EPTFuture of Games? - Page 3
| Forum Index > General Forum |
|
HiOT
Sweden1000 Posts
| ||
|
azndsh
United States4447 Posts
the one benefit I can see is that it'll allow access to a much larger variety of games, but other than that you won't see any improvements in games that already have good servers or don't use a centralized server to run their games (ie starcraft) | ||
|
pyrogenetix
China5098 Posts
| ||
|
cava
United States1035 Posts
| ||
|
armed_
Canada443 Posts
On March 25 2009 08:22 azndsh wrote: this will be good for multiplayer games, but it'll be definitely be worse for single-player games unless your internet connection is considerably better than your computer (relatively speaking) Lag in multiplayer games would actually be probably be much more noticeable, because as Luhh mentioned, there's lag between your own input and the client. Most modern netcode for games can hide smaller amounts of latency; even if the game client doesn't have all the information necessary to instantly resolve a player's actions, through a combination of extrapolation and other fancy stuff it can make it seems as if the reaction is instant. When lag becomes slightly higher, rather than the player's actions being delayed onscreen, you'll just have some inconsistencies between the client and the server; ie. the player's bullets appear to hit on-screen but the they didn't on the server so no damage is registered. There's no similar way to hide lag between your own input and the client; it's simply impossible. That lag will always be noticeable, in addition to the normal lag between two clients/the server and the clients. Although that latter bit could probably be eliminated if all players were connected to the same Onlive server farm. | ||
|
McCrank
204 Posts
| ||
|
CubEdIn
Romania5359 Posts
On March 25 2009 08:29 cava wrote: This uses the internet. Connecting to anything over the internet has delay regardless of bandwidth capabilities simply due to distance and hops required over numerous amounts of routers. There is no such thing as lagless gaming over the internet. Yet. But see, we manage to play relatively lagless gaming on the internet right now. Even if it's not 100% lag free. Have you tried quake live? It's great, even if it's a FPS. Also, if someone were to tell you 10 years ago what the internet will be like in 2009 you wouldn't believe him/her. Technology and internet speeds are improving each year. This kind of thing may not look so insane next year. And maybe in 3-4-5 year it will start taking over consoles and so on. Have some faith. There's no other we're able to evolve if people aren't at least TRYING new things.  | ||
|
D00dles
Cambodia217 Posts
| ||
|
Random()
Kyrgyz Republic1462 Posts
On March 25 2009 07:14 Day[9] wrote: ok i was totally wrong about something: there will be some latency that's inherent with your own internet connection. shouldn't be more than b.net though! should be really sick! This should be quite obvious, especially for someone related to math/computing like you No hyper-duper-compression is able to eliminate signal latency.It seems to me that this is one of those things that are really cool on paper, but fail in practice. A very important implication of such setup is that you will have considerable input delay, which is quite different from in-game network delay, especially in fast-paced games such as FPS. Even in laggy network conditions these games emulate immediate response to your input, e.g. in Quake/CS the weapon firing animation happens instantly whenever you press the fire button. In addition, a great deal of motion prediction logic is used in network games, which eliminates visible lag even more. Starcraft is a good example too. Even if the network delay is 100ms, it is very much playable. However, would it be as much playable if the mouse cursor responded to your mouse movements with a 100ms delay? Plus, this input delay would in effect be added to the network delay if you were playing a multiplayer game, making the game experience very bad. | ||
|
Jovan
Canada65 Posts
Much respect for your podcasts, but, are you trolling? 1. The concept is not new, and there's nothing to revolutionize Client-server, dumb-terminal, cloud-computing models have been around for 30 years. To any programmers here, RPC and how often it's used should be familiar. (Cloud computing is a marketing droid term that's come up recently) I read on a game physics programmer's blog how he found it idiotic that every time the games industry gets something new, it's heralded as the dawn of man, when it's in fact been around in the software industry for a long time. This is more to deal with programming than these types of high level client-server models. I will try to dig up his blog. 2. Latency will kill this and I kind of chuckled when you threw in: ok i was totally wrong about something: there will be some latency that's inherent with your own internet connection. shouldn't be more than b.net though! should be really sick! This is THE REASON why this hasn't been done, and why it will fail. Compression or not, if you scroll your mouse around, you will get latency. I am surprised you haven't considered this yourself and started asking questions right away regarding this. Not worse than bnet latency? You shouldn't be new to players who prefer to play at more than 60 frames per second when they play FPS games. You can tout this megalomaniac compression algorithm, but you can't tell me that all the data will arrive, uncorrupted. Some of the data will be dropped and it will have be resent. It happens in Starcraft, Warcraft, Quake, Unreal Tournament and even in your very own browser. It's nothing new, its just that games like Quake have to go through extra means of simulating what's not there. Can a computer without any hardware simulate up to 200ms of Crysis? Not at all. 3. North American connections suck Awhile back the government handed out $200B to telecommunication companies to upgrade the infrastructure and lay down the fiber-optic cable, but they never did. Word is they blew it on coke, hookers, and fast cars. I wouldn't doubt it. More information on the subject. As you are no doubt aware most of us still have, at the most, up to 10 Mbps, and are restricted to less than 100 GB bandwidth per month. I'm interested in as to how what kind of algorithm they've developed to compress this so much that I can play 1920x1200x60 game for about 7 hours, every day, for a month, and not run out of bandwidth. Conclusion You're being scammed and you won't get to play Crysis without buying your own hardware. I left out some other parts, but those three are the big three. Edit The basement dwellers over at Slashdot have a copy of the story. | ||
