCity 2013. They could miraculously fix the engine to handle bigger maps and more stuff, but... my suspicion is that they won't bother wasting lots of money on that. I believe that these engine fixes may come from brilliant community modders, should the tools be given to them, in time just like modders have managed to create 128 player servers for Battlefield 2, and even then without an SDK.

The thing that strikes me though is that engine limitations are an awfully big elephant to be ignoring. Its not something you can brush aside. It poses serious implications for the future of this particular title, at least in terms of those who want to see things get bigger and at least return to SC4 size.

I'm curious to see how this hashes out, and whether the developers will be honest about it or BS their way around it like the SWTOR devs have been doing (in terms of firmly saying absolutely no).

Good post!

And I agree 100% Like I said in one of my previous post, I would have liked to be a fly on the wall during some of those development meetings.

And from what we've seen so far, it looks like they are going to try and BS their way around the engine limitations.

But I have faith in our modding community... so who knows maybe in 3 years I'll be able to play SC:2013

[croxis 150]

I laugh, hard, at the use of the phrase "engine limitations." A well designed game is composed of several specialized engines. Sometimes the design of the engine creates limits. The graphics engine of homeworld 1 will start choking past what we would now call a low polygon count despite the modern hardware you run the game on, because the engine was designed for the hardware from 13 years ago. They never considered supporting the millions of polygons we can now render on screen, or HD textures. They, rightly, didn't spend the time or money to have the software handle technology a decade in the future, so structured the code to handle the polygon counts for the day and the memory management for the texture sizes of the day.

That is not the case here.

The limitation is a hardware limitation. They chose not to do large cities not because the engine itself can't handle it, but because the computer resources needed would be too steep. The game already requires 4 gigs of ram which is still considered a steep requirement, especially if you look at the more hard core Steam hardware profiles. Is it possible that Sandbox is inefficient for the detail that it is running? Sure that is possible. It is also possible that the performance is as efficient as reasonably expected for the resolution of the simulation. Glassbox's objective is to abstract nothing into general statistical formulas - simulate everything using agents. I accept the possibility that our hardware just isn't quite there yet and that the engine can continue to scale breath and depth as our hardware continues to improve over time.

[Ocram's Razr 963]

The original Tropico had 2D (isometric) graphics but its people were all simulated from cradle to grave. It had many limitations because of the immense processing power needed to simulate many individuals with wants, needs, and political affiliations. The 3D graphics are not a limiting aspect, the Glass Box engine is. Most games are in full 3D now so it only makes sense that SimCity 2013 would as well.

--Ocram

[Dewm 177]

I laugh, hard, at the use of the phrase "engine limitations." A well designed game is composed of several specialized engines. Sometimes the design of the engine creates limits. The graphics engine of homeworld 1 will start choking past what we would now call a low polygon count despite the modern hardware you run the game on, because the engine was designed for the hardware from 13 years ago. They never considered supporting the millions of polygons we can now render on screen, or HD textures. They, rightly, didn't spend the time or money to have the software handle technology a decade in the future, so structured the code to handle the polygon counts for the day and the memory management for the texture sizes of the day.

I'm not really understanding here... you are saying "the engine isn't at fault, its the hardware"

...so your saying that no 3D engine in the world could do what SC4 has done?....if that is the case we are WAY to early for this tech.

Truth is, glassbox is a pretty poor engine from everything I've seen.

If minecraft can pull off a map that is twice the size of the earth...on JAVA... then a good engine could pull off a large 3D map from SC4

[P*Funk 12]

Is it possible that Sandbox is inefficient for the detail that it is running? Sure that is possible. It is also possible that the performance is as efficient as reasonably expected for the resolution of the simulation. Glassbox's objective is to abstract nothing into general statistical formulas - simulate everything using agents. I accept the possibility that our hardware just isn't quite there yet and that the engine can continue to scale breath and depth as our hardware continues to improve over time.

I don't buy this line of reasoning. We're firmly into the 64 bit era. We have moved past the 4 gig Ram limitation and we're getting to a stage of development where its more than just easy but expected that games are able to handle and utilize multi-core architecture to full effect. Any failure to embrace these changes are just uninspired design.

