Your first paragraph is inaccurate. You are trying to critique quantum theory after watching a science fiction movie about time travel. Advanced emulators like mame32k .108+ overclock framerate to 2-4x when sending savestates etc. So they achive rates around 120-240+ fps. And their games run at normal rates. So if your connection is on excelent, only half of their savestates get transfered? Dont take my word for it if you think I am talking shit. Find it out yourself.
And how can connection setting relate to accuracy? Your MAME does not know or care about what connection setting your client is running. Like I said, dont take my word for it…find it out yourself if you think im talking shit. It runs constantly at 60fps all the time and does not even know what ping you have. It can’t know what ping or connection setting you have. Even if MAME somehow hacked your kaillera client and did know, there would have been no point. Dont take my word for it if you think I am talking shit. Find it out yourself. But don’t talk shit!
Connection setting does not mean anything different to the the emulator and its gameplay. It has nothing to do with your ping. If it did, that part would have been left out or would never have been introduced. You dont get different pings based on your connection setting. To the server and client, it refers to how many new frames to enclose per packet so it can do with sending less no of packets. To the user, independently with ping, it merely refers to delay. Nothing more and nothing less.
Sorry for not being active in this since yesterday…
From my understanding, the primary issue with 3d game emulation being slow is that Mame puts all emulated processing on the CPU. If you’ve ever seen a CPU try 3d emulation, it is not pretty, nor is it exceptionally fast, in terms of FPS. The MAME guys have acknowledged this issue, and responded that trying to make MAME offput video processing to video cars would take a much greater amount of time and energy than any of them are willing to commit. This has been a major deterrent of unofficial mame build hackers, too, I beleive. Can’t say I blame them on that. I believe there was also the issue of not being able to emulate accurately, if offputting to 3d cards, but rather tweaking the handling of the graphic processing…and that goes against the official MAME guys’ credo and moral validation for keeping the project alive. But, there are emus, such as Zinc, which do handle 3d processing differently than Mame, and games like SFEX and other 3d’ers run great on it…but, Zinc at least, does not integrate online play, and its emulation is limited. And, of course, there’s Chankast, but the guys that do that have already said they will not implement online play. Of course, you also have to consider that the greater the graphic quality, the greater the bandwidth needed to play online.
I think there are plenty of things in the pipe, in terms of technology, that will one day allow us to play the mentioned games basically in real time, but for better or worse, it isn’t happening right now.
Yo dumbass, nobody said your ping changes depending on the connection type. You can use all the numbers you want, but common sense beats everything else. Simply put, it’s much easier to get moves out and react on excellent than it is on good. You are, of course, too ignorant to understand what the numbers mean, and you are apparently too weak of a player to notice the difference. Again, stupid people like you don’t understand simple things. Just misinterpreting random data doesn’t mean you’re right, it just means you are wrong and alse deluded.
EDIT- If you believe there is no difference, then why argue and complain? Just switch to excellent and don’t bitch. Simple.
I’m just wondering 0746, how are you currently involved in the creation of kaillera? Were you involved in it’s programming before, or any current open source projects, etc?
According to him, can’t be, since he says it’s all the same (and then goes on to post what you pointed out). I think he’s just a troll. How can anyone think those two connection types are the same?
Lol you’re crazy man. I’ve had people beat me online and they couldnt really score any wins on me offline. It’s a completely different game dude. Oh and xbox live runs way better than kaillera imo. Actually it should really be fact. In terms of input delay live blows kaillera away.
So pretty much, Hitman and 0746, i dont even have to say much to either one of you anymore. You both really have no one that agrees with you. haha, telling people they need to “L2P”. If ya’ll both actually "L2P"ed yourselves then you’d know exactly where and how the lag and delay in Good and 64 effect things for worse.
I wouldnt even be surprised if both Hitman and 0746 were the same user posting under both those names.
Having all that delay and whatnot just has you learning how play guessing games. AND it shows you how to abuse things that dont fly, cause you dont really get punished as hard as you should be.
EDIT: people that REALLY know how to combo big in the lag, tear me up too cause i cant defend the obvious start ups to it. It’s always to their benefit to just keep at it.
