1266 ED Crashes

Case number:845829-2002660
Topic:Crash/Hang
Opened by:Susume
Status:Open
Type:Bug
Opened on:Tuesday, August 2, 2016 - 04:26
Last modified:Saturday, August 6, 2016 - 04:44

Lots of people are crashing on 1266; let's have a thread! Why not keep it all in the puzzle comments? Because you can't upload a log.txt file in the puzzle comments.

This crash: clicked "Center Protein on Density." Nothing happened for a while, then it crashed. log.txt attached - as usual, I cut several MB of nulls out of the file before uploading. There is an error message before the unhandled exception, but the unhandled exception itself is... dun dun dun... an OpenAL function call. (There was another client open; I assume that one wrote the lines after the exception.)

AttachmentSize
log - Copy (16).txt94.21 KB
(Tue, 08/02/2016 - 04:26  |  11 comments)


dbuske's picture
User is online Online
Joined: 09/22/2011
Groups: Beta Folders

Yep same hear. I bet your machine is on win 10 susume!

Susume's picture
User offline. Last seen 5 days 17 hours ago. Offline
Joined: 10/02/2011

Nope, Win 7. I never had the crashing-every-five-minutes problem that the Win 10 users have had - just an odd openAL error here or there.

20160728-c4d88693d5-win_x86-devprev

bertro's picture
User offline. Last seen 6 weeks 4 days ago. Offline
Joined: 05/02/2011
Groups: Beta Folders
Topic: General » Crash/Hang
Type: Suggestion » Bug

I agree this 1266 is unplayable.

Even crashed just doing a wiggle@1.0 of 65-69 on Auto.

bkoep's picture
User offline. Last seen 2 days 18 hours ago. Offline
Joined: 11/15/2012
Groups: None

Thanks for letting us know! There have been a lot of problems with Puzzle 1266, so we've made it a non-competitive puzzle (worth zero points), but still available to anyone who wants to tinker with it.

I have an idea about the issue with Center Protein on Density. The log file will help to confirm, but it's definitely unrelated to the OpenAL exceptions.

@bertro, your wiggle crash appears to be related to an ongoing symmetry bug that has been really difficult to crack. It's helpful to know that this bug is also rearing its head with crystal symmetry.

bertro's picture
User offline. Last seen 6 weeks 4 days ago. Offline
Joined: 05/02/2011
Groups: Beta Folders

@bkoep: Are you able to duplicate the crash?

bkoep's picture
User offline. Last seen 2 days 18 hours ago. Offline
Joined: 11/15/2012
Groups: None

Kind of... We've been able to generate similar crashes, but not reliably or reproducibly. It appears that Foldit can sometimes "break symmetry," but the event seems to be non-deterministic—or at least very finicky. We can repeat the same steps many times, but only rarely does the problem occur.

If you have a case in which this asymmetry error occurs reliably (e.g. "load this solution, wiggle all, crash on tenth iteration"), that would be a huge help to us.

bertro's picture
User offline. Last seen 6 weeks 4 days ago. Offline
Joined: 05/02/2011
Groups: Beta Folders

you mean this error I guess: "asymmetry introduced by ActionRepack"?

I'll sure share if it happens.

bertro's picture
User offline. Last seen 6 weeks 4 days ago. Offline
Joined: 05/02/2011
Groups: Beta Folders

Sorry I cannot replicate.

However, this may help or not:

It was crashing everytime (as stated before here) while I had, in the same foldit directory, a previous version of devprev running puzzle 1264 at the same time. Meaning, 1264 was running a script. From that same directory, I started a new client but it updated to the latest devprev and I opened 1266 which stared crashing doing only a wiggle on the selected segments on auto.

I don't know if it is relevant. I cannot test this since I lost the previous devprev...

Susume's picture
User offline. Last seen 5 days 17 hours ago. Offline
Joined: 10/02/2011

Maybe some or all of the openAL exceptions in the various log files are a red herring. Maybe openAL throws exceptions all the time but does not crash the client. Then some other error crashes the client, and when the error handler looks up what the last error was, the new error has not left data in the error stack and the most recent data is left over from the latest openAL error, so that gets reported in the log.

Or maybe the client is trying to look up the appropriate sound for the latest event, but the latest event is an error, and passing that to the openAL lookup function crashes openAL.

In the former case, disabling openAL will change what gets reported to the log (possibly to no report, or possibly to some other red herring) but will not remove whatever error is actually crashing the client. In the latter case, disabling openAL may reduce the crashing, or may change what is reported to the log (as the original error might be reported), or both.

Joined: 06/24/2008
Groups: Void Crushers

1266 wont stay launched; reading the crash logs makes me think it has something to do with chat.
MAC OS 10.11.6

Attaching Fold log and MAC Crash Log

Joined: 05/26/2008
Groups: Hold My Beer

I have been getting the symmetry crash on occasion. Other than being incredibly slow to do tasks. Sometimes waiting 5 mins for a rebuild to change a section of 5-6 segments its been fairly stable.

I did not play the original

I have noticed however foldit "stops responding" at times probably because of memory loading

The puzzle has been using about 1.2-1.5GB of ram for me

i5 win10

Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons