Description of the problem: I basically did something very rudimentary because I did not find a better solution thus far. I want to have multiple versions of the same task one for each session (I know I could have all sessions in one but I prefer it this way as it is easier to check if participants have completed the tasks). I programmed the task in psychopy half a year ago and it does not work anymore when I upload it to gitlab. So I basically uploaded everything and then copied the js code from a version that I had in there from a previous study that works, but it is giving me a weird issue now: the size of stimuli is extremely small and, even though I make changes to it, nothing really happens. Is there a better way of doing this? I know I could fork the tasks but that would be for a different task, right? Is there any other way of copyin a task to have multiple versions of it?
’ I programmed the task in psychopy half a year ago and it does not work anymore when I upload it to gitlab’ - If that could be an underlying issue for the current problem please could you explain what didn’t work so that we can figure out the problem there.
‘So I basically uploaded everything and then copied the js code from a version that I had in there from a previous study that works’ - do you mean that you copied over the .js file from a different project? Just so that you are aware the changes will not be reflected in the builder experiment and may make it difficult to make edits moving forward.
You are correct forking might not quite be what you need as gitlab doesn’t allow multiple forks in the same namespace, so I think at the moment it wold be a case of downloading the repository, deleting the online associated files and pushing it from psychopy as a new project (which I think is what you have done is that correct?)
I am not sure why the task is not currently working when I push it from psychopy. I think I made some adjustements to it the first time I coded it directly in the JS code to make it work as some things were not translating well to JS at the time. I decided to push the task multiple times under different names so I could have different versions of it and then I copied the content of the JS file of the task that works to the new directory. I am not concerned that the builder is not up to date as I the knowledge I have of js is little but enough to adapt the task when I need to (I only do minor changes). The problem is that the stimuli is showing extremely small even though the js code is the same as before. Is there a problem with copying? I have done it before with other tasks and it worked well.
Unfortunately since my JS knowledge is not huge I am not sure I am best to advise on this one (I could help if this was python/builder!)
I will flag this to the JS members of our team @thomas_pronk@sotiri So that we can help figure out the why your stimuli appear small, please can you make your repository public and share that URL with us so that we can also see the code.
Something must be different, but the question is what exactly Here are two options:
Maybe there is some difference in the JS after all? Consider a diff tool for that
Maybe different JS libraries get loaded? A quick-and-dirty way to examine that is by taking a look in the browser console, network tab, and compare the size of the libraries.
Well, I copied and pasted without making any changes. All text looks normal except for the stimuli and even when I change the size of stimuli it does not change when I run the experiment. I wonder if there’s a bug or something.
Here are two options: