Stuttering 1.1 problems What is going on at ED??

[GGTharos]

Okay, let's concentrate on the client-side rowboat effect, eh? That's what started all this if I recall correctly - not lag effects, but the entire screen actually freezing for a fraction of a second, very regularely, on client machines.

 

Lag issues will be lag issues.

[Prophet_169th]

Thats what I was referring to with the AWACS transmissions. That causes the pausing. everytime. Though I havent seen anything else yet cause it.

[coldcrew]

you're wrong. The rowboat effect doesn't just appear on the client. I've been "stutter" free since this issue was brought up, but I experienced the rowboat effect while hosting for about a minute and then it went away.

 

I had completed the mission and about halfway through the final landing I saw the rowboat effect. I asked the other player if he was experiencing it too and he said no. There was another player who was trying to join at that time (you know that MP bug where you join a game and you get no plane? and you have to eject to get one?).

 

Anyways after I completed the landing and went to parking I ejected, got a new plane and the effect went away. Never experienced it again even when landing on the same mission.

[zeufman]

Okay, let's concentrate on the client-side rowboat effect, eh? That's what started all this if I recall correctly - not lag effects, but the entire screen actually freezing for a fraction of a second, very regularely, on client machines

 

you're right, we are talking here about sutter with multiplayer client.

We know all what is lag or less ping, and this not the same effect.

[Prophet_169th]

The lag issue I am speaking of, isnt necessarily lag. Even with 100ms ping, the effect is seen. So I am not sure if this is the stuttering effect. Or if they are seperate issues. The only time I have seen 'stutter', like the game pausing kinda, is when AWACS transmitts.

 

And I use enditall2, so there is nothing besides whats needed to play. OS stuff, trackir software, Saitek software, soundcard software.

 

Could it be a sound issue? You guys updated some of the sounds, did you also update the code in which the sounds were deployed/processed?

[Kula66]

Is it lag when sounds actually load? Do you get it with other voice calls? Are you tight on RAM? Do you see disk activity at the same time?

[Prophet_169th]

If you are asking me..........

 

only stutters with AWACS transmittion. I use the tower all the time, and nothing.

 

No issues with RAM, 1GB, and the most I have seen LOMAC use is about 650MB.

 

And no disk activity all the time.

[LawnDart]

Sounds like you guys are talking about a different kind of issue. Look at the PS screenshot in this thread (reconstruction of the stuttering) a few pages back on p12. This is as close as anything to the airplane stuttering we see in multiplayer. It's different from lag, warping, sound loading, graphics (terrain stutter). It's the airplane model jiggy! ...and for us, this particular thing only happens as client in a multiplayer game and PinPeriod (config.lua file) does has a big influence on how much/often it gets that jiggy!

 

LawnDart

[GGTharos]

Okay, what this issue is not:

 

-Lag

-Warping

-The AWACS thing (but it needs to be addressed as well)

 

What this thing IS:

 

Very regular micropauses of the stuff displayed on monitor apparently somehow related to pingperiod, and nly seems ot happen on client machines, and only in MP.

These do not cause warping for the most part.

 

If you have a DIFFERENT stutter issue please don't confuse this thread with it.

[LawnDart]

Nice summary GGTharos!

 

These stutters don't slow the game down or pause the whole simulation, it just targets the other aircraft/objects. I can have high framerates, no lag, no warping, no sudden game freezes (sound or graphics hickups) and still get this micro-stuttering of the other planes as client...

[GGTharos]

Ahh, ok Lawndart. So your stuttering IS different than the one I described. It's important you see 'cause if we don't make this very clear we will onyl confuse the devs.

 

So for you, the landscape doesn't freeze? It does for me.

[coldcrew]

just bc someone said it never happened while hosting doesn't mean it doesn't. ED needs to have a general jist of the problem to be able to locate it in code. Eliminating symptoms people are reporting does no good when trying to find bugs.. maybe that's why the game was released buggy in the first place. Besides all this is moot if ED doesn't have developers testing it out and trying to find it in the code.

[LawnDart]

Ahh, ok Lawndart. So your stuttering IS different than the one I described. It's important you see 'cause if we don't make this very clear we will onyl confuse the devs.

 

So for you, the landscape doesn't freeze? It does for me.

 

Ok, my landscape stutters too, however that seems to me to be a different cause and not at the same time interval as the plane stutter. Changing the pingPeriod only impacts the plane stuttering for us. I'm not ruling out that the two could be combined though! It's the object stuttering that is really annoying and if I had to describe it I'd say:

 

Plane stutter looks like something caused by a background operation or coding... (Although turning off ALL background ops using "EndItAll" didn't eliminate the stutter). The landscape looks more like a rendering issue if you ask me. Burner can probably tell you more or give you a better description, since he's flown mostly as client with me hosting during our tests!

 

LawnDart

[Burner_Tbird2]

GGTharos,

 

Good summary, I would like to add to the symptoms that it seems that when ping period is closer to 3 the overall studder is less but the host plane seems to jerk around more (jerk around = random jerkiness that looks close to lag but is not at any regular interval at all and therefore not coinciding with the studder). The closer it is to 1 the more frequent and noticable the studder becomes and the host plane does not jerk at all rather the host plane is part of the studder like the ground.