Lets take the upcoming game Planetary Annihilation for instance. This is an RTS game which seeks to create a sense of scale on a galactic level but in an RTS combat environment. This means planets with moons orbiting them in solar systems with multiple planets. Once you conquer your planet you can go fight the winners on their planets and so on. This is a game that has the potential to scale to meet insane demands of dozens of players playing 12 to 24 hour games if they want. All of this I might add involves hundreds, even thousands of robots executing complex AI driven behavior, all of which is also being developed to be smarter than in the past. They're designing it around a client-server model so that big games can be hosted on game servers either made by the developer or by someone with a powerful box to support it. Now, this sounds exciting but what if you just want to play a small game with some friends, without a big server? What if you just want to play a LAN party with 4 people on a modest map? You can do this too. The game is designed to scale to the demands of the players.

The game engine is being built to deal with multiple cores, 64 bit architecture, basically to take anything you can throw at it, but also to scale down for people of modest hardware capability. This is the sensible thinking that all PC games should be designed around, but they tend not to because of a lowest common denominator mentality that is more in keeping with these big publisher models. I firmly believe that they intentionally don't create these scalable potentials within games because people feel like they're missing out on something if they can't experience what someone else can.

If I have 8 cores and 12 gigs of ram why shouldn't I be able to build as big and illustrious a city as I want? Why should I be limited by the hardware a bunch of broke mid-core college laptop gamers have available? I would think that given the direct agent based model they're using they'd want to let me scale it up to handle as much as possible, to push the game as far as I can. Its a lack of creative vision that I've come to expect from big budget publishers. If this is purely about hardware requirements then the game is holding itself back intentionally. If its really just the engine though, not being able to handle the scale no matter the hardware, then thats again closer to what I was talking about in my first post.

Also, for the record Planetary Annihilation is a kickstarter project with no publisher and a budget of around $2.2 million. If they can do all that (and we're going to find out inside a year I believe) then the limitations of Sim City 2013 may just be incomprehensible and laughable in short order.

[BLANKBLANK 4,414]

Truth is, glassbox is a pretty poor engine from everything I've seen.

If minecraft can pull off a map that is twice the size of the earth...on JAVA... then a good engine could pull off a large 3D map from SC4

But minecraft has only a fraction of the complexity (regarding the simulation and programming) as SC4 (and hopefully SC 2013 to), each point on the map is covered by the simulator and when it's really taxed you got that lovely pop-up in SC4 and SC3. Personally I wish 4*4 was the minimum size of city tile for this game but maybe such tiles (and larger ones) will come in expansion packs... Even so 2*2 is a decent amount of space but if that is the maximum and there are no plans for expanding the size, well that would be a real disappointment...

[croxis 150]

I'm not really understanding here... you are saying "the engine isn't at fault, its the hardware"

...so your saying that no 3D engine in the world could do what SC4 has done?....if that is the case we are WAY to early for this tech.

Truth is, glassbox is a pretty poor engine from everything I've seen.

If minecraft can pull off a map that is twice the size of the earth...on JAVA... then a good engine could pull off a large 3D map from SC4

You need to get this whole 3d monolithic engine idea out of your head. It is just incorrect.

I referenced the graphics engine of homeworld to illustrate what a software-bound bottleneck looks like, where no matter what advance hardware you put the game on you just can't get modern polygon counts or textures. Each software has a bottleneck somewhere. For most games the bottleneck is the graphics card. For dwarf fortress the bottle neck is the speed of one core of a cpu. For minecraft the bottle neck is either cpu, ram, or if hosting a server its the network speed.

In interviews the developers have said the 3d graphics engine of Sim City is not the bottle neck of the game. They could add even more visuals if they wanted. The bottle neck is the simulation engine -- which they call glassbox -- because it requires a great deal of ram and cpu power to simulate a city at that fidelity. THEY COULD USE TOP DOWN 2D GRAPHICS FOR GLASSBOX AND IT WOULD HAVE THE SAME CPU AND RAM REQUIREMENTS.

Also -- minecraft can *generate* a map with a large area, but it does not simulate it all at once. As any serious server operator knows, the game only loads about 100 - 150 chunks around a player at any given moment, which can be adjusted with bukkit. On my poor 1 gig of ram vps I can only handle about 8 spread out players before running out of ram.

