SimCity: Multiplayer Discussion

[City Planner 121]

City Planner @I guess there always hope.....

as long as they don’t focus on multipayer too much the game will be alright.

And as long as they give multiplayer the resources that it should deserve then it should be alright as well. I can remember back when SC4 was announced and they called it multiplayer as well, but once the game came out, it showed little multiplayer in the traditional sense.

I'm not really one to care much about multiplayer in CB games, it's more of a novelty to me, and I'll try it with a few friends from my website, but in general in the past, those times that we've played a CB style game, we rarely ever get beyond maybe 2 to 5 sessions before we depart to go solo again.

For me, SC's new multiplayer will have to be something so spectacular to draw me into it, but which I have some serious doubts on just how involving it will actually be. I have a hard time imagining them putting anything together in SC13 that would be interesting enough to pull me out of solo play and having complete (100%) control over my city with no adverse infections from other players, be it that they are trying to vandalize my city or just having a little fun at my expense.

  Edited March 21, 2012 by City Planner  

[A Nonny Moose 8,261]

The system as I have envisioned it would be very flexible. The Server-Back-end can be one machine. It does not have to be a PC, simply connected through the Internet to the client PC's. As volume grows, the central system can be expanded, moving the actual processing away from the communications server to something more capable. In the end you could have a rather interesting cluster of machines, and perhaps not all in one location. This would, of course, be income and margin dependent.

The client PC's could be very small if they didn't have to run the simulation code. They would be smart workstations which opens the ability to have a very nice graphics system. It may very well be that a dichotomy of this type would work very well indeed.

This would, in effect, spin off a new profit center for EA/Maxis, and it would have to be a subscription system that provided regular income from all users. Perhaps the initial price includes a year of server service, but after that an annual or quarterly license fee should be expected.

The whole business plan for this would mean that they could start on a relatively small scale and grow with success. It also means a totally different business model than has perhaps been really seen before.

EA/Maxis could easily become a very large global corporation. And why not? The initial reaction might be poor, but if people really want this level of city simulator, there it is. I suspect if they really expand this out of the "game" situation into more RL capability, there will be other customers besides gamers. The gaming introduction makes a nice opening ploy however. With continuing success, the planning departments of several municipalities could well come on board. A corporate client group would help keep fees down for the individual user.

I guess what you hear talking is forty-odd years of main frame experience with distributed systems.

[Mark Waybill 242]

I always thought when I contemplated whether SC5 would ever be made or not for the past ten years that if it ever was it would surely be part of some evolution in technology like this. Indeed when I was working on Virtucity using Game Maker and it's object based 'programming' I even considered something similar to GlassBox but like Game Maker, and the concept of an 'agent', 'map' etc were in my mind a long time ago. But not having any programming skill myself, and knowing such a project would require a lot of workers and effort and so forth, naturally that remained a dream. To see them create this engine is indeed quite fantastic, as it is exactly the sort of thing I imagined might be created if SC was ever to return.

But getting back to Multiplayer, The concept of client PCs being smart stations and the Server handling the actual simulation and code sounds very good. I always felt that for a prioper integration of the Sims and SimCity and for the true potential of the game to be realised this sort of server/client model should be used.

As to the subscription while normally I shy away from any spending and am like Scrooge in most areas, this is about the one true hobby I have besides coin collecting so as long as the fee was reasonable I would do my bit for the cause.

[croxis 150]

But getting back to Multiplayer, The concept of client PCs being smart stations and the Server handling the actual simulation and code sounds very good. I always felt that for a prioper integration of the Sims and SimCity and for the true potential of the game to be realised this sort of server/client model should be used.

As to the subscription while normally I shy away from any spending and am like Scrooge in most areas, this is about the one true hobby I have besides coin collecting so as long as the fee was reasonable I would do my bit for the cause.

This is massive multiplayer games work. The client is a dumb terminal with enough information of the game state to display it and maybe some client side predections. Most client-server games (tf2, counter strike, etc) work this way as well. We know however that SimCity 2013 does not work this way, all simulation is client side and the servers just store save games and other global information.

