Closed emilygoodwin closed 6 months ago
This should be fixed in the latest release (1.9.6). You should be able to upgrade your project and run with a tajriba.json file created with an older version, and you should see all the data. This is a rewrite of export, so please tell me if you notice any inconsistencies with the previous version or any other problems.
Is there an existing issue for this?
What happened?
Finished experiments seem not to have all the information saved to the players.csv file. I am not sure if they're "in" the tajriba file (is it an export problem or an actual data-saving problem) but re-exporting the same tajriba file yields the same .csv.
Steps To Reproduce
Running 1.9.2, start 1 batch with 1 game. Complete that (will work fine).
Finish game, scp tajriba file to local, export locally. players.csv file will be missing participant identifiers, exit survey data for usually at least one participant.
When using earlier versions, I found that if I ran multiple games, it was only 1 participant from each game that was missing data. However, when running 1.9.2, I cannot run more than one game (without restarting the server and erasing the tajriba file) so I can't verify if that's true in 1.9.2. I believe the 'can't run more than one game' problem is unrelated- I will comment on the other relevant issues.
Empirica Version
What OS are you seeing the problem on?
macOS, Other
What browser are you seeing the problem on?
Chrome, Safari
Relevant log output
Anything else?
Ubuntu on the remote server (where the experiment is served) and MacOS 11.7.10 locally (where it is bundled). I included the log out put that I got, which has errors that are I believe unrelated to the data-dropping issue- I get data dropped even from the first game in the batch, which produced no errors.
Code of Conduct