Anonymous | Login | Signup for a new account | 2024-03-29 04:39 UTC |
My View | View Issues | Change Log | Roadmap | Zandronum Issue Support Ranking | Rules | My Account |
View Issue Details [ Jump to Notes ] | [ Issue History ] [ Print ] | ||||||||
ID | Project | Category | View Status | Date Submitted | Last Update | ||||
0003787 | Zandronum | [All Projects] Bug | public | 2020-05-05 23:18 | 2022-02-24 05:09 | ||||
Reporter | mifu | ||||||||
Assigned To | |||||||||
Priority | high | Severity | major | Reproducibility | random | ||||
Status | closed | Resolution | fixed | ||||||
Platform | Microsoft | OS | Windows | OS Version | XP/Vista/7 | ||||
Product Version | 3.1-beta | ||||||||
Target Version | Fixed in Version | 3.1 | |||||||
Summary | 0003787: Delays in certain player actions online (delay in weapon firing, opening doors, pressing switches) | ||||||||
Description | Players have reported in some delays in some player actions when playing with the 3.1 alpha build. These actions include (but possibly more): - Opening doors (the door opens 300ms later after pressing the use key) - Delay in firing weapons - Delay in switch activation once a switch is pressed - Slight delay in jumps It seems to occur at random and players will have different symptoms. For example Zakken who originally reported the issue at'https://zandronum.com/forum/viewtopic.php?f=55&t=9937#p115762 [^]' has advised that he is getting delays on any kind of player action. This also happens on a LAN server for him where he gets 0 ping As for me, I seem to get only a delay when firing the plasma rifle while playing online. It will take atleast a second to fire it. | ||||||||
Steps To Reproduce | Hard to reproduce as some players will get this, others will get mild symptoms and others wont get this at all. | ||||||||
Additional Information | Edward-San advised its possible that CPU usage could be spiking up causing these delays. I will be monitoring TSPG's CPU usage for a bit to see if there is any issues with this. | ||||||||
Attached Files | marc_qcdetestwads.png [^] (17,567 bytes) 2020-05-24 17:33
ticbufferdebuglog.c8017e10c1ab636c91feae75700a8c5e.txt [^] (995,328 bytes) 2020-09-06 08:09 | ||||||||
Notes | |
(0021324) EnsaladaDeTomate (reporter) 2020-05-07 00:48 edited on: 2020-05-07 00:50 |
Can confirm that this happens at random. IDK if this is related to this issue, but I remember playing normally on a server, then after pressing the jump button, i was stuck in mid-air with the "connection interrupted" message, i couldnt move or do anything but chat normally (such thing lasted for a minute or more, cant remember correctly) |
(0021326) lordtestes (reporter) 2020-05-07 17:22 |
When hosting a server through TSPG and selecting the Zandronum 3.1 build, I noticed some performance issues as well. This came in the form of delayed actions and lack of smoothness, which didn't happen to me in Zandronum 3.0. I noticed upon hosting a server that I wasn't able to join at first, and was required to have Brutal Doom. I'm not sure if Zandronum 3.1 uses Brutal Doom as a resource or not, but I know that it causes excessive load through servers/personal computers. I hope this helps! |
(0021329) Edward-san (developer) 2020-05-07 22:29 |
If you can reproduce this consistently on a local LAN server, the only thing I can suggest is to build and bisect the changes between 3.0 and 3.1 and see which change regressed. |
(0021332) Edward-san (developer) 2020-05-10 19:50 |
Just to be sure: these problems didn't happen in previous betas, like in'https://zandronum.com/forum/viewtopic.php?f=55&t=9340 [^]' , did they? |
(0021333) EnsaladaDeTomate (reporter) 2020-05-10 22:32 |
I played a lot 180901-1533 and it didnt had such issues |
(0021334) Torr Samaho (administrator) 2020-05-11 06:16 |
Quote from Edward-san One thing to keep in mind is that some servers don't run "true" 3.0, but 3.0 with some backports from 3.1. So when comparing the behavior of 3.0 to 3.1, make sure to use a "true" 3.0 for testing. |
(0021336) Cyber (reporter) 2020-05-13 04:24 |