I don't think a client-server model would work well either. That is a LOT of data to transmit from the server to client. Every building and agent. Eve-online chokes up on space battles with only a few hundered players/agents (that is mostly cpu limits, but the bandwidth costs are not anything to sneeze at either, but still worth discussing*). If you are simulating a large, full city that is a lot of data to be sending to the clients. For a dumb client the postition of every car and the changes of every cell in the water table has to be send. To easy up on the bandwidth the client can be smart and simulate some of it client side. So you simulate traffic and water and before we know it we are esentally simulating almost all the game on the client, so why bother doing it on a server?

*Lets assume that with the multicore design an average city simulation tile takes up the processing power of one cpu core. Even if 100,000 people play the game at any given time that is 100,000 cpu cores needed. That is A LOT. If you look at the online user graph at steam it fluctuates a lot during a 24 hour period. EA/Maxis would have to anticipate enough hardware for peak conditions and then have it sitting idle during off hours. And what happens if active simulation count gets above the server farm's capacity? Do you REALLY want want to wait in 40 minute queues on launch day, which is what tends to happen with MMOGs?

[squeaky024 14]

But getting back to Multiplayer, The concept of client PCs being smart stations and the Server handling the actual simulation and code sounds very good. I always felt that for a prioper integration of the Sims and SimCity and for the true potential of the game to be realised this sort of server/client model should be used.

As to the subscription while normally I shy away from any spending and am like Scrooge in most areas, this is about the one true hobby I have besides coin collecting so as long as the fee was reasonable I would do my bit for the cause.

This is massive multiplayer games work. The client is a dumb terminal with enough information of the game state to display it and maybe some client side predections. Most client-server games (tf2, counter strike, etc) work this way as well. We know however that SimCity 2013 does not work this way, all simulation is client side and the servers just store save games and other global information.

I don't think a client-server model would work well either. That is a LOT of data to transmit from the server to client. Every building and agent. Eve-online chokes up on space battles with only a few hundered players/agents (that is mostly cpu limits, but the bandwidth costs are not anything to sneeze at either, but still worth discussing*). If you are simulating a large, full city that is a lot of data to be sending to the clients. For a dumb client the postition of every car and the changes of every cell in the water table has to be send. To easy up on the bandwidth the client can be smart and simulate some of it client side. So you simulate traffic and water and before we know it we are esentally simulating almost all the game on the client, so why bother doing it on a server?

*Lets assume that with the multicore design an average city simulation tile takes up the processing power of one cpu core. Even if 100,000 people play the game at any given time that is 100,000 cpu cores needed. That is A LOT. If you look at the online user graph at steam it fluctuates a lot during a 24 hour period. EA/Maxis would have to anticipate enough hardware for peak conditions and then have it sitting idle during off hours. And what happens if active simulation count gets above the server farm's capacity? Do you REALLY want want to wait in 40 minute queues on launch day, which is what tends to happen with MMOGs?

But what if you don't need to send the location of absolutely everything in your city? You only need the neighbors to know how much traffic goes in/out of what road, (same with utilities, ports, airports, power, pollution) It's just a number that is updated once in a while. The client runs their own simulation, so It uses the data "1000 cars IN on road X" and it does that, 1000 cars come in, and people go to work. You don't need the server to know where Susy is, how much coffee she drank, where she works, and and what's on her mind. The Client-Server connection can be simplified a lot without sacrificing on the actual connectivity of the two cities. I am certain you do not need a central machine to process everything and people will be able to host their own server and have maybe 2 or 3 friends join them (or rent a server for some money to host more if they don't have enough bandwidth, like in any other multiplayer game)

Another point is what if sometime down the road (many years from now), maxis leaves this to rot like sc4. They shut down the website, shut down the beloved central server, and now we all cannot play multiplayer? It is important to decentralize the multiplayer feature and allow people to play in a self contained multiplayer region that is unaffected by outside influences (other regions). Perhaps there should be an option if you want resources from outside your region, but If someone wants to play completely separate from "the rest of the world" and play with their friends and their friends only, they should be able to do so, and maybe substitute in their own resource quantities and prices as a simulated "global" market. Personally I like to play as if I am on a completely separate world with my friends and I don't want to have prices or availability of resources that don't match how I would like them to be.

edit: Basically what I am trying to say in the first paragraph is that you don't need to run someone else's city, so you don't need to know all the aspects of it. Each person simulates their own city, and only certain outputs that are necessary for the simulation to function are sent out to other players and maybe a "snapshot" of the city to keep the game updated.

