<!--quoteo(post=1797017:date=Aug 31 2010, 11:27 AM:name=Asraniel)--><div class='quotetop'>QUOTE (Asraniel @ Aug 31 2010, 11:27 AM) <a href="index.php?act=findpost&pid=1797017"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->bringing the bullet physics engine up to date. will it replace physx? if i understand right physx was part of the lag problems?
it's actually great news. bullet can be accelerated by opencl, which is cross platform and cross GPU. :)<!--QuoteEnd--></div><!--QuoteEEnd-->
By up-to-date do you mean bullet physics like drop and travel time? Because I really don't care about those, and the distances are so short it's negligible anyways.
Heh I was confused for a moment too... but then I realized. <a href="http://bulletphysics.org/wordpress/" target="_blank">http://bulletphysics.org/wordpress/</a>
<!--quoteo(post=1797005:date=Aug 31 2010, 08:47 AM:name=FocusedWolf)--><div class='quotetop'>QUOTE (FocusedWolf @ Aug 31 2010, 08:47 AM) <a href="index.php?act=findpost&pid=1797005"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I thought about what you and others have been saying for a while now, and came to the conclusion that the majority of people here expect a playable, even if only a few features are implemented at the very start, <b>game</b> for the alpha? Otherwise it feels to much like (unpaid) work to methodically test the games features in single player like the actual developers? Think i'll have to agree then that if the netcode is what is holding people back from <u>playing</u> (as opposed to hardcore-boring testing) the alpha, then that definitely should be the priority of the devs. I mean it will result in more people observing the game while its running, and that means more bug reports overall.<!--QuoteEnd--></div><!--QuoteEEnd-->
It's not about a distinction between work and fun. They have internal playtesters for making sure that a feature, in the broadest sense, "works" without something that actually breaks the engine happening. Maybe some of the more elusive gamebreakers will only be revealed on larger-scale testing, but even so, having more features available for the public to test in individual, singleplayer scenarios really wouldn't help with that regardless.
The purpose of larger-scale multiplayer testing isn't just to make sure that individual features work on some basic level. It's to make sure that they work well and seamlessly in a multiplayer environment, with all the nuances of latency mechanics, etc., factored into that. If the latency mechanics don't even exist in any finalized form because clientside prediction has everyone snapping around all over the place, the point of multiplayer testing is moot.
<!--quoteo(post=1797017:date=Aug 31 2010, 01:27 PM:name=Asraniel)--><div class='quotetop'>QUOTE (Asraniel @ Aug 31 2010, 01:27 PM) <a href="index.php?act=findpost&pid=1797017"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->bringing the bullet physics engine up to date. will it replace physx? if i understand right physx was part of the lag problems?
it's actually great news. bullet can be accelerated by opencl, which is cross platform and cross GPU. :)<!--QuoteEnd--></div><!--QuoteEEnd-->
I don't think Bullet is replacing anything. Nvidia PhysX just runs in software if you don't have an Nvidia card, so I would imagine Bullet is being fostered for use when hardware-accelerated PhysX is not available.
<!--quoteo(post=1797052:date=Aug 31 2010, 08:09 PM:name=scorpydude)--><div class='quotetop'>QUOTE (scorpydude @ Aug 31 2010, 08:09 PM) <a href="index.php?act=findpost&pid=1797052"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->What do you guys think this means on the progress page?
"Build simplified network test game mode"<!--QuoteEnd--></div><!--QuoteEEnd-->
Shell of a game with collision/hit detection and network only, I'd guess. Lets you test that you can run around, collide with stuff and hit stuff, and clientside prediction isn't causing a lot of snapping. Lets them isolate core network issues from other things that might be going wrong.
remiremedy [blu.knight]Join Date: 2003-11-18Member: 23112Members, Super Administrators, Forum Admins, NS2 Developer, NS2 Playtester
<!--quoteo(post=1797201:date=Sep 1 2010, 06:33 PM:name=wackybrak)--><div class='quotetop'>QUOTE (wackybrak @ Sep 1 2010, 06:33 PM) <a href="index.php?act=findpost&pid=1797201"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->looks like the netcode is a ###### to fix<!--QuoteEnd--></div><!--QuoteEEnd--> Yeah, it really does look like that one is a bit of a puzzle for them. It's a bad sign when they need to build new tests to diagnose a problem, but on the other hand it will also probably help them to fix any other network problems that crop up, as well as allow them to have a way to keep track of network performance so they know if adding a new gun impacted the latency.
Hopefully Max will figure it out soon; I have faith he has the ability to. I can't wait to play some silky smooth NS2!
