City restraints due to Online DRM?

[Catchin22 299]

The more I have been thinking about the features left out such as terraforming and small city size the more I realize this has less to do with glassbox and more to do with Online DRM. If a city size of 2k x 2k takes up X amount of resources then doubling those allowances would double that requirement. Being that I am in the hosting business I know there are only so many customers I can fit on a shared hosting server. Not only do I have to be concerned with bandwidth per user but resources used per customer as well.

I think a good share of our $60 entrance fee is paying for those resources and fitting as many possible customers on a server signifigantly lowers the OH needed. Developers have said that Glassbox is not the reason for the small city size, I believe it. After all, they could have given us multiple sized regions to work with if larger sizes were too much for some of our systems to handle. Terraforming would have likely greatly increased both system resources and bandwidth needed.

This is why I am also sure there will be add on content, microstores perhaps or DLC. Expect something along the lines of the Sims Franchise. Our $60 is only going to go so far for so long before addon content pays for extended play, and eventually that's going to run dry and and keeping those servers running will cost them money.

Thoughts?

[Ma8thew 77]

You don't have any evidence for this. In interviews it has been stated by Maxis employees that the reason for the size limitation is computational resources. You are accusing them of lying.

[Catchin22 299]

Are you trolling? My "evidence" Is logic and being in the hosting industry for 12 years. I've heard they have said just the opposite, that city size limits were not due to customers PCs not being able to handle it. If they really said "computational resources" that does not have to mean on the customers side.

[halicem 16]

If a city size of 2k x 2k takes up X amount of resources then doubling those allowances would double that requirement.

Actually, if you double the size, you'd be quadrupling the computational requirements(at least), as in you can fit 4 2k x 2k cities in a 4k x 4k map.

Now that's just pertaining to quadrupling the number of objects within the game. Then you have to take into account that the relationships between those objects, eg: a sim looking for a job, is like a permutation function - and quadrupling the number of available objects to relate to will more than quadruple the number of available relationships possible.

I have no doubt the simulation engine can handle it, and that modern computers can handle it. Cloud hosting has also been getting cheaper and cheaper(seems like AWS keeps lowering their prices every 6 months!). So the only thing I can think of right now is the actual bandwidth, both inbound to their servers, and upload bandwidth from players' homes(which ISPs don't really care about and most people don't know that it's separate from the down).

I forgot to check during the weekend but has anyone actually checked how much bandwidth simcity consumed during gameplay?

[Ma8thew 77]

Are you trolling? My "evidence" Is logic and being in the hosting industry for 12 years. I've heard they have said just the opposite, that city size limits were not due to customers PCs not being able to handle it. If they really said "computational resources" that does not have to mean on the customers side.

Where have you heard that? Here's a source for the performance explanation:

“The game has to run smoothly on ordinary computers,” said Quigley. “And I decided that it’s better to make cities with dense activity and visual detail over cities that are sprawling, but low resolution and inert.”

And there's reason to be hopeful too:

City size. Yes, its smaller than I'd like. Please understand that we're at the beginning of a completely new system and give us some time.

[Fornix 4]

I don't think so.

Thing is, you could unplug your connection during beta and keep on playing for quite some time. This suggests that client-server interaction is limited to once every few minutes, and although we're not exactly sure on what levels it does interact with the server, I wouldn't be surprised for it to just be very basic figures of data.

And to that, it doesn't really matter much how great the city size is.

For example telling the server you have 13000 workers, 800 barrels of oil and 200 tons of coal or telling the server you have 150000 worers, 8000 barrels of oil and 20000 tons of coal will make very little difference. SimCity also barely used any bandwidth during my hours I spent in it during beta, it's a matter of megabytes over several hours. Playing a game such as BF3 for example uses much more bandwidth than that.

Also online storing of your city states, is unlikely to form a real issue. Sure they'll need to invest in storage space, but plot save sizes are generally not too big in games. At 5 MB per plot size (which could be considered large), with 16 plots available in 10 regions, it's 800 MB per person. In other words, a 4TB drive may store the data for over 5000 players. Even if those don't all pay on Origin and EA sees only 25$ per game or so, such a HDD isn't going to cost the 125000$ gained there to not allow paying for the other expenses.

What I consider much more likely to be the cause for the current decision is client based lack of computational resources. In particular CPU usage. If people can reach 500k sims on current scales, that's a lot of units moving around. Add to that garbage, waste, power, water, vehicles, etc. and you have a high number of items to be tracked by the CPU. That's resource intensive. If you scale up from 2x2 to 4x4 you do not just quadruple that figure, but it exponentially grows as city density increases, more room for work, more room for residents, more power and water needed, more garbage, waste, etc. The average computer will turn into a slug, and even high-end systems may find difficulties.

