Posts by Zhav

    Thanks for the talk, Ted.

    That is plain false. WebGl is much more resource demanding hog. Another proof that you are clueless of everything that you are saying.

    If you understand the underlying technology behind HTML5 and why companies are making such a big push on it, then you'd understand why.


    Still words without results.

    Hella: Your game is nothing if I can't play it
    Me: Do you want to play? Also, I was talking about what Aeria Games can do.
    Hella: Your game is nothing if I can't play it

    Well as good as this discussion is I'm gonna give you the information that I have. Wich leads to a pointless discussion I guess.


    First of all gamigo bought aeria games and aeria games "own" S4 League. But aeria games has actually no developer team. That means some gamigo developers are making the changes in S4 League, but they work on several games. So they took the cheapest way (except doing nothing for the game) what leads to endless taking patches that mostly fail because the developer team doesn't work with the meta all the time. And the S4 League Team probably can't do anything about it. They can give the developer team the ideas but they are dependent on them. And because gamigo bought aeria games, aeria games is on the bottom of the regime. Wich means if a patch fails it is gonna take 6 weeks(?) again until maybe the problems get solved. Because gamigo games > aeria games games.

    Only to give ya the picture of it (Based on the information that I have. I have no insight.)

    This is super interesting. So that would mean Gamingo purchased Aeria from ProSiebenSat.1 -> Unless I'm missing a lot more information.

    Those who started playing the game 10 years ago will. Also DX9 is not as obsolete as you may think. Even some recent games still supports it.

    Sure, I'll give you that. An S4 remake built on an out of the box version of UE4 will not run on 10 year old hardware that was low-end at the time. There's always Unity.


    No. I am just giving example of why older versions are better. Also the WebGl is pretty bad compared to the web player... but why have one when you can have both?

    You my friend, are an id*ot. You're going to completely ignore all of the other performance increases in almost every area? You're going to compare WebGL to old and bloated WebPlayer? WebGL is so much more performant and easier to work with.... and they dropped WebPlayer because the technology used was being phased out.


    I have read it. Everyone can talk without having done or shown anything.

    If you want to play it, I'd be delighted to organize a match. But once again, since you're so slow at understanding things, we're talking about what Aeria can do to make the game better. It really seems like you're ignoring anything that would make Aeria look bad in anyway... there's no reason for you to have ignored the rest of my sentence and chose to focus on that alone.

    U4 doesn't even support DX9. Opheliafacepalm

    You understand DX9 was being phased out for years now, right? I can't imagine anyone would still want to play on Windows XP.


    It doesn't have the web player anymore. The way how sound is handled in newer versions is also terrible as well as some other things which don't remember right now.

    This is what's making it hard for you to develop? The lack of some features in the built in sound system (even though you can now extend it even more)? WebPlayer was phased out for WebGL...


    We can talk when it happens.

    Ignore the rest of the sentence, as usual.

    Old hardware just won't run U4 and the frame rat will be terrible even if it somehow does.

    The cryengine seems like a better option imo as long as accessibility goes (and it still looks impressive even today).

    You're not serious right... Hella please tell me it ain't so. You've just made it clear what your experience in game development actually is. I'm literally crying on the inside from what you just said. What about UE4 (an open source engine) that makes it unable to run on old hardware?


    An old one but on purpose (which is from where most of the compatibility issues comes from). The newer ones however are heavier and lacks features + more compatibility issues if I have to reimport again.

    The newer versions of Unity are heavier and lacks features? What?! I'm bewildered.

    Are you referring to the Package Manager as lacking features?


    Then maybe you can publish? If you don't want to get into legal stuff, you can get models from the unity store you know.

    I've got my own models done, and I will publish lt. But we're speaking about what Aeria can do with a game they have the rights for. I'm pointing out that it shouldn't be hard for them.... contradictory to what you said.

    Agree but there is no guarantee everyone will fall in love with a reboot. But even if we say it is coming, it still will take longer to happen than just fixing the game. May as well just have the game fixed while waiting if it is going to be a thing.

