Hm, maybe the script was running in the first place. It looks fine now by calling systemctl enable athsense_startup
where the Pd patch receives data from the Python script when I reboot the Bela. I'm sorry, but I don't know what I did wrong and you were proposing different debugging methods while there was possibly no problem in the first place...
Coming back to the subject of this thread though, I do get these kinds of messages 4 mode switches detected on the audio thread!
when the Pd patch runs, plus lots of blocks dropped, even though most audio objects are hidden in subpatches with a [switