Problem signature: |
ID:239405
Feb 2 2012, 3:34 pm
|
|||||||||||||
| |||||||||||||
This any good for diagnosing a crash?
| |||||||||||||
#1 Feb 3 2012, 3:29 am
|
|
Please don't use bug report posts for conversations. The information in this report may give us some clue as to where the process crashed in our code, which is the most important piece in resolving it.
| |
#2 Feb 7 2012, 12:17 pm
|
||
Here is another one, from a different game being hosted on the same machine. The only difference seems to be the "Exception Offset". This game is not using pixel movement. The previous version of the game was stable, but the updates were so massive that I have no idea what could have lead to the crash.
| ||
#3 Feb 17 2012, 1:18 am
|
|
This second crash seems to have been due to large amounts of changes to turfs (Repeatedly changing the name and adding text overlays to display that name on a turf). Using turfs in this manner was causing various other issues as well. Changing the /turfs to /objs seems to have resolved all of the issues there.
| |
#4 Mar 22 2012, 1:05 pm
|
|
I have traced both crashes and they appear to be completely independent, though without more info I can't rule out that some kind of corruption caused the first.
I've added a sanity check to the routine responsible for the first crash, which just concatenates two strings. The other crash is in a routine that reads a cell from the map, so corrupted data from a turf operation would fit your theory. Unfortunately it doesn't point out the actual source of the problem, which will require more digging, but if it's just a matter of changing the name and appearance frequently it might be possible to reproduce. | |
#5 Mar 22 2012, 8:31 pm
|
|
The sanity check for the first issue will be in 494, but the second crash really needs a test case. I tried to reproduce this but failed. More info is needed.
| |
#6 Mar 22 2012, 8:31 pm
|
|
Lummox JR changed status to 'Unverified'
| |