I have a chronic, sporadic, yet annoyingly common issue where without warning my client's display will turn into a slideshow, display about 3 frames 2 seconds apart, then freeze while java.exe pegs whatever CPU it's running on at 100% CPU time. At this point, java.exe invariably needs killed. After reconnecting my character is usually stuck in place, rubberbanded in place, or stuck eternally re-enacting whatever commands I was issuing when the game froze. This invariably needs either a logout or extended GM assistance to clear up.
Here's a log of what this looks like from the game client's perspective:
2013/09/16 18:13:58:640 INFO J$a.b: Logging to 'C:\Program Files (x86)\Steam\steamapps\common\Spiral Knights\.\projectx.log'.
2013/09/16 18:14:07:465 INFO J$a.b: Steam initialized.
2013/09/16 18:14:11:650 INFO J$a.b: Connecting [host=game.spiralknights.com/184.73.171.156, port=47624]
2013/09/16 18:14:11:665 INFO DefaultControllerEnvironment.getControllers: Loading: net.java.games.input.DirectAndRawInputEnvironmentPlugin
2013/09/16 18:14:12:035 INFO J$a.b: IME created [Thread=18]
2013/09/16 18:14:12:794 INFO J$a.b: Datagram connection established [port=47624]
2013/09/16 18:14:17:028 INFO J$a.b: Moving at request of server [zoneId=83886080, sceneId=2147483647]
2013/09/16 18:14:17:137 INFO J$a.b: Issuing zoned moveTo(5:0, 2147483647, 1).
2013/09/16 18:14:29:961 WARNING J$a.b: Long invoker unit [unit=News Reader (class com.threerings.projectx.client.cu), time=754ms].
2013/09/16 18:14:58:390 INFO J$a.b: Connecting [host=/184.73.171.156, port=47626]
2013/09/16 18:14:58:693 INFO J$a.b: Datagram connection established [port=47626]
2013/09/16 18:14:59:323 INFO J$a.b: Issuing zoned moveTo(2:346, 351, 64).
2013/09/16 18:14:59:378 INFO J$a.b: Updating outgoing message throttle [msgsPerSec=45]
2013/09/16 18:16:02:741 INFO J$a.b: Moving at request of server [zoneId=33554432, sceneId=1073741890]
2013/09/16 18:16:02:746 INFO J$a.b: Issuing zoned moveTo(2:0, 1073741890, 0).
2013/09/16 18:16:03:069 INFO J$a.b: Updating outgoing message throttle [msgsPerSec=45]
2013/09/16 18:16:03:309 INFO J$a.b: Got updated scene model [placeId=2578, config=[type=data.LevelSceneConfig, aAi=null, aAj=true], scene=1073741890/%dungeon:m.candlestick_keep_vanilla_whipping_mishandling/1]
2013/09/16 18:30:15:301 INFO J$a.b: Moving at request of server [sceneId=1073741894]
2013/09/16 18:30:15:318 INFO J$a.b: Issuing moveTo(1073741894, 0).
2013/09/16 18:30:15:836 INFO J$a.b: Updating outgoing message throttle [msgsPerSec=45]
2013/09/16 18:30:15:942 INFO J$a.b: Got updated scene model [placeId=2740, config=[type=data.LevelSceneConfig, aAi=null, aAj=true], scene=1073741894/%dungeon:m.ct_fiend_whipping_mishandling/1]
2013/09/16 18:31:48:554 WARNING J$a.b: Frame index went negative! [anim=throw, fidx=-11, accum=-0.015999794, elapsed=-0.459, frames=-11]
2013/09/16 18:31:50:026 WARNING J$a.b: Frame index went negative! [anim=throw, fidx=-10, accum=-0.94400024, elapsed=-0.456, frames=-10]
2013/09/16 18:40:58:516 INFO J$a.b: Idle state changed. [idle=true]
2013/09/16 18:40:58:517 INFO J$a.b: Idle state changed. [idle=false]
(If it helps, I was throwing some Oil Pots at a Howlitzer before moving to engage Kats when the crash occurred.
After going through the logs, I've hypothesized that the issue that's causing this mess is linked to three log warnings: "Frame index went negative!", "Someone call Einstein! The clock is running backwards", and "ControllerEnvironment.log: Failed to poll device: Failed to poll device (8007000c)".
It would be nice if someone actually read this and did their job.