Scratch

From what I've seen, about 2 seconds, so multiplayers will be possible.

TorbyFork234 wrote:

From what I've seen, about 2 seconds, so multiplayers will be possible.

Of course, no one's using the 2.0 system yet.  So, who knows what the real performance is once people start using it?  We don't know.

*and, by "no one" I really mean only a handful of alpha testers.

Last edited by BoltBait (2012-11-05 13:17:17)

scimonster wrote:

About a second for me, using Linux. It used to be 2 seconds on Windows, but i think there were some recent speed improvements.

YES!

Only problem is, if there is a fighting game...
2 second lag where you can whack people. 

It's definitely not enough for anything that isn't turn-based (RTS, fighting, the list goes on). But hey, it's a service provided free of charge, you can't expect much. 

I think it'd be cool if you could use your own servers for this kind of stuff. I don't know exactly how online play works, but I suppose opening and listening to sockets as an hidden advanced feature would work for this purpose.

You could then program the server in any other language that supports sockets - even Scratch itself in the rare case it does. c: Any unauthorized socket request would be caught by the firewall in most cases.

BoltBait wrote:

TorbyFork234 wrote:

From what I've seen, about 2 seconds, so multiplayers will be possible.

Of course, no one's using the 2.0 system yet.  So, who knows what the real performance is once people start using it?  We don't know.

*and, by "no one" I really mean only a handful of alpha testers.

And anyone who was on scratch during scratch day 

technoguyx wrote:

It's definitely not enough for anything that isn't turn-based (RTS, fighting, the list goes on). But hey, it's a service provided free of charge, you can't expect much.   

I think it'd be cool if you could use your own servers for this kind of stuff. I don't know exactly how online play works, but I suppose opening and listening to sockets as an hidden advanced feature would work for this purpose.

You could then program the server in any other language that supports sockets - even Scratch itself in the rare case it does. c: Any unauthorized socket request would be caught by the firewall in most cases.

Actually, for practically anything I've done via a server (FPS, minecraft, gmail chat, etc...), it's about 1-4 seconds delay. You can still make the multiplayer games and have them not be turn based. Just whenever you have something done, also post the time with the nifty new Scratch-time block (from what I've heard from other scratchers, and possibly the ST, i forget who), and whoever did it earlier gets right of action.

Chainmanner wrote:

Two seconds is rather too little. Is there any way in improving this?

Too little? You'd rather it be larger, to be 5 seconds ? 

I don't really understand why you would need it to be faster. No information can travel instantaneously between computers.

I'm testing. It looks like it is very fast, even when our internet is slow for some time (0.1 mbps up) it takes 0.5 secs, less if the up is normal. (6mbps+)


Edit: do note there are like 100 testers as of now, so it will be slower when finished unless more servers are added.

Last edited by Laternenpfahl (2012-12-04 00:38:23)

veggieman001 wrote:

MathWizz wrote:

My question is, what is the system behind the variables. Is is a socket/server based thing, or a HTTP request for every update? If it's polling, I'll really be disappointed. 

I think it's HTTP, but I'm not 100% sure on that.

Whhhhhhyyy???