  Edited March 22, 2012 by squeaky024  

[Mark Waybill 242]

Perhaps there should be an option if you want resources from outside your region, but If someone wants to play completely separate from "the rest of the world" and play with their friends and their friends only, they should be able to do so, and maybe substitute in their own resource quantities and prices as a simulated "global" market. Personally I like to play as if I am on a completely separate world with my friends and I don't want to have prices or availability of resources that don't match how I would like them to be.

I agree. I'd like the option of creating a new 'world' for a region rather than using the default SimEarth accessed online.

[DazzaDazza 29]

The system as I have envisioned it would be very flexible. The Server-Back-end can be one machine. It does not have to be a PC, simply connected through the Internet to the client PC's. As volume grows, the central system can be expanded, moving the actual processing away from the communications server to something more capable. In the end you could have a rather interesting cluster of machines, and perhaps not all in one location. This would, of course, be income and margin dependent.

The client PC's could be very small if they didn't have to run the simulation code. They would be smart workstations which opens the ability to have a very nice graphics system. It may very well be that a dichotomy of this type would work very well indeed.

This would, in effect, spin off a new profit center for EA/Maxis, and it would have to be a subscription system that provided regular income from all users. Perhaps the initial price includes a year of server service, but after that an annual or quarterly license fee should be expected.

The whole business plan for this would mean that they could start on a relatively small scale and grow with success. It also means a totally different business model than has perhaps been really seen before.

EA/Maxis could easily become a very large global corporation. And why not? The initial reaction might be poor, but if people really want this level of city simulator, there it is. I

suspect if they really expand this out of the "game" situation into more RL capability, there will be other customers besides gamers. The gaming introduction makes a nice opening ploy however. With continuing success, the planning departments of several municipalities could well come on board. A corporate client group would help keep fees down for the individual user.

I guess what you hear talking is forty-odd years of main frame experience with distributed systems.

Ummm that sounds like CXL plant offer(are you shore you didn't work for Monte Cristo), I think is safe to say that was a dismal failure, CXL was also a heavily focus mailtyplayer game, the game cant relay function without it, it was as mailtyplayer as a city building game could get, and people still didn't wont to pay for it, because the majority of the time your playing a single player game, slapping a simcity name on it wont make people pay to play online, at the end of the day your only relay trading statistics and values, it's a lot of work for very little interaction (in my opinion) .

[croxis 150]

Ummm that sounds like CXL plant offer(are you shore you didn't work for Monte Cristo), I think is safe to say that was a dismal failure, CXL was also a heavily focus mailtyplayer game, the game cant relay function without it, it was as mailtyplayer as a city building game could get, and people still didn't wont to pay for it, because the majority of the time your playing a single player game, slapping a simcity name on it wont make people pay to play online, at the end of the day your only relay trading statistics and values, it's a lot of work for very little interaction (in my opinion) .

That wasn't how cities xl worked. The very *basic* design of cities xl was that the simulation was run client side and the server just hosted the save game and the world market. SC2013 is going to work in *basicly* the same way.

[A Nonny Moose 8,261]

IMHO and considerable experience in both configuring and running transaction processor systems, I think that running the simulation client side is too damned expensive. It reduces the server to being a journalizer and makes synchronization in multi-player impossibly difficult.

Think about this for a second:

We know that EA/Maxis is very sensitive about releasing their code to anyone. How much better if all they have to provide is the client-side stub?

A single server or cluster of machines can run as many instances of the main as needed, so that many (16) user regions can be active concurrently on as many copies of the base code as is required to maintain response time.

Small XML packets can be exchanged to update the graphics and accept keyboard/mouse inputs. In an interactive game like this, not much changes at a time.

The main problem is the queuing and response at the service end. However, this is resolved though careful line loading.

