Online 3D Fighter

Discussion in 'General' started by martialfanatic, Dec 27, 2003.

  1. KiwE

    KiwE Well-Known Member

    [ QUOTE ]
    If someone tries a low swipe to beat your block, it's slow enough (Edit: That's the game's choice, this is no proof of anything generalized for FPS as it's not even one of the most popular FPS's, it would be like someone saying reversals are to easy in fightinggames after playing DOA) that you can jump or evade because you see it coming. Nothing requires split second timing (Edit: Try another game). Low ping still gives advantage(Edit: Yes it does, but you can still get your ass handed to you in one of those games that require "split second timing" by someone with worse ping then you due to him being better - even with a instant hit weapon). Melee mods allow for punching and kicking along with use of sabres. Hit detection is probably generous(Edit: Huh? That would be cause of everybody having the same hitbox to not give some special skins advantage, you hit the hitbox you get a hit). Fighting? Yes. VF style fighting? No.

    [/ QUOTE ]


    [ QUOTE ]
    FPS are designed with lag in mind.

    [/ QUOTE ]

    Please explain how the game would differ in experience / design if it wasn't intended for multiplay / designed "with lag in mind" (as I felt it was a very generic and vague).

    [ QUOTE ]
    I don't think CVS2 re-written from the ground up with networking code present from the beginning would come out any different than what you have now.

    [/ QUOTE ]
    I do. I think you're totally stuck on your "speed of light" connection and think that's all that matters. Different games do have differently well working network codes, honestly. VF2 works online and that's a <font color="yellow">PEER TO PEER</font> game with no server interpolating OR synching off the data (and old and not optimized). How odd right?
    The truth to be told, imo, is that the reason why fightinggames hasn't made it online as of yet (or have been tried) is cause of fightinggames never beeing popular on the PC which is the natural home for games played over a network - not much else. And online they will go and I'll be one of the first to great them with open arms and a cake.

    Over and Out (imho on all of this of cource etc). /versus/images/graemlins/grin.gif
     
  2. KTallguy

    KTallguy Well-Known Member

    FPS shooters are designed with lag in mind, and so are RTS games. For example, say there was a weapon in an FTS that was shot at someone, and right as it hits them you can hit the button again and do more damage. However, it has to be right as it hits. That would never work on a server with even a little lag.

    But that right there, is the very definition of a catch throw in VF, which does require split-second timing.

    If you can name one online game that relys on split second timing in an internet environment and does well, then I'll change my mind. But with current technology, playable VF (in my opinion, playable means completely functional) is not a possibility. VF2 being perfectly lag free seems impossible to me, whether it's 'playable' or it 'works' is really up to the players involved, but even a small amount of lag would probably ruin the game for most 'serious' VF players.

    People who think that a little lag is ok as long as they have comp fail to realize that the comp they get will not help them get better, and it probably won't be very fun when they are lagging and killed because their catch throw doesn't connect, or because their triple throw escape doesn't work. I'm just saying that playable online VF is pretty far off, and most good players will not take it seriously if they begin losing because of lag, and then it will degenerate into what I described in my first post.

    No one is getting high off of their 'speed of light' connection, but face it, online VF is a good 5 to 10 years away. Hopefully by then we can move from joysticks and controllers and start controlling games through comprable VR.

    PS: I take that comment about swapping vidoes back, that wouldn't be a bad thing =) Spectating a fight would be nice too, but I think swapping replays is easier, as a VF fight isn't really a longer game like an RTS, unless you just wanna sit in on a series of fights.
     
  3. Pai_Garu

    Pai_Garu Well-Known Member

    I've never said online gaming would be a bad idea. In fact, if there was a way to bring the arcade experience home through any sort of online feature, I would pay a premium to use it. Case in point here though, is that, so far, it is not possible to do so with VF. Simply for the lag factor. I do understand that in a country such as the US, online gameplay is direly needed to introduce competition. Just because players in the US need a feature like this, doesn't mean anything for developers in Asia to consider this as necessary or the next logical step in gaming. Face it, in Japan, Taiwan, Korea, HK, etc. the necessity for an online fighting game is simply non existent. They don't suffer from a lack of competition in game centers, and people in each of these countries regularly and can easily travel to any other country to compete or just to play. You can read about Ice-9's travel log to HK in one of the other threads about players in HK regularly travel to japan to play and to register cards for their friends etc.. I can also confirm this fact based upon the time that I've lived in Taiwan when I was younger. You may even be able to find old VF2 competition videos where people from all over the world traveling ot Japan for the world VF2 championship. For Asian countries, they can get exposed to the best in the world without an inferior online feature. Regardless of the lag problem, a serious online fighter just doesn't seem to be upon the horizon anytime soon. It's just conceptually unnecessary and technically nonsensical.

    In the end though, it would be nice to have this feature as I am currently living in a place without a healthy dose of competition. Regardless of what it may seem like, countries like Korea and Japan are mostly concerned about their home market. They take the game business seriously there, and players there are gifted with like minded people to play against. The only thing you can blame is having to live in a place where pc competition games are valued and because of that, the attention is simply not placed upon your hobby. It'd still be kind of fun to have it just for kicks, but as other people have stated, it'd be a fun thing to have, but will not be taken seriously.

    To KiWe:
    You obviously don't understand how fps compensates for lag. Here's an example, your machine goes out of sync with everyone else's. You continue to play normally even though a lag indicator comes on. You shoot someone and they get hit, then two or three seconds later the game resyncs, the guy you killed didn't die, but you suddenly die right away as the game resyncs. Do you know what happened there? This is partly the reason why instant hit weapons still work in a laggy environment. The sync'ed game continues to play ignoring lagged machines. It simply continues the last command done, which is why when you hit a moving target even when it's somewhat lagged, the hit will still register, even though sometimes your movement is ignored. The sync'ed game interpolates the shot as a hit. This same engine would definitely detract from playing VF imo, because inputs will be screwed up, and things will be delayed more that it actually is. In situations where 1 frame is crucial to decide what happens, such as a -8 throw guaranteed move, even 1 frame of delay can immensely change the outcome of the exchange. The cable example is simply used to illustrate the large scale effect of global internet connectivity. The lag in pressing a button on a joystick connected to your ps2 still exists, but so little to the point it can be neglected. Most people, however, would still agree that a good stick that registers inputs well is better than a stick that tends to "lag" ie bad buttons, or bad spring, etc...
     
