Really? Given what Notch plans (2,000 DCPUs per machine - which is certainly doable), let's assume he's running a high-cpu instance (large) at Amazon for those 2,000 DCPUs. That's $500 per month for the machine. If he can't get that much money out of 2,000 players, he's doing it wrong.
Assuming you allow for eventual instead of atomic consistency, and allow for loss of CPU for several seconds if an instance goes kablooey, the load distribution is really not that hard. (And your spaceship really doesn't need five 9's. It's a game, so just blame it on "an ion storm". Or "a space monster" if that floats your boat ;)
Partitioning does not need to be too clever if you can constrain the number of ships per region - certainly something you can impose via game design.
Lag only matters if you interact twitch based. Notch's plans lend themselves to indirect agency via the simulated ship's computer instead, so if everybody perceives the world as it was a few seconds ago, not much is lost.
It's certainly a doable task. Notch's mind is in the right place to pull it off. No doubt, it'll be hard - but it'll be interesting to see what he comes up with.
That would obliterate the point of the MMO aspect though.
EDIT: One more thing regarding the ion storm or whatever - gamers aren't that forgiving about those things.
You could try and blame it on space monsters and what not, but gamers know pretty quick when something is borked. And they hate it hard. - Anything that isn't within their locus of control which negatively impacts game play is easily hated.
I can't stress enough how painful flamewars from pissed off gamers are to deal with.
I'm not sure how much you are aware of current occurrences in EVE online, but for what its worth - they too had the similar lag issues (while still being called excel in space).
They recently deployed a time dilation mechanic to make it possible for people to play the game and enjoy massive fleet battles, which has increased satisfaction amongst the player base immensely.
Well a first person shooter engine only really needs an (x, y, look-at) tuple for each player, certainly putting thousands of players on one map is doable. You're dreaming.
Assume for a moment a team of developers spend several months if not years doing infrastructure build-out to support the imagined CPU simulation system. Now they just have to implement the rest of the engine, the part powering the actual game, in which all players, CPUs, and other elements are active and interacting with one another at all times.
I'm hoping Notch succeeds, but lets be realistic -- the closest anyone has got to offline simulation is effectively "event queue and timer" and not for lack of trying.
> Assume for a moment a team of developers spend several months if not years doing infrastructure build-out to support the imagined CPU simulation system
Huh? A CPU simulation is not that complicated a task. If it takes Notch several developers and years of time, he's definitely doing it wrong.
> active and interacting with one another at all times.
Yes. That's a solved issue, mostly. It's a large task, but it's not an unsolved problem. (See e.g. social networking sites. Humongous amounts of people interacting with each other)
The point is that you'll need to make some concessions to the realities of large scale when it comes to the game design.
> the closest anyone has got to offline simulation is effectively "event queue and timer" and not for lack of trying.
Funny. And here I thought Havok just gave a talk on physics in MMOs. (GDC China).
I think you might be missing the point, all simulation continues while players are offline. Implementing consistent physics in an MMO is (very) hard, but having all players simulated even when they are logged off is something else entirely. The state of the art there is things like mail systems, auction houses, and skill queues.
I wouldn't be surprised if the offline simulation curtails your abilities somewhat, enabling a much lower CPU-effort simulation. We'll see, I guess :)
(There's also the point that a monthly subscription fee of $15 buys you a nice VPS slice these days. If you're willing to cut into the - significant - profit margin of MMOs, you have a lot of performance available for offline simulation. I'll stand by my judgment that it's a hard, but solvable, problem)
Assuming you allow for eventual instead of atomic consistency, and allow for loss of CPU for several seconds if an instance goes kablooey, the load distribution is really not that hard. (And your spaceship really doesn't need five 9's. It's a game, so just blame it on "an ion storm". Or "a space monster" if that floats your boat ;)
Partitioning does not need to be too clever if you can constrain the number of ships per region - certainly something you can impose via game design.
Lag only matters if you interact twitch based. Notch's plans lend themselves to indirect agency via the simulated ship's computer instead, so if everybody perceives the world as it was a few seconds ago, not much is lost.
It's certainly a doable task. Notch's mind is in the right place to pull it off. No doubt, it'll be hard - but it'll be interesting to see what he comes up with.