Bad press and bad word of mouth. What will be your average player's reaction when they play on a region that is a little too large for their system? "Oh this game suzors its buggy poo!" and the proceed to tell everyone they know online. A decade ago with SC4 I would wager that when this happened it happened on forums, and the more thoughtful of us would respond with what you said, that their computer sucks. Now it happens through facebook and skype (of all the things to get popular, skype?!) where there is less of a chance that there is a rational mind to set the younggens strait.

According to the steam hardware survey only 36% of steam systems have more than 4 gigs of ram. 53% are at 4, and 9% have 2 or less. 14% are on windows 7 32 bit. Assume that the steam hardware reports are biased to the hard core, then those stats are even worse. I would love it if we were in the 64 bit era. Stupid hardware companies holding us back. The fact that 10% of the market is being excluded is actually surprising to me.

According to past interviews they were talking about maxing out the active agent count at 200,000. A lot more than what PA is going for. I wouldn't sneeze at the ai for these either as at least some of these agents will be using fairly robust path finding when simulating cars. Assuming a 1:1 relationship between agent count and ram (which we both know is false, but its the best approximation I got) then if it takes 4 gigs of ram to simulate a 2x2km area, then it will take 16 gigs of ram to simulate 4x4 (double the side of a square you quadrupedal the area). Do some people have that? Sure. Is it worthwhile to spend developer hours on it? Probably not.

[Dewm 177]

Well before you judge you should know, that I play MC quite a bit, including hosting my own server, I am the lead tech at a IT company that does everything from security systems to microwave systems for our instate communications company.

I've run network cabling in buildings, terminated fiber, installed IP phone system, built a couple of computers

...I wouldn't say I'm a "novice" when it comes to modern tech.

And I've dabbled in C++ programing, HTML, Java, worked with Blender, 3D max, and game maker.

And yes I know MC doesn't keep the whole world loaded at once, but its smart programming like that, that Maxis should have implemented. What I was saying is this.. and I'll try to state it as simply (for your sake) as possible.

it may not be due to graphics, but when you have to sacrifice LARGE portions of the franchise, because your engine can't deliver with current hardware tech. Then you have the wrong engine.. plain and simple..

[Jasoncw 11,840]

There's no point to making another SimCity if it's just going to be 2.5D. What could it do that SimCity 4 doesn't already do? Other than a very short SC4 wishlist that modders can't implement because it's stuff in the .exe.

[fatbuster 44]

How is Minecraft's engine in any comparable to SimCity's? As croxis pointed out, It's not constantly pushing anywhere near the same amount of data that Simcity does. The only time minecraft has a large amounts of data moving around is either when someone creates a gigantic explosion or spawns a large number of mobs, both of which can bring even powerful servers to their knees very quickly.

If you want to compare it to another game at least compare to one using the same type of simulation behavior, like Tropico 4. Just to compare one aspect of both games, Tropico 4 has a hard population limit around 1,500. That may sound laughably low but you can get rid of it with a mod. Though, the funny thing is that even with the cap removed people still struggle to get over 20k without their games slowing to a crawl. Now, SimCity claims to not have any population cap and we've heard reports of cities topping 1mil.

That may not sound like it means anything but the thing to remember is that low population limits in agent based simulations like Tropico were pretty much the norm until now. The fact that SimCity can handle that many more times the information than other games in the same genre, that were released recently and remain playable, is actually pretty amazing.

Now, I'm not a fan of the limitations the game has and I'm willing to agree that Glassbox may still be too much for our PC's right now to be used in this game, but by no means would I consider the Glassbox engine to be a crappy engine from everything I've heard and seen so far.

[StoNe220022 159]

I laugh, hard, at the use of the phrase "engine limitations." A well designed game is composed of several specialized engines.

That just simply isn't true, it is incorrect.

The engine creates the limitation, not the hardware. Software isn't designed to scale, it's designed to a limit based on the existing design.

I design and develop software for a living; the engine has foundations and these foundations are made up of algorithms and constraints. To change the software is to change the foundations of the very software engine; which in turn...will change your initial infrastructure requirements.

Now data however is another story, software is designed to scale with the data...but not hardware, the software architecture and infrastructure are predetermined based on the software requirements.

The limitation is a hardware limitation. They chose not to do large cities not because the engine itself can't handle it, but because the computer resources needed would be too steep.

There are several cases where this can be proven false...take SC4 for example, run mods to expand the viewing area to a 1,600 x 1,200 screen res...guess what, it can run slow on a large city, even using modern day computers. Shoddy hardware? Or old out of date engine?