  4. KiwE

    KiwE Well-Known Member

    [ QUOTE ]
    FPS shooters are designed with lag in mind, and so are RTS games. For example, say there was a weapon in an FTS that was shot at someone, and right as it hits them you can hit the button again and do more damage. However, it has to be right as it hits. That would never work on a server with even a little lag.

    [/ QUOTE ]

    You bringing up there not being a presence of a hitthrow weapon as an example of how FPS's are designed with lag in mind when there are instant hit weapons is truelly amazing. Even in vf, many people just react upon the animation of your own character in these situations (as Akiras palm>elbow), audioclues or hell - even buffer them anyways as you loose nothing in buffering them every time. Is that the definition of making a desicion with splitsecond timing?

    [ QUOTE ]
    If you can name one online game that relys on split second timing in an internet environment and does well,

    [/ QUOTE ]

    Every fps game ever made since quake1.

    [ QUOTE ]
    Here's an example, your machine goes out of sync with everyone else's. You continue to play normally even though a lag indicator comes on. You shoot someone and they get hit, then two or three seconds later the game resyncs, the guy you killed didn't die, but you suddenly die right away as the game resyncs.

    [/ QUOTE ]

    WHAT?

    Please tell me of the game this has happened to you Srider or stop talking out of your ass.

    Imo, you don't know what you're talking about (no offence). Here's are some examples of how good FPS's compensate for lag:

    <font color="yellow">Timenudging data</font>: You can push yourself before (or after) the servertime. This is probably the highest factor why a person with high ping (or semi high) can use an instant hit weapon and dominate someone using the same with almost no lag on the server. You can have tell the server to interpet all your shots as coming -20 for example to compensate for your poor network traffic. In a vf enviroment (and I'm just listing examples, not solutions) you aren't thrown into a situation in which you have to react asap. You gaurd a big move to get that TC situation and while you're gaurding it (until you get control back of your character), you have options (and buffering something is one of them). The down side to nudging data is that the opponents movements can feel choppy (but that can be worked around, locally, thru smoothing out the client).
    <font color="yellow">Snaps/Worlddata</font>: In current FPS games you have commands like this to customize for your connection (unless the network code is shite) snaps is simply put how many refreshes of the world you get from the server (your world will not lag as it's drawn locally). If you have many failed packages you can turn this down.
    <font color="yellow">Maxpackets</font>: The maximum amount of packages of network traffic you recieve from your server.
    You can manipulate data in many many ways - these are just some examples to contrast against Sriders lightningspeed theory, and at the end of the day the amount of data that would have to be sent (I'm talking about the size of the packages describing what is going on) of a fightinggame is so much much smaller then that of a fps. Yes, vf is a extremly nice game, but stop kidding yourselfs - it's not impossible online due to it's delicate / conniseur nature. Here's an interesting math puzzle for you to think about:

    Ping is measured in ms. One frame is 1/60th of a second. A person with broadband pings 10 (that is, data back and forth) against a server. What will the problem be?

    A) Speed of light theorem (made by Srider and the others in cojunction with Einstein).
    B) Packetloss.
    C) Spikes in the network.


    Now, I'm not saying another word until all the weird nay sayers fucking answer why vf2 works online as of now as it keeps getting ignored.

    P.S; Even if VF had a long line at the arcade down here, I wouldn't want to blow 1000$ on getting a high rank at it, afraid of trying out a new character, playing basicly the same "good competition" over and over if I had an online option. But none of that matters as I'm not living in the East, just like most of the people who come here.

    Yours Sincerly - KiwE (FPS fanboy / Blind with hope).
     
  5. Pai_Garu

    Pai_Garu Well-Known Member

    KiWe, all those compensation options you listed clearly shows that the "actual" game played in the online environment is not what is actually going on. Against players in a laggy environment, the actual sync'ed game is not always truly represented on your machine. Note I say not always, because in many cases when variable are favorable, things do go along nicely in sync. If you have never experienced the things I've listed in my previous post, it's obvious to see that you have not played enough FPS to talk about this issue. Regardless of what you may think, that's just the way the internet protocol in games behave. Get any directx game networking book and you'll see what I mean. Regarding online vf2... I don't know how it plays like, but my assumption is that no one will be able to do movements similar to minami stepping etc when playing online in a non favorable condition. Have you tried this yourself KiWe? When you say ping of 10, that is a highly favorable condition, and you know very well that happens very rarely. Usually you get that from a LAN environment, and I have said that of course it's doable in that environment, but then what's the difference of playing LAN and just regular vs mode or in the arcades? When you talk about playing the game globally, hell, even international phone calls have a big lag sometimes. You can hear your own voice after you say it in the other person's side.

    On the issue of data being sent, no one said you have to send over a massive amount of data, the only thing is the time it take for the data to travel that's the key. Again, any simple online game programming book will tell you what's actually sent over the internet when you play a game.

    If you really want to get into the technical details, I claim that I am not aware of all the most advanced programming methods or network protocols out today. The obvious fact though, is that all the online console games out right now, and even online PC games, lags in a global or national connection level. Simply from this fact, you should be able to understand why many people believe the way they do.

    Anyways... All the things I've said are simply facts to explain why all online game lags when things don't go smoothly. In the end it just comes down to what you consider to be playable. If you don't mind things going haywire every once in a while, then I have nothing to say. The bottom line is that vs mode as it is now, can't be reproduced in a global online environment.
     
  6. KiwE

    KiwE Well-Known Member

    Srider; do you understand how insane this statement sounds if you have basic knowledge of internet gaming man? :

    [ QUOTE ]
    Here's an example, your machine goes out of sync with everyone else's. You continue to play normally even though a lag indicator comes on. You shoot someone and they get hit, then two or three seconds later the game resyncs, the guy you killed didn't die, but you suddenly die right away as the game resyncs.

