0001131: Client mispredicts its movement if moving on floors/slopes with negative height values

Notes
(0005127) Torr Samaho (administrator) 2012-10-15 20:51	Is the behavior in 1.0 and 98d identical if 1.0 is used with cl_ticsperupdate 3?

(0005129) unknownna (updater) 2012-10-15 22:23 edited on: 2012-10-16 03:43	Quote from Torr Samaho Is the behavior in 1.0 and 98d identical if 1.0 is used with cl_ticsperupdate 3? When compared side by side, both with an emulated ping of 302, they seem to be nearly identical to one another, but I'm not too sure. I also noticed while testing IGPack that bridge things seem to be affected by this issue as well. If the floor/slope beneath them has a negative height, the client will start to jitter when jumping on the bridge things. I created an example WAD. Only the leftmost platform with a negative floor height will cause the client to mispredict its movement. Edit: It seems that the client will mispredict its movement on the rightmost platform as well when jumping on the bridge things.

(0005142) Torr Samaho (administrator) 2012-10-20 19:25 edited on: 2012-10-20 19:41	I've got the impression that the jumping on the bridge things in your example wad got worse with the recent changes for rocket jumping on bridge things. Can you confirm this? Do the other effects only appear if slopes are involved and the floor heights are negative? EDIT: The new jumping problems where possibly caused by another change that was not in the last binary I sent you.

(0005148) unknownna (updater) 2012-10-21 01:39 edited on: 2012-10-21 02:04	Quote from Torr Samaho I've got the impression that the jumping on the bridge things in your example wad got worse with the recent changes for rocket jumping on bridge things. Can you confirm this? I can't confirm that it got worse, but I can confirm that the client mispredicts its movement in 1.0. Quote from Torr Samaho Do the other effects only appear if slopes are involved and the floor heights are negative? It also jitters on slopes when the heights are positive, but it's a lot worse if they're negative. In slope_jitter_01b.wad, the heights are positive and it doesn't jitter nearly as much as it does in slope_jitter_01.wad when compared side by side. Quote from Torr Samaho The new jumping problems where possibly caused by another change that was not in the last binary I sent you. I noticed in this binary that the client now jitters when moving from real floors to bridge things.

(0005152) Torr Samaho (administrator) 2012-10-21 17:10 edited on: 2012-10-21 17:10	Quote from unknownna I noticed in this binary that the client now jitters when moving from real floors to bridge things. That's bad, can you make a demo of this? As predicted we seem to have come to a point where any changes to the prediction break things that used to work. I may have to revert the rocket jump on bridge things changes then.

(0005156) unknownna (updater) 2012-10-21 17:44 edited on: 2012-10-21 18:00	Quote from Torr Samaho That's bad, can you make a demo of this? I recorded a demo with this binary. zandronum.exe -file player_bridge_desync_test_04.wad -playdemo 2012.10.21_19.41.31_player_bridge_desync_test_04.cld player_bridge_desync_test_04.wad Quote from Torr Samaho As predicted we seem to have come to a point where any changes to the prediction break things that used to work. I think the new fixes brought fundamentally broken bridge behavior to the surface more than anything. Quote from Torr Samaho I may have to revert the rocket jump on bridge things changes then. I don't like this attitude, but it's your call.

(0005160) Torr Samaho (administrator) 2012-10-21 18:29	Quote from unknownna I recorded a demo with this binary. Is this any better? Quote from unknownna I don't like this attitude, but it's your call. I spent many hours during the last weeks on the prediction stuff and if I see that with every fixed issue a new issue pops up, I'm questioning myself if all the time was well spent. If the "effort / impact" ratio is too bad, it's not worth to do something as long as there are other things to do. I'm sure that the prediction issues can be fixed, but I have no idea how much work it'll take and I also don't know what the tipping point is.

(0005162) unknownna (updater) 2012-10-21 18:59	Quote from Torr Samaho Is this any better? Yes, that took care of it. Quote from Torr Samaho If the "effort / impact" ratio is too bad, it's not worth to do something as long as there are other things to do. Considering that these issues were never dealt with until now, I consider the ratio to be very good. So far we fixed all the known side effects but one.

(0005167) unknownna (updater) 2012-10-21 19:57	Quote from Torr Samaho I've got the impression that the jumping on the bridge things in your example wad got worse with the recent changes for rocket jumping on bridge things. Can you confirm this? In this binary I can confirm that it got a bit worse when moving/jumping on the bridge things above the slopes in slope_jitter_02.wad. But I still feel that we only see the symptoms of fundamental issues since they were there before as well.

(0005196) unknownna (updater) 2012-10-25 01:24 edited on: 2015-06-07 01:26	Quote from Torr Samaho Do the other effects only appear if slopes are involved and the floor heights are negative? ~~It seems that they appear on regular floors as well, but it got a bit worse in the newer binaries. I recorded a small demo of the jitter. We have to take care of this issue.~~ zandronum -playdemo 2012.10.25_03.20.23_doom2.cld Edit: This is strange. For some reason I'm unable to reproduce the jitter on regular floors now.

(0012575) unknownna (updater) 2015-06-07 01:25 edited on: 2015-06-07 01:37	Edward-san, do you have any idea why this behavior is present? It's weird that the jitter is affected by the floor height. What's also weird is that moving and jumping on bridge things above slopes is affected by the issue, as I noticed in IGPack (and demonstrated in slope_jitter_02.wad and now in slope_jitter_03.wad).

