psychopy.org | Reference | Downloads | Github

Editable TextBox component does not register keypresses after updating to version 2021.1.2

Online experiments
,
#1

URL of experiment: Pavlovia

Description of the problem:
i am using the editable function in the textbox component and it worked nicely online in version 2020.2.10
but when i updated to version 2021.1.1 and 2021.1.2 it does not register keypresses

[Editable TextBox2 component does does not allow me to type any text (no letter appears on the screen). Offline, I can type anything and it appears on the screen perfectly fine.

i checked (Editable TextBox2 component does not register keypresses) solved by @Becca

and i added a code component to each frame : textbox.editable = true;

but it does not seems to help

thank you

#2

This the same issue you reported in the thread about the announcement about the new PsychoPy release right? Once you shared the gitlab repo I’ll continue with this topic in this thread

#3

yes it is the same :slight_smile: i have an update that i just noticed now.
the oddest thing, if i click on the keyboard for longer than normal it will register keypresses
for example i can only type ‘yyyyeeeessss’ instead of ‘yes’

this is my gitlab repo
thank you!

#4

i added you as a developer

#5

Could you tell me the name of the routine & component where the problem is at?

#6

this is in all my routines that contain routine

e.g. texbox component in TextBox_typing routine in the brainDump loop :\
sorry for being so complicated

#7

In this clone: Thomas Pronk / test_stroop_experiment · GitLab

I removed all of the other routines and loops, and then the editable textbox works fine on my browser (Win 10 + Firefox). I did notice you’ve got quite a lot of code components doing rather funky stuff. My guess is they cause a memory and CPU overload, making the experiment less responsive.

#8

interesting…and i wonder why it worked fine before

yes i do have a complex experiment and my JS skills are very poor but i am not sure about the funky stuff since my experiment does run quite smoothly :grimacing:
is there a way to overcome this issue?

thank you for trying!

#9

In another experiment i removed all of the other routines and loops too and it still gives me the same issue.
I really don’t understand how to approach this issue

#10

I’d recommend a big cleaning; the code components probably do one thing that you want to achieve, but also have a whole bunch of undesirable side effects.

1 Like
#11

sorry for bothering but i have created the simplest task with only beta editable textbox and the issue still appeared

#12

That wasn’t a link to a GitLab repo, could you link it? I’ll check it out

#13

sorry and thanks againstroop_writing.psyexp (18.6 KB)

#14

Thanks for the update! And sorry for the inconvenience. I tried some stuff out and I think it’s because of the routine before the textbox that registers a keyboard response. If I remove that one, I can type into the textbox. Of course this should just work, with and without, so I’ll ask around to see if there is a solution

2 Likes
#15

thank you! ill wait for you reply

#16

My report went through some curating. Made a GitHub issue about it: Textbox doesn't register input after a keyboard response · Issue #287 · psychopy/psychojs · GitHub

1 Like
#17

hi @thomas_pronk do you have any clue if someone has seen the issue?

#18

No news on a bug-fix yet, I’m afraid. I will be adding a test to our test-suite though, so that once it’s fixed, we’ll keep checking whether it stays fixed with every update.

1 Like
#19

Thank you. i tried updating to 2021.1.3 with no luck
i have removed the routing before the text box inside the loop but it still did not help.
i am still trying to figure out 2 things:

1.why e2e_textbox works smoothly even though there is a keyboard response in it.
i copied the routine to my experiment and still nothing.

  1. why long click on the keyboard does registers a keyboard response (gggg ffff etc) in the textbox.

maybe inspecting it will give me some answers, but since pavlovia is currently down i will wait

#20

I’ve been discussing it with the team. The bug is quite nuanced; by fixing one thing (let’s call it bug1) we broke something else (bug2; that’s what’s bothering you now). Then we thought up a fix for bug2, but that caused yet another bug (bug3). Could take a while before we got a proper fix in place, especially since most web-devs will still be caught up in getting Pavlovia back online again.

I do have some good news. The automated test I built, inspired on your report, is in place, so less risk these bugs will slip through the cracks.