    [/ QUOTE ]

    [ QUOTE ]
    If you have never experienced the things I've listed in my previous post, it's obvious to see that you have not played enough FPS to talk about this issue.

    [/ QUOTE ]

    Still waiting for what game that happened to you - DO TELL. I want to know of what game that resynchs client/server every two or 3 seconds / where you can have a confirmed kill on your screen and then learn you didn't get it etc. Please share this info, like really please with sugar on top.

    [ QUOTE ]
    Regarding online vf2... I don't know how it plays like, but my assumption is that no one will be able to do movements similar to minami stepping etc when playing online in a non favorable condition

    [/ QUOTE ]

    Well, duh. Minamistepping isn't something you could even do in VF2 (step in 3d). Assumption is the mother of all fuckups. Nice one man. Frame dependent moves are still framedependent moves though and constant movement / updates aren't something new to online play.

    [ QUOTE ]
    Get any directx game networking book and you'll see what I mean.... Again, any simple online game programming book will tell you what's actually sent over the internet when you play a game.

    [/ QUOTE ]

    Which book that you have are you reffering to (I just might have it or have access to it) and what has directX got to do with tcp/ip traffic? Maybe you should read it some more since you say " I claim that I am not aware of all the most advanced programming methods or network protocols out today." before recommending it to others though.

    [ QUOTE ]
    When you talk about playing the game globally, hell, even international phone calls have a big lag sometimes. You can hear your own voice after you say it in the other person's side.

    [/ QUOTE ]

    You really don't understand the difference in network traffic thru tcp/ip on a broadband connection and using your phone to send uncompressed sound thru a system that has been around for decades with little change do you? Hell, phoneline limits (such as if a node / line goes your conversation goes) are why ip networks were created in the firstplace. And no, having around 10 ping to a server on broadband connection isn't unheard of. You normally have around 40 with a adsl modem (70 with an isdn but none of that matters - latency does). It depends on how close / good the server is you're at. Of cource you cannot expect to connect to a server in asia and experience a lagfree game but I, living in Sweden, can play on a Danish server for example.

    'Lag' can be caused by many things: incorrect Window settings, incorrect modem settings, crummy Firewall or Proxy settings, loose physical wire connections (causing packets to be sent oncemore), software programs loading your CPU while you are also playing games, software programs also using bandwidth over the internet during games, <font color="yellow"> speed of light </font> *smiles*, not enough RAM in your computer, older/slower computer's CPU speed (game not optimized for it), your modem/modem settings, IRQ conflicts, etc, etc. and trying to synch to computers who are not alike is one reson for 'lag'. None of these problems would arise with two, exactly alike, systems.

    VF online, that is, a homeconsole version of VF5 (probably) is about 3 years forward in the future atleast. 3 years ago a 56k connection was standard to many. Think ahead ffs man. Peace out Srider, I love reading your posts (you've made great ones on vfdc and you're always objective) but the 2-3second resynch thing and etc just sounded of my bs alarm so I started posting (the picture was an ironic humor thing - not a Shang I'm a dick thing) - you perhaps shouldn't post strong statements as VF online requires speed of light etc though if you're (imo) clueless to network traffic.
    This isn't a optimistic vs pessimistic (or as I feel, nostalgic "please don't change vf from it's arcade culture!") point of view. Fightinggames are online today and VF is going online - just stick it out.

    "An imperfect plan implemented immediately and violently will always succeed better than a perfect plan." George S. Patton
     
  7. Pai_Garu

    Pai_Garu Well-Known Member

    I don't feel like wasting the time to pick through the post and replay to everything you've said, since I think by now it's clear that you have your own ideas on how the world works and refuses to look at things in anyother way. I'll just touch on some of the things that I feel needs to be cleared up. You don't seem to understand how my examples relate to what I'm saying. You are only picking on the surface and that only make it as if you can't understand the ideas I'm trying to convey. Anyways, how is that what I've said about lag in fps insane? In every fps i've ever played, when things go out of sync, you can sit there and shoot at other players who are just standing there, and they are not really there. Meanwhile, you could have been killed and you don't realize it until your game resyncs/stops lagging. I am sure you know what I am talking about. Quake 3, CS, Raven Shield, etc.

    Directx has everything to do with online gaming in the majority of the games out today. Do you know why most PC games ask you to install directx? Because the game uses that layer to interface with the hardware. Now of course there's no way for me to know if there is some customized code to expand upon what's included in directx for tcp/ip gaming, most likely there is. However, I can safely says that if your directx installation is messed up, the internet game function of your game simply won't operate. If you know how to use regsvr32, try removing things like dpnet.dll or dpwsockx.dll and see what happens to your online game.

    VF2 issue, the example of minami stepping is used to simply say that anything requiring complex and quick inputs won't be possible. I don't quite understand what you mean when you say frame dependent will be frame dependent.. It all comes down to latency. You seem to have a good grasp on what causes lag. Here is something though that clearly shows you still don't fully understand what causes games to lag playing over the internet.

    [ QUOTE ]
    ...trying to synch to computers who are not alike is one reson for 'lag'. None of these problems would arise with two, exactly alike, systems.


    [/ QUOTE ]

    So tell me this, if two people bought the exact same model of computer, let say from falcon northwest, that uses the same parts and everything. (I'm using a mass manufactured (falcon) example to reinforce the fact they would be the same like two people buying ps2's.) These two people, one in Japan, one in Germany, decided to play quake 3 together. Both using DSL line. So what you are saying is that if that's the case, there would be no lag?? You know as well as I do that won't be the case.

