opensesamerun: Crash after fixed number of sequence runs / TCP pakets
Hi
We're running into a strange problem in our current study using iMotions and OpenSesame.
The setup looks like this:
1. Start Study in iMotions, which consists only of a Screen Recording
2. Screen Recording launches OpenSesame experiment through a batch file with the command "C:\Program Files (x86)\OpenSesame\opensesamerun.exe" %path%\MakeAChoice.osexp -s %IMotionsRespondent% -l %path%\logs\subject%IMotionsRespondent%.csv -f
3. OpenSesame connects to external API and Control API of iMotions using TCP
4. Respondents plays a game with 40 repetitions, each repetition sends the following 8 events through the API:
- M;2;;;FixStim_x;;N;I
- M;2;;;Stim_x;;N;I
- E;1;MouseInput;;;;;MouseStimulus;x
- M;2;;;FixResult_x;;N;I
- M;2;;;Result_x;;N;I
- E;1;ResultInput;;;;;ResultInput;x
- M;2;;;Scalex6;;N;I
- E;1;Valence;;;;;Valence;x
The problem is then, that everytime after a fixed amount of sequence runs (in the 12. run) and the 93. TCP-packet (we snooped the traffic using Wireshark) OpenSesame crashes with this error message:
The error message is not very common, there are some people saying that this occures if a python socket wants to connect again after it was closed before. However this does not happen at any point our experiment.
Most importantly, the error only occures if the Experiment is launched using said opensesamerun-Command, but not if the experiment is run from inside OpenSesame.
We also substituted the iMotions TCP-Server with mocked ones (using ncat -l 127.0.0.1 9087
), this also had no effect on the error. We even tried killing the Windows Firewall (no other firewalls installed) and Antivirus, no effect either. Same thing happens on a second computer.
From our point of view it looks like opensesamerun.exe is buged. Do you have any idea how to proceed?
Thank you
Stefan
Comments
To follow up on this error: I'm also seeing this message too (3.1.9 Jazzy James), using the opensesamerun.exe in combination with an Eyetribe eyetracker.
However, in my case it is unpredictable when it occurs for our experiments. It is rather inconvenient though, since file naming occurs using a small VBscript which calls opensesamerun.exe with all the parameters...
Wow, I have long forgotten about this one.
Sorry I can't help you, we've change the experimental setup and couldn't resolve the problem as far as I remember.
It's very likely you are experiencing the exact same bug as we did, as the code of opensesamerun seemingly hasn't changed since the original post...