You say I am too ignorent to understand while you dont understand it at all. I am not misinterpreting randomly.
I know I am not and I know you are because there was a time I thaught its like that. It was a good enough understanding at abasic level but most of the things i coded based on this assumption did not work. Does not make sense for emulators to send less data because of connection setting…it would mean you would run the game at half the speed. Thanks to Moosehead clearing things up for me and I am doing exactly the same thing. Someone said it in this thread or a previous one that you do reaction based gaming. The reason its easier to do moves or I should say it feels easier to do moves on excelent than good is because there is less delay and hence is more responsive. Not because it goes at 30 fps instead of 20. There are still people who will do long infinites and a whole lot of other shit on good connection setting. Well, they just vomet them and dont care as long as the first hit got through. Their shit will work equally good if they play on bad. They are just used to the delay.
Not the same thing. All games in MAME always run at 60 frames per second, both offline and online all the time. Let me tell you that he does not understand what connection setting means. Your games would run fine on 120 fps connection if thats how the emulator worked without any loosing any integrity irrespective whatever connection setting he picked.
If instead of saying frame rate, if I said input frames or something like that, would that change anything? I don’t mean the rate at which the frames are displayed on screen or anything like that. I just mean how inputs are recognized or accepted. I would not have reacted the way I did if you hadn’t attacked me, but now that you seem to actually want to discuss things I’m cool with that.
I’m not used to the delay at ALL.
I refuse to compensate cause that hurts me when im offline cause then i’d have to change again.
Still i think that the older mame versions are better cause they wouldnt slow down at all. I used to play in private servers on LAN with mame .61 for capcom and may 3 mame .60 for snk.
It was great then.
Delay hurts you, i dont care what anyone that only plays on kal says. When you get that last sec reversal when youre watching your opponents every move, and you spot them change up. You get that reversal and you turn the MOMENTUM, and can clearly see the pace of matches you’ll know what im saying.
Kal feels too much like a random guessing fight in the filth.
Aye. One of the worst things about lag is that it can totally kill someone’s momentum.
Like sometimes I’ll be watching someone make a huge comback against his opponent, and then it’ll lag and no one’ll get hit, but you know that guy’s probably gonna lose now, because his momentum/rhythm is gone.
Sometimes, all your opponent needs is that extra half a second to compose himself.
But you said Kaillera is better than Zbattle, when in reality it’s the other way around. You make it SOUND like you know what you’re talking about, but in actuality you really don’t.
I’ll try to give you the scenario. Consider 2 players on a server with 0 ping. One is on LAN and the other is on bad. When the game starts, both of the emus load up the game and reads keyvalue from your keyboard and controllers and gives it to kaillera client. When client receives the first keyframe from your emulator, it knows your game is all loaded up. It sends a singnal to server saying I’m loaded up. When everyone is loaded up, server sends another signal to all players at once saying your game is loaded up. Since both of them have 0 ping, both of them recieve that signal almost immidiately. Thats when the game begins. The client that is on LAN connection setting sends the data it has immidiately after it receives the ready signal. Why? Because It assumes that mame goes at 60fps and sending the frame now wont alter the no of packets sent. It sends it immidiately and gives execution control back to the emulator. But for the guy who is on bad, he is only supposed to transfer 10 packets per frame. If he sends his frame immidiately like the guy on LAN, he will end up sending 60 packets per second as MAME runs at 60 frames per second and thats not what the user wants. So it does not send the frame and just gives execusion control back to MAME.
On the second frame, MAME on both player’s end does the same thing in reading the person’s keyboard and controller button states and gives it to the clients. Client of the person on LAN does the same thing as before and sends the second keyframe to server because it assumes MAME runs at 60 frames per second and it wants to send 60 packets and there is no problem sending data each frame. But it has 0ms ping. The game is running at 60 frames per second and each frame takes 1000ms per second / 60 frames per second = 16.6666667 ms per frame. I.e. mame gives frames to client at that interval. It has been more than 0 ms (the client’s ping) since it sent his keyframes to server last time. So it expects server to have sent a response to his previously sent keyframes. And thats when it checks if there is any data and if there is none, it waits for the data until it receives it. Once it receives the data it gives it to MAME and thats the maximum delay for that player who is on LAN connection setting with 0 ping. As long as he pings less than 16.6666ms, that will be his maximum delay. He is on delay tier 1 for LAN connection setting. If he pings 20 ms, he is on delay tier 2 and client no longer expects the frame sent now to be available on the next frame. It expects them another frame later. And it will not wait on the second frame for the data sent on first frame. Because if it does, it will take more than 16.666667ms + MAME has to do a lot of processing too and hence the next frame will be sent later than in another 16.6667ms and it will create discontinuity…When it becomes discontinuous, the effect will echo on and create huge choppy laggy and unplayable start.