    Oh, the speed of light idea and how phone call lags, they are used simply to illustrate the point that data lags more as distance between two points increases. They are in no way related to what causes lag, besides the fact that there is a peak point where the speed in which data travels. Regardless, you also assume that I do not understand what tcp/ip is and why it is the standard used today. I'm not going to go into details, but tcp ip are two completely separate things. tcp is slower and more reliable, and ip is the opposite. You also assume that better ping = better tcp/ip performance, which is not the case. Regardless, besides going into technical details, just look at the obvious fact. There are quite a few online games on consoles nowadays using tcp/ip protocol. If things were the way you say it is, how come they all still lag? Even xbox live which requires broadband suffer from the same problem. It's true that things are getting better and better, but how much have things progressed in the last 4-5 years? Regardless of how much things are getting better, games are still not lag free, partly due to the increasing complexity of games. So in 3 or 4 years down the road, maybe there will be a miracle solution, but as things stands now, it doesn't seem likely to change much.

    p.s. If you don't think the complexity of game has to do with online performance even in the smallest way, then you are making a huge mistake.
     
  8. KiwE

    KiwE Well-Known Member

    [ QUOTE ]
    Anyways, how is that what I've said about lag in fps insane?

    [/ QUOTE ]

    Well simply put cause you think that the game resynches with the server every 2-3rd second for starters in your first post (just think about that for a second) and you claimed to have killed someone on your screen (not hit, there's a difference in confirmation) and then they magically weren't dead. You did say those things and they are insane / bs claims.

    You have some weird views on directX btw. Do you think you can't copy files over a network etc without direct X? DirectX are done multimedia libraries (with limited ability of change) that can be used as shortcuts when programming for access to lowlevel functions - they totally sucked compared to OpenGL etc until just recently. Do you think no other machines then PC's with directX use the internet / can communicate over a network? Do you think that PS3 will use direct X for online play (if it has online play included, which it sure as hell should)?

    [ QUOTE ]
    However, I can safely says that if your directx installation is messed up, the internet game function of your game simply won't operate.

    [/ QUOTE ]

    Try playing the old Quake1 in freakin DOS for reference.

    [ QUOTE ]
    These two people, one in Japan, one in Germany, decided to play quake 3 together. Both using DSL line. So what you are saying is that if that's the case, there would be no lag?? You know as well as I do that won't be the case.

    [/ QUOTE ]

    Stop being silly Srider and read my post again. Hardware and software can also induce lag (and my thumb!). There will be less lag / better working software and hardware if everything can be optimzed for one specific machine without variations (I gave 100 examples of lag not just being dependent on the connection speed ffs) - it's you that believe in speed of light - not I.

    [ QUOTE ]
    Oh, the speed of light idea and how phone call lags, they are used simply to illustrate the point that data lags more as distance between two points increases. They are in no way related to what causes lag, besides the fact that there is a peak point where the speed in which data travels.

    [/ QUOTE ]

    No Srider, that was you talking out of your ass / fucking up. Your choice of words / argumentation make that clear. Let's have a look shall we?

    [ QUOTE ]
    When you talk about playing the game globally, hell, even international phone calls have a big lag sometimes.

    [/ QUOTE ]

    (How's that for picking apart a post man?)

    [ QUOTE ]
    You seem to have a good grasp on what causes lag.

    [/ QUOTE ]

    Gee, thanks. /versus/images/graemlins/smile.gif

    [ QUOTE ]
    Regardless, besides going into technical details, just look at the obvious fact.

    [/ QUOTE ]

    No, let's talk technical - cause you're still fronting and think you can say something vague and quickly move to what you feel more familiar / comfortable with. There are counters to evades /versus/images/graemlins/tongue.gif

    [ QUOTE ]
    I'm not going to go into details, but tcp ip are two completely separate things. tcp is slower and more reliable, and ip is the opposite. You also assume that better ping = better tcp/ip performance, which is not the case.


    [/ QUOTE ]

    TCP/IP is a network protocol that is composed of layers, the IP part is responsible for moving datapackages from node to node based on the adress and the TCP part verifies that the delivery of the delivery of that data from client to server is ok (and not for example lost). TCP functions as support to see if there's an errors or data has been lost and then triggers retransmission until the data is correctly and completely received.
    Yes btw, I assume that the ping is some sort of indication of how good a tcp/ip performance is and has a correlation to it - if you don't you should write to all serverbrowser programs and even the creator of mirc etc. Btw what did you mean then, once again, with " but tcp ip are two completely separate things. tcp is slower and more reliable, and ip is the opposite". How are they eachothers opposites etc?

    Btw, dpnet.dll and dpwsockx.dll are apart of direct play (if it's being used) but they are sockets and don't have anything to do with data / speed of data. Sockets are names given to the package of subroutines that provide access to network protocolls on some systems (and fun to attack). Good luck finding them in your basic programming for directX book.

    [ QUOTE ]
    There are quite a few online games on consoles nowadays using tcp/ip protocol. If things were the way you say it is, how come they all still lag? Even xbox live which requires broadband suffer from the same problem. It's true that things are getting better and better, but how much have things progressed in the last 4-5 years?

    [/ QUOTE ]

    Do you consider an adsl modem to be broadband? That's what most users who use Xbox live are using. And as to " but how much have things progressed in the last 4-5 years?" the answer is tremendously. Commands like timenudge and client smoothing didn't even exist until the launch of Quake3 and connection/speedwise it has tenfolded for a normal user. Hell - playing console games online would be laughed at ffs.

    Now fuck this technology pr0n / who dominates who discussion - I promise you, 1/60th of a second isn't something impossible to beat - specially in 3 years time. I hope both of us still play VF then so one of us can say "I told you so".
     
  9. Painty_J

    Painty_J Well-Known Member

    Okay...staying out of the flame-war that's goin on.

    I saw someone point out that catch throws would be nigh-impossible in an online match. I'd like to propose an example of how they could make catch throws work online. Using the method I'm going to outline would work for more than just catch-throws, it would help with every situation that requires lightning-fast inputs.

