View Issue Details
| ID | Project | Category | View Status | Date Submitted | Last Update |
|---|---|---|---|---|---|
| 0006222 | Spring engine | General | public | 2019-05-17 16:59 | 2019-06-13 23:25 |
| Reporter | Google_Frog | Assigned To | Kloot | ||
| Priority | normal | Severity | major | Reproducibility | always |
| Status | resolved | Resolution | fixed | ||
| Product Version | 104.0 +git | ||||
| Fixed in Version | 104.0 +git | ||||
| Summary | 0006222: 104.0.1-1239-g12caea1 maintenance Circuit AI does not do anything in ssf loaded games | ||||
| Description | Circuit AI seems to stop or not be initialized in ssf save games. In slsf it worked, although it did not perfectly pick up from where it left off (and at this point I don't expect it to). I'm reporting it here because, even if the fix in AI-side, lamer would need to know what is required for a fix. | ||||
| Tags | No tags attached. | ||||
| Attached Files | |||||
| Checked infolog.txt for Errors | Yes | ||||
|
|
https://github.com/ZeroK-RTS/CrashReports/files/3192035/movetest.zip |
|
|
See https://github.com/rlcevg/CircuitAI/issues/101 |
|
|
IMHO it's AI-side issue: with slsf Zero-K recreates units and AI receives AI::UnitCreated and AI::UnitFinished events for each unit before AI::Load event; while with ssf AI receives only AI::Load event (and it can ask engine for the list of own units). |
|
|
CircuitAI actually should receive unit events when loading an ssf since it claims loadSupported=no (https://github.com/rlcevg/CircuitAI/blob/master/data/AIInfo.lua#L42), see https://github.com/spring/spring/blob/develop/rts/ExternalAI/SkirmishAIWrapper.cpp#L140. Can you test this in 1248-g5b914dd and attach an infolog? |
|
|
104.0.1-1249-g9ed1ae2 maintenance didn't work, infolog attached. [f=-000001] [AIWrapper::InitLibrary][AI=0 team=1] postLoad=0 loadSupported=0 numUnits=0 numUnits==0 for some reason at that stage. |
|
|
hmm yes: host AI's are currently not even PostLoaded (EngineOutHandler just CR_IGNORE's them) and regular postLoad=0 AI creation precedes game state restoration (hence numUnits=0), solving this will take one additional commit. |
| Date Modified | Username | Field | Change |
|---|---|---|---|
| 2019-05-17 16:59 | Google_Frog | New Issue | |
| 2019-05-17 16:59 | Google_Frog | File Added: infolog_full.txt | |
| 2019-05-17 16:59 | Google_Frog | File Added: movetest.lua | |
| 2019-05-17 17:01 | Google_Frog | Note Added: 0019991 | |
| 2019-06-04 18:08 | Google_Frog | Note Added: 0020004 | |
| 2019-06-05 18:32 | lamer | Note Added: 0020005 | |
| 2019-06-11 20:12 | Kloot | Note Added: 0020006 | |
| 2019-06-12 17:52 | lamer | File Added: infolog.zip | |
| 2019-06-12 17:52 | lamer | Note Added: 0020008 | |
| 2019-06-12 21:49 | Kloot | Note Added: 0020009 | |
| 2019-06-12 22:51 | Kloot | Note Edited: 0020009 | |
| 2019-06-13 23:25 | Kloot | Assigned To | => Kloot |
| 2019-06-13 23:25 | Kloot | Status | new => resolved |
| 2019-06-13 23:25 | Kloot | Resolution | open => fixed |
| 2019-06-13 23:25 | Kloot | Fixed in Version | => 104.0 +git |
