Question - Crashes 10x per day (7 Viewers)

[raw]

One of my 8 toons crashes regularly while hunting.
I can't find a pattern.
However, he is the only one that uses RGmercs. The rest are running CWTN plugins.



---------------------------
EverQuest Crash Detected
---------------------------
MacroQuest caught a crash:

Process ID: 45516
Version: 3.1.52471.15
Location: RaiseException+0 (C:\WINDOWS\System32\KERNELBASE.dll+00000000000C1AE0)

CrashID: 7c963c77-a014-4fee-995c-134ab8bc25fc

You can either:
* [RETRY] Continue execution and hope for the best.
* [CANCEL] Write a crash dump and terminate EverQuest.

Copy the contents of this dialog to your clipboard by pressing Ctrl+C

---------------------------
Retry Cancel
---------------------------

 

[Algar]

What build number, which class, which config loaded?

 

[Sic]

unfortunately i don't see a crash dmp attached to that crashID.

the summary that the error reporting guesses at points towards a coroutine issue, but without the .dmp (and some eyes like brain or knightly) not gonna have any good suggestions for you

 

[Derple]

Could be another exception inside an exception issue in MQ that @brainiac would need to look at.

 

[Redbot]

can you check your macroquest.ini in veryvanilla for

[Crash Handler]
EnableCrashSubmissions=1

 

[Algar]

Not 100% sure about any of this, the below definitely contains some presumption, but, food for thought:

In the past, the coroutine in a similar case or two was actually the rgmercs "thread" in mq2lua, and the typical offenders were non-bool mq delays.

I'm auditing the delays again now (every time we feel we got them it seems we find one more, and even if we have, it doesn't hurt).
It is important, however, that you are on recent codebase, and haven't brought forward any past bugs by using older custom configs.

 

[evianwater]

Not 100% sure about any of this, the below definitely contains some presumption, but, food for thought:

In the past, the coroutine in a similar case or two was actually the rgmercs "thread" in mq2lua, and the typical offenders were non-bool mq delays.

I'm auditing the delays again now (every time we feel we got them it seems we find one more, and even if we have, it doesn't hurt).
It is important, however, that you are on recent codebase, and haven't brought forward any past bugs by using older custom configs.

Are non-bool mq delays a known issue?

 

[Algar]

Are non-bool mq delays a known issue?

To be clear, I am not sure that it's what is occurring here, but yes.

Before the change the devs made to mq2lua a few months back, they would silently fail.
Since then, I've seen it end scripts, and I've seen it causes crashes. There has been some iteration, and I'm not sure of what would specifically trigger one versus the other, however.

For the record, I prefer the script ending over silent 10 second delays that can never return true/etc