About performance and releases...
Pawndemonium
Join Date: 2010-07-24 Member: 72725Members
Join Date: 2010-07-24 Member: 72725Members
I didn't really want to create a topic for that, but as moderators tend to always come up with that release card, I'm quite curious if it's more than just talk.
I've followed the development of NS2 for over 2 (or 3? no idea) years now through beta participation and random observing, and it's still the question which haunts me. The game really evolved a lot, and the whole UWE team has proven its strong bond with the community, which is commendable these days (sorry for not participating in any way, but tbh I kinda was too pissed =P).
But in all honesty, how feasible is it to assume, that this game ever runs on a mediocre rig with reasonable graphic settings? Let's say a dual core at like 3Ghz and a mid tier GPU (or top shelf 3-4 year old stuff).
I still think it was a waste of time to have a whole new engine for a single game, but that's just my opinion, as I don't care about nifty eye candy, which could have been implemented on different existing choices in that time anyway (yes even dynamic infestation...), but whatever, in the end you had your reasons, as it's quite the experience to create something from scratch, I can accept that.
Now it's about people playing the game though, while this game offers a lot, it doesn't matter when the engine doesn't work in most people's favour.
So can please someone who really knows what's going on in the regard clarify that? By that I don't really mean mods who come up with stuff they have no idea of (not blaming them in any way, they just don't work on the engine, they can't know).
The engine looks fine, but doesn't offer that much to justify those hardware demands.
tl;dr: Does more performance at release mean, the game looks like a 1998 game to run on 2008 stuff, or are we talking about improvements to the engine itself to not require more than necessary?
Cheers.
I've followed the development of NS2 for over 2 (or 3? no idea) years now through beta participation and random observing, and it's still the question which haunts me. The game really evolved a lot, and the whole UWE team has proven its strong bond with the community, which is commendable these days (sorry for not participating in any way, but tbh I kinda was too pissed =P).
But in all honesty, how feasible is it to assume, that this game ever runs on a mediocre rig with reasonable graphic settings? Let's say a dual core at like 3Ghz and a mid tier GPU (or top shelf 3-4 year old stuff).
I still think it was a waste of time to have a whole new engine for a single game, but that's just my opinion, as I don't care about nifty eye candy, which could have been implemented on different existing choices in that time anyway (yes even dynamic infestation...), but whatever, in the end you had your reasons, as it's quite the experience to create something from scratch, I can accept that.
Now it's about people playing the game though, while this game offers a lot, it doesn't matter when the engine doesn't work in most people's favour.
So can please someone who really knows what's going on in the regard clarify that? By that I don't really mean mods who come up with stuff they have no idea of (not blaming them in any way, they just don't work on the engine, they can't know).
The engine looks fine, but doesn't offer that much to justify those hardware demands.
tl;dr: Does more performance at release mean, the game looks like a 1998 game to run on 2008 stuff, or are we talking about improvements to the engine itself to not require more than necessary?
Cheers.
Comments
But in all honesty, how feasible is it to assume, that this game ever runs on a mediocre rig with reasonable graphic settings? Let's say a dual core at like 3Ghz and a mid tier GPU (or top shelf 3-4 year old stuff).
..<!--QuoteEnd--></div><!--QuoteEEnd-->
<!--quoteo(post=1770991:date=May 12 2010, 04:55 PM:name=Max)--><div class='quotetop'>QUOTE (Max @ May 12 2010, 04:55 PM) <a href="index.php?act=findpost&pid=1770991"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->By the way, my desktop also has a GeForce 8800 GTS 512.<!--QuoteEnd--></div><!--QuoteEEnd-->
I think Max (UWE Technical Designer) uses a pretty mediocre (oldish) rig I think.
I suppose most of the performance cost goes into the account of Lua, since the complete gameplay part is realised with it. So they traded extreme modifiability and "easier" development for performance. And depending from the point of view, this does justify the hardware demands somehow.
With easier development i mean, that its just much easier to hire a developer that knows Lua on a certain level, than a c++ developer (or an equivalent language). But this also comes with the tradeoff that Lua is by far not as efficient as c++.
Im on to you Unknownworlds... all those pizza parties and late nights, lol.
Finally, the architecture is designed to be resistant to client-side hacks while still allowing a great deal of client-side prediction, which also adds quite a bit to the server load.
That being said, performance will get better as the engine evolves.
Will there be enough players to justify time invested in modding a new engine that has such high hardware requirements as to limit the playerbase.
The UDK is already a popular choice, and it runs on a suprisingly high number of machines with acceptable visuals , Unity is also very popular to use for 3D titles bypassing a lot of the rendering expertise required and bringing decent 3D capabilites to modders and indies alike, then there is Source which is famous for low requirements (but is getting old on the old eye candy department)... to name a few options a modder could take.
