Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

Error that messages that data files are too big to commit when they are not

Hi everyone, I'm having trouble where the participants arrive at the final debriefing screen but the study does not finish as it normally does; there are error messages appearing in the console as if the data file is too big to commit.

That is not the case, though, because the resulting files are 3.0 MB and seem to be saved just fine, everything is complete, as you can see in the next screenshot.

This has only been happening recently with our last few participants – before it was finishing fine and saving the 3.0 MB files as usual, and I haven’t made any recent changes to the script. So, I just wondered if there’s anything that you’re aware of that could avoid this issue in the future.

Thanks in advance!

Vanessa

Comments

  • Hi Vanessa,

    We solved this problem via email. But I want to answer here in the forum to have it documented and it might others.

    Sorry for taking so long but I had to be sure and check and test it thoroughly. It's a bug in how JATOS checks for the maximum result data size for jatos.submitResultData. For jatos.appendResultData it works fine. But for jatos.submitResultData it takes the currently stored result data into account and adds the newly submitted data on top. This way you can reach the 5MB limit by submitting 3MB twice (although it will be only 3MB in the database). I fixed it (https://github.com/JATOS/JATOS/commit/953eaceb780dc0033e468138e303ea2e826e0a30) and it will be part of the upcoming release 3.8.1 (out in a couple days).

    There was actually a second bug that allowed you, under certain circumstances, to circumvent the max result data size limit. The fix handles that too.

    Best,

    Kristian

Sign In or Register to comment.