View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0003590 | Spring engine | public | 2013-03-16 23:43 | 2013-03-31 01:03 | |
| Reporter | silentwings | Assigned To | zerver | ||
| Priority | normal | Severity | minor | Reproducibility | have not tried |
| Status | closed | Resolution | unable to reproduce | ||
| Product Version | 93.2.1+git | ||||
| Fixed in Version | 94.1.1+git | ||||
| Summary | 0003590: desync with 93.2.1-25 | ||||
| Description | see demo, map is throne with ba 7.72. | ||||
| Tags | No tags attached. | ||||
| Attached Files | |||||
| Checked infolog.txt for Errors | |||||
|
|
[f=0000000] -> Version: 93.2.1-35-ge1440a1 develop (OMP) [f=0000000] -> Version: 93.2.1-35-ge1440a1 develop (GML ASIM OMP) |
|
|
ouch: Spring 93.2.1-35-ge1440a1 develop (GML ASIM OMP) [f=0000000] Connecting to 94.36.249.236:8452 using name [Fx]Bluestone [f=0000000] Connection attempt from Jarzynka [f=0000000] -> Version: 93.2.1-35-ge1440a1 develop (GML ASIM OMP) [f=0016335] Sync error for Jarzynka in frame 16328 (got da4b3704, correct is 4062480b) [f=0016736] Sync error for [Fx]Bluestone in frame 16729 (got 74090da9, correct is ebf7213b) -> GML desynced itself? or is the desync message from the server? |
|
|
Looks like a normal desync to me [f=0013940] Sync error for Jarzynka, [Fx]Bluestone in frame 13922 (got 101f8ae7, correct is 84dc11f3) and a deterministic one too, so should be easy to fix. |
|
|
Hmm, the GML clients have a different NETWORK_VERSION :) Serious fuckup... [f=0000000] Connection attempt rejected: Wrong network version: 6, required version: 7 (Message ID: 65 Network version: 0 Datalength: 52) [f=0000000] Connection attempt from [Fx]Bluestone [f=0000000] -> Version: 93.2.1-35-ge1440a1 develop (GML ASIM OMP) [f=0000000] -> Connection established (given id 4) [f=0000000] Connection attempt rejected: Wrong network version: 6, required version: 7 (Message ID: 65 Network version: 0 Datalength: 52) [f=0000000] Connection attempt from Jarzynka [f=0000000] -> Version: 93.2.1-35-ge1440a1 develop (GML ASIM OMP) [f=0000000] -> Connection established (given id 2) |
|
|
I don't know enough about the engine to understand "the GML clients have a different NETWORK_VERSIO" but this might help: We had lots of people trying to help us test who didn't know how to download a dev-build, portable install, disableversioncheck, etc. We tried to get as many of them as could working but didn't get all. I think the lines that read "[f=0000000] Connection attempt rejected: Wrong network version: 6, required version: 7 (Message ID: 65 Network version: 0 Datalength: 52)" came from people trying to join the game with the wrong engine version; these ones didnt make it in to the game afaik. We also tried a long game (1/2 hour, 3 people ffa) with all of us having the singlethreaded exe and that game played really well; major bugs from 93.1/2 seemed massively improved and we no errors at all. |
|
|
Silent: You are right, it was a coincidence that the network version spam started at the same time as MT clients connected. Abma: what about Ninja? I have some suspicion that it causes this desync. Did you make any other changes on the build slave? |
|
|
no commit that i know about. only commit that imo could cause problems like this i remember is: https://github.com/spring/spring/commit/af9a86b5dac9bf533c66075e697bb90b831d2cf4 -> base content possible is different (but i tested like three times if it works correctly) edit: no thats not the cause: i grabbed linux static x32 & win32 build, extracted basecontent and did a diff -r... no difference. also i'm using ninja since half a year or so and never got a desync in network games. i would think more, the cause could be switching between branches on the buildslaves, cmake often has problems with some cached vars. (this is why some weird errors appear, like .cpp file missing, undefined symbols, etc) but just, if its not MT, which seems be more likely, as ST vs ST seems to sync. |
|
|
maybe as note: on the buildbot everything was changed, it was setup from scratch. not sure which release exactly, but it was at the time the linux static builds where added. a new gcc version was/is used as well, so many possible sources for bugs. |
|
|
AFAIK 94.0 runs fine, impossible to say what caused this |
|
|
maybe a reused path-cache... but yep, we can't reproduce it. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2013-03-16 23:43 | silentwings | New Issue | |
| 2013-03-16 23:43 | silentwings | File Added: 20130316_222755_Throne v1_93.2.1-35-ge1440a1 develop.sdf | |
| 2013-03-16 23:43 | silentwings | File Added: infolog.txt | |
| 2013-03-16 23:59 | Kloot | Note Added: 0010092 | |
| 2013-03-16 23:59 | Kloot | Note Edited: 0010092 | |
| 2013-03-17 00:29 | zerver | Assigned To | => zerver |
| 2013-03-17 00:29 | zerver | Status | new => assigned |
| 2013-03-17 02:44 | abma | Note Added: 0010093 | |
| 2013-03-17 11:30 | zerver | Note Added: 0010099 | |
| 2013-03-17 11:37 | zerver | Note Added: 0010100 | |
| 2013-03-17 11:57 | silentwings | Note Added: 0010101 | |
| 2013-03-17 16:33 | zerver | Note Added: 0010103 | |
| 2013-03-17 17:24 | abma | Note Added: 0010104 | |
| 2013-03-17 17:24 | abma | Note Edited: 0010104 | |
| 2013-03-17 17:25 | abma | Note Edited: 0010104 | |
| 2013-03-17 17:29 | abma | Note Edited: 0010104 | |
| 2013-03-17 17:33 | abma | Note Edited: 0010104 | |
| 2013-03-17 17:36 | abma | Note Edited: 0010104 | |
| 2013-03-17 17:37 | abma | Note Edited: 0010104 | |
| 2013-03-18 11:38 | abma | Note Added: 0010109 | |
| 2013-03-31 00:16 | zerver | Note Added: 0010330 | |
| 2013-03-31 00:16 | zerver | Status | assigned => closed |
| 2013-03-31 00:17 | zerver | Resolution | open => unable to reproduce |
| 2013-03-31 00:17 | zerver | Fixed in Version | => 94.1.1+git |
| 2013-03-31 01:03 | abma | Note Added: 0010331 | |
| 2014-01-16 16:06 | abma | Category | Spring MT => (No Category) |