The client machine need not be even able to run the full simulation, removing more than half the code. If all you have is a standard graphics encoding in, say, OpenGL, and the usual loose nut handlers, the machine could be running Android. The plus for the market extending to just about anything makes this more attractive to EA, because the current market for 'Droid is growing like a weed.

It also solves any debugging and fixing problems in the main program, because they can be done once, and propagated only to the local server libraries. Only error-fixes in the client-side need be propagated, and this could happen when a client logged in.

I really am not in favor of putting umpteen copies of the main software into the general market.

[DazzaDazza 29]

Ummm that sounds like CXL plant offer(are you shore you didn't work for Monte Cristo), I think is safe to say that was a dismal failure, CXL was also a heavily focus mailtyplayer game, the game cant relay function without it, it was as mailtyplayer as a city building game could get, and people still didn't wont to pay for it, because the majority of the time your playing a single player game, slapping a simcity name on it wont make people pay to play online, at the end of the day your only relay trading statistics and values, it's a lot of work for very little interaction (in my opinion) .

That wasn't how cities xl worked. The very *basic* design of cities xl was that the simulation was run client side and the server just hosted the save game and the world market. SC2013 is going to work in *basicly* the same way.

How the games runs on online mode compared to each other doesn’t really matter, if they charge people for the privilege of trading it’s the same thing and they will end up with the same result.

[croxis 150]

Ummm that sounds like CXL plant offer(are you shore you didn't work for Monte Cristo), I think is safe to say that was a dismal failure, CXL was also a heavily focus mailtyplayer game, the game cant relay function without it, it was as mailtyplayer as a city building game could get, and people still didn't wont to pay for it, because the majority of the time your playing a single player game, slapping a simcity name on it wont make people pay to play online, at the end of the day your only relay trading statistics and values, it's a lot of work for very little interaction (in my opinion) .

That wasn't how cities xl worked. The very *basic* design of cities xl was that the simulation was run client side and the server just hosted the save game and the world market. SC2013 is going to work in *basicly* the same way.

How the games runs on online mode compared to each other doesn’t really matter, if they charge people for the privilege of trading it’s the same thing and they will end up with the same result.

Well you did say that! You are right though. We have 0 evidence so far of any of this mmo-serverside monthly fee. Only that saved games are stored online, global supply/demand, leaderboards, and optional multiplayer regions that operate asyncriously.

[A Nonny Moose 8,261]

Well you did say that! You are right though. We have 0 evidence so far of any of this mmo-serverside monthly fee. Only that saved games are stored online, global supply/demand, leaderboards, and optional multi-player regions that operate asynchronously.

Sure does sound like a server/client system, though, doesn't it. A lot simpler at the EA/Maxis end if it is a massive transaction processor, rather than a simple server. The individual regions must be synchronized in multi-player, but they can have as many instances of the same region with different players running totally out of touch with the others. Integrating the "global supply/demand" will be a rather large undertaking, especially as the universe of discourse grows and they get into multiple servers.

This kind of ambitious talk pretty much puts me in mind of a certain three tier system we installed over 30 years ago that had two pairs of shadow main frames in two locations connected to a set of redundant regional servers that communicated with not PC's but mini-computers that drove 'dumb' terminals. Replace the last two components with smart PC's and you've got a rather nice design. Eventually this expanded to allow connection of WiFi workstations on the road.

Frankly, the less code on the remote stations, the easier the system is to maintain and debug. The big problem with this design will be the update propagation delays for the graphics over the Internet. People who are not on broadband will suffer.

[croxis 150]

Sure does sound like a server/client system, though, doesn't it.

No it really doesn't. I have no idea where you are drawing this conclusion from. Please show us your evidence, because right now to a lot of the people who are already nervous about the mp aspects this is borderline FUD.

[Mootilda 76]

As far as I can tell, this is just "wishful thinking" on the part of A Nonny Moose. Let's hope that he's wrong, so that we have a hope of continuing to play the game after EA abandons it.