At the moment Spark offers Source level visuals but with Crysis 2 / BF 3 level hardware requirements... not only does that limit the playerbase but it makes modders doubt if they could get anything out the engine if the original programmers cant get the requirements down.
Hey, I aint against anyone here... just putting out my opinion and confusion about the insistance on sticking with LUA and making the engine so modder friendly at the immense performance cost.
I think the engine is a marvel for a small team, mostly Max's hard work and I salute the skill required to make it (hell I am still making tile engines in 2d using SDL)... but business doesnt care about that, and business is what will pay the bills.
Yes, "We're so open!" said people who made perfectly closed C++ game.
I don't know many people that would like to mod game/engine that gets 15 FPS out of pretty good hardware. UDK is far more open, far more stable, far better at everything.
matso: I don't know what you're smoking but it's got to have serious side effects. No game/mod ever no matter how many features got you that few FPS as NS2 does.
"It also allows you to write your games using Lua rather than C++ which speeds up development at the cost of extra CPU" - that's half-truth, you can develop good games with C++ alone. You can reload code on the fly too. NS2's Lua setup is just horribly slow, allocates where it shouldn't and is bad technical choice unless one of these can be fixed. I don't care how much easier it is to reload Lua code, it simply isn't an acceptable choice for game language using UWE's method.
I don't see how prediction has to do anything with NS2 being just slow. If movement code run few times more doesn't fit into 16/33ms budget then IT IS TOO SLOW. Go look at HL2 prediction. Even mods with fancy features in them don't take so much time (prediction in +showbudget is like 1%).
" think the engine is a marvel for a small team" - you mean it sucks and you accept it because of brown nosing/emotional investment or other excuses. It clearly is not a marvel, amount of people playing at once stayed more or less constant - 100-150 people with spikes around updates. Many more people preordered, from what I asked people I know they all stopped playing "beta" to wait for something that is actually playable.
Twitter/Youtube/Show hype doesn't do anything for me playing the game. I can understand overhyping because funds are running out or something like that but all I care about at the end of the day is stable 30+ FPS. For any given time period where it's not true usability/fun function is simply 0, it doesn't matter how many features there are. Excuses like "beta" are unproductive, they don't change my FPS.
Am not going to bother playing until the actual release because its not enjoyable.
I have been playing CS GO since release and it runs perfectly and the graphics look amazing.
Am hoping NS2 will have the same finished polish.
I lol'd
I don't really see the big point, UWE decided to do this, and the made some choices along the way, that people like a lot, and ofcourse some changes that people argue about.
However I think it's more fun to make something that is impossible, and can't be done, and then actually get it done, instead of just being another random mod or game that runs on one of the big systems. When I look at my work and think back, the fun times was when I made something that was hard, not when I just did what everybody expected.
Right now, the game is unfinished. Performance is not where it needs to be - I agree.
<b>Yet</b>, I've played games made by AAA studios, with millions and millions of budget, which I enjoyed exponentially less than I do NS2 right now. Sure, this won't be the same for everyone, but I'm just saying that it's quite possible to please people with what they've got. And I have faith that it'll improve further.
I fully understand the question, the reason why you want to know whether the game will be as 'perfect' as it should be. I've got a very high-end PC myself, which can run the latest games on the highest of settings - I like quality.
However, it's impossible to say where things will end. The devs try to make things work as well as possible, obviously. They can't be a 100% sure where things end up. In my opinion, this is where the faith comes in. As much as people hate to read it, it <b>is</b> still beta, and we'll just have to wait for UWE to fix everything. There's not much else we can do, right?
I don't find that very convincing. Joint Operations had 150 player battles with vehicles and huge maps in 2004. The HL2 mod Empires had similar elements like NS2 (commander, upgrades, buildings, vehicles) in 2007. Same goes for Nuclear Dawn, which came out last year. What exactly is so much more complex about NS2, that makes it challenge top range hardware in 2012?
<!--quoteo(post=1969414:date=Aug 31 2012, 10:16 PM:name=Perestroika)--><div class='quotetop'>QUOTE (Perestroika @ Aug 31 2012, 10:16 PM) <a href="index.php?act=findpost&pid=1969414"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I lol'd<!--QuoteEnd--></div><!--QuoteEEnd-->
Well, it's not <i>that</i> much better.
Oh now let us be fair, we have what... 3 weeks to release? Don't you think desperately holding on to beta-status is starting to get a little silly?
Those games were all <b>programmed</b>. If you wanted to make a game in the source engine, you acquired the source code, in C++, and began <b>programming </b>your game. Because it was <b>programmed</b>, it was precompiled and made much easier to execute by a machine, and thus much faster. <b>Programming</b>, as opposed to <b>scripting</b>, is arguably "more difficult", and usually involves formal education or experience. <b>Scripting </b>can typically be picked up much more quickly and easily. This is because <b>scripting </b>takes place in a controlled environment with a limited number of considerations which need to be taken into account. A <b>script </b>is interpreted by a <b>program</b>, and does not need to be precompiled. This means that the code needs to go through more steps in order to be understood by a machine, and thus takes longer to execute.