    At this point, you're just looking for every reason to justify not remaking the game. Why would people not fall in love with the same game that's more optimized, and less bugs (if they do it right)? I'm sure the dev team would also be 10x happier. They'd be working with something that's easier for them, and not be limited by an in-house engine made in 2006-2007. It's a win for everyone.

    I wouldn't mind losing everything on the account if that's the cost of having a reboot for S4, if we take in consideration they would develop a less p2w system (old players don't really have problems anymore with it, but newer ones do still struggle with it)

    In addition, it's not like they can't convert the data from the old database files into better structures. How do you think they transferred from Alaplaya to Aeria?

    It is not just the code but how the engine works. There is a lot of clunky stuff and it is not very good optimized too. It has its advantages but i just don't like using it.

    What specifically is wrong with Unity that gives you problems? Also, what version are you using?


    Maybe they are working on one right now but new engine in general means a lot of things will just have to be made from scratch like the netcode for example and other things may not be possible to replicate in the same way.

    I've done a fully authoritative version of it, with client-side prediction, and lag compensation. It feels like S4 almost 1:1 (without skills), with a few of my own twists and changes here and there. Match servers can run either on client computers, or started in headless mode on AWS clusters, with a fully working user account API (excluding inventories, shops and item switching -> which is relatively easy since it's using Authoritative Game Logic (server controls everything, client predicts and replicates)).


    I did that by myself, so I would expect more experienced programmers and a whole company to be able to do it.
    EDIT: I forgot to point out that it only took me 8 months to do so. All while juggling other tasks including getting animations to test with, models to use, and managing a team.


    I don't want to make an assumption, but I'm sure UE4 was more of an example.

    And PCs that could run S4 can definitely run a UE4 game if they put their resources into optimization. UE4's overhead can definitely be reduced a lot more as well by an experienced C++ dev.

    In terms of models, yeah sure they'll still be blocky, but tons of issues will be solved... And you can use shader tricks to make those blockly models look a lot better.

    If you wanted to make a game, nothing would stop you though.

    Can you stick to a single point and stop jumping all over the place? Yeah, nothing is stopping me from making a game. That's not what we're talking about. We were just talking about making something using S4's assets.


    It is never just type of thing. One time it is a compatibility issue. Other times something trivial becomes a show stopper because of being dependent on something else etc.

    I mean if it's a fundamental issue with your own code, then it shouldn't be hard to fix, more tedious. Of course if you're willing to accept help, I would be happy.

    That is not how things work. No company will give you resources to make a game without knowing what you are capable of. Which means you should have something ready to show before requiring any resources from them.

    I'm quite aware of that. Not like it matters to Aeria anyway. It's clear that they've given up on putting much resources into S4.


    Not to mention, I've heard they don't even accept help from private server devs.


    S3D takes time but for all for different reasons. Mostly because under the hood issues that requires time consuming workarounds.

    What're the under the hood issues?

    EDIT: Are they Unity specific issues, or just fundamental issues from when you first begun to learn?

    That is far from finished game and it probably still took him a lot to make. Not to mention, a lot of the things seems like taken from the unity store and put together so he already had most of it done.


    You probably have never made a game if you think it is so fast and easy.

    Well, I've got 5+ years of Unity C# experience, specialized in gameplay programming and networking. I'm not talking out of my *ss.


    And I think you're still misunderstanding what I'm trying to point out. Recreating S4 with what's already been done wouldn't be hard. If it were legal for me to use the S4 assets (models, music, animations, etc.), and Aeria provided an API to handle shops, user accounts, etc. I would have done it for free.


    And I really hope you're not suggesting that S3D was hard to make. S42D is understandable.

    This project is abandoned.

    I'm aware. But the point still stands that someone, even as seemingly inexperienced as that developer, can create a solid S4 base. Of course I'm ignoring all of the other stuff that game needs like game mode logic, weapon logic, client->server communication, etc.... but as I said, that stuff really isn't hard to do for an experienced developer.

