Can we focus on performance now?!
PoNeH
Join Date: 2006-12-01 Member: 58801Members, NS2 Playtester, NS2 Map Tester
Join Date: 2006-12-01 Member: 58801Members, NS2 Playtester, NS2 Map Tester
I've been trying to test out the BETA that I invested in and it's not been possible. I have a M11xR2 laptop from AlienWare with Core i7, 4GB DDR3 RAM, 256GB SSD, and 1GB Nvidia GT335M video card with Optimus. I know these are not the latest and best for today's standards, but it isn't bad either. Hell, some guy just posted 36 FPS avg. on the Crysis 2 benchmark with the same setup as mine less the SSD hard drive I have. Also, I get 150+ FPS constantly on CS:S with all my settings maxed out (literally... AA, resolution, everything). I don't expect NS2 to run at 100+ FPS, but a nice 40+ FPS avg with my system is more than reasonable.
I know that there are still lots of kinks to work out in the game still, but I feel like you guys are going on tangents and not focusing on performance. I understand how this happens when you guys are all testing and developing on top-of-the-line machines, but that isn't the case with the majority of the people who will buy and have bought this game. I just feel that details such as alien vision (which you guys just stumbled upon) can come later.
I want to contribute as much as I can to the community, but honestly I haven't been able to get more than 5 minutes of horrible 10-15 FPS gameplay before calling it quits.
I know that there are still lots of kinks to work out in the game still, but I feel like you guys are going on tangents and not focusing on performance. I understand how this happens when you guys are all testing and developing on top-of-the-line machines, but that isn't the case with the majority of the people who will buy and have bought this game. I just feel that details such as alien vision (which you guys just stumbled upon) can come later.
I want to contribute as much as I can to the community, but honestly I haven't been able to get more than 5 minutes of horrible 10-15 FPS gameplay before calling it quits.
Comments
Or Bee-Sebbinteen Bawmaz.
Seemed like garbage collection was supposed to be fixed for this patch but it never was. I'm pretty sure this is one of the more major issues causing hitches. Next patch needs to fix this. I bet the entire community would be happy with a GC fix that included not a single other fix with it. From what I was reading posted by others playing with their GC settings, they were able to significantly improve their performance.
Even if holding off on extra gameplay features is counter-productive, it might be a good idea to develop them and just not add into our patches until performance is fixed. People are clearly getting the wrong message. Even those who understand the dilemma are starting to as well. Not good.
I've seen this mentioned multiple times, and its a complete fallacy. We are not developing on top-of-the-line machines. Most of us are running machines that are 3 years old. I get about 10 FPS when I run the game on my office computer, and I know at least several of the other guys do, as well, so its just as frustrating for us playtesting and running the game as it is for you.
Performance remains a focus, but a lot of the larger improvements are bigger tasks that are a work in progress. If we have bug fixes and some cool new features we don't want to hold up the release of a patch, just because some of the optimization tasks aren't done yet. Some patches will have optimization improvements, some won't, but it doesn't mean performance isn't being continuously being worked on.
--Cory
Gameplay is not the thing we are supposed to test right now and performance is somethin...
!
Why do I keep trying to bother, in a few days another one of these threads will pop up :/
Here guys have the fully featured search engine for these here forums:
<a href="http://www.unknownworlds.com/forums/index.php?s=&act=Search&mode=adv&f=0" target="_blank">http://www.unknownworlds.com/forums/index....ode=adv&f=0</a>
Any smart game designer would work in that same method or pattern, to work on performance while still implementing a tremendous amount of new content/material isn't very wise.
Let them do their thing, it's beta guys not the final version of the game, clearly, half the abilities are still not implemented yet and honestly, the game is still very playable, If you want to see a mess of a beta or even launch, you should have played
Age of Conan.
The was byfar the worse experience I have ever had in a video game and it's still like playing a beta even after 2 years of release. NS2 is on the right track, with the money they have, the small development team, these guys are sick at what they do, it's inspiring actually.
Keep up the good work NS2, you guys are on the right track, don't water the game down for newbs, keep it skill/team based and keep those intriguing creative idea coming that you guys have been doing, the world is dying for a next gen game with old gen gameplay!
I miss CS 1.3, NS1, Duke Nukem 3D days :(
DON'T LET US DOWN!
<img src="http://primates.ximian.com/~miguel/pictures/well_actually_trollcat.jpg" border="0" class="linked-image" />
Well, actually it's not true at all. They have fundamental problems and not things like drawing 10-100k tris over limit. They're working on it as they said.
I see this argument pop up more frequently as time passes. Do keep in mind that while the statement is true, sometimes performing optimization is crucial to figure out what works, and more importantly, what does not. Case in point is Flash (which has been largely replaced by a Lua-interface and will probably be gone completely by release). As an interesting side-note, I do wonder whether they would've gone with Lua had they the chance to start over completely (as it's proving to be quite a performance-bottleneck).