<!--quoteo(post=1797228:date=Sep 2 2010, 07:10 AM:name=saltybp53)--><div class='quotetop'>QUOTE (saltybp53 @ Sep 2 2010, 07:10 AM) <a href="index.php?act=findpost&pid=1797228"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Me dissapoint for no jetpacks D: but we got bots now :D but there gonna be buggy D: but we still get onos sized patch :D<!--QuoteEnd--></div><!--QuoteEEnd-->
Yeah, and we're not getting the patch until the netcode is working properly, which means when we get it, no more lag. On top of that, Charlie is just freakin cruising along adding gameplay and features in, so the difference between 151 and the next patch is going to be insane. It might even almost feel and play like a game. I already get good enough FPS, so I'm set.
From the way that the engine and editor feel i am thinking the original operating specs are not going to be realistic for the final release. It's not video that kills me, It's a tached out AMD dual core thats bogging me down.
<!--quoteo(post=1797234:date=Sep 2 2010, 08:53 AM:name=SN.Wolf)--><div class='quotetop'>QUOTE (SN.Wolf @ Sep 2 2010, 08:53 AM) <a href="index.php?act=findpost&pid=1797234"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->@ Avalon - What are the specs on your set-up?
From the way that the engine and editor feel i am thinking the original operating specs are not going to be realistic for the final release. It's not video that kills me, It's a tached out AMD dual core thats bogging me down.<!--QuoteEnd--></div><!--QuoteEEnd-->
Agreed on the original specs not being realistic. I had said that a while back and got some flack for it, but meh. The game will still be playable on a wide range of hardware.
My system isn't anything too special...running a Core2 E8400 overclocked to 4Ghz, 4GB RAM overclocked, a 512MB Radeon 4850 overclocked, a 1TB Western Digital black edition drive as my primary boot drive, and Win7 64bit. I run the game at native resolution of 1680x1050, high details, and get about 25-40fps. I really think it's the overclocking that helps, because I notice people with much better CPUs and video cards getting FPS in the teens at similar resolutions. Not sure if the game is optimized for multi core yet.
I could investigate the effects of clock speed on the game's FPS with FRAPs, if anyone wants?
Please add the definition (and order) of each status to the actual <a href="http://www.unknownworlds.com/ns2/progress" target="_blank">progress page</a>. Also, please add an easy-to-find link to the progress page from the forums (or preferably the <a href="http://www.unknownworlds.com/ns2/" target="_blank">main page</a>).
(Below is my understanding of the statuses, based on info in this thread, and from the <a href="http://www.pivotaltracker.com/help" target="_blank">PT help FAQ</a>.)
NS2 progress statuses: 1. Started - Development has started, but work is not yet been completed. 2. Finished - Development work completed, but has not yet been deployed to internal test environment. 3. Delivered - Feature has been deployed to internal test environment, but testing is not yet complete. 4. Accepted - Testing is complete, and the feature is working. The next public release will contain the feature.
Other statuses (that we do not see): Unstarted - Feature has been accepted for work, but development has not yet started. Rejected - Testing is complete, and the feature is not working. The next public release will not contain the feature.
<!--quoteo(post=1797217:date=Sep 2 2010, 01:20 AM:name=Psyke)--><div class='quotetop'>QUOTE (Psyke @ Sep 2 2010, 01:20 AM) <a href="index.php?act=findpost&pid=1797217"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Yeah, it really does look like that one is a bit of a puzzle for them. It's a bad sign when they need to build new tests to diagnose a problem, but on the other hand it will also probably help them to fix any other network problems that crop up, as well as allow them to have a way to keep track of network performance so they know if adding a new gun impacted the latency.
Hopefully Max will figure it out soon; I have faith he has the ability to. I can't wait to play some silky smooth NS2!<!--QuoteEnd--></div><!--QuoteEEnd-->
I wonder if they could just license the network code from another commercial game. :P
<!--quoteo(post=1797318:date=Sep 2 2010, 08:49 PM:name=FocusedWolf)--><div class='quotetop'>QUOTE (FocusedWolf @ Sep 2 2010, 08:49 PM) <a href="index.php?act=findpost&pid=1797318"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I wonder if they could just license the network code from another commercial game. :P<!--QuoteEnd--></div><!--QuoteEEnd-->
<!--quoteo(post=1797316:date=Sep 2 2010, 01:40 PM:name=Ezekiel)--><div class='quotetop'>QUOTE (Ezekiel @ Sep 2 2010, 01:40 PM) <a href="index.php?act=findpost&pid=1797316"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->oh come on, has been discussed and explained already. anyone with half a brain can work out what they mean.<!--QuoteEnd--></div><!--QuoteEEnd--> The values are not at all obvious if you don't have any context. This is exactly the scenario that keys are designed for.