    Have you played Half-Life online any time within the past 2 years, or so? I'm not trying to be a smartass when I ask, because 2 years ago Valve did something very big with Half-life's netcode. About 2 years ago, Valve put out a controversial (at least, among its players) new piece of code that basically allowed your client to determine its own hit detection. It was Valve's way of defeating lag, and IMHO it works extremely well. The way it works is kinda like this:

    You have 2 guys playin a game, Mr. LPB with a ping of 20ms and mr HPB with a ping of 1000ms. The server is goin along, doin that thing servers do. Anyways, Mr LPB runs down a hallway and past an open doorway. Mr HPB comes to the other side of the hallway at the same time in reality, but because his lag is so high he hasn't gotten the message from the server yet saying that LPB has passed the door. 1020ms later, mr HPB sees mr LPB run past the door, and shoots. Now, according to the OLD way servers and netcode was written, mr HPB will miss, no questions asked. But with Valve's NEW netcode, mr HPB's computer tells the server that he hit mr LPB, so he actually lands all of his shots, killing mr LPB.

    Now, back on mr LPB's screen, 1 second later, he suddenly dies. There was nobody around, nothing that he saw that could kill him, but he died. That's because the server saw that mr HPB's shots did indeed hit him, and it checks the packets from mr HPB to see if he indeed did fire at where mr LPB was a second ago. If it all checks out, then BLAM. LPB dead.



    So...now you may be wondering what all of this means. Well, basically with similar netcode, Sega/AM2 could write a server wherein you could do these same things on the internet. Sure, the system's not entirely perfect, but it's still alot better than a basic server. This way, when you enter the command for the first part of a catch throw, and you see it connect and input the second part, your game will tell the server that it connected. The server will check it out to make sure, and if it did then you'll pull off the catch throw.

    Using a system like this, it actually rewards high-ping players to be more aggressive, and rewards low-ping players to be more choosy about their attacks. But now I'm starting to digress more than a bit, so with that last piece I'm goin to sleep. Gnight
     
  10. Pai_Garu

    Pai_Garu Well-Known Member

    Ok ok fine, just like how people in the 20th century envisions flying cars in the 21th century, yeah all the current obstacles of online gaming will be solved in the next few years and VF will be arcade perfect online. Yay. /versus/images/graemlins/laugh.gif
     
  11. KiwE

    KiwE Well-Known Member

    Yeah, exactly like that *thumbs up*
     
  12. KTallguy

    KTallguy Well-Known Member

    Yea, the Half-Life thing is interesting but it does leave for some difficulties... For example, it's difficult to balance. Lets say the 'laggy' person's computer shows that he fired at me, but on my computer (with my high ping connection), I saw him before and ducked behind a box. If his computer says that he hit me, but my computer says that I ducked, how do you fix that ? I believe (and I'm just trying to remember this...) that there were problems where you'd duck behind a box before shots were fired, but still get hit and still die, because of that new Half-Life idea.

    In a game like VF, stuff like that just would be so difficult to balance. If I duck someone's high, but on their game it says that it hits, do I immediately go from a ducked position to a stunned standing position? It's kind of a pickle.

    KiwE:

    1. Instant hit weapons, such as the lightning gun in Quake 3, work fine if you can lead the opponent depending on your lag. Sucks if the gun fires 1/60th of a second late and you don't lead, as you might miss. It also would suck if your catch throw didn't connect because you missed the frame because of lag, therefore leaving you wide open.

    2. Splitsecond timing isn't as crucial in FPS as it is in VF. Example: FPS, you have 100 ping, your shots come out a little late, so you compensate. VF, you do a TFT with Kage, and then miss the Knee into the combo because of lag. Your whole combo is screwed.

    3. Have you ever had a connection problem in an online game? Happens once in a while. Exactly how Srider explained it. You keep moving but the rest of the game stays still, because you're lagged. You think you're firing and attacking someone, but you're not. Then when you re-sync, you're dead because you were stuck in one place while you're lagging. Come on man, regardless of all the new 'solutions' for solving lag, in a VF game, it's too fast. All the timenudging in the world won't help you if you ate a knee because when you pressed block, it came out 10 frames late.

    4. Arcade communities MAKE you want to play, because it's better than playing faceless morons online in the first place. That's why so many people join FPS clans, because of the community, and I'm sure you can find very tight nit communities at asian arcades, it's a shame that in the US we don't have that luxury.

    5. If you increase the number of inputs necessary to do a move... ie. minami step, it's just more data, and there's more of a probability for some data to be lost. Minami step requires split second timing of several inputs at once. If one input is off, the move won't work at all in the first place. Most weapons in FPS games require a single mouse click to fire.

    6. Your thumb may cause lag, but that's just a direct reflection of your skill at the game. Lag because of hardware and connection problems is not a reflection of skill, which means it's unfair to both players. I don't get pissed off at people because I can't click fast enough when playing UT. If there are identical machines, it won't eliminate lag, even if there are identical connections, lag won't necessarily be eliminated, because spikes in the server or connection occur. It's not reliable enough.

    7. Distance takes time to traverse, IE Srider's phone call example. Until there are connection speeds that can almost eliminate that traveling time, and technology that allows a steady connection without lag, VF won't be playable. A few frames will make a difference in the gameplay for anyone who takes the game seriously. Buffering won't work if your inputs are late, and neither will catch throws, evade-throw escape-guard, etc. They'd have to change the entire game (ruin the entire game) to make the frames more lenient for online play to work. I'd much rather talk on these forums and play with real people. You make better friends that way.
     
  13. Pai_Garu

    Pai_Garu Well-Known Member

    I just reread some of the stuff and just wanted to defend myself a little without expanding the argument.

    [ QUOTE ]
    No, let's talk technical - cause you're still fronting and think you can say something vague and quickly move to what you feel more familiar / comfortable with. There are counters to evades

    [/ QUOTE ]

    Ok fine, I'll explain things in detail since people think it's vague and I lack the ability/knowledge to respond to them. Here goes.