You should also consider that this game is no longer in-house, but out in the open, which kind of forces them to perform a bit more optimization than they normally would have. Simply put, people expect something that runs better than this, it might not be reasonable, but that's the way it is (spoiled by SC2\TF2's definition of 'beta' perhaps?). You could've seen that coming from miles away when the beta was first announced, and acting like they're somehow bumbling idiots isn't the answer.
So, let's all just keep crying about it on the forums with our tears and hopefully they can make a cup full of them and make a toast before they drink them.
So many losers, so little time.
Maybe some answers from Max could explain things a bit ? Im pretty sure the answer ain't as easy as it might sound.
<!--quoteo(post=1833438:date=Feb 18 2011, 09:05 PM:name=sypro)--><div class='quotetop'>QUOTE (sypro @ Feb 18 2011, 09:05 PM) <a href="index.php?act=findpost&pid=1833438"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Maybe some answers from Max could explain things a bit ? Im pretty sure the answer ain't as easy as it might sound.<!--QuoteEnd--></div><!--QuoteEEnd-->
I should think the answers are mostly related to him simply not enough hours in the day for the poor man and UWE not having the budget to hire a full team to work on the engine.
While I am along with everyone else wanting to play a fully working game as soon as possible, I endorse "when it's done" unless we are talking valve time / 3DR take on this resulting in a teenage game.
The human eye will accept any performance that's around about 12fps as movement. 24 fps is acceptible (used in film) and anything over 30fps is usually perceived as smooth.
If you're getting above 24 fps, quit complaining. It will get better. Or, if you cant wait, buy a faster computer. If you cant afford to buy a faster computer, quit complaining. You knew it was a game that was going to be in development for a long time, like NS1 was.
In other news, "Integrate stack based C++ objects into Lua" marks the beginning of Max's development of a complete interface between Lua and C++. This corresponds to the opening of the 6th seal and prophecy states that only 1 Mayan-calendar long count year remains until the 7th seal is opened, upon which the entire game will be re-coded in C++.
The human eye will accept any performance that's around about 12fps as movement. 24 fps is acceptible (used in film) and anything over 30fps is usually perceived as smooth.
If you're getting above 24 fps, quit complaining. It will get better. Or, if you cant wait, buy a faster computer. If you cant afford to buy a faster computer, quit complaining. You knew it was a game that was going to be in development for a long time, like NS1 was.<!--QuoteEnd--></div><!--QuoteEEnd-->
This is not entirely true. Film, you see, uses motion blur, which smooths out the frames. You need many more frames per second with video games than with film to see it 'smooth'.
<a href="http://www.100fps.com/how_many_frames_can_humans_see.htm" target="_blank">Reference.</a>
juice: you lost me.
why do some ppl here still argue where the priority is on atm? everythings on priority as there are to-do's in different aspects of the game, which are divided amongst the team according to their profession.
getting tired of all the whine about "pleez get my probs fixed at first i have no clue what exactly you are doin in san franziscoez but u gotta halp me furst"
I understand that you must first get all your pieces lined up before putting your product together (I am an engineer), but I feel like they would better utilize the community if more of us were able to partake in addressing bigger issues. Believe it or not, no matter what you think, the balance and gameplay aspects of the game will be resolved, if ever, much after the game has been optimized to run as it should.
In a different note, I am glad that Cory popped in and expressed his situation. I feel like they sympathize with some of us. Now let's hope if something gets done sooner than later. :)
I Have a few questions about the performance of the current build, my computer can run some powerful games and yet when i start a server or join a game the FPS is terrible.
Now i don't want to read 98 pages so excuse me if it has already been mentioned in this thread but is there a reason the performance dips so low, i was looking around and realized that the maps use alot of realtime lights and i recently found out that these if used improperly and too often can really lower the performance of the levels.
I'm not sure how the Spark engine handles the lights so i could be completely wrong but the engine that i use at work requires a little bit more thought when the level designer/level artist are placing the lights. we use a combination of baked lighting and real-time lighting to help performance.
I would really love to know why the performance is so bad and if it has something to do with the lighting or if there is another issue occuring that could be a problem.
The reason im so interested is that i would really like to make a custom map but i would like to know some of the limitations when creating one.
Cheers,
Arman
"What, like we haven't been already for the past two months?"
The answer is that the tasks are split. Artists don't help with the optimisation, game coders don't help with the optimisation (well they can, but it's limited, and Max has stated that he wants to write a system to eliminate this need), as it is out of their area of expertise and knowledge. The engine/technical coder (Max) does do the optimisation and lately it's pretty much all he does. The only other two things from memory that he's done recently is add decals and add alien vision mode.
My post wasn't an order or a request, just my personal (and noobish) point of view on the perf situation :)