 

I'd say the host is always part of the studder but when PP is at 3 its hardly noticable compaired with the jerkiness.

 

PS- this is still a problem, any word from the Devs on this issue, tell me it'll be addressed post haste in the next patch and I'll shutup

[Slammin]

Slammin, are you using an intel chip? Is your hyperthreading on?

 

This is all I have to go on right now - it certianly looks like there's a thread blocking the video rendering every now and then, associated with the ping ... so I don't think it's network flow being affected, but somehow either the bus or osmething else is being temporarily - and unintentionally - kidnapped.

 

Okay, can't be the bus, no warping ... but maybe just the AGP, for some silly reason?

 

 

 

Intel HT=ON here. But I do think this is tied to the local system, specifically, the hosting system. Didn't someone say that the net code has been changed from synchronous to asynchronous in 1.1? This to me would mean that the host machine has a higher workload now since it has to be able to service clients simultaneously instead of one at a time.

 

Question: I know Ice runs a standalone server, and he says the problem does not exist there. Could this be something that just affects non-stand alone servers? I have a stand-alone server and can put it up anytime if someone wants to test. Server is a dual P-3 1Ghz, 768mb RDRAM. Fresh 1.1 install but I haven't built any custom missions yet.

[Slammin]

The lag issue I am speaking of, isnt necessarily lag. Even with 100ms ping, the effect is seen. So I am not sure if this is the stuttering effect. Or if they are seperate issues. The only time I have seen 'stutter', like the game pausing kinda, is when AWACS transmitts.

 

And I use enditall2, so there is nothing besides whats needed to play. OS stuff, trackir software, Saitek software, soundcard software.

 

Could it be a sound issue? You guys updated some of the sounds, did you also update the code in which the sounds were deployed/processed?

 

 

Most likely just a sound issue. In IL-2, you would get a nice pause the first time you fired your guns and to avoid it happening in combat, you could either take DX sound Accel down a notch, or just fire a short burst of rounds before taking off when you join a server.

[Burner_Tbird2]

Hey Slammin, I think we'd be willing to test your Ded, I as well as LawnDart will be on tomorrow for a good bit of the day. My MSN is RacingFB@hotmail.com

 

Hit me up if you see me online, and we'll put this theory to the test.

[Slammin]

Server is up. Fresh install (no changes made to default config files), running in windowed mode with lowest possible gfx settings selected. Mission: Cranes and Eagles.

[LawnDart]

Server is up. Fresh install (no changes made to default config files), running in windowed mode with lowest possible gfx settings selected. Mission: Cranes and Eagles.

 

What's the addy to the server (or can you email it to me:lawndart@virtualthunderbirds.com) Also, could you please make a barebones mission with only a few planes (clients) starting on the ramp in it (preferably F-15's) so we can really test this stuttering up close in formation without all hell breaking loose with SAM's and engagements!?

 

LawnDart

[Slammin]

I just made a new map with only a few planes. Bases are very close together and it is up on HyperLobby. My current IP is 65.69.91.75.

[Burner_Tbird2]

LawnDart and I just tried your Ded Server and got worse studdering than in LawnDart's server. Oh well, thanks for helping us eliminate another possibility on the list.

[LawnDart]

I just made a new map with only a few planes. Bases are very close together and it is up on HyperLobby. My current IP is 65.69.91.75.

 

Well, I'm sorry to say that the stutter is as noticable as ever! Me and Burner joined your server and flew for only a few minutes (that's all it took). I could tell it was stuttering bad even before taking off. Burner had a pingPeriod of 2 and I had the default setting of 1 to test and see the differences if any. Both of us agree that the stutter was really bad, even for him with a pingPeriod of 2 which usually reduces it. For me it ranked right among the worst I've seen so far and totally unflyable in close formation. It's not your fault and a big thanks for putting up the server! Really appreciate it! This is definitely something ED needs to look into however...

 

Just out of curiosity for further testing. You did say this computer ran this server only with no other background ops and set up only to host the game, not actually fly in the mission right!? What were your puter specs again?

 

I also believe one of our other team members joined Ice's server the other night and said he saw stuttering there too...

 

LawnDart

[Slammin]

Yes, the server has nothing else running, is a minimal install of WinXP Pro running on a dual P3 1GHZ box with 768 RDRAM. My DSL line is 6mb down, 608kb up and easily hosts 9 players with 1.02.

 

I flew for a few minutes this morning in close formation and things looked pretty normal to me, but my ping was only 25 too. I'm going to set the server to pingperiod=2 and see what happens.

[Burner_Tbird2]

Changing host PingPeriod seems to do nothing, however, tweeking your upload/download speed does change things, LD used 12000 up and down with good result.

[LawnDart]

 

I flew for a few minutes this morning in close formation and things looked pretty normal to me, but my ping was only 25 too. I'm going to set the server to pingperiod=2 and see what happens.

 

Did you fly as a client on the server? The low ping could have something to do with it for sure, but we had 70-80 pings and got major stuttering. Also, the pingPeriod only has impact on each players own settings, the server could have any other setting in his config.lua file and not change a thing for each client...