jsPsych version 6.3.0:
Just starting today we have been having trouble with results of experiments not saving to the server. We’re getting:
[jspsych-pavlovia plugin 2020.4] Error
Essentially the same information is also in the browser console without any additional information to help troubleshoot unfortunately. This is a task that has been running and collecting data fine previously, and no changes have been made to the task.
Has something changed server side that we need to account for, or is it possible there’s some temporary server-side issue that is causing this?
Has this sorted itself out of are you still having issues?
I just tested it this morning and unfortunately it is still happening.
I’ve let @apitiot know and he’s going to take a look. However, I note that there was a pull request to the plugin repo last week here: Commits · jspsych/jsPsych · GitHub by becky-gilbert and Jolg42 which may have caused the issue.
Any further information on this? There have been no git commits to this task (outside of collected data) for 6 months, so definitely no changes on our end, and it had been working fine prior to about 2 weeks ago.
Weirdly, we have an older version of the task that this one was modified from that also has had no changes and is using the same version of the plugin (jspsych-pavlovia plugin 2020.4) to upload data and when I tested that task last week it was working, so it’s really baffling why this one has suddenly stopped.
Just wanted to bump this post (I’m from the same study team) and was wondering if anyone had any ideas on how to fix this issue; it’s still happening. Thank you!
Just wanted to update here again. Since we have not made any changes to this task, and I have other similar tasks running on the same jsPsych version and same plugin version, I still suspect this is some weird server-side issue. To that end, I copied the task exactly into a new repository, keeping everything identical with the exception of deleting all the files in the data folder. I then made this task active and just ran a test of it, and it ran and submitted results just fine with no errors. So as I thought, this is definitely some weird server issue, or possibly some issue with the data in the data folder on the original task.
Happy to continue talking to someone if there’s anyone at pavlovia that would like to troubleshoot this with us to figure out the actual issue with the orignal experiment/repository, but I guess for our purposes, we’ll just use the new identical task that is working.
I’m glad you’ve found a workaround. Also, since the author of the other thread (@sahkaur ) is from the same study team, then I guess that means this was an isolated issue rather than (as it appeared) two independent labs having identical issues at the same time.
We don’t have any OST staff who use jsPsych, so I don’t think we’ll have time to troubleshoot. However, I wonder if the issue was actually that the repository itself became “locked” (which can happen to shelf entries).
I’m not sure what you mean by the repository becoming locked. How would I check that to see if that is the issue? I just tested and can commit things to the original problematic repository without issue.