The problem is comparible to the issue openTTD has whilst playing on larger maps with a high number of vehicles. Top 750 vehicles and many systems start to struggle. Sure, that game isn't completely optimized. However, even after optimizing it further it will run into its limits at some point again once the number of units simply turns high enough.

And SimCity is just a game in which numbers can easily turn high enough.

If EA decided to go for having SimCity just like the previous games by working with the simple radius systems for coverage and have residents just be randomly moving sprites which appear and vanish, then there'd be very little for the CPU to do, and technically there wouldn't be any reason not to be able to handle 4x4, 8x8 or heck even larger if they'd go for seamless loading. 64x64 would even lie in the realm of possibilities. As simply you're not seeing all of that detail anyway! However, as they didn't go down that route, you may not see the units, but the CPU still has to do the processing for every single one of them.

[magicrobbie 77]

Yes, I also think that most limitations can be traced back to server-backed gaming. And this makes me angry because server-based gaming is just another form of DRM. There is no need for Sim City to be online all the time other than for the multiplayer component - and this is a new and arguably optional component.

I'm pretty sure offline singleplayer could handle the glassbox calculations for large city tiles and cities side-by-side. And without the need to balance and control multiplayer, modding, mapmaking, terraforming, etc. should also be possible.

[DmainEvent 195]

I don't think so.

Thing is, you could unplug your connection during beta and keep on playing for quite some time. This suggests that client-server interaction is limited to once every few minutes, and although we're not exactly sure on what levels it does interact with the server, I wouldn't be surprised for it to just be very basic figures of data.

And to that, it doesn't really matter much how great the city size is.

For example telling the server you have 13000 workers, 800 barrels of oil and 200 tons of coal or telling the server you have 150000 worers, 8000 barrels of oil and 20000 tons of coal will make very little difference. SimCity also barely used any bandwidth during my hours I spent in it during beta, it's a matter of megabytes over several hours. Playing a game such as BF3 for example uses much more bandwidth than that.

Also online storing of your city states, is unlikely to form a real issue. Sure they'll need to invest in storage space, but plot save sizes are generally not too big in games. At 5 MB per plot size (which could be considered large), with 16 plots available in 10 regions, it's 800 MB per person. In other words, a 4TB drive may store the data for over 5000 players. Even if those don't all pay on Origin and EA sees only 25$ per game or so, such a HDD isn't going to cost the 125000$ gained there to not allow paying for the other expenses.

What I consider much more likely to be the cause for the current decision is client based lack of computational resources. In particular CPU usage. If people can reach 500k sims on current scales, that's a lot of units moving around. Add to that garbage, waste, power, water, vehicles, etc. and you have a high number of items to be tracked by the CPU. That's resource intensive. If you scale up from 2x2 to 4x4 you do not just quadruple that figure, but it exponentially grows as city density increases, more room for work, more room for residents, more power and water needed, more garbage, waste, etc. The average computer will turn into a slug, and even high-end systems may find difficulties.

The problem is comparible to the issue openTTD has whilst playing on larger maps with a high number of vehicles. Top 750 vehicles and many systems start to struggle. Sure, that game isn't completely optimized. However, even after optimizing it further it will run into its limits at some point again once the number of units simply turns high enough.

And SimCity is just a game in which numbers can easily turn high enough.

If EA decided to go for having SimCity just like the previous games by working with the simple radius systems for coverage and have residents just be randomly moving sprites which appear and vanish, then there'd be very little for the CPU to do, and technically there wouldn't be any reason not to be able to handle 4x4, 8x8 or heck even larger if they'd go for seamless loading. 64x64 would even lie in the realm of possibilities. As simply you're not seeing all of that detail anyway! However, as they didn't go down that route, you may not see the units, but the CPU still has to do the processing for every single one of them.

I honestly hope they designed glassbox to be able to handle resources better than that. That would mean that the relationship between city size and resources is exponential. Which is one of the worse Big O of n run times possible. Any software with that design should receive a failing grade, and I know if I turned in anything like that i would likely loose my job.

[fatbuster 44]

I don't think so.

Thing is, you could unplug your connection during beta and keep on playing for quite some time. This suggests that client-server interaction is limited to once every few minutes, and although we're not exactly sure on what levels it does interact with the server, I wouldn't be surprised for it to just be very basic figures of data.

And to that, it doesn't really matter much how great the city size is.

For example telling the server you have 13000 workers, 800 barrels of oil and 200 tons of coal or telling the server you have 150000 worers, 8000 barrels of oil and 20000 tons of coal will make very little difference. SimCity also barely used any bandwidth during my hours I spent in it during beta, it's a matter of megabytes over several hours. Playing a game such as BF3 for example uses much more bandwidth than that.

Also online storing of your city states, is unlikely to form a real issue. Sure they'll need to invest in storage space, but plot save sizes are generally not too big in games. At 5 MB per plot size (which could be considered large), with 16 plots available in 10 regions, it's 800 MB per person. In other words, a 4TB drive may store the data for over 5000 players. Even if those don't all pay on Origin and EA sees only 25$ per game or so, such a HDD isn't going to cost the 125000$ gained there to not allow paying for the other expenses.