(0012581) Edward-san (developer) 2015-06-07 16:36 edited on: 2015-06-07 16:36	Quote from unknownna Edward-san, do you have any idea why this behavior is present? It's weird that the jitter is affected by the floor height. Sorry, can you make a demo with zan 2.0? I don't see any jittery going on in slope_jitter_01 and slope_jitter_01b, even with emulated ping of 302..

(0012582) unknownna (updater) 2015-06-07 16:52 edited on: 2015-06-07 16:55	Sure, check it out. It apparently starts to jitter when moving from slopes to floors and vice versa. zandronum -file slope_jitter_01.wad -playdemo jitter_01.cld Edit: And here's a demo of the bridge thing issue zandronum -file slope_jitter_03.wad -playdemo jitter_02.cld

(0012583) Edward-san (developer) 2015-06-07 17:11	Sorry, can you redo both the demos with this latest 2.1-beta build:'https://www.dropbox.com/s/76gxuvlsq9yhb9k/zandronum-2.1-150604-1440-9b91f0c-windows.zip?dl=0 [^]' ?

(0012584) unknownna (updater) 2015-06-07 17:24	Sure, here you go: zandronum -file slope_jitter_01.wad -playdemo jitter_01b.cld zandronum -file slope_jitter_03.wad -playdemo jitter_02b.cld

Only registered users can voice their support. Click here to register, or here to log in.
Supporters:	unknownna Dark-Assassin
Opponents:	No one explicitly opposes this issue yet.

Issue History
Date Modified	Username	Field	Change
2012-10-15 15:16	unknownna	New Issue
2012-10-15 15:16	unknownna	File Added: slope_jitter_01.wad
2012-10-15 15:16	unknownna	File Added: 2012.10.15_17.12.25_slope_jitter_01.cld
2012-10-15 15:16	unknownna	Status	new => confirmed
2012-10-15 15:34	unknownna	Severity	minor => major
2012-10-15 20:51	Torr Samaho	Note Added: 0005127
2012-10-15 22:22	unknownna	File Added: slope_jitter_02.wad
2012-10-15 22:23	unknownna	Note Added: 0005129
2012-10-16 00:37	unknownna	Note Edited: 0005129	View Revisions
2012-10-16 03:41	unknownna	Note Edited: 0005129	View Revisions
2012-10-16 03:43	unknownna	Note Edited: 0005129	View Revisions
2012-10-20 19:25	Torr Samaho	Note Added: 0005142
2012-10-20 19:41	Torr Samaho	Note Edited: 0005142	View Revisions
2012-10-21 01:39	unknownna	Note Added: 0005148
2012-10-21 01:40	unknownna	File Added: slope_jitter_01b.wad
2012-10-21 02:04	unknownna	Note Edited: 0005148	View Revisions
2012-10-21 17:10	Torr Samaho	Note Added: 0005152
2012-10-21 17:10	Torr Samaho	Note Edited: 0005152	View Revisions
2012-10-21 17:11	Torr Samaho	Note Revision Dropped: 5152: 0002815
2012-10-21 17:44	unknownna	File Added: 2012.10.21_19.41.31_player_bridge_desync_test_04.cld
2012-10-21 17:44	unknownna	Note Added: 0005156
2012-10-21 18:00	unknownna	Note Edited: 0005156	View Revisions
2012-10-21 18:29	Torr Samaho	Note Added: 0005160
2012-10-21 18:59	unknownna	Note Added: 0005162
2012-10-21 19:57	unknownna	Note Added: 0005167
2012-10-25 01:24	unknownna	Note Added: 0005196
2012-10-25 01:25	unknownna	File Added: 2012.10.25_03.20.23_doom2.cld
2012-10-25 05:51	unknownna	Note Edited: 0005196	View Revisions
2015-06-07 01:25	unknownna	Note Added: 0012575
2015-06-07 01:25	unknownna	File Added: slope_jitter_03.wad
2015-06-07 01:26	unknownna	Status	confirmed => feedback
2015-06-07 01:26	unknownna	Note Edited: 0005196	View Revisions
2015-06-07 01:37	unknownna	Note Edited: 0012575	View Revisions
2015-06-07 16:36	Edward-san	Note Added: 0012581
2015-06-07 16:36	Edward-san	Note Edited: 0012581	View Revisions
2015-06-07 16:52	unknownna	Note Added: 0012582
2015-06-07 16:52	unknownna	Status	feedback => new
2015-06-07 16:52	unknownna	File Added: jitter_01.cld
2015-06-07 16:55	unknownna	Note Edited: 0012582	View Revisions
2015-06-07 16:56	unknownna	File Added: jitter_02.cld
2015-06-07 17:11	Edward-san	Note Added: 0012583
2015-06-07 17:11	Edward-san	Status	new => feedback
2015-06-07 17:24	unknownna	Note Added: 0012584
2015-06-07 17:24	unknownna	Status	feedback => new
2015-06-07 17:24	unknownna	File Added: jitter_01b.cld
2015-06-07 17:24	unknownna	File Added: jitter_02b.cld
2018-08-23 05:09	unknownna	Relationship added	related to 0002294