Tested in JMX server using 3.1 Follow the steps below and this should reproduce the delay bug. During this test I was in OpenGL. Start in windowed mode. Let the game load in on a server. Join game, then alt enter into Fullscreen mode. At this point your shots should be more delayed than normal. I've tested it the reverse way and I've had something similar except I'd have "connection interrupted", and yet my messages would show to the other players with no problems. |
(0021343) Marcaek (reporter) 2020-05-24 17:35 |
'https://www.dropbox.com/s/elm81mwvwfmezht/Zandronum_05.05.2020_23.54.39.cld?dl=1 [^]' 'https://www.dropbox.com/s/rdnz7vyei7e972x/Zandronum_06.05.2020_22.53.40.cld?dl=1 [^]' 'https://www.dropbox.com/s/vgo2n0ledo2y0jw/marc_qcdetestwads.png [^]' A pair of QC:DE games using 3.1-alpha-r200501-1847 with reported delays. |
(0021344) Torr Samaho (administrator) 2020-05-24 19:30 |
Can you check if setting "sv_useticbuffer 0" on the server fixes the issues? |
(0021348) Cutman (reporter) 2020-06-03 01:12 |
I definitely am noticing some big difference in delays between versions. I don't have any vanilla Doom footage yet unfortunately (hard to notice on my end) but one of the 8BDM players came up with this gif using the same map and conditions on TSPG (vanilla 8bdm):'https://i.imgur.com/BGxn1iE.gif [^]' Setting "sv_useticbuffer 0" DID fix the issue for this person, on both a TSPG server and a server of their own. |
(0021462) Marcaek (reporter) 2020-06-10 20:39 edited on: 2020-06-10 20:40 |
Push demos using 3.0 and 3.1: 'https://www.dropbox.com/s/k09ikq0u2f5em0c/Zandronum_25.05.2020_00.50.47.cld?dl=1 [^]' 'https://www.dropbox.com/s/k8j5ton58uj622s/Zandronum_25.05.2020_01.22.28.cld?dl=1 [^]' Required WADS:'https://www.dropbox.com/s/qof122lmucanros/pushwads.png [^]' Related: I'm thinking the tracker should have some kind of upload space for demos tbh. |
(0021469) Torr Samaho (administrator) 2020-06-21 19:23 |
So looks like we need to investigate the movement regulator used by the tic buffer, but for this we need "field data". Can you apply the following patch to the server (clients need no changes) 'https://bitbucket.org/zandronum/zandronum-sandbox-stable/commits/84867c708c9a69a61a3d595659eca11413ed26c2 [^]' start the server with "+sv_logticbufferdebuginfo 1" and send the resulting ticbufferdebuglog.txt the server should write? Note that this will create the file ticbufferdebuglog.txt in the server's working directory. So don't activate this on multiple server sharing the same working dir. If the output name needs to be configurable, let me know. |
(0021492) Cutman (reporter) 2020-07-11 00:30 |
I still don't know how to make executables for these things. Since there's been no update for ages, could someone provide one for me? Also do you think it's worth making a discord server or something for those who have time to developing/testing Zandronum builds and patches? It seems like progress is always at a standstill when things like this need to bested. I'm sure it would be nice to be able to grab a group of people to try things out quickly. I have a couple of people who I can grab to test things with but I don't always know what needs doing. |
(0021500) Torr Samaho (administrator) 2020-07-12 19:57 |
Quote from Cutman For which OS do you need binaries? Unless it's Windows, it would be simpler if I show you how to compile them yourself. Quote from Cutman Anything that increases the amount of testing is highly welcome. |
(0021501) mifu (reporter) 2020-07-12 22:42 edited on: 2020-07-12 23:38 |
TSPG should have that particular build as a test build so you can host a server with now with that change. The issue is getting people to test. People have been unwilling. If you need some people the TSPG has a channel dedicated to bugs/issues in general #tspg-bug-reports and two of the zandro devs are in there. Highly suggest you use that instead of creating a new discord. EDIT: On that note, Torr when are YOU going to finally come online in discord? Would be awesome if you were visible there too. |
(0021503) Cutman (reporter) 2020-07-14 10:33 edited on: 2020-07-14 10:34 |