    EDIT: I know you were just pointing out it's abandoned. Just wanted to clear up anything else.

    I have played countless MULTIPLAYER games from COD to flyff and hacks has never been as prevelant as it was on S4. Maybe im just lucky who knows.

    Downside of peer to peer

    I have been tolerating the drops in some things. Hacks are a part of every game so I can even take that. Some of the problems, I feel as seen from both a player and the staff is that angry people are unpredictable, toxic, and makes things worse. And that is something you can't patch out >:

    inb4chatcensor

    What? I though TCP was better than UDP?

    Not for a game like S4, where players are moving around really quickly. The issue with TCP is that it's slow... Since it requires a response that you received the packet. With UDP, we just fire the packet, with no need to wait to see if it was received. For events like Melee swinging, or kill feeds, we can instead use another transport layer called Reliable UDP, which does something similar to TCP (keeping packet order and packet acknowledgement).

    Understandably, it would be hard to wrap your head around this if you don't understand all of the networking models used in games, and how they differ.


    I'm going to use what Hella said and say that S4 is purely P2P, no relays, just each player directly connecting to each other. This would mean that each client is in charge of their own simulation... and we know this because everything client-side can be hacked.


    Now that we know that, we can create some scenarios. Let's set some locations and bandwidth limits first:

    (For this we're going to assume there's 0% packet loss, since that would just be wayyyyy to complex for a little forum post)


    Zhav lives in the Caribbean, with an upload speed of 0.5mbps and download speed of 0.1mbps, and his ISP routes directly to Miami, USA (no other routes).

    Hella lives in Miami, USA with an upload speed of 10mbps and download speed of 20mbps, and his ISP routes all over the world (he's near a hub).

    Rezilia lives in Scotland, with an upload speed of 20mbps and download speed of 0.1mbps, and his ISP also routes all over the world (he's also near a hub).


    All three of us are in a match of S4, sending packets directly to each other.


    Zhav's packets have to travel to Miami before routing to Rezilia. It takes 150 milliseconds on average for Zhav's packet to reach to Rezilia, vice versa.

    Hella's packets don't have to travel far to get to Rezilia and Zhav. It takes 40ms for Hella's packet to reach Zhav, and 80 to reach Rezilia, vice versa.

    Rezilia's packets don't have to travel far to get to Hella, but takes extra long to get to Zhav.


    We also need to assume the average size of a packet and how often it sends. So let's say it's 120bytes per packet, being sent 20 times a second. That's (0.12kb*20)^n where n is the number of players in a game. So in our case, that's 33kb/s, in other words 0.264mbps.... And now we already have a problem..


    Two of our players have less upload speed than required, so what happens to those packets? They never get sent at the right time, and since S4 uses TCP (I think), it tries to guarantee that they get sent, so they will arrive at a later time. To Hella, both Zhav and Rezilia would appear to be out of sync, with choppy movement, or teleports, or just act weird, because our packets arrive late. To Zhav and Rezilia, Hella looks fine, since we have enough download speed to receive her packets perfectly, and she has enough upload speed to send packets perfectly.

    Actions Per Minute? What does that have to do with anything?

    Whatever the delay is for sprinting, single tap is faster anyway. Even if you can double tap a key with a frequency of 20 Hz (which I doubt is even possible for a human), it's still 0.05 seconds slower than a single press.

    But that's not considering that double tapping could just feel more natural, at least for S4. You guys already solved the problem by having in either Double Tapping or Single Pressing.

    As for double tap to sprint, I think it's fine for S4. I think it's something you get used to quickly, and it's still common in Korean games.


    I'm going to assume WS is wall shooting and say, the heavy machine gun doesn't have wall shooting................... No idea why other weapons still have it.

    I mean, if we're remaking S4, the thing that needs to be updated the most is the networking model. Authoritative game design and networking is a must if we want to have a secure, (mostly)hack-free, fair, gameplay.