    I said that TCP/IP are two completely separate things. It's obvious even from what KiWe said, that they are in deed two separate things. Basically IP is used to move packets inbetween nodes and TCP is used to verify the packets. One is used to locate where the packet needs to go in a network and TCP make sure it went to the right place and error checks the packets. You could consider them to be one thing since people say TCP/IP together. But really, if you trace the history of networking, IP networks were created without the "TCP component" (Figuratively) inorder to maintain communication if nodes are to go down. So really, they are in some ways mutually exclusive, but in today's common internet settings, since it is standardized to be the TCP/IP standard, it can be easily categorized to be one package. So I don't see why I was wrong in saying that they are two complete different things with different performance speed? It is possible to have an IP network which is more robust, but more error prone. When you add the TCP layer to it, it's much more reliable, but very slow when recoverying from a line or node failure. It's obvious to understand they are different things if you look at UDP/IP protocol, since the TCP layer is not used there.

    On the topic of ping, ping is part of ICMP, which is part of TCP/IP. It exists within the network layer (IP), rather than the transport layer (TCP). All ping is is an echo request, it is related to network speed performance, but it doesn't necessarily mean better TCP/IP performance. You can easily see this when you sent packets of specified sizes. Try pinging your gateway with different sized packets. You'll find out that even if you send different sized packets, the ping won't change. So to optimize TCP/IP performance, you need to figure out the maximum trasmit unit possible within the ping time. For example, say a delivery truck that drives between two warehouses takes 5 mins each way. You can either put 4 boxes in the truck, or 5 boxes (max size) in the truck. So that means, within the same 5 mins (ping) , the truck can either carry 4 boxes or 5 boxes (mtu). The difference may seem small at first, but let say the truck goes back and forth for 300 mins, and later 2400 mins.

    300/5 = 60 * 4 = 240 2400/5 = 480 * 1920
    300/5 = 60 * 5 = 300 2400/5 = 480 * 2400

    As you can see, the difference increase as more and more the traffic increases. So yes, ping is important, but really it's only part of optimizing a network.

    On the part of directx. Initially, the context called for examples given in fps. So for me to use my experiences in programming, I had to relate it to how most games are structured, which is using directx. DirectX is NOT only a "multimedia library". Yes it does include prebuild function to which developers can directly access the hardware. In that sense, it is a "multimedia library." Fundamentally, it is a software layer that provides high level code to directly interface with the hardware. Do you remember in the old days, such as doom, you had to pick which sound card you have in the setup? Ever wonder why you don't have to do that now? This is cause before, developers had to write custom code to support each soundcard because they all had different features/functions. But now, with directx, it provides a common layer for developers to code for instead of having to write different versions of the sound code for each soundcard. About direct play, really, direct play is just a layer to which the hardware will provide events to the software from a remote source. In order for a game to run, it must receive inputs to process. These are turned into event by the directx layer for the game to process, whether it be from your keyboard, or data coming through the network. So those sockets, which IS part of direct play, interfaces the game code to the network. How can you say they are not part of it when 1) they have dp as a prefix. 2) deleting them or unregistering them will prevent direct play games to work? This is in no way a claim that directx is NEEDED for all online games in the future. It's obvious that ps2 doesn't use directx. The issue here is that when event generation from the networked sources is lagged, the local machine doesn't receive input. However, the game will still run despite the delayed input. This is why the game can still run on your machine even when it's out of sync'ed. To further illustrate this example, take any game. Theoretically, once a game is started, inputs are needed for the game to progress. So why is it that if you don't press any buttons, the game doesn't freeze up? This is because the game code will run it's main loop despite no 'events' (input) happening. So the game doesn't freeze up and wait for inputs. You can see the opposite at work in many online java games. The game will freeze until you make an input, this is cause the game doesn't use that structure. This is why even when lags happens, your game doesn't always/fully freeze up. (It happens sometimes because the game simply cannot continue to the next part of the code without completing the current routine.) To see how this actually works in code, this is why I suggested directx network game books. Because it will illustrate all that stuff in gory detail.

    On the speed of light issue, if you would reread the posts. I never stated that as one of my examples. It does however, still support my claim that there is a peak in which data travels though a wire. So stop saying it's me who stated that. Also, what do you have against that claim? Can you make data travel faster than light? Have you done something to make data travel faster than what is possible right now? Speed of light is simply something to show a current natural limitation that no one can overcome right now. It is extreme, maybe even too extreme, this is why I did not specifically stated that myself, only supporting it validity and adding that fact into my argument. I used a much more related example, such as computer cable/wires, and phone line. Mind you, many people are still on 56k (maybe not for gaming), and DSL connection are still traveling on phone lines. Yes DSL is considered to be broadband.

    Anyways.... this is not meant to further this discussion. It's not meant to challenge anyone. These things are said simply to show what I know and why I have said what I said. I just don't want people to get the impression that i'm talking out of my ass. Anyways, back on the original topic, just because someone is making DOA online, doesn't mean someone figured out a way to make a lag free online fighter. Either way, we'll find out how much better things are once it comes out, if at all. On a side note, everything is relative. I for one, as an example, doesn't think the mobile phone technology have advanced very much here in America in the past few years. Maybe many people think otherwise, but I wish I can show them what all is possible in Asia in that area. VF.net anyone? Maybe I'm just pessimistic.

    /versus/images/graemlins/laugh.gif
     
  14. Shadowdean

    Shadowdean Well-Known Member

    An arcade is a wonderful idea...if you have them. Online VF would be a MUCH welcomed addition to the series for people that do not have local competition and can't drive hundreds of miles to play people.
     
  15. Painty_J

    Painty_J Well-Known Member

    Just for the record:

    Speed of light < speed of electricity.

    Well, the halflife example I gave above is just an example of a way they can make lag better. Yeah, it's still gonna have kinks in it, and it'll still kinda suck. But the example I gave shows how they can make it better than just allowing both machines to play at whatever packet rate they can do.

    Besides, I dont think lag would be able to cause you to miss stuff like Kage's knee after the TFT. Using the example above, ytour machine would still tell the server that you pulled the knee off at [THIS TIME], accurate to the same time it would be if you were playin solo. So you should be able to pull it off.

    Basically, in a well-designed game system, lag doesn't affect YOUR OWN inputs. Whatever you enter, you enter, and you'll see the effects of it right away. What lag does affect is the timings and visibility of your opponent's actions. Lag won't suddenly make you unable to do Akira's knee, or the [P][P][P][K] combo. It'll just make it harder to react to your opponent's actions.