[Mark Waybill 242]

I foresee in the not too distant future a world where terminals do not handle much at all, and giant mainframe centres are set up to do all the grunt work, and our pcs are just input devices. Then again I used to read a lot of Verne and Huxley and so forth as a kid.

[DazzaDazza 29]

-Snip-

How the games runs on online mode compared to each other doesn’t really matter, if they charge people for the privilege of trading it’s the same thing and they will end up with the same result.

Well you did say that! You are right though. We have 0 evidence so far of any of this mmo-serverside monthly fee. Only that saved games are stored online, global supply/demand, leaderboards, and optional multiplayer regions that operate asyncriously.

lets hope that we still have 0 evidence in a years time,

Sill buy storing Saved games "online" is still costing some one money, whether its 50cents or $50 it's still costing someone something And they will wont to get that back somewhere,

, like this site, it may seem free, but its not free, it is still costing someone something to run

arrrrrrrrrrrre I love speculating

[A Nonny Moose 8,261]

The only proof is in the pudding. Take a look again at all the hype from the announcement and the conversations with the developers. I merely took the most extreme path based on what is possible in my experience.

1. A server must be present to support all this global economy this and that. They say saves will be on the server.
2. It is more economic to run a central system with one copy of the code that can be instantiated multiple times on demand rather than distribute the whole program to every PC on the network.
3. It is also better for the owners of the code to know where it is at all times and have a single point of maintenance.
4. A client stub in the local smart terminal (PC) is less external code to have fully debugged. It also reduces the need for a drastically powerful PC station. Frame buffers do not have to be replaced holus bolus. Changes can be made incrementally.
5. A central repository for DLC is better than everyone working alone. It implies verification of contributions.
6. Stand-alone, off-line support becomes collateral damage in this kind of system. Stand-alone on-line support remains possible.
7. People with slow Internet lines may want to upgrade, but if the packets are small enough should not be much affected.

All this fear of not being able to run off-line is just parochialism. Everyone wants the world wide web, but at the same time some people want to have their cake and eat it too. From the number of MMOS games around I fail to see why anyone should be over excited by my speculations in this regard.

To put in all the features that everyone wants, the practical solution is a central point. Access to optional features and DLC material then becomes a matter of granting permissions rather than transmitting massive amounts of program text. This scenario also improves the chances for successful operation on a laptop or other small devices.

One should keep the main chance in view.

[Gamma Stardust 50]

You might be right, and in my perspective unfortunately right. It is exacly because of some of those hints, but other ones too, that I fear we won't be given an option to play Simcity offline, and the very reason of the poll i have put up.

The change, trend is definitely in that direction: online, online, online, online... you just gave 7 examples of online upon online; one reads the previews and interviews... more online and online; social this and social that.

Simply put, we are NOT talking about the same genre of gameplay Simcity originaly brought us. There is of course leaway for change, but specially improvement when one thinks of a sequel to whatever the game. Now we're being hinted at something that diverges more and more from what a traditional Simcity player came to love (building cities and manage them - despite as a lone process). But this is not the apologia of "tradition" per se, this is in defense of what the game genre actualy is (was) essentialy, despite of natural network limitations at the time of the originals (not because of), which shaped them that way.

If you mean "parochialism" in the sense of backwardness against change, i doubt anyone here would worry about having all those additions as true additions but not as a replacement, even if imposed by technicalities, which many, as players (not developers), one couldn't give a dime.

If you mean "parochialism" in sense of "evangelization" by raising awareness and "persuading" potential players that the game might not fulfill certain basic expectations, so be it, specially since i am one of those potential players and hope not to be hyped into an unfortunate but increasingly probable final frustration.

If you mean "parochialism" as opposed to "universalism" by one whishing to play alone whithout the need to worry about multiple cities interdependence at the hands of a third party raising the issues of healthy (or not) cooperation, competitiveness and/or more generaly what comes about as griefers, that detract from enjoyment, in many multiplayer games today. On the spot!

But i also have to wonder who exacly is in the position of "having their cake and eating it too" in this regard because whishing both to play offline and online depending on one's mood does not have to be mutually exclusive options, many games have dealt with this, what you present are indeed technicalities imposed by an online model of gameplay which make it difficult to allow for simultaneous offline possibility to play the original genre. But these are not in the base expectations of city building players' community (as i see it of course). So if anyone here wants to "have their cake" (have an online mode and what it implies) and "eat it too" by still allowing one to play offline, i don't plainly see it.

