bjg wrote:Marauders are partially broken in the 32 version too. Seems like a regression. Have uploaded few errors. Also uploaded game_857. Just hit the "next turn" to reproduce.
This is actually an entirely different error. The Marauder ships at Azimir are leftovers from a base that was conquered 30 turns ago -- but, something about the way it happened left them in an odd state. Thanks for the uploads.
Interesting. It was an AI that did it, perhaps due to another bug. I'm glad I've found a new error (or two). Is there a way I can continue that game or it's screwed?
bjg wrote:Interesting. It was an AI that did it, perhaps due to another bug. I'm glad I've found a new error (or two). Is there a way I can continue that game or it's screwed?
Just posted a patch to stable that should work around the issue (r14765). Interactions with those Marauders may still trigger bugs -- they really shouldn't exist in their present state
sven wrote:Just posted a patch to stable that should work around the issue (r14765). Interactions with those Marauders may still trigger bugs -- they really shouldn't exist in their present state
Thanks! Is there something you'd not like me to do or is it just a disclaimer?
sven wrote:Just posted a patch to stable that should work around the issue (r14765). Interactions with those Marauders may still trigger bugs -- they really shouldn't exist in their present state
Thanks! Is there something you'd not like me to do or is it just a disclaimer?
Just a disclaimer -- I haven't really "fixed" the problem yet -- just replaced the error with a warning.
mute wrote:SIS will hang from time to time in combat, I haven't been able to 100% reproduce the hang but it is associated with the following elements though not necessarily in this order:
Fleet size greater than ~5 combatants
Selecting all ships
Issuing a simultaneous move order
Toggling weapon settings
Ending the turn
It doesn't stop my machine, but the SIS becomes completely unresponsive. When I check on taskmanager I don't see a memory leak, but I'll check it through Process Explorer going forward and see if that sheds any light on things.
Thanks for this report, and the associated log uploads. My best hunch, at this point, is that bug may be somehow specific to SLI mode. We have other GeForce GTX 400 series in the beta (I actually have one in my development box), and this is the first time I've seen these sorts of OpenGL exceptions.
Could you try disabling SLI while running SiS, then seeing if multi-ship moves are still prone to crashes?
Doomed wrote:If I remember correctly the marauder ship retreated from a battle in another system
I got this to reproduce before too, with the key seeming to be that the retreating fleet was the last of its faction of marauders.
Yeah, thanks to both of you for reporting this one. There is indeed some strangeness going on with these sorts of "orphaned" marauder fleets. It's not a high priority bug (as it happens rarely, and it's often harmless), but, I can replicate the behavior, and should get around to fixing it eventually (probably as part of the big diplomacy system update).
It happens on every map, but now I can't reproduce it myself (loading fresh from the save). There are probably some steps I do that confuse UI into thinking the panel is there than it's not. Sorry for an unreliable report.
bjg wrote:It happens on every map, but now I can't reproduce it myself (loading fresh from the save). There are probably some steps I do that confuse UI into thinking the panel is there than it's not. Sorry for an unreliable report.
Hrm. Well, just knowing that it can happen at all is useful. If you can figure out how to replicate it, let me know.
Think I know how to reproduce it. Click on any Transport, click "start trading" (the panel dissapears) - the ships in the lower third become "unclickable".