What I consider much more likely to be the cause for the current decision is client based lack of computational resources. In particular CPU usage. If people can reach 500k sims on current scales, that's a lot of units moving around. Add to that garbage, waste, power, water, vehicles, etc. and you have a high number of items to be tracked by the CPU. That's resource intensive. If you scale up from 2x2 to 4x4 you do not just quadruple that figure, but it exponentially grows as city density increases, more room for work, more room for residents, more power and water needed, more garbage, waste, etc. The average computer will turn into a slug, and even high-end systems may find difficulties.

The problem is comparible to the issue openTTD has whilst playing on larger maps with a high number of vehicles. Top 750 vehicles and many systems start to struggle. Sure, that game isn't completely optimized. However, even after optimizing it further it will run into its limits at some point again once the number of units simply turns high enough.

And SimCity is just a game in which numbers can easily turn high enough.

If EA decided to go for having SimCity just like the previous games by working with the simple radius systems for coverage and have residents just be randomly moving sprites which appear and vanish, then there'd be very little for the CPU to do, and technically there wouldn't be any reason not to be able to handle 4x4, 8x8 or heck even larger if they'd go for seamless loading. 64x64 would even lie in the realm of possibilities. As simply you're not seeing all of that detail anyway! However, as they didn't go down that route, you may not see the units, but the CPU still has to do the processing for every single one of them.

I honestly hope they designed glassbox to be able to handle resources better than that. That would mean that the relationship between city size and resources is exponential. Which is one of the worse Big O of n run times possible. Any software with that design should receive a failing grade, and I know if I turned in anything like that i would likely loose my job.

You'll probably be interested in reading this (specifically at the ryani's second post). This is the most technical explanation from a maxis dev I've found so far regarding both city sizes & adjacent cities.

This topic has already been talked about at length though so I won't add too much. Personally, considering the publisher in this case, I think the online requirement most likely would have existed regardless what city sizes were finally chosen.

[Dewm 177]

well first off... If I wanted to play a MMO I would buy a MMO... but not I wanted to play a simcity.

Second I think you are probably incorrect in most of your assumptions (OP), as others have stated you can unplug for periods of time, which means that very little is actually held on the server (probably a region snapshot and a few economic statistics, so it can interact with the other people on that region)

Lastly, if you look at a standard MMO today most of the servers can handle 10,000 people.. I really doubt that Maxis servers (or EA's) are limited because they have (up to) 16 people on the server.

[Catchin22 299]

Saves are not local, so I am right that the servers are handeling a lot of data. I'll explain this the best I can since I am a web host, I have ran one of the oldest hosting companies to still exist today since the late 90s.

With any normal MMO you are paying a monthly cost, sans GW2 which is supported by a micro store. Server costs may appear to be getting lower but in reality server hardware is expensive and shared hosting gets cheaper because of a few things.

1. Outsourcing Support

2. Competition

3. Overselling

There is no host that is selling it's resources at the correct costs with exception to dedicated setups. So yes you will see lower and lower hosting costs but server hardware is much, much more expensive than PC hardware and only so many clients can fit on a server. Either those clients have to pay a high price for semi-dedicated or a ton of lower paying clients (shared hosting) have to be paying more than they are using. Most hosting customers think they are going to run a website with 500+mb and demand it, in reality most customers use 50mb or less. A lot of customers only use a few mb of space. Hosts sell huge amounts of space as a marketing ploy, they hope you wont ever use it, and except for the occasional forum or specialised site its not, and when it is those sites get shut down or moved to a higher paying account / dedicated or semi dedicated / VPS do to system resource usage.

So no, running servers is not cheap, and with a game pumping terabytes of data 24/7 through millions of players the costs would add up fairly quickly. That has to be paid for, servers are expensive, make no mistake that's in the budget. The larger the data is on the server side the more bandwidth needs to be transfered as well. Any way you look at it, doubling the plot sizes and quadrupling the needed system resources would cost them 4 times as much on the server end.

So we can assume that at $60 a pop they know exactly how long that will pay for services. I suspect the Beta had more to do with testing BW / Server resources than anything else, in fact probably mostly only had to do with that. If they want the game to keep paying for those server costs then they have to sell DLC / Add-ons.

[DmainEvent 195]

You'll probably be interested in reading this (specifically at the ryani's second post). This is the most technical explanation from a maxis dev I've found so far regarding both city sizes & adjacent cities.

This topic has already been talked about at length though so I won't add too much. Personally, considering the publisher in this case, I think the online requirement most likely would have existed regardless what city sizes were finally chosen.

This is what I expected...

They said they would have gone for it if it weren't for this...