On one side i am not sure Maxis is actually interested in providing an offline mode anymore (no doubts about online), on another side a good portion cares sufficiently to be skeptic about mantaining interest in the game, on yet another side we have many which would probably like the best of both worlds but are still doubtfull about commiting and apparently show themselves able to compromise in favour of multiplayer features in detriment to an offline option.

All in all, only if a player was aware of the technical limitations imposed by the apparent online model choosen (as if it had to as a player) would he be in that position of "wanting the best of both worlds". Only if Maxis had come forward and plainly stated that it was in their intentions to provide offline mode would also the same thing apply.

So... we'll have to wait, of course the more we do and Maxis keeps the secret behind doors, I am inclined to take it that we are just being concealed of the truth which would certainly hammer down whatever hyped marketing campaign may be produced. And that would ultimately affect a certain bottom line, can i censor them? no, of course not, it's their game afterall.

What I can assure anyone here is that i won't part my "hard earned" because of some "clever", "bate and switch" marketing hype deception, if any. ("Bate" being the original series one came to love and "Switch" being the final shiped product which might come at odds with the original). Specially since lately this is becoming a very common and fraudulent business practice in this industry.

And believe me I definitely whish I am finally proven wrong, because i love this game so much and kept my self waiting for so long.

sorry for no TL;DR

  Edited March 24, 2012 by Gamma Stardust  

[Mark Waybill 242]

I have nothing against SC being 'online' for the reasons Moose gave. Also I have a laptop, so I'd rather not fork out $$$ for some fancy gaming PC just so I can use it for one game (I'm not a gamer in the usual sense of the term, and SC is pretty much the only recreational thing I use my pc for) Centralisation is imo the way forward with these things. As Moose said it makes maintenance a lot easier, allows client terminals to be less hi tech and so forth, and user contributed content like BAT files can be verified and tested before being made available. Also this facilitates if Maxis/EA decide to have a user alliance to some extent as regards expansion packs and user contributed material.

My main personal concern is the cost, as playing SC4 on Steam takes up a lot of credit. I am unsure exactly what the cost would be for this type of game so, I hope it will not be excessive to the point I need to recharge my internet every week. I love city building but I am also a pauper.

[Moskva 75]

Simcity, by definition, cannot be just multiplayer. It would destroy the game. Meanwhile they confirmed some SP version like it has been said, we right don't know about offline, but i've hope it'll be offline... I am amazed how the gaming industry began literally punching hardcore fans and casual gamers. First with the DLCs, then with non-dedicated servers, then with paying monthly fees to play online... this in various videogames i've seen and few played in recent years. There really should be no need for the casual gamer to pay to play online. I've bought World in Conflict and CoD4 in 2007 and never spent a cent to play them online but they're still there, because of course someone is renting a server. It could work the same way, some people rent servers and make Clan/Public regions. Or better, we have some Massgate-like system from World in Conflict, where some servers are set up by the gaming companies. However i believe the most feasible is the first one, you pay for a server and you're the master of it, with random people coming in or friends. However it also could be disaster like CXL, where you have to pay monthly fees to play online. After all EA is always looking forward to make most money possible. They wont care about hardcore fans unless they've impact on their economy. Good to know the hardcore fans in SC sure have weight. There's too much money raping today, i wish things were like back in 2007: You buy the game, have SP and MP versions, no additional costs, patching, etc.

[Ocram's Razr 963]

Anno 2070 is online oriented but you can still play single player offline.

[Catmando 7]

Dear God, I had a dream last night that Maxis announced a fully featured offline single player mode. I'm not sure if this is a good omen or a sign I'm completely obsessed with the new game.

[meowza 230]

My subconscious tells me the same thing.

We can't deny it, we all want offline mode.

[A Nonny Moose 8,261]

My subconscious tells me the same thing.

We can't deny it, we all want offline mode.