    And I still want to stress what shadowdean said above: Online VF would still beat the hell out of nothing, if you dont have any local competition, any arcades within a 100 mile radius, and dont live in Japan.
     
  16. KiwE

    KiwE Well-Known Member

    Thank you fucking god for being one of the few sane people in this thread, still thinking about if I should write a mother of all posts again, will have to be when I get home from Uni. L8r.

    Edit: Atleast Srider has tried to do his homework before posting this time - lol.
     
  17. Fishie

    Fishie Well-Known Member

    [ QUOTE ]
    Pntsnfng_Joe said:

    Just for the record:

    Speed of light < speed of electricity.

    Well, the halflife example I gave above is just an example of a way they can make lag better. Yeah, it's still gonna have kinks in it, and it'll still kinda suck. But the example I gave shows how they can make it better than just allowing both machines to play at whatever packet rate they can do.

    Besides, I dont think lag would be able to cause you to miss stuff like Kage's knee after the TFT. Using the example above, ytour machine would still tell the server that you pulled the knee off at [THIS TIME], accurate to the same time it would be if you were playin solo. So you should be able to pull it off.

    Basically, in a well-designed game system, lag doesn't affect YOUR OWN inputs. Whatever you enter, you enter, and you'll see the effects of it right away. What lag does affect is the timings and visibility of your opponent's actions. Lag won't suddenly make you unable to do Akira's knee, or the [P][P][P][K] combo. It'll just make it harder to react to your opponent's actions.

    And I still want to stress what shadowdean said above: Online VF would still beat the hell out of nothing, if you dont have any local competition, any arcades within a 100 mile radius, and dont live in Japan.

    [/ QUOTE ]

    Bulshit

    Speed of light=speed of electricity.
     
  18. akai

    akai Moderator Staff Member Bronze Supporter

    PSN:
    Akai_JC
    XBL:
    Akai JC
    [ QUOTE ]
    Speed of light=speed of electricity.

    [/ QUOTE ]

    in IDEAL conditions...yes
     
  19. KiwE

    KiwE Well-Known Member

    Had the most boring lecture (if one can call it that) ever of my life and now I'm sitting and waiting for the double Naruto ep to be subbed like a cheap 2 dollar whore so let's do this (I see that Srider has taken the time to explain ICMP to add posture to his post, fun stuff).
    If someone is, like, sensitive or something don't look at this as I'm not pulling punches here (cause why the fuck are we talking DirectX when discussing online Virtua fighter online etc?) Srider has atleast done alittle webpage (imho with sugar on top) looking before this post but fails to see big things and does misstakes constantly (yes, talks out of his ass).

    [ QUOTE ]
    I said that TCP/IP are two completely separate things.

    [/ QUOTE ]

    No you didn't. You said TCP and IP are eachothers opposites and that tcp is slower (wtf?)
    Is this another one of those phone things where you don't really say what you say? TCP/IP is one thing (a network protocol). Ip is a network layer and tcp is a transport layer which function is to verify date (that I explained). How can TCP exist on its own? How can it be the opposite of IP? Wtf are you smoking?

    [ QUOTE ]
    All ping is is an echo request, it is related to network speed performance, but it doesn't necessarily mean better TCP/IP performance.]

    [/ QUOTE ]

    Well pinging gameservers is what people actually do in the real world do to make a quick judgement about how their performence will be on the server isn't it? What do you want, a 30min package swap test to check latency? Has it or has it not a correlation with better tcp/ip performance Srider?

    [ QUOTE ]
    You'll find out that even if you send different sized packets, the ping won't change.

    [/ QUOTE ]

    UTTER, COMPLETE, BULLSHIT. You should look up your "any Direct X networking game book" Srider cause it fucking sucks. Try ripping your friends ip (and be nice to it or just get any ip) and then starting a command prompt and pinging it along the way of:

    <font color="yellow"> ping ip -l -fxxxx </font>

    Then vary xxxx, the packet size, just between the intervall of 800 and 1200 and see what happens.
    Who the fuck knows of ICMP, says shit like that and then tries to describe a basic MTU ping test (and contadicting himself) if not a BS'er / Someone talking out of his ass Srider?

    Let's have a look at some of the, oh so funny, directX statements:

    [ QUOTE ]
    Fundamentally, it is a software layer that provides high level code to directly interface with the hardware.

    [/ QUOTE ]

    Direct X gives low level access to your systems hardware ffs. And it's code is generalised and slow (even Direct Sound sucks compared to Asio on a good soundcard but we'll get back to that).

    [ QUOTE ]
    Do you remember in the old days, such as doom, you had to pick which sound card you have in the setup? Ever wonder why you don't have to do that now? This is cause before, developers had to write custom code to support each soundcard because they all had different features/functions.

    [/ QUOTE ]

    Or maybe it's cause of you're within the Windows interface and it's easy to check hardware for a program doing a SysId check? I just played a .wav file on my OS without initialising direct sound routines or telling the program my IRQ and DMA?!! OMG SOMEONE CALL BILL GATES!!! RUN OVER HILLS OF FIRE AND YES YES, SLAY THE DRAGONS IF YOU MUST BUT YOU MUST GET THE MESSAGE ACROSS!

    [ QUOTE ]
    So why is it that if you don't press any buttons, the game doesn't freeze up? This is because the game code will run it's main loop despite no 'events' (input) happening. So the game doesn't freeze up and wait for inputs. You can see the opposite at work in many online java games. The game will freeze until you make an input, this is cause the game doesn't use that structure.

    [/ QUOTE ]

    Java is based out of C scripting language and yeah, it can have has recursive loops just the same (correct me if I'm wrong?). Just out of curiosity Srider - what are your "experiences in programming" (if you say Direct X I think I'll laugh my knob of) and were they ever beyond highschool level m8? No I don't consider adsl modems broadband (and never will), it's just a marketing ploy imo (that sells to senior citizens due to the likes of a "multimedia computer" 5 years ago) but hey, you're free on that opinion atleast (see, I'm not always an ass).