|
Random()
Kyrgyz Republic1462 Posts
On March 25 2009 08:41 CubEdIn wrote: Yet. But see, we manage to play relatively lagless gaming on the internet right now. Even if it's not 100% lag free. Have you tried quake live? It's great, even if it's a FPS. Also, if someone were to tell you 10 years ago what the internet will be like in 2009 you wouldn't believe him/her. Technology and internet speeds are improving each year. This kind of thing may not look so insane next year. And maybe in 3-4-5 year it will start taking over consoles and so on. Have some faith. There's no other we're able to evolve if people aren't at least TRYING new things. A signal cannot travel faster than light. Even under ideal conditions, i.e. speed of signal propagation = speed of light (which is obviously impossible when there is a number of mid-points, or "hops" that require the data packet to be processed and routed accordingly), the minimum time it would take for a signal to travel from say England to Eastern China would be ~50 ms. Bandwidth will continue to improve, but latency not as much. | ||
|
armed_
Canada443 Posts
No, there's simply no way to completely eliminate lag. Internet speed has already ceased to be a major limiting factor in terms of latency; the bandwidth available to the average person is more than enough for a videogame. The problem is simply moving the data from point A to point B, and even if you're doing so at the speed of light straight across the circumference of Earth there is still latency. Of course it's not very noticeable then unless you're pretty much on opposite sides of the planet, but taking physical topography, the cost of simply building the physical infrastructure necessary, and the transfer of data between different routes into account, lowering latency to the point where it's not noticeable just isn't realistic. Edit: Damnit Random(). At least I beat you on the other point. ;P | ||
|
ShadowDrgn
United States2497 Posts
On March 25 2009 08:21 Lobbo wrote: Whaaa, scary scenario hacking will be to easy and god forbid if someone gets a single trojan or something onto their networks servers :O Blackhatftw! Actually, one huge advantage of this is that hacking will be virtually impossible because your computer/console won't have any game state information to read or manipulate. That's pretty awesome. In fact, this solution has always been the holy grail of cheat-free gaming; it just hasn't been feasible. | ||
|
HiOT
Sweden1000 Posts
On March 25 2009 09:01 ShadowDrgn wrote: Actually, one huge advantage of this is that hacking will be virtually impossible because your computer/console won't have any game state information to read or manipulate. That's pretty awesome. In fact, this solution has always been the holy grail of cheat-free gaming; it just hasn't been feasible. cheating free yeah, but you still have to connect to their servers, and if there is a trigger happy virus on there you will also have it transferred back. | ||
|
armed_
Canada443 Posts
On March 25 2009 09:04 Lobbo wrote: cheating free yeah, but you still have to connect to their servers, and if there is a trigger happy virus on there you will also have it transferred back. No, if you think this you lack a basic understanding of how this works(and I suspect of how virus' work in general.) | ||
|
Jayme
United States5866 Posts
On March 25 2009 08:29 cava wrote: This uses the internet. Connecting to anything over the internet has delay regardless of bandwidth capabilities simply due to distance and hops required over numerous amounts of routers. There is no such thing as lagless gaming over the internet. This is true to an extent but do keep in mind that the internet is consistently getting better and better. I wouldn't be surprised if we saw a very near sub 5ms internet connection within the next 5 years. The closer to the speed of light information can travel and all. | ||
|
ShadowDrgn
United States2497 Posts
On March 25 2009 09:07 armed_ wrote: No, if you think this you lack a basic understanding of how this works(and I suspect of how virus' work in general.) Someone could hack into the Onlive servers, alter their compression program to encode a virus into the video stream, and then exploit a bug in the local decompression algorithm to execute the virus on your machine. This would only be a few orders of magnitude more difficult than spreading a virus in a normal method.  | ||
|
blapsd
England121 Posts
On March 25 2009 08:57 Random() wrote: A signal cannot travel faster than light. Even under ideal conditions, i.e. speed of signal propagation = speed of light (which is obviously impossible when there is a number of mid-points, or "hops" that require the data packet to be processed and routed accordingly), the minimum time it would take for a signal to travel from say England to Eastern China would be ~50 ms. Bandwidth will continue to improve, but latency not as much. why on earth do you think people in europe will have to play on servers in eastern china? You will play on your closest server. I can typically play left 4 dead at less than 5ms lag in the uk to a uk server (i have a reasonably good connection). I'm sure that would be fine no? And as for playing other people online, couldnt your just play people also connected to the server you are already streaming from? Hence giving you no additional lag? I'm no expert by any means and if someone else could tell me where im wrong please do, i just dont see why ppl argue about the whole lagging issue with playing ppl the other side of the world. The answer is you simply dont. | ||
|
Pseudo_Utopia
Canada827 Posts
Seemed like a great idea when I started reading the thread, then p3 is not so great :O... | ||
|
armed_
Canada443 Posts
On March 25 2009 09:15 blapsd wrote: I can typically play left 4 dead at less than 5ms lag in the uk to a uk server (i have a reasonably good connection). I'm sure that would be fine no? If you're lucky enough to live next door to the Onlive server farm, sure. On March 25 2009 09:12 Jayme wrote: I wouldn't be surprised if we saw a very near sub 5ms internet connection within the next 5 years. The closer to the speed of light information can travel and all. Would you really? Because even if you're transferring data at the speed of light in a straight line between two computers, you wouldn't be able to get a 5ms ping from anything more than ~750 km away. | ||
| ||