Edits to address points in this topic and thread:

- Yes I think glassbox is very much so the limition!

- Yes my personal feeling is they should have tonned down the simulation aspect of the new design as it came at too great a cost

- No I don't belive graphics and engine are independant: locked isometric with the current engine would behave the same, it's the engine designed with the free-form view in mind...which in turn is the limitation.

- If the engine wasn't drastically redesigned; what would be the diff between 2013 and SC4? Simple, 10 years of game design evolution

[croxis 150]

That just simply isn't true, it is incorrect.

..snip...

Maybe I'm exhausted, but your first section is making no sense to me at all. I think you might of misinterprited what I said. I did not mean to say that every and all performance issues are because of hardware. I even referenced Homeworld 1. Run it on a GForce 660 and it will still choke with a few thousand polys because of the nature of the engine. Same thing with sc4. Its (probably) a single thread and the draw system just wasn't made to work at that resolution.

The point I was making is that the performance of SimCity2013 will hit a bottle neck. According to EAMAXIS it is because anything much larger than 200,000 agents results in too much Ram being used for most people's systems. I am confident they playtested and prototyped like crazy to figure out how many agents and containers should be simulated per sq km, and that limit seems to be about 50,000 agents per sq km. Next question, which no one has the answer to except for EA, is the nature of this specific limit at this moment in time. Is it like the render system of SC4 and HW1 which is that the software is already maxed out? Or does glasbox still have room to scale and consumer hardware is the limit?

Now was agent based simulation an appropriate choice for features vs cost? That is an opinion. Raph Koster (theoretical game dev) wrote a bit on this. As a player plays a game they develop a model of how the game works in their head. In a FPS if you wrote an AI with 12 emotional states, but only made a happy, sad, and angry texture that you shoehorn the 12 emotions into, the players internal model of the game is that the AI only has 3 emotions. So don't spend the time coding 12 emotions if the player only "experiences" 3 of them. The opposite is true too. A player can develop a very complex model of how the game works, when really the game just does some statistical handwaving.

SC1-4 has had both problems. On one hand they used statistical hand waving to create the illusion of complexity but the actual simulation was simple. But there was also some deeper simulation that they player didn't experience directly or visually, so as far as the player was concered it didn't exist. This time around maxis is going right at the target. The player will see everything that is simulated so the player's psyche has a correct model of how the game works and no time is wasted on features that the player wont experience. I am personally a fan and I feel the cost of city size is worth it. But I also know that is my opinion and the part of games I enjoy most.

[P*Funk 12]

Bad press and bad word of mouth. What will be your average player's reaction when they play on a region that is a little too large for their system? "Oh this game suzors its buggy poo!" and the proceed to tell everyone they know online. A decade ago with SC4 I would wager that when this happened it happened on forums, and the more thoughtful of us would respond with what you said, that their computer sucks. Now it happens through facebook and skype (of all the things to get popular, skype?!) where there is less of a chance that there is a rational mind to set the younggens strait.

Basically you're agreeing with me, but for different reasons, users who don't understand the software they're using. Basically you're saying that we have to assume that all users of software are dumb enough to not understand the limitations of their own hardware. When I was a teenager even I had the aptitude to appreciate the fact that my computer wasn't bleeding edge and that if I set the graphics and any other scalable performance hitting features too high and I got snail frames or even a game that wouldn't load that my computer just wasn't up to snuff. I dunno, maybe its because I was bred as a gamer on PCs and not on consoles that I can understand this. Maybe the console generation and all those people that play low fidelity mobile/facebook games don't understand this concept.

According to the steam hardware survey only 36% of steam systems have more than 4 gigs of ram. 53% are at 4, and 9% have 2 or less. 14% are on windows 7 32 bit. Assume that the steam hardware reports are biased to the hard core, then those stats are even worse. I would love it if we were in the 64 bit era. Hardware companies holding us back. The fact that 10% of the market is being excluded is actually surprising to me.

