Initial Beta Thoughts
morphman
Join Date: 2010-11-29 Member: 75341Members
Join Date: 2010-11-29 Member: 75341Members
So I purchased the game to test out the beta a few weeks ago and was finally able to play consistently after the last patch that went in before Thanksgiving. First off I want to say great work guys, NS1 was one of my all time favorite games and I'm really looking forward to NS2.
My observations thus far is that spending more time on optimizations so that the game is playable in > 6 player games would be the best bang for our buck. Adding more features wouldn't even matter as unplayable as the game is at the moment in terms of lag and performance. Anyway, that's just my 2 cents.
My observations thus far is that spending more time on optimizations so that the game is playable in > 6 player games would be the best bang for our buck. Adding more features wouldn't even matter as unplayable as the game is at the moment in terms of lag and performance. Anyway, that's just my 2 cents.
Comments
My observations thus far is that spending more time on optimizations so that the game is playable in > 6 player games would be the best bang for our buck. Adding more features wouldn't even matter as unplayable as the game is at the moment in terms of lag and performance. Anyway, that's just my 2 cents.<!--QuoteEnd--></div><!--QuoteEEnd-->
I'm not sure I agree completely. At least for me the game is running (albeit barely) sufficiently well up to 10 players, as long as there's not too much turret spam. My biggest hindrance to fun atm is the flawed Marine tier 2 dynamic, where the game is decided at the point where Marines get to tier 2, or when Aliens manage to prevent that, which happens in the first 5 minutes. After that it's getting dull.
So not only the technical side of the game needs focus right now, the game play side needs it too. Of course that's already known and UWE approaches it in the right way.
Optimising existing features doesn't affect your ability to add new ones, and the exisitng ones will still need optimising whether you do it now or later.
it's not so much about optimizing "features" as it is about optimizing how data is pathed through the entire game engine pipeline. so optimizing individual pieces, to the extent that is even possible, before the entire system/pipeline/pathway is complete is basically a waste of time aside from the PR factor since it will be changing by definition as new systems are added mandating new data transmission pathways.
I like how we are both using the term "optimization" totally absent of any significant meaning. It's really a problem with the OP in this thread posing a false dichotomy between completeness and how well the net code works. again, i'm pretty sure UW is prioritizing fixing the net code in order to make you guys happy, so I woldn't worry about it personally.
I like how we are both using the term "optimization" totally absent of any significant meaning. It's really a problem with the OP in this thread posing a false dichotomy between completeness and how well the net code works. again, i'm pretty sure UW is prioritizing fixing the net code in order to make you guys happy, so I woldn't worry about it personally.<!--QuoteEnd--></div><!--QuoteEEnd-->
If they add new models and maps and some new code to control them, it's all still going to run through the same renderer and netcode and run on the same server.
If running a server is eating massive amounts of CPU time, it needs fixing, and if you fix it now it will stay fixed more or less later. You might need to fix each new feature you add a bit, but unless you're going to largely remake the stuff you already fixed, it will stay fixed.
By optimise I mean 'change something from a very inefficient system which provides the end result, to a much more efficient system that provides the same end result.
Like you could do calculations every fram to see if something is happening and then act on it, or you could transmit a signal when it happens and remove the check per frame, freeing up resources, both achieve the same end which is that you know when something happens, but one is much more efficient, more optimal, more optimised.
this should hopefully lower CPU usage and give some more fps :)
Anyone who thinks otherwise doesn't know how to develop large software projects efficiently.
this should hopefully lower CPU usage and give some more fps :)<!--QuoteEnd--></div><!--QuoteEEnd-->
It did :)
Anyone who thinks otherwise doesn't know how to develop large software projects efficiently.<!--QuoteEnd--></div><!--QuoteEEnd-->
I agree with you wholeheartedly. Where we might disagree is the subjective notion of when is "premature". Premature in my mind means trying to make "optimizations" without really knowing that there is a problem. However, I don't think it is premature to work on performance when you know that a good chunk of your users consider it unplayable because of performance. Just 2 more of my cents. That and 75 more cents will buy you a cup of coffee.
what do you mean? are you a playtester?
i don't think so tim. some of the flash is still there in 156
"Replace weapon displays with Lua GUIs"
Puzl's quote pretty much says it all, but I feel the need to point out that "models and maps" are not "features." Features are gameplay elements, which constitute new information that needs to be transmitted between servers and players, which again is relevant to the data transmission pipeline which will be constantly in flux until the game is feature-complete at which point optimization of the entire system will be most effective/efficient. Duplicating work sucks, i.e. spending days fixing something only to have it broken again when it becomes dependent on a newer system or is altered to facilitate other game elements.
Because I thought it was common knowledge... don't know how exactly I got the idea.
Pandora want's his box back....
To an extent.
You get to a point where optimization needs to be done to make the game more playable, as its important to have a more stable experience sooner rather than later and a horrible position to be in would be spending the last few weeks crunching on performance before the publication date of the game.
I think they are at the point where its important to be working on performance as a major priority; which is what I think they're doing now, in accordance with other engine features and of course gameplay/mapping/assets etc, its definitely on the radar.
false dichotomy
it's like the machinery of a car made at the factory, and the interior made by Pimp My Ride...
so optimizing by exterminating flash, improving lighting rendering, adding inferred rendering,
optimizing <i>player</i>-related netcode, all that will stay as is.
maybe some specific entity related netcode (hydra, sentry) might need to be changed, but i think the devs are fully competent in handling this without their fans second-guessing them all the time!
No second-guessing here! Merely stating that if I, as a user, got to pick between new game mechanics or the ability to play the existing game with less "lag" I would surely pick "less lag".
and i just said, they can work on performance AND features pretty much independently
157 will deal mostly with client hitchups and balance it seems (and hopefully a little performance)
158 i can imagine will be performance oriented on the technical side, and LUA features rolling on as they come