"The local simulation stops at the edge of your box. Everything outside your box is part of the region, simulated on the server (for this discussion, the region highway you see coming in is actually part of your box, but in general nothing else is). It looked odd to have cars vanishing into the neighboring city, and you expected when your tourists commuted to the neighboring city that you could see where they went, but your city doesn't really know about the neighboring city."

So they have two levels of focus one of them is simulated on the server. So don't expect to be able to mod this thing. Some of the simulation occurs on the server, and I am pretty sure that they don't want people uploading files to their servers.

But that asside, I think that decision was incorrect. I think Sim city players would not mind, having a dead city next to theirs. It is a step up from what we use to have, no cities next to each other except when in the region view.

If we think on my idea for a minute, if they added one more focus level (lets call it a district focus level... They have already listed City and Region), when in that focus level, you could see paths being traversed by smaller autonoma (cars, bicycles, people etc... You could see larger autonoma like trains and plains and so forth in the region and city focus levels). You would simply load 4 contiguous districts and calculate all the paths for autonoma occuring in those districts (4 districts would be about the size and complexity of a city-assuming we don't mess with city sizes for this thought experiment). When traversing through your creation in this focus you would load the next district as you approach it. They could use graphical depth-of field technique (Tiltshift I think), to hide dead cities (or districts) while they are not directly being simulated (you could see the general outlines but not fully rendered). They then wouldn't have to shade the buildings (keeping the graphical calculations to a minimum) and you wouldn't have to simulate it.

This all could have been hashed out in a quick design meeting if they didn't think of it at first. If they publish the game as is, then I am afraid that this will never occur until Simcity 2015 or something like that.

Somebody from Maxis should have hired me as an architect for this project... I am going to send this id to Maxis. I want to see what they think and why they don't do something similar to this.

Saves are not local, so I am right that the servers are handeling a lot of data. I'll explain this the best I can since I am a web host, I have ran one of the oldest hosting companies to still exist today since the late 90s.

With any normal MMO you are paying a monthly cost, sans GW2 which is supported by a micro store. Server costs may appear to be getting lower but in reality server hardware is expensive and shared hosting gets cheaper because of a few things.

1. Outsourcing Support

2. Competition

3. Overselling

There is no host that is selling it's resources at the correct costs with exception to dedicated setups. So yes you will see lower and lower hosting costs but server hardware is much, much more expensive than PC hardware and only so many clients can fit on a server. Either those clients have to pay a high price for semi-dedicated or a ton of lower paying clients (shared hosting) have to be paying more than they are using. Most hosting customers think they are going to run a website with 500+mb and demand it, in reality most customers use 50mb or less. A lot of customers only use a few mb of space. Hosts sell huge amounts of space as a marketing ploy, they hope you wont ever use it, and except for the occasional forum or specialised site its not, and when it is those sites get shut down or moved to a higher paying account / dedicated or semi dedicated / VPS do to system resource usage.

So no, running servers is not cheap, and with a game pumping terabytes of data 24/7 through millions of players the costs would add up fairly quickly. That has to be paid for, servers are expensive, make no mistake that's in the budget. The larger the data is on the server side the more bandwidth needs to be transfered as well. Any way you look at it, doubling the plot sizes and quadrupling the needed system resources would cost them 4 times as much on the server end.

So we can assume that at $60 a pop they know exactly how long that will pay for services. I suspect the Beta had more to do with testing BW / Server resources than anything else, in fact probably mostly only had to do with that. If they want the game to keep paying for those server costs then they have to sell DLC / Add-ons.

This might have been cheaper for them (and more fun for us) if they went with the Xbox Live model of multiplayer... Server by committee sort of. The guy with the best connection is essentially the server, host the Region on his machine.

[magicrobbie 77]

Has there been an article on how the intercity simulation works? What we are learning from this link (http://www.reddit.com/r/SimCity/comments/17cjbl/impressions_from_a_long_time_sim_city_fan/c84b7l1) is that a) there is no Glassbox between cities, b) our city boxes are calculated on our computers, c) just simulating a 2x2 km tile eats up a lot of resources (=reason for small city size), and d) the reason why there are no adjacent cities is that there is no Glassbox between cities and it would have looked odd to have unsynced cities side-by-side. To me, it seems that the inter-city simulation must be somewhat similar to the one of Sim City 4 and primarily aesthetical reasons are limiting us from having adjacent city tiles. Based on the link, I have three questions: 1) Why not let the player decide on city size based on computing power available to him/her? 2) since our city boxes are calculated on our computers - is server-based gameplay really a requirement? and 3) would it be possible to somehow overcome aesthetical issues and provide us with adjacent city tiles?

[croxis 150]

I don't think there is a lot of simulation going on, mostly telling cities what resources there are from other cities and the world, and moving said resources around.

1) I think EA wants an even playing field no matter what system you are running. Think the leaderboard.