Sorry should have specified, I'm on Windows. Also mifu, since this test poops out a .txt file I won't be able to access it using TSPG, so it would be easier for me to be able to grab my own executables. Would be handy to have you around for live testing Torr, even if you just tell us what to do and we feed you back results. It's a lot faster than waiting on random users on the tracker for important issues. |
(0021504) mifu (reporter) 2020-07-14 11:04 |
I or one of the devs of tspg can extract the file upon request |
(0021505) DrinkyBird (developer) 2020-07-15 21:37 edited on: 2020-07-15 21:37 |
On TSPG the log can be acquired from https://static.allfearthesentinel.net/crash/ticbufferdebuglog.<usid>.txt where <usid> is your server's USID as displayed on the Server Info page |
(0021506) Cutman (reporter) 2020-07-16 19:58 edited on: 2020-07-16 19:59 |
'https://static.allfearthesentinel.net/crash/ticbufferdebuglog.0aead7a6fe596df6d220a92f06dc694c.txt [^]' here it is Played with a partner, ran around shooting and trying out weapons and inventory items. |
(0021507) Torr Samaho (administrator) 2020-07-19 19:20 |
Quote from Cutman Do you still need Windows binaries? Looks like you use TSPG for the logging now. Quote from Cutman We could try to organize a live testing session, but right now I just need debug logs created with "+sv_logticbufferdebuginfo 1" that were recorded while the problematic delays are happening. So, the only task I can give you right now is play to reproduce the delays while recording the extra debug info. Quote from Cutman Thanks! Did you perceive any problematic delays while testing? According to the logs, "player 0" was sometimes delayed by 4 instead of 2 tics, but that's the only suspicious thing I see in the log. |
(0021509) Cutman (reporter) 2020-07-22 00:45 edited on: 2020-07-22 00:46 |
The delay was most noticeable when using inventory items. One in particular fires out a projectile, and it would fire at an angle and pitch of where we were aiming almost a full second ago. I'm up for doing more testing but it's tricky to tell what's delayed and what is just normal lag. Do you just need more random sample games? Should we be playing vanilla doom2 or mods with tons of functions being called? Edit: Oh and I don't need the binaries now since we can get the logs from TSPG |
(0021511) Torr Samaho (administrator) 2020-07-22 19:44 |
Quote from Cutman Hmm, the log you linked would only explain an additional delay of 4 tics, i.e. about 4/35 seconds. Nothing close to a full second delay. Quote from Cutman Yes, you can play whatever you like. I just need logs recorded while the unusual delays happen. |
(0021531) Torr Samaho (administrator) 2020-08-30 19:14 |
Did anybody have a chance to create more logs yet? Unfortunately, I don't see any way to fix this without additional data. |
(0021532) mifu (reporter) 2020-08-31 08:51 |
Not yet it seems. I think everyone has been busy or are simply not willing to test. I have been trying to get a game started to test this but it seems its hard to get willing people to test out 3.1. I know Cutman has been busy also. I'd like to remind you that you guys (the devs) have access to use tspg to throw up a test server any time to test stuff out like this if no one else is providing the data you need. Our technical staff can also provide you with the files. I would strongly suggest doing this if we cant get any games started. |
(0021535) EpicTyphlosion (reporter) 2020-09-06 08:10 |
'Righty then. Since nobody has brought any data up in almost two months, allow me to be of assistance. I just finished a roughly hour-long deathmatch with a few people from the TSPG Discord server. This was done from maps 1 thru 7 of DWANGO5. All the flags were kept similar to how they were in ZCA matches (save for the monsters and regen invincibility, the former being changed on map 3). I also made sure to use the SSG most of the time, which seemed to be causing lots of other trouble as far as dropped shots go. In total, this is roughly an hour's worth of data. 'https://static.allfearthesentinel.net/crash/ticbufferdebuglog.c8017e10c1ab636c91feae75700a8c5e.txt [^]' I'm leaving the server up, so I've attached the text file as it was just after I left the server. Finally, here's a link to the demo if anyone wants to take a look (was to big for me to attach it). 'https://www.dropbox.com/s/ag57thmjtq6qlj6/2020.09.05_23.55.26_doom2.dwango5.cld?dl=0 [^]' |
