Notes |
|
|
I don't see any floor waggling in the example wad. |
|
|
|
It's the red side. I just used a slow waggle to make it easier to jump on. |
|
|
|
Uploaded a new test wad with a stronger waggle although that was not needed at all but since people seemed to not care once again.. |
|
|
|
What do you mean with "since people seemed to not care once again"? |
|
|
|
"since people seemed to not care once again"
"Supporters: ibm5155"
'-'
I can't help, the only thing I know for now is copy and paste commits from zdoom to zan and fix some simple problems :B |
|
|
(0008538)
|
Arco
|
2014-04-11 02:12
|
|
I cannot reproduce the jittery lag as written in the ticket. Are you doing something else that may be doing this? |
|
|
|
I am sorry that it took me so long to respond.
I re-tested it and apparently it doesn't happen as much as it did before.
But it's still there. I suggest hosting a server with the second test-wad I made (It seems to be more likely on this one) and connecting to it with gamersproxy with just an emulated ping of 100 and then simply holding +jump on one of the waggling floors. Within 10 seconds at least I get the desync. |
|
|
|
Uploaded an even better testing wad (WaggleTest3.wad)
Just load the map on a server and keep pressing +jump and the effect immediately happens.
Switch between an emulated 0/200 ping on your local server and you will notice the difference and it's even more noticeable with chasecam on.
Also it seems that this is caused by the client's movement prediction because when setting cl_predict_players to 0, the effect disappears completely. |
|
|
|
Just thought I'd post to confirm this, numerous people have complained about very jittery waggle floors when ping is more than 100 or so, particularly problematic on jumpmaze. |
|
|
|
I can confirm that WaggleTest3.wad with an emulated ping of 200 allows to easily reproduce this effect. |
|
|
|
I had a look and I'm pretty sure I know what's going wrong. While predicting, the client assumes that sector heights do not change during the prediction window. There are some additions that take care of moving floors, but these workarounds cannot cover the case where a moving floor changes its direction during the prediction window.
So this is more of a missing prediction feature than a bug. Fixing requires to take into account the effect of the waggle during the prediction. Unfortunately, this is not trivial. |
|