Search the Community
Showing results for tags 'plugin corrupt papyrus stack'.
-
Been working with the CK tutorial recently and found some very weird issues. The first appears to be keeping the CK loaded for a long time (in between crashes of course) It appears to function properly, but if left in focus when either bringing the machine out of standby or running a memory intensive prog (like Skyrim) the render window freezes- and unfortunately crashes the rest of the CK. Not only that, but over time the papyrus logs from game runs would blow up to a scary 4 mb with every possible error in the book. Quit the CK and load the corrupt plugin we find a more "respectable" log of about 70k-340k. These would contain curious errors of the "truncated stack" kind: "USKPRetroactive121Script.psc" Line ? [USKPRetroactive12 (020339CE)].uskponload05script.Process() - "USKPOnLoad05Script.psc" Line ? [USKPRetroactive11 (0202EE20)].uskponload04ascript.Process() - "USKPOnLoad04aScript.psc" Line ? [USKPRetroactive103 (0202C823)].uskponload02script.Process() - "USKPOnload02Script.psc" Line ? [USKPVersionTracking (0200D8B9)].USKP_VersionTrackingScript.ProcessRetroScripts() - "USKP_VersionTrackingScript.psc" Line ? <truncated stack> [12/20/2014 - 01:54:18PM] Error: Cannot call IsDead() on a None object, aborting function call stack: Which bring to mind Arthmoor's post on the civil war bugs. The plugin I have used is corrupt because the linked ref from the draugr to the defaultActivateSelfTRIG in the ambush exercise got broken. A symptom of this is the draugr spawn outside the coffin instead of lying within the coffin marker. In my case it got worse as the defaultActivateSelfTRIG failed to activate at all. But this corruption follows through all similar setups in the game as the culprit plugin is loaded last. Suppose we load (the unedited) WarehouseAmbushes (via COC) with the corrupt plugin activated. The draugr in the ambush setup at the south wall spawns outside the coffin. De-activate the plugin and he/she spawns inside correctly. Of course, Xedit doesn't pick up any of this problem. But why is the CK so tetchy about catching this spawn-outside-coffin bug? Might there exist a bug latency in editing some other unrelated object in the cell? Must have been a pain for the game designers. A solution is to employ the use of an overriding script confining the recalcitrant undead to their tombs. Would be slow as it has to detect thousands of sets of properties that match ambush. Oh well. As an aside, a performance booster (minus increased loading times) was the use of fPostLoadUpdateTimeMS=1000 Probably a good idea to boost this while the civil war scripts run. Also increasing fExtraTaskletBudgetMS a little from 1.2 might be beneficial in the CWl but haven't tried that yet. Getting in the habit of PCB before COCing anywhere thinking it might make a difference. Doubt it though. And clearing the memory before each run of the game (and CK?) with this.