    ---------------------------------------------------------------------------

    Ktallguy; Neither I or Pntsfng_joe are network experts / programmers but low life gamers. Understand that for every try of an explanation we make someone compitent and actually programming (as the Tecmo guys right now) these things will probably have better solutions.

    Vf doesn't work instantly and it's not exclusive with framedependent moves (do you know how many justframe things that exist in tekken?). I think you see the whole thing kinda wrongly though to be honest.
    Like, why wouldn't your knee come out? Your world and your moves are made locally, as Pntsfng_joe said - the problem, if any would be reacting to the other persons move, not ever your own. Nobody is reacting and performing moves on 1/60th of a second. Not even close.

    Here's a typical example of what, imo, happens in a standard vf game:

    *P1's move starts after input>p1move is considered active and hitting>p2blocks>p1looses control of his character>p2gets blockstunned>p2 gets control back of his character first>p2 attacks>p2s move starts up>p2s move is considered active.

    Note that even the fastest move in vf is at 8 frames in that counterattack (punch is 12 normally), there is time to verify what happens after the gaurd etc (say player 1 does a lowpunch against p2's throw attempt) and time to see if the move hit.

    Now, take a hitthrow example and translate and you should see that if you hit someone with an attack in vf, they will have a rather large hitstun during which your machine could send that you (locally) did a hitthrow correctly on your side and then perform it globally (specially if you consider the time it would take for him to regain control over his character + actually do something else). If anything problems would occure in flaky animation etc (but not really if clients are smoothed out and you are at a fast connection - take Akiras knee for example).
    This is basic shit and it's just a theory - not a solution.

    As to the reason I though many were strongly against VF:Online without reason (afraid of changing from the arcade culture) - that's just a matter of opinion.
    I have met many friends while playing a game online, rts and fps who I have later on met irl and become very close to so I don't consider it faceless (moron) competition (but then again I don't play games with 100 ping like you seem to do).

    ---------------------------------------------------------------------------

    Fishie:

    [ QUOTE ]
    Speed of light=speed of electricity.

    [/ QUOTE ]

    Yeah, I'm not so familiar with physics on a high level point but there are differences between the two aswell if I'm not misstaking. For example, speed of light, c, is lights speed in a vacuum (not really common) while electricitys speed, similar to that is simply defined as thru a wire (which is highly probable). It's all theory though but he didn't use any formula / symbols but basic words. Could probably argue alot about if lightsspeed is a constant variable (it's not imo) and the end of all limits. Electrical pulses have recently broken the speed of light record though.

    Yours Sincerly - KiwE (Badmouse/BadPad/BadComp/BadPing/Imbalence/YourCheating/BuggyGame/YouhavenoLife/ILetYouWin/FriendsAcc/BrotherPlaying/CrappyServer/HaventSlept/KicktYouIRL).
     
  20. Pai_Garu

    Pai_Garu Well-Known Member

    I like how you twist my words into incorrect things while at the same time the corrections you give are clearly just the same things I have said.
    Such as
    [ QUOTE ]
    Direct X gives low level access to your systems hardware ffs.

    [/ QUOTE ]

    I said
    [ QUOTE ]
    Fundamentally, it is a software layer that provides high level code to directly interface with the hardware.

    [/ QUOTE ]
    That statement means directx provides a direct interface between high level code and low level hardware. Even if you think I meant it's high leveled, it is pretty high leveled compared to assembly. Even you had said before that directx is just "multimedia libraries", which would be high level code, nice of you to now agree with me that it's code to provide low level api.

    I've always been just saying the same things over and over again, just more and more into detail as you have challenged me to do so. It's not revelations gained from looking things up, since my stance and argument have been the same since square one. For example, how come now you agree with me that ping is not a true representation of TCP/IP performance when you said they are earlier. Nevertheless, it's pretty obvious what's going on. Someone just doesn't want to let up on this argument and need to constantly change things to make it something else. Oh well. How about this, give us some technical details on why you think online VF would work? You've been claiming how my arguments are nonsense, so why not give us/me some insight into how it would work then? I am personally curious as well, as if it is possible, then it's news to me. Please shed some light into this issue.

    Also, I have to defend myself again here since you said one of things I said as b.s. So here's the proof on why it's not. Here are some data from pinging my gateway.

    Pinging 192.168.1.1 with 1400 bytes of data:
    Reply from 192.168.1.1: bytes=1400 time=8ms TTL=150
    Reply from 192.168.1.1: bytes=1400 time=11ms TTL=150
    Reply from 192.168.1.1: bytes=1400 time=9ms TTL=150
    Reply from 192.168.1.1: bytes=1400 time=9ms TTL=150

    Minimum = 8ms, Maximum = 11ms, Average = 9ms


    Pinging 192.168.1.1 with 1405 bytes of data:
    Reply from 192.168.1.1: bytes=1405 time=8ms TTL=150
    Reply from 192.168.1.1: bytes=1405 time=11ms TTL=150
    Reply from 192.168.1.1: bytes=1405 time=8ms TTL=150
    Reply from 192.168.1.1: bytes=1405 time=9ms TTL=150

    Minimum = 8ms, Maximum = 11ms, Average = 9ms

    About the same ping, but one with more data??? This 5bytes of difference is enough to be a huge difference when the traffic increases. You can even try up to 200-300 bytes of difference and get about the same ping, depending on your network.


    Also, to joe, yes lag doesn't affect your own local operations. However, the game doesn't just run according to inputs from one source. In your scenario it's just like playing against the cpu. What matters is the inputs from the other player, if that lags, then the two individual local processes won't be in sync. Like, if one guy does TFT, and the other guy jabs. The first guys's machine says TFT happened since the jab input was a little lagged. The other guy's machine says his jab landed before the TFT happened since the TFT input was a little lagged, so which machine is right and how should it be decided what actually happened? That's what's keeping online VF from happening.
     

Share This Page

  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.
    Dismiss Notice