2) It wouldn't surprise me that EA mandated it. However the devs did say they also want the next sim city to be very connected with the world to reflect the more modern times. To be honest I think Maxis did a really good job making the online system meaningful despite the limitations.

3) It would be hard from a graphical and gameplay perspective. First you would at least quadrupal the number of things to render. Eight times if you count the corners. Then the simulation. If a sim enters from a neighboring city and you have 8 road connections, how should the simulation decide which road the sim enters on? Sure one sim doesn't matter, but if there is a lot it may not be in a realistic pattern.

[lotherius 16]

As far as computational resources, did anyone observe during the beta if the game is actually multi-core enabled, ie, takes advantage of multiple CPU cores? Or is it only sucking down 1 core of our modern processors, and thus "limited by resources" due to lack of being properly multicore aware?

Myself, I didn't get to play the Beta.... I ran the installer, it said it was installed, had to go to work. Then Sunday I tried to play the game and it said it had to "update"..... and the update had been going for 6 hours before I gave up and had to go to bed.

EDIT:

Looked at the suggested link... and Sir I AM Disappoint..

They targetted 32 bit hardware. Which means maximum 2gb addressable memory, and likely even if it is multicore aware, it is designed to assume a lack of multicore processing when determing "resources"..

So basically.... they built a game to be played on granny's netbook.

[noxiousdog 6]

You don't have any evidence for this. In interviews it has been stated by Maxis employees that the reason for the size limitation is computational resources. You are accusing them of lying.

I am.

It's a ridiculous excuse and it's a travesty nobody has called them out on it. They would have us believe that their servers are thousands of times more powerful than a home pc above and beyond any issues introduced by remote computing.

Do the math. Ignoring any operational costs (electricity, cooling) and any labor (development, support), it would take 100 copies of the game purchased directly from EA to buy a single medium range server ($100 * 50).

I'm not sure what they are doing on the servers beyond hosting and matchmaking, but it isn't computational power they need.

[Ma8thew 77]

You don't have any evidence for this. In interviews it has been stated by Maxis employees that the reason for the size limitation is computational resources. You are accusing them of lying.

I am.

It's a ridiculous excuse and it's a travesty nobody has called them out on it. They would have us believe that their servers are thousands of times more powerful than a home pc above and beyond any issues introduced by remote computing.

Do the math. Ignoring any operational costs (electricity, cooling) and any labor (development, support), it would take 100 copies of the game purchased directly from EA to buy a single medium range server ($100 * 50).

I'm not sure what they are doing on the servers beyond hosting and matchmaking, but it isn't computational power they need.

Your argument does not make any logical sense. I can't see how you are disproving that the SimCity is taxing to personal computers.

[Ocram's Razr 963]

Has there been an article on how the intercity simulation works?</p>What we are learning from this link (http://www.reddit.com/r/SimCity/comments/17cjbl/impressions_from_a_long_time_sim_city_fan/c84b7l1) is that a) there is no Glassbox between cities, b) our city boxes are calculated on our computers, c) just simulating a 2x2 km tile eats up a lot of resources (=reason for small city size), and d) the reason why there are no adjacent cities is that there is no Glassbox between cities and it would have looked odd to have unsynced cities side-by-side. </p>To me, it seems that the inter-city simulation must be somewhat similar to the one of Sim City 4 and primarily aesthetical reasons are limiting us from having adjacent city tiles. Based on the link, I have three questions: 1) Why not let the player decide on city size based on computing power available to him/her? 2) since our city boxes are calculated on our computers - is server-based gameplay really a requirement? and 3) would it be possible to somehow overcome aesthetical issues and provide us with adjacent city tiles?</p>

Running out of resources for cities you are not playing was a design "feature" of Cities XL's Planet Offer. It even had a website to make trade deals. Fortunately, you can turn on "import from Global Market" to keep your city running.

--Ocram

[fatbuster 44]

As far as computational resources, did anyone observe during the beta if the game is actually multi-core enabled, ie, takes advantage of multiple CPU cores? Or is it only sucking down 1 core of our modern processors, and thus "limited by resources" due to lack of being properly multicore aware?

Myself, I didn't get to play the Beta.... I ran the installer, it said it was installed, had to go to work. Then Sunday I tried to play the game and it said it had to "update"..... and the update had been going for 6 hours before I gave up and had to go to bed.

EDIT:

Looked at the suggested link... and Sir I AM Disappoint..

They targetted 32 bit hardware. Which means maximum 2gb addressable memory, and likely even if it is multicore aware, it is designed to assume a lack of multicore processing when determing "resources"..

So basically.... they built a game to be played on granny's netbook.

Yeah, looks like the game would require an extensive re-write if they even want to think about raising the city size. Least of which would mean converting the game into 64-bit, which I'm sure is more work than anyone would think. If they tried raising the size now I would imagine many people either start having their games crash or have major slowdowns as the game reached the 2GB limit. Unfortunately I don't see a fix for a very long time.