Agreed. However we have no defense against being forced on-line if saved games can only be on the server. The defense against not having an off-line, free-standing game with full features is to keep SC4 going forever, and ever, and ever. Amen.

[JanH 5]

The only proof is in the pudding. Take a look again at all the hype from the announcement and the conversations with the developers. I merely took the most extreme path based on what is possible in my experience. [...]

Well, I look, and they have indeed been careful not to make any definitive statement on an offline-only mode, but, nevertheless, there is definitive evidence pointing against a server-based game:

1. The announcement on Rock, Paper, Shotgun where they state that "If the internet goes out for a bit, you can still play."

2. This slide from their GDC presentation, which says "Asynchronous server model: No reliance on dedicated live server running to support your play session, graceful degradation if we have server issues (highlighting by myself)

[A Nonny Moose 8,261]

By graceful degradation they probably means your game will go slowly static and eventually you can do some kind of suspend until they get back up. They do say that saves are on the server.

Since the Internet is unlikely to go down, it is more likely that one or the other end may crash. Your ISP may also go down, but most of these guys have a 24/7 policy. If you have a reliable firewall, your end is probably safe, but their end is subject to all kinds of nasties including hackers. I hope their security is as good as mine.

[KeanoManu 42]

But would it really be such a good idea to store single-player saves in the cloud? Have they really said that ALL saves will be stored at their servers or just implied that multi player saves will?

If all saves, single and multi, will be saved in their servers it will take up an incredible amount of space for them. I have so many different SC4 regions that i can't keep track of them anymore but i still have them all saved on my computer (somewhere) and would expect Maxis to keep them all too even if i havn't played a particular region in seven years time.

I also regularly play a game called Football Manager, which is one of these games that has a new edition each year (which is a concept i prefer every day of the week over GDC and other micropayment solutions). I also keep back-ups of my progress from time to time. In last years edition the size of my back-up folder of saved games ended at 8,8 GB each save file are around 100-200 mb). And i'm just one player.

There's no reasoanable explanation to why they should keep single player saves in the cloud. I can see why they would keep multi player there though.

So... Have they really said that both single and multi will be saved on their servers?

[A Nonny Moose 8,261]

But would it really be such a good idea to store single-player saves in the cloud? Have they really said that ALL saves will be stored at their servers or just implied that multi player saves will?

If all saves, single and multi, will be saved in their servers it will take up an incredible amount of space for them. I have so many different SC4 regions that i can't keep track of them anymore but i still have them all saved on my computer (somewhere) and would expect Maxis to keep them all too even if i havn't played a particular region in seven years time.

I also regularly play a game called Football Manager, which is one of these games that has a new edition each year (which is a concept i prefer every day of the week over GDC and other micropayment solutions). I also keep back-ups of my progress from time to time. In last years edition the size of my back-up folder of saved games ended at 8,8 GB each save file are around 100-200 mb). And i'm just one player.

There's no reasoanable explanation to why they should keep single player saves in the cloud. I can see why they would keep multi player there though.

So... Have they really said that both single and multi will be saved on their servers?

Mostly because it is a pain in the ass to have two technologies when one will do.

One of the cheapest resources around today is storage, and a back-ended server can have thousands of terabytes of storage. My current on-line storage for SC4 Regions is 259.5 MB for 10 regions. I don't keep old regions around forever, but I do have backups in compressed files off-line. For my average region of 26MB, that's .000026 TB. A drop in a bucket. Even if the necessary storage increases by an order of magnitude, it will be in the hundreds of MB range. I currently have 1.8 TB total storage on my machine if everything is on-line. I don't see why a cloud (echh! I hate that term) should be smaller than 500 TB.

Many people don't think in the terms of distributed computing, but I do. It used to be my business. File servers, communications servers, separate special purpose processors, smart stations, dumb stations, etc. By 1990 this kind of thing could be done with PCs with a little expansion to a multi-bus architecture.

  Edited March 26, 2012 by A Nonny Moose  

[Mootilda 76]

Sure, but have they actually said that single player games will be stored online only, or is this just speculation?

[jacksunny 741]

This is all just speculation, we know very little about this game.