Why can't more devs make fighting games with good online?

omgad, your fucking reading comprehension is GOD awful.

I said no net code can fix a high lag situation, you’re abusing the connection which isn’t going to work properly FOR ANY net code.

hey this waffle maker doesn’t make for a good computer I must be right!!! when you use something its not intended to do, like play in high lag situations, it won’t be good. Thats fucking obvious. However if you were to give 1 game both versions of roll back\input delay, roll back hands down wins every time under ideal conditions. When GGPO has ideal conditions, it is a magic fix. Almost everyone in the world could play with those condition requirements and have a god like connection. That’s why I call it a magic fix where as input delay isn’t even playable with people in my own damn city

I didn’t change anything in my quote from the last page, please learn to read. Rather than argue your moronic perspective yet again.


I honestly think you’re 100% scarred because of your debacle with mike Z about this topic. So now you take any stance possible against GGPO to spite him or ponder. Its pretty clear that you don’t actually understand how GGPO works and why its so fucking good online. How many fighters have actually been online? its well over 20 @ this point if you count bad games and NONE of them are even decent unless they copied GGPO technology or used GGPO itself.

so of all the 20 or so fighters to be online, only the GGPO ones allow tourney strategies to work online. hmmmm, is that just a coincidence or it could it actually be good?

nah, its not good because that means you would be WRONG and you’re never wrong because you don’t travel or play in tournaments.

Wasn’t the SFxT Sound glitch just the music gouing out? So anyone bitching about sound cues need to be quiet.

Not sure if you realize this but that’s how every game works or supposed to work, if it was programmed well. The only thing GGPO gives you is control. If what you said is true then there wouldn’t be so much lag. But obviously there is a lot more information being passed around. Simple buttons inputs are the least of a programmers worries for online. GGPO barely gives you an explanation of how it works. Being peer-to-peer is a stupid mistake. It’s the worst of the possible online solutions. Allows for hacking quite easily. My reference to LAN was more in general gaming than anything.

Why is p2p bad other than easier to hack?
Funny thing is not many here know how to use a debugger.

http://mauve.mizuumi.net/2012/07/05/understanding-fighting-game-networking/

It wasn’t just music:

[media=youtube]rasbm11e3d4[/media]

Peer-to-peer was chosen because handling 1 connection (between both peers) and syncing 2 copies of the game was deemed more efficient than having to handle 2 connections (client to server to client) and syncing 3 copies of the game. Even if you do a simplified setup where one player is the server and the other the client, this would mean that the player on the server always has an advantage due to the game always syncing to their game.

Also, IIRC most rollback netcode implementations simply create a simulation of what should be happening were the game offline so the only thing that needs to be passed around are inputs and timestamps for those inputs.

EDIT: Or just read the link jed posted.
http://mauve.mizuumi.net/2012/07/05/understanding-fighting-game-networking/

Not really. GGPO gives a really clear explanation of how it works. It gives a lot more than control. And P2P is the best choice for fighting game net code for a variety of reasons.

How would you explain every release of SNK’s having an unplayable netcode? 98, 02, Garou, etc are all unplayable online & I know they aren’t taxing the system more than SF4.

And with TTT2’s netcode no game has an excuse.

They are cheap ports and SNK didn’t devote resources to network engineering. They (probably rightfully) assumed that network performance wouldn’t impact sales. Those who were interested in the games would buy them, regardless of performance (I did).

Even if they did devote enough resources, we don’t know if licensing GGPO would be cost effective for that project. There might be issues with their emulation… who knows. Without being on the project, it’s hard to tell what sort of challenges that had. For all I know, it would’ve been fairly easy, but the resources just weren’t there.

Namco is a huge company with enormous resources. They also did some awesome tricks with TTT2 online play.

Still though, not every project is the same, and it’s never easy. I agree TTT2 has a fantastic online experience, but to expect every other game to meet or exceed that is unrealistic.

Rollback is all well and good but honestly the only way fighting games will be great online is if the US gets better internet infrastructure.

I live in Hong Kong and have a 100Mbps fiber optic cable coming out of my wall. It’s the cheapest internet plan here. I pay like, 20 USD a month for it. All my games (even KOF XIII!) work great online.

Most places in the US are using shitty comcast 10Mbps cable internet. It’s sad.

It’s a consequence of bad regulations that limit competition and infrastructure investment.

Also, Hong Kong is a single city, but the US is pretty vast and not all cities or regions have a population density that might let companies make money on higher speed internet.

Oh I’m well aware of the problems. I’m just saying that whether you use rollback or other technologies isn’t going to help much for a lot of people living in the US with the way things are now. =/

How much could it possibly cost to license GGPO? I don’t get the impression that Ponder is in it for the money. If he was in it for that, then he wouldn’t have a free client up for 4-5 years now. I wouldn’t expect him to charge nothing either, but let’s be realistic. I’d guess about $10-20 Grand.

The free client can be used as a demo of his product. He’s not going to make much money charging some dudes a monthly fee to play on his client, especially since GGPO has largely maintained a majority of older games with smaller communities. He can use it to showcase it to developers, who are willing to pay him a good chunk of money in order to license their game as having GGPO online functionality.

And it’s not like he’s maintained the GGPO much beyond just letting it run on without him and doing occasional changes to the servers. He’s not getting paid to run GGPO, therefore why should he work to update the client?

So you’re saying SNK hasn’t realized how important online play is? Even after all the complaints with every title they release about the same thing? That still begs the question, if they know they don’t have the resources to create a good online experience why waste & resources on online play to begin with?

I don’t expect expect every game to meet TTT2’s netcode but there’s no excuse for me to be able to play people in different countries on that game I can’t even play people in other cities in UMvC3.

leads one to think that there must be some other reasons they don’t use it. A team of network engineers costs waaaaaaaaaaaaaaaay more.

Misunderstood what I said. I know he cannot make GGPO pay to play, Capcom/SNKP would be on his ass. I thought that was pretty obvious. The point I was trying to make but I don’t think I made it clear enough. Is that if he was in it for the money -I.E. charging god knows how much for GGPO- he would not continue a free client to this day that costs at least a few hundred every now and then. So therefore, the cost of a COMPANY obtaining GGPO would be relatively low if he his willing to sacrifice a few hundred dollars and personal time to maintain a free service.

The proof is in the pudding. If they thought that investing $300k in network engineers or licensing would make them $500k more in sales, they would do it. It really is that simple.

Namco has more money to spend and engineer a new back-end which is tolerant of lag. Capcom is in hot water, so they reuse older tech. I agree that MVC3’s online experience is terrible, but the game wasn’t exactly AAA, so eh.