As for the guy on bad, he cant send data on second frame because if he sends data on every second frame, he will end up sending 30 packets per second. User only wants 10 packets transferred per second. So client just saves the frame on its local memory and gives execusion control back to mame. This goes on till the 6th frame. If it sends data every 6th frame, it will end up sending 10 packets per second and thats exactly what the user wants. And thats exactly what the client does and sends all 6 frames at once. So you have that obvious 6 frames delay at the start. But then when it gets to the server on the next frame, server will wait for 6 frames until it has enough data from player 1 to send to it. So the client who is on bad cant expect frames from server before another 6 frames. So he gets the first frame he gave to client back on the 12th frame. Thats 0 ping optimim condition. But what if the guy who is on bad has a ping of 60 ms? It will take his data 30 ms to get to server (thats what kaillera assumes) and then that 6 frames delay so the server has enough frames to send to him and then another 30 ms for data to travel back to him. so he has that intrinsic 180ish ms delay just for choosing bad with 60 ping. This system even works with dialup regardless of connection setting because users are discriminated based on their ping.
Now why did I say Emulinker does not work? Because server does all this calculation and tells client what delay tier they belong to. Client simply follows orders. And emulinker has that delay tier value hardcoded in it. It always says you are on the second delay tier (i was wrong before saying it was on first but ive confirmed with emulinker coders some time after that.) So the person on bad with 60 ms, instead of forcibly waiting for data on the 12th frame, does it on 18th frame (it only does the waiting stuff when its time to send). That means the game starts out real smooth with that extra bit of delay.
This also mean if you belong to a delay tier greater than 2 and you are using lan, for example if you ping more than 2 * 16.6666666667ms = bit more than 33ms, it will forcibly stall on the second frame to wait for data from server and will send the next frame later and server will get the next frame later and send it to you even later and this will continue on and the game will fuck up with emulinker.
Suprafast made a nice little mod for emulinker to set this delay tier value which would allow you to set the delay to less or more if you have less or more ping. But the last time i tried, it seemed to be game specific. That doesnot work either. Setting the delay to more just means you will lagg more and to less means you will chop to death. It just means you will be able play with hisher ping with more ping. Even if it was made userspecific, it would not work. Because emulinker’s internal architecture discriminates between black or white and not based on ping. It requires that everyone must play on the same connection setting regardless of their ping. And regardless of their ping, server does not send responses to clients before it has keyframe values from all the players. So your ping is 13 ms and you are playing with this 70 ms dude who is trying lan out, server wont have keyvalues before 30 ms and you are stalling on second frame not sending any value and it just gets worse and worse like a chain reaction. Thats not how its meant to me and of course…I dont disagree that its unplayable…on emulinker servers. But then it works worse than kaillera server in this respect and you guys have been playing on emulinker server since the last lot of script kiddies. My argument is, it does not make sense for it to be generalised for all of kaillera. Thats not how kaillera is meant to work and thats not how it works. You don’t say all games suck because you came across one shit one. And you should not argue all games suck with your experiance of playing only one game. Its mo different from those people who tries their first fighter game and gets smashed heavily and never plays it again and says stuff like “fighters are ***.”
Kyokuji> I’ve been to GW less than 5 times in total (counting occations rather than no of logins) and never once did I speak about kaillera other than asking how to use the new setdelay feature works once.
Blame your losses on online tactics is just weak. First time I played against Interceptor he crossed me up roundhouse standing fierce jab dragon punch 100% of the times, I have no idea why I couldn’t block such a simple combo, but he landed it like 100 times in a row. Whenever I tried to do it on him he blocked it 100% of the times.