(0021543) EpicTyphlosion (reporter) 2020-09-27 20:54 |
Um, do I have to host another experimental deathmatch or something? I would've expected a response by now. |
(0021547) Torr Samaho (administrator) 2020-10-04 19:31 |
Sorry for the late response and thanks for your logs! These show some problems with the movement regulator (it's definitely too strict in its current form), we need to take care of before further testing is needed. |
(0021691) Kaminsky (developer) 2021-08-08 12:05 edited on: 2021-08-08 12:07 |
Update: the movement regulator was removed from 3.1 a few months ago, as it was virtually impossible to improve it so that the artificial delays wouldn't become a problem for the clients. It has now been replaced with the skip correction, which has the same goal as the movement regulator (smooths the movement of lagging players) but doesn't add delays. It still needs to be tested thoroughly to make sure it's working properly. |
This issue is already marked as resolved. If you feel that is not the case, please reopen it and explain why. |
|
Supporters: | Soul EnsaladaDeTomate Cutman EpicTyphlosion mifu Kaminsky |
Opponents: | Filystea |
Issue History | |||
Date Modified | Username | Field | Change |
2020-05-05 23:18 | mifu | New Issue | |
2020-05-07 00:48 | EnsaladaDeTomate | Note Added: 0021324 | |
2020-05-07 00:50 | EnsaladaDeTomate | Note Edited: 0021324 | View Revisions |
2020-05-07 17:22 | lordtestes | Note Added: 0021326 | |
2020-05-07 22:29 | Edward-san | Note Added: 0021329 | |
2020-05-10 19:50 | Edward-san | Note Added: 0021332 | |
2020-05-10 19:50 | Edward-san | Status | new => feedback |
2020-05-10 22:32 | EnsaladaDeTomate | Note Added: 0021333 | |
2020-05-11 06:16 | Torr Samaho | Note Added: 0021334 | |
2020-05-13 04:24 | Cyber | Note Added: 0021336 | |
2020-05-24 17:33 | Marcaek | File Added: marc_qcdetestwads.png | |
2020-05-24 17:35 | Marcaek | Note Added: 0021343 | |
2020-05-24 19:30 | Torr Samaho | Note Added: 0021344 | |
2020-06-03 01:12 | Cutman | Note Added: 0021348 | |
2020-06-10 20:39 | Marcaek | Note Added: 0021462 | |
2020-06-10 20:40 | Marcaek | Note Edited: 0021462 | View Revisions |
2020-06-21 19:23 | Torr Samaho | Note Added: 0021469 | |
2020-07-11 00:30 | Cutman | Note Added: 0021492 | |
2020-07-12 19:57 | Torr Samaho | Note Added: 0021500 | |
2020-07-12 22:42 | mifu | Note Added: 0021501 | |
2020-07-12 22:42 | mifu | Status | feedback => new |
2020-07-12 23:38 | mifu | Note Edited: 0021501 | View Revisions |
2020-07-14 10:33 | Cutman | Note Added: 0021503 | |
2020-07-14 10:34 | Cutman | Note Edited: 0021503 | View Revisions |
2020-07-14 11:04 | mifu | Note Added: 0021504 | |
2020-07-15 21:37 | DrinkyBird | Note Added: 0021505 | |
2020-07-15 21:37 | DrinkyBird | Note Edited: 0021505 | View Revisions |
2020-07-16 19:58 | Cutman | Note Added: 0021506 | |
2020-07-16 19:59 | Cutman | Note Edited: 0021506 | View Revisions |
2020-07-19 19:20 | Torr Samaho | Note Added: 0021507 | |
2020-07-22 00:45 | Cutman | Note Added: 0021509 | |
2020-07-22 00:46 | Cutman | Note Edited: 0021509 | View Revisions |
2020-07-22 19:44 | Torr Samaho | Note Added: 0021511 | |
2020-08-30 19:14 | Torr Samaho | Note Added: 0021531 | |
2020-08-31 08:51 | mifu | Note Added: 0021532 | |
2020-09-06 08:09 | EpicTyphlosion | File Added: ticbufferdebuglog.c8017e10c1ab636c91feae75700a8c5e.txt | |
2020-09-06 08:10 | EpicTyphlosion | Note Added: 0021535 | |
2020-09-27 20:54 | EpicTyphlosion | Note Added: 0021543 | |
2020-10-04 19:31 | Torr Samaho | Note Added: 0021547 | |
2021-08-08 12:05 | Kaminsky | Note Added: 0021691 | |
2021-08-08 12:05 | Kaminsky | Status | new => feedback |
2021-08-08 12:07 | Kaminsky | Note Edited: 0021691 | View Revisions |
2022-02-24 05:09 | Kaminsky | Status | feedback => closed |
2022-02-24 05:09 | Kaminsky | Resolution | open => fixed |
2022-02-24 05:09 | Kaminsky | Fixed in Version | => 3.1 |
Copyright © 2000 - 2024 MantisBT Team |