Don't be surprised if adjacent cities as a workaround is starting to look really good to them right about now.

[DmainEvent 195]

As far as computational resources, did anyone observe during the beta if the game is actually multi-core enabled, ie, takes advantage of multiple CPU cores? Or is it only sucking down 1 core of our modern processors, and thus "limited by resources" due to lack of being properly multicore aware?

Myself, I didn't get to play the Beta.... I ran the installer, it said it was installed, had to go to work. Then Sunday I tried to play the game and it said it had to "update"..... and the update had been going for 6 hours before I gave up and had to go to bed.

EDIT:

Looked at the suggested link... and Sir I AM Disappoint..

They targetted 32 bit hardware. Which means maximum 2gb addressable memory, and likely even if it is multicore aware, it is designed to assume a lack of multicore processing when determing "resources"..

So basically.... they built a game to be played on granny's netbook.

Yeah, looks like the game would require an extensive re-write if they even want to think about raising the city size. Least of which would mean converting the game into 64-bit, which I'm sure is more work than anyone would think. If they tried raising the size now I would imagine many people either start having their games crash or have major slowdowns as the game reached the 2GB limit. Unfortunately I don't see a fix for a very long time.

Don't be surprised if adjacent cities as a workaround is starting to look really good to them right about now.

A good dev team would have a 64 bit version waiting in the wings. At least the foundation of a code base. They probably do not, but yes, that is extremely frustrating. They basically targeted 32 bit single core systems. My question is why?!?! If they had targeted computers with high system specs, and scaled it back for those with lower end machines we would be in a completely different discussion. I know, what some of you are going to say-Self righteous Maxis-Apologists-"They are targeting medium range systems to get the largest possible customer base". Yes, that is fine. But if I am making a correct inference from that reddit post, they DESIGNED the game for middle range machines in mind. So they basically set a ceilling on possible performance that is much lower than what is technically possible! So there is very little room for growth!!!!!!!!!!! So, basically they can say they want to address this all they want, but if what that reddit post say's is true, there is a definite ceiling for city sizes and/or contiguous cities, that even upgrading your machine won't fix. If they have to rewrite the code base to acommodate 64 bit machines, that may be costly. I am not certain however. I know when I want to switch code from 32 bit, to 64 bit, I just have to flip a setting on my compiler, but my code runs on a technical virtual machine, with basically uniform requirements across the board. To be as efficient as possible, they probably wrote the Glassbox engine in C++. I hope they have precompiler options to switch all 32 bit specific variables and such to 64 bit.

Boy is this frustrating.

[macvirt 361]

You don't have any evidence for this. In interviews it has been stated by Maxis employees that the reason for the size limitation is computational resources. You are accusing them of lying.

Actually, we do. People with access to the GDC Vault can look up the various interviews and panels and presentations with awesome EA folks explaining the evolution of anti-piracy and DLC strategies into a combined format supported by the commercial interests of profiling research. The result of such ventures is translated in to revised strategy and that is ultimately translated in to directives for projects and subsidiaries. Don't except a hard working game developer to even be in the loop by the way, this is the corporate side of just another industry. Directives are translated into research targets, from there we move to design targets, and so forth. It's simple.

It's old stuff by now, to be perfectly honest. And, it is not like EA is the only one to pursue such roads. It's business, it pays, ergo it is done.

[noxiousdog 6]

Your argument does not make any logical sense. I can't see how you are disproving that the SimCity is taxing to personal computers.

Let's say you need 100 computations per second (for scale purposes) to play Sim City.

Let's say the average home computer can do 80. That means 20% must be offloaded to the EA Servers. In this situation for every 5 players (500 computations per second), EA would need 1 server to support them.

If you allow the home computer to do more work (90 cps), you can decrease the ratio down to 10:1, which is still impossible.

If you allow the home pc to do 99 cps, you can decrease the ratio to 100:1, which is still not cost effective and is only a 1% slow down on your home pc.

Do you follow now?

[Ma8thew 77]

Your argument does not make any logical sense. I can't see how you are disproving that the SimCity is taxing to personal computers.

Let's say you need 100 computations per second (for scale purposes) to play Sim City.

Let's say the average home computer can do 80. That means 20% must be offloaded to the EA Servers. In this situation for every 5 players (500 computations per second), EA would need 1 server to support them.

If you allow the home computer to do more work (90 cps), you can decrease the ratio down to 10:1, which is still impossible.

If you allow the home pc to do 99 cps, you can decrease the ratio to 100:1, which is still not cost effective and is only a 1% slow down on your home pc.

Do you follow now?

Your assumption is that the EA server is doing a significant amount of the work. I believe this not to be the case.

You don't have any evidence for this. In interviews it has been stated by Maxis employees that the reason for the size limitation is computational resources. You are accusing them of lying.