Obviously, it is <i>possible</i> to figure their meanings out. But, don't assume everyone wants to spend their time wading through forum threads.
it is patently obvious. it has been discussed repeatedly. no wading, or even going into the forums needed to work out what they mean, a tiny bit of actual thinking.
<!--quoteo(post=1797318:date=Sep 2 2010, 02:49 PM:name=FocusedWolf)--><div class='quotetop'>QUOTE (FocusedWolf @ Sep 2 2010, 02:49 PM) <a href="index.php?act=findpost&pid=1797318"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->I wonder if they could just license the network code from another commercial game. :P<!--QuoteEnd--></div><!--QuoteEEnd-->
They could have, but if you listen to some of the more recent interviews, you'll notice that they're really running on fumes in terms of budget at this point. Netcode is (relatively) uncomplicated to roll out for yourself, so it wasn't a bad choice.
I know that doesn't sound right, considering all the problems we're having, but tweaking a good balance of clientside prediction when developing netcode is always very touchy. You want good, smooth physics, etc., but you also don't want things snapping into place with server syncs. Anyhow, it sounds like the fix is well under way. I'm sure the next iteration of the netcode will be much smoother, and the one after that, etc.
<!--quoteo(post=1797306:date=Sep 2 2010, 09:01 PM:name=Racer1)--><div class='quotetop'>QUOTE (Racer1 @ Sep 2 2010, 09:01 PM) <a href="index.php?act=findpost&pid=1797306"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->Please add the definition (and order) of each status to the actual <a href="http://www.unknownworlds.com/ns2/progress" target="_blank">progress page</a>. Also, please add an easy-to-find link to the progress page from the forums (or preferably the <a href="http://www.unknownworlds.com/ns2/" target="_blank">main page</a>).
(Below is my understanding of the statuses, based on info in this thread, and from the <a href="http://www.pivotaltracker.com/help" target="_blank">PT help FAQ</a>.)
NS2 progress statuses: 1. Started - Development has started, but work is not yet been completed. 2. Finished - Development work completed, but has not yet been deployed to internal test environment. 3. Delivered - Feature has been deployed to internal test environment, but testing is not yet complete. 4. Accepted - Testing is complete, and the feature is working. The next public release will contain the feature.
Other statuses (that we do not see): Unstarted - Feature has been accepted for work, but development has not yet started. Rejected - Testing is complete, and the feature is not working. The next public release will not contain the feature.<!--QuoteEnd--></div><!--QuoteEEnd-->
There is a link from the main page. News -> Progress
Comments
it's actually great news. bullet can be accelerated by opencl, which is cross platform and cross GPU. :)
it's actually great news. bullet can be accelerated by opencl, which is cross platform and cross GPU. :)<!--QuoteEnd--></div><!--QuoteEEnd-->
By up-to-date do you mean bullet physics like drop and travel time? Because I really don't care about those, and the distances are so short it's negligible anyways.
<a href="http://bulletphysics.org/wordpress/" target="_blank">http://bulletphysics.org/wordpress/</a>
It's not about a distinction between work and fun. They have internal playtesters for making sure that a feature, in the broadest sense, "works" without something that actually breaks the engine happening. Maybe some of the more elusive gamebreakers will only be revealed on larger-scale testing, but even so, having more features available for the public to test in individual, singleplayer scenarios really wouldn't help with that regardless.
The purpose of larger-scale multiplayer testing isn't just to make sure that individual features work on some basic level. It's to make sure that they work well and seamlessly in a multiplayer environment, with all the nuances of latency mechanics, etc., factored into that. If the latency mechanics don't even exist in any finalized form because clientside prediction has everyone snapping around all over the place, the point of multiplayer testing is moot.
<!--quoteo(post=1797017:date=Aug 31 2010, 01:27 PM:name=Asraniel)--><div class='quotetop'>QUOTE (Asraniel @ Aug 31 2010, 01:27 PM) <a href="index.php?act=findpost&pid=1797017"><{POST_SNAPBACK}></a></div><div class='quotemain'><!--quotec-->bringing the bullet physics engine up to date. will it replace physx? if i understand right physx was part of the lag problems?
it's actually great news. bullet can be accelerated by opencl, which is cross platform and cross GPU. :)<!--QuoteEnd--></div><!--QuoteEEnd-->
I don't think Bullet is replacing anything. Nvidia PhysX just runs in software if you don't have an Nvidia card, so I would imagine Bullet is being fostered for use when hardware-accelerated PhysX is not available.
so this sounds to me like the physx implementation had some serious issues and they go back to bullet (they used bullet before).