None of that actually explains to me why things should be scaled back to meet the hardware limitations of the 14% on Win7 32 bit. What are you doing blaming hard ware companies for holding us back? My friend has owned a 12 core Intel CPU for 2 years, granted he paid ridiculous amounts of money for it, but if thats the upper limit of ridiculous then where's the average? Fact is that quad core is the new standard. Not the future, its here now. If you buy a computer today its a quad core. The only thing that was holding back the quad core shift was the fact that SOFTWARE wasn't able to handle multiple cores for the longest time so dual cores were generally favourable for performance obsessed people because 2 cores at 3.0 was better than 4 at 2.3 because the software would only use one core at a time. Thats not the case anymore, not as much anyway. But we'll never be in the 64 bit era and the multi core era properly until software is correctly designed to take advantage of where we're going, not where we are now. There's nothing wrong with having the game meet the needs of lower level hardware, but that doesn't explain why the engine wouldn't be able to scale up for multiple cores. It doesn't make sense that a game would meet 4 gigs as a minimum but not be scalable up to 8 or 12 gigs, same with cores. Its just a matter of expanding the demands on the system. Limitless resources usually means the upper end is firmly embedded in the design of the software. With multi core software support there's no logical reason you can't scale the game up, unless the engine itself falls apart when you throw more stuff at it regardless of the hardware running it.

According to past interviews they were talking about maxing out the active agent count at 200,000. A lot more than what PA is going for. I wouldn't sneeze at the ai for these either as at least some of these agents will be using fairly robust path finding when simulating cars. Assuming a 1:1 relationship between agent count and ram (which we both know is false, but its the best approximation I got) then if it takes 4 gigs of ram to simulate a 2x2km area, then it will take 16 gigs of ram to simulate 4x4 (double the side of a square you quadrupedal the area). Do some people have that? Sure. Is it worthwhile to spend developer hours on it? Probably not.

I don't now what the upper limit of PA's agent load is. From what I've read theoretically there is no limit, as long as you have the hardware to support it, server side. Unit AI also with regards to pathfinding is pretty much the main focus in any RTS. We'll see how that works out, but I still think that hard limits imposed by antiquated hardware is a stupid way to design things. Frankly you still haven't explained why there can't be scalability other than some lame marketing oriented limited goal nonsense. Its one thing to acknowledge this as the reason, its another to say its a good reason. Everyone knows that big box publishers are morons with no ambition who milk the market for what its got without doing any of the labour to move it forward. The work of pushing the limits will come from kickstarter projects and independent thinkers who'll hit it big then get bought out by the EA's and THEN we'll see mainstream games embrace the new paradigm, built by the fringe thinkers.

So yea, I get it, we have limitations for this reason or that And if consumer perception of the product is such a big deal I don't know why they can't just come up with normal "Detected best setting for your machine" stuff with some "Warning, gameplay settings exceed recommended hardware minimum requirements" or something. In any Total War game you can scale the size of the units on the battlefield. Plenty of games cope with this. The only thing that this makes me think is that they're targeting such a passive and uncommitted gaming market that we can't even expect them to visit one forum, one web site or anything to acquire knowledge about the game and its limitations or its potential growth, which would be ironic given this is supposed to be some emerging game within the social media revolution.

[yummysinsemilla 16]

^this

This holy grail "glassbox" is the biggest waist of time IMO.

Until you realize this is the most realistic way to simulate a city.

I don't really agree with the city size being small to work with computers of age, though hopefully in the future, Maxis will be able to make it more efficient and allow some CPU process customization for the user to make the most out of it. I'm not sure, but I'd say that a decent quad core box (with Maxis making use of that technology) would run a city three times the size of the 2x2km limit, but not everyone has a decent quad core, yeah?

There are several cases where this can be proven false...take SC4 for example, run mods to expand the viewing area to a 1,600 x 1,200 screen res...guess what, it can run slow on a large city, even using modern day computers. Shoddy hardware? Or old out of date engine?

You do know that SC4 only uses one core, correct? I don't even think dual cores even came out until like 2005 or so. So, having a 2.8ghz quad core will actually run SC4 slower than a 3.4ghz single core. (given no background processes infererring)

[CaptCity 4,752]

Just stepping in to remind posters that throwing around words like stupid, lazy, ignorant, and moron, whether specifically or generally, is not the way to encourage positive discussion. If this is not you, "nothing to see here"; if it is, please take heed.

[jcarlilesiu 37]

^this

This holy grail "glassbox" is the biggest waist of time IMO.

Yes, because in Simcity 4, having a traffic jam 14 blocks long in a small city when the street one block over with the same terminus is completely empty is a better alternative.

/sarcasm