Actually, we do. People with access to the GDC Vault can look up the various interviews and panels and presentations with awesome EA folks explaining the evolution of anti-piracy and DLC strategies into a combined format supported by the commercial interests of profiling research. The result of such ventures is translated in to revised strategy and that is ultimately translated in to directives for projects and subsidiaries. Don't except a hard working game developer to even be in the loop by the way, this is the corporate side of just another industry. Directives are translated into research targets, from there we move to design targets, and so forth. It's simple.

It's old stuff by now, to be perfectly honest. And, it is not like EA is the only one to pursue such roads. It's business, it pays, ergo it is done.

How does a reduced city size play into DRM?

[noxiousdog 6]

Your assumption is that the EA server is doing a significant amount of the work. I believe this not to be the case.

Right. That is the whole point. There's no way the EA server can be doing much of the work. Ergo, computational power is a fake excuse.

[Ocram's Razr 963]

1. I believe that SC4 can be made to use 3GB, maybe 4GB of RAM, why wouldn't the new SimCity be able to do so as well (on 64 bit computers). It did say that 4GB of RAM was the minimum (or was it recommended?) requirements.

2. Regional Minor Works or large Great Works that are 4x4km would be a major boon and would not tax the computers or servers as hard as 4x4km skyscraper cities.

3. They said that all simulation in "empty" regional space and Great Works are simulated by the servers.

--Ocram

[Ma8thew 77]

Your assumption is that the EA server is doing a significant amount of the work. I believe this not to be the case.

Right. That is the whole point. There's no way the EA server can be doing much of the work. Ergo, computational power is a fake excuse.

I am completely lost now. The city size is restricted because of computational limitations on the user's computer. This has been stated by several Maxis developers.

[macvirt 361]

Your assumption is that the EA server is doing a significant amount of the work. I believe this not to be the case.

Right. That is the whole point. There's no way the EA server can be doing much of the work. Ergo, computational power is a fake excuse.

I am completely lost now. The city size is restricted because of computational limitations on the user's computer. This has been stated by several Maxis developers.

Marketing. In simple terms: it is part truth, and part bullshit. These are matters of design choices resulting from strategic directives. Most of the time people who actually work on a game work with goals set for them and approach those on the basis of what is provided to them. They rarely get exposed to the steps from strategic policy to directives for studios and projects. That is an executive level. Such statements are not necessarily wrong, but they are based on conditions created out of decisions elsewhere. If other decisions are made, other conditions reign for design goals and project targets, and the technical foundation for these are suddenly quite different. That is why EA is so strict with directives.

Hardware and platform requirements are always a topic, absolutely. But the project orientation towards managing those can take very different shapes and forms. EA has a clear strategy, the consequences of which leave no other path than the online play requirement. That developers fill in design requirements within the EA directives should not be a surprise. So ofcourse as a consequence they code and design to cater to the feature sets made required. That can actually be very beneficial, on a tactical level. We should however not forget that it is a consequence of decisions elsewhere.

EA has made clear over the past years, both in media and at developer & industry conventions that prt of their strategic focus is the evolution of anti-piracy, user profiling and other elements towards an integrated platform service model. Online play combined with multiplay combined with profiling and DRM in one go, conditional to the strategic goals of profiling and DRM. They also made clear that this strategy is executed in directives to the various subsiduaries and studios as part of project scoping and planning.

To be blunt, I heard the "story" of making the pitch for marketing and sales of this evolution several years ago already, as part of a public panel at a convention on the case of Spore and as part of a presentation during a GDC by an EA rep. The talk there was in every respect identical to the marketing talk we are given now as consumers. I'm still grinning over the statement of how a required online play model is the best course for a least resistance model of anti-piracy methods with the added foundation bonus for furthering the potential of user profiling for research and commercial interests. After which the guy actually winked..

Simply put: we say online play is required for technical reasons, we say how that enhances the experience of the user, we use that to strengthen DRM control and to deepen our insight into what people do, when & where and how, so we can use that for R&D and sell the data.

It's not hard to punch through that. It really isn't. After all, this is a logical and commercially very viable evolution of strategy. The business interests are substantial to say the least. And it does make sense. Not just on the sales side, but also on the costing side of project development.

[Ma8thew 77]

Your assumption is that the EA server is doing a significant amount of the work. I believe this not to be the case.

Right. That is the whole point. There's no way the EA server can be doing much of the work. Ergo, computational power is a fake excuse.

I am completely lost now. The city size is restricted because of computational limitations on the user's computer. This has been stated by several Maxis developers.

Marketing. In simple terms: it is part truth, and part bullshit. These are matters of design choices resulting from strategic directives. Most of the time people who actually work on a game work with goals set for them and approach those on the basis of what is provided to them. They rarely get exposed to the steps from strategic policy to directives for studios and projects. That is an executive level. Such statements are not necessarily wrong, but they are based on conditions created out of decisions elsewhere. If other decisions are made, other conditions reign for design goals and project targets, and the technical foundation for these are suddenly quite different. That is why EA is so strict with directives.