"Build simplified network test game mode"
"Build simplified network test game mode"<!--QuoteEnd--></div><!--QuoteEEnd-->
Shell of a game with collision/hit detection and network only, I'd guess. Lets you test that you can run around, collide with stuff and hit stuff, and clientside prediction isn't causing a lot of snapping. Lets them isolate core network issues from other things that might be going wrong.
Yeah, it really does look like that one is a bit of a puzzle for them. It's a bad sign when they need to build new tests to diagnose a problem, but on the other hand it will also probably help them to fix any other network problems that crop up, as well as allow them to have a way to keep track of network performance so they know if adding a new gun impacted the latency.
Hopefully Max will figure it out soon; I have faith he has the ability to. I can't wait to play some silky smooth NS2!
Same here.
Yeah, and we're not getting the patch until the netcode is working properly, which means when we get it, no more lag. On top of that, Charlie is just freakin cruising along adding gameplay and features in, so the difference between 151 and the next patch is going to be insane. It might even almost feel and play like a game. I already get good enough FPS, so I'm set.
From the way that the engine and editor feel i am thinking the original operating specs are not going to be realistic for the final release. It's not video that kills me, It's a tached out AMD dual core thats bogging me down.
'We all float down here! And you'll float too!'
o_O
From the way that the engine and editor feel i am thinking the original operating specs are not going to be realistic for the final release. It's not video that kills me, It's a tached out AMD dual core thats bogging me down.<!--QuoteEnd--></div><!--QuoteEEnd-->
Agreed on the original specs not being realistic. I had said that a while back and got some flack for it, but meh. The game will still be playable on a wide range of hardware.
My system isn't anything too special...running a Core2 E8400 overclocked to 4Ghz, 4GB RAM overclocked, a 512MB Radeon 4850 overclocked, a 1TB Western Digital black edition drive as my primary boot drive, and Win7 64bit. I run the game at native resolution of 1680x1050, high details, and get about 25-40fps. I really think it's the overclocking that helps, because I notice people with much better CPUs and video cards getting FPS in the teens at similar resolutions. Not sure if the game is optimized for multi core yet.
I could investigate the effects of clock speed on the game's FPS with FRAPs, if anyone wants?
(Below is my understanding of the statuses, based on info in this thread, and from the <a href="http://www.pivotaltracker.com/help" target="_blank">PT help FAQ</a>.)
NS2 progress statuses:
1. Started - Development has started, but work is not yet been completed.
2. Finished - Development work completed, but has not yet been deployed to internal test environment.
3. Delivered - Feature has been deployed to internal test environment, but testing is not yet complete.
4. Accepted - Testing is complete, and the feature is working. The next public release will contain the feature.
Other statuses (that we do not see):
Unstarted - Feature has been accepted for work, but development has not yet started.
Rejected - Testing is complete, and the feature is not working. The next public release will not contain the feature.
Hopefully Max will figure it out soon; I have faith he has the ability to. I can't wait to play some silky smooth NS2!<!--QuoteEnd--></div><!--QuoteEEnd-->
I wonder if they could just license the network code from another commercial game. :P
Sigh, I hope this is a joke.
The values are not at all obvious if you don't have any context. This is exactly the scenario that keys are designed for.
Obviously, it is <i>possible</i> to figure their meanings out. But, don't assume everyone wants to spend their time wading through forum threads.
They could have, but if you listen to some of the more recent interviews, you'll notice that they're really running on fumes in terms of budget at this point. Netcode is (relatively) uncomplicated to roll out for yourself, so it wasn't a bad choice.
I know that doesn't sound right, considering all the problems we're having, but tweaking a good balance of clientside prediction when developing netcode is always very touchy. You want good, smooth physics, etc., but you also don't want things snapping into place with server syncs. Anyhow, it sounds like the fix is well under way. I'm sure the next iteration of the netcode will be much smoother, and the one after that, etc.
(Below is my understanding of the statuses, based on info in this thread, and from the <a href="http://www.pivotaltracker.com/help" target="_blank">PT help FAQ</a>.)
NS2 progress statuses:
1. Started - Development has started, but work is not yet been completed.
2. Finished - Development work completed, but has not yet been deployed to internal test environment.
3. Delivered - Feature has been deployed to internal test environment, but testing is not yet complete.
4. Accepted - Testing is complete, and the feature is working. The next public release will contain the feature.
Other statuses (that we do not see):
Unstarted - Feature has been accepted for work, but development has not yet started.
Rejected - Testing is complete, and the feature is not working. The next public release will not contain the feature.<!--QuoteEnd--></div><!--QuoteEEnd-->
There is a link from the main page. News -> Progress