What this means is the following; a game programmed by the average Joe would run slower than a game scripted by the average Joe. A game programmed by an educated or experienced professional would run faster than a game scripted by an educated or experienced professional.
also, any single patch can cause a very noticable increase in fps, so even 3 weeks (who said that btw?) can still matter a lot: features are in and ready for performance polish, servers can still become more stabilized in their tick rates.
Made my day! :D
Back around 2000, the modding scene was booming. There were so many amazing half life mods, and I rate them all far higher than most triple A game releases now. Rocket Crowbar, Vampire Slayer, Action Half Life, The Specialists, Half Life Bumper Cars, Science and Industry... there's another one whose name I forget... Oh yes, the crowning achievement of the Goldsource engine, the finest modification to ever exist, Natural Selection. These were the golden years of gaming, mainly because (not to sound elitist here) "casual gamers" didn't really exist. For the most part, if you were playing games on a computer at this time, you weren't casual, and were pretty dedicated to them. Gaming in its entirety was nowhere near as big as it is now, and computer gaming was still a bigger industry than console gaming. So you had all these dedicated people, who loved the games they were playing, but were getting kind of bored. So what did they do? They created their own games, if for nothing else, they did it for ###### and giggles. They may not have been "balanced" or "well designed", but they were really goddamn fun, because they were made by people who liked to play games.
Unfortunately, the appeal of gaming would be its downfall (in some sense, an elitist sense I guess). As the gaming industry got bigger and bigger, people began to realise that you could make a whole lot of money by making a game. As I said, I rate a lot of half life mods higher than some current triple A releases, and I've paid money for current triple A releases, and damn right I would have paid money to play those half life mods. Unfortunately, when money is involved, you have risk assessment, you have quality assurance, you have investors, you have share holders, you have a lot of bureaucracy, which slows the whole process down, and kills innovation.
Worse than that, you have things like consoles and mobile devices. Devices used by hordes of casual gamers, who only want their attention diverted for a brief few minutes. People who will pay for angry birds or the like. Games which take no time to make at all, time which is much better spent than making some free modification for a proprietary piece of software. Now this isn't all bad news, not at all. I tend on making a lot of money off these devices, people, and games, but it is bad news for mods.
DayZ shone a ray of light down on the mod scene, giving hope of resuscitation. It's creator spent a great deal of his time, with no promise of any kind of return, creating a game exactly how he wanted. It was a game that didn't consider any of the modern day design heuristics like balance, fairness, or usability, but which tens of thousands of gamers flocked to play. However, its success was quickly identified. Regardless of how little promise of return there was to begin with, there's no doubt that the games designer is now rolling in "return", and is no longer an independent modder.
I pray the spark engine signifies the second coming of the mod era in gaming. It would be a fitting achievement for a game that started as the best modification <b>ever</b>. But with gaming as it is now, I don't see people with the inclination or skill to create games doing it for free. I certainly won't be.
Imagine NS2 being able to run with 60 FPS constant on medium rigs, the accessability would skyrocket. Only then can you actually enjoy better movement code and so on (meaning only then you can actually beta-test). The game just became feature complete and immediatly goes final. I think a lot of potential is wasted that way, both performance and gameplay wise. Though, it's still a good game...
I suppose they are more aiming for small indie studios or even single developers. Since the "old mod-scene" we had 10 years ago is just not existent any more.
Imho the look of spark is also not that bad - actually i find it pretty decent (i don't know which graphics settings some users in here are using to play this game). This and the fact that you can write the gameplay-part completely in Lua probably makes it attractive in some scenarios (by which i don't mean explicitly a 16vs16 Multiplayer Shooter).
What you may have on your hands is a much more ambitious version of the Unity engine.
There are a couple of areas of multi-threading already, actually. Some physics works (ragdolls), packaging
network data for the clients and client-side rendering already runs on a separate thread.
There are some interesting potential to use multi-threading built-into the engine, actually. The server runs
each client input handling as a full game locally, which is inherently parallel, and updating the game state is
also a theoretically parallel work. The word here is "theoretical" though; there is no chance of it happening
before 1.0.
Multithreading really is not that hard, but adding it to a program made to run on one core? Well that's is a BIT harder.
Why did they not think of Multithreading in 2004? well I have no idea!!! :P
All jokes aside, there is a lot more they can do too make this game run well, I mean a LOT. Really it comes down to how much they can do with there funds, and the time they have to spend. All anyone can say for sure at this point is it will not run as well as BF3 but will run much better then minecraft!
Could you elaborate on that matso? I'm not sure if you're saying that the engine has a greater than usual potential for client-side prediction, or just that the work done handling player movement on the server is expensive but stops obvious client hacks. (Is this anything to do with the server using the client time step to handle movement?)
It may require a Direct X level higher than 9 though...