Hardware and platform requirements are always a topic, absolutely. But the project orientation towards managing those can take very different shapes and forms. EA has a clear strategy, the consequences of which leave no other path than the online play requirement. That developers fill in design requirements within the EA directives should not be a surprise. So ofcourse as a consequence they code and design to cater to the feature sets made required. That can actually be very beneficial, on a tactical level. We should however not forget that it is a consequence of decisions elsewhere.

EA has made clear over the past years, both in media and at developer & industry conventions that prt of their strategic focus is the evolution of anti-piracy, user profiling and other elements towards an integrated platform service model. Online play combined with multiplay combined with profiling and DRM in one go, conditional to the strategic goals of profiling and DRM. They also made clear that this strategy is executed in directives to the various subsiduaries and studios as part of project scoping and planning.

To be blunt, I heard the "story" of making the pitch for marketing and sales of this evolution several years ago already, as part of a public panel at a convention on the case of Spore and as part of a presentation during a GDC by an EA rep. The talk there was in every respect identical to the marketing talk we are given now as consumers. I'm still grinning over the statement of how a required online play model is the best course for a least resistance model of anti-piracy methods with the added foundation bonus for furthering the potential of user profiling for research and commercial interests. After which the guy actually winked..

Simply put: we say online play is required for technical reasons, we say how that enhances the experience of the user, we use that to strengthen DRM control and to deepen our insight into what people do, when & where and how, so we can use that for R&D and sell the data.

It's not hard to punch through that. It really isn't. After all, this is a logical and commercially very viable evolution of strategy. The business interests are substantial to say the least. And it does make sense. Not just on the sales side, but also on the costing side of project development.

I'm not disagreeing with you. It is obvious to me that the intertwining of the internet into SimCity is in order to eliminate piracy. Developers have said in answer to this question that there is stuff that is calculated by EA's servers. This I would not consider an outright lie, but an economy of the truth. I would prefer that they could say that the extent of PC piracy has required these drastic measures from publishers, but I'm sure this has been forbidden by the EA bureaucracy.

However, the city size limit is not, I believe, related to this. I don't think a developer would have chosen to respond a discussion about this limitation with an explanation of time complexity if this were a limitation imposed by server limitations. I think he would have chosen not to respond at all.

[macvirt 361]

Honestly, all due respect for those who do the real work. But extremely rarely do they have any notion of why they are basing what work with which requirements for what purposes. This has been a fundamental shift within this industry the past 10 years, what many of those developers call the "crappy corporate side" of things. The games industry became just that, an industry. Look at EA's own evolution, as what exactly did it start out, how the heck did it become an investment firm

Devs tend to be completely honest, but this is unfortunately always a case of "statements are based on the information resulting from decisions elsewhere while working in the trenches of the job and not on the cloud of the executive elsewhere".

Design choices are derivative in their nature. This is unfortunate, but it cannot be avoided. It is principle to game design as part of the venture management.

Look at the situation around the online play requirement. We've seen the Maxis statements The devs are honest in their statements. Whether they have any idea of how EA's presentations on strategic policies & directives are translated into what become their conditions for goal based work is a completely open question. From my own experience I'd say that most tend to really have no idea of how these things come in to play. A substiantial amount tends to have some idea, and bitches about it but accepts it as part of the work being in an industry. Fresh meat tends to have no idea, they come in believing the marketing of the industry But on the online play requirement topic, we know there is a higher reality above that of Maxis. They work with what is set. Server limitations? What are the specifications for that part of the project management .... faciliation on a feature level there is a derivative of other decisons resulting from directives. What is the primary focus of online play? DRM and user profiling. Ergo.

Too much of all this really is just marketing. Even when people are truthful, it still tends to be secondary to other things resulting from completely different perspectives and priorities.

[Riithi 0]

City restraints are coming from two problems:

1. The agent system behind glassbox scales badly when you increase the map (more agents, and each agent will require more CPU for pathing due tot the larger map)

2. Since we are supposed to play in multiplayer you can't have different map sizes, it would make the game unfair.

So from their perspective it will be like: either grannies laptop can run 4x4 or nobody will.

And as mentioned above the online requirement is there for anti-piracy, the tech is designed to require online servers.

For me it's a two-edged sword, on one side it's good that it might persuade pirates to buy the game, meaning more $$$ for future simcity content, and an incentive for EA to keep making games like these.

On the other hand it leaves me as a consumer at their mercy, they can shut down the servers rendering my product useless. In general 'always-online' adds more negative things compared to positive.

The city restraint is primarily a scaling problem with the glassbox engine, and unrelated to DRM.

But since were all supposed to play together it will be hard to give the extra space to computers who could run bigger maps.