Our connection wasn’t ideal, Super Turbo Speed 3 online is too fast IMO, I couldn’t get jab dragon punch out reliably and I had to resort to crouch fierce’s, yet he had no problem whatsoever jab dp me out of pretty much everything I threw at him.
Sometimes it is called ADMITTING YOU ARE PLAYING AGAINST A SUPERIOR PLAYER who is MORE ADAPTABLE than you are. Is online a different game than offline? Of course! But I will guarantee Interceptor would beat the vast majority of you offline as well. Unless you are a V-ism player (why don’t you just play A-ism or stick to simplest VCs and STFU about lag?) or something you don’t magically beat someone going from online to offline.
Super turbo on speed 3 is very very very very fast, almost too fast. Even the tourney players will rarely ever wake up DP against jump ins. I guarantee you offline play for majority of the scrubs like me will favor jump-ins as well. Ooooo and the godly “mix-ups” aka tick throws, it is again possible because of the retarded game speed, I have seen countless tourney players trapped in a corner and ticked throw to death just like a Kaillera match. I would say ST is THE most heavily luck-based tourney game out there. Just watch Jwong in Evo East finals, second round, three tick throws from Balrog. Or heck mighty Alex Valle ticked throwed 4 times in same round. Seriously the Valle match looks just like a Kaillera match.
For the last time, for a dead genre, it is almost impossible to find consistent offline competition. A lot of us have gave up on offline play COMPLETELY, there is ONLY online play, thank you very much.
Any given day there is far more people playing on XBL and Kaillera than offline. Online is the future, offline is only if you are serious enough to play in tournaments and don’t want to lose your timing or something. That is like less than 0.5% of the population that plays fighters at least once a month.
Uh, actually Ryu’s about the same as he is buddy, and I play Intercepter quite frequently and we go back and forth most of the time. Like I said before, guys like Jinrai and DGV are much better.
What are you, his agent or something? Nobody was complaining about anything. Just making a statement about his playing style. It wasn’t out of malice or anything. Just a statement of fact.
Maybe in your area and mine, but not in a lot of other places.
And why the hell did you suddenly switch gears and start talking about Alpha 3 mid-paragraph? Intercepter doesn’t even really play that game.
You’re pulling a lot of figures out of your ass, and making a lot of poorly reasoned blanket statements.
0746, I don’t care how good your netcode is. Fighters are not playable at 150ms period. Maybe for basic stuff, but there’s no way you’re going to be able to do VC’s or links or anything like that at that kind of ping.
Lol…HOW is it weak to blame losses on online tactics when it is usually the online tactic that results in the loss? I don’t understand your mentality at all. People are complaining about things that CLEARLY exist and clearly affect the outcome of the match to the point where the lesser skilled players have their gameplay improved due to lag. Noobs online can toss out random shit for no reason and partially get away with it. This is the reality.
Look, I played you in kof2k2 and you just mashed. You didn’t seem to even know the importance of reaction gameplay…how can you make comments related to reaction gameplay if you aren’t a reaction player?
If someone says “I know I hit the attack button before I saw the opponent’s animation make contact with my character, and the move(or any other move) I was trying to do never came”, then who are you to say they are making up excuses? How would you know? How can you say I’m making excuses when I know that DOES happen to me.
Why would you admit something like that if it wasn’t true? First off, we aren’t even talking about just people that end up beating others online and not offline. We are talking about tactics that just end up working better online and offline. These tactics are mainlyl used by beginneres who don’t even know what they are doing. It’s not that they are adapting, it’s that they don’t even know that what they are doing would get annihilated in the reality.
Why not…explain yourself.
Where did you come up with these statistics? How are you going to know how many people are walking into arcades around the US and Canada? Give me the source for your statistics…
I find it laughable that someone thinks online players outnumber offline players. Just because you don’t know offline players doesn’t mean they don’t exist. In arcades, college campuses, apartments and houses across the country people will meet up to play games. Most of these people do not enter tournaments. The online player base is extremely small. If Capcom and these other companies depended on online players, they’d have gone out of business a long time ago. There are maybe 300 people on kaillera that play fighting games, and that’s a very generous estimate. I don’t know how many people are on Xbox Live, but for online players to outnumber offline players, the player base would have to be over 50% of the game sales. And if that were true, the developers would have made virtually no money from the sale of the game, as they’d only move very limited numbers of the game. As someone that used to go to arcades and gatherings of people, I can attest to the fact that of every 20 or so people at any gathering, maybe 1 or 2 would play online.
It also sounds to me that you just suck period, so I don’t understand why you’ve been chosen as the official talent evaluator. You’re like those scrubs at the arcade that refuse to play, and when you do play, no matter to who, you’ve lost. You base your entire talent judgment based on who beat you more badly, or whoever you see beating people by more. I’ve always thought it stupid when someone says A beat C, and C beat D, so A must be able to beat D. This is not sound logic. Fighting games are about matchups and playstyles, a person that barely beats everyone is a much better player than someone who destroys half the people he plays.
And as Kyo pointed out, why are you switching mid-stream from ST to A3? I don’t know who interceptor is, but it sounds like he’s not actually that good. There’s an easy way to test though, just have the guy step up onto ZBattle, where the lag isn’t nearly as stupid, and see how he does. I don’t think ZBattle is as good as offline play, but it sure as fuck is a hell of a lot closer than kaillera is.
I don’t play KOF (except 94, the only version I am strong in), you can clearly I didn’t know what low jumps are or max is or CD counters are and barely knew the move set for the characters, and I got destroyed, just as I would be in offline play. And it is hard to improve in KOF when everyone I play against pretty much either OCV me or win by 1.5 bar or more. The learning curve is too steep.
Play CE or A2, these are my strongest games. I never claimed to be good, but I assure you in those games if you can’t beat me online, you won’t beat me offline. At this day and age, unless I played a certain fighter a lot when I grew up, I won’t be good at it nor I would spend the neccessary time to get good at it.
None of the 2D fighters played at Evo still sell today. And most people do not go to those gatherings you talk about. They play online, XBL or Kaillera. Look at the PS2 (a much much more popular system) board for Street Fighter Anniversary:
Zero posts, that is how popular the game is. I am sure PS2 version sold more copies too. But how many people still play it outside of those that read SRK.com?
At least for Xbox there is still interest, due to online play, even if the system has a far smaller installed base than PS2.
DOA4 is easily the most popular fighting game in the U.S. today (everywhere except SRK, just like soccer is the most popular sports everywhere except the US, just like KOF is more popular than SF everywhere else except the US, you can see my point, just because it is not popular in SRK doesn’t mean it is not popular). Used in pro gaming tournaments like WCG, only fighting game to have sponsored players, and the main reason is online play.
BTW, while Kaillera is not nearly as big as XBL (nor it ever will be, the whole emulator/rom thing is still not legit afterall), it is far bigger than you give it credit for. KOF97 is the biggest game on Kaillera by a HUUUUUUUUUUUUGE margin, yet you wouldn’t even see a single game hosted for it if you look at the U.S. servers. The number of players you see on godweapon represent a tiny fraction of the player base for Kaillera. For something completely free, completely home brew, completely fan supported, it is pretty damn good.
because if he sends data on every second frame, he will end up sending 30 packets per second. User only wants 10 packets transferred per second. So client just saves the frame on its local memory and gives execusion control back to mame. This goes on till the 6th frame. If it sends data every 6th frame, it will end up sending 10 packets per second and thats exactly what the user wants. And thats exactly what the client does and sends all 6 frames at once. So you have that obvious 6 frames delay at the start. But then when it gets to the server on the next frame, server will wait for 6 frames until it has enough data from player 1 to send to it. So the client who is on bad cant expect frames from server before another 6 frames. So he gets the first frame he gave to client back on the 12th frame. Thats 0 ping optimim condition. But what if the guy who is on bad has a ping of 60 ms? It will take his data 30 ms to get to server (thats what kaillera assumes) and then that 6 frames delay so the server has enough frames to send to him and then another 30 ms for data to travel back to him. so he has that intrinsic 180ish ms delay just for choosing bad with 60 ping. This system even works with dialup regardless of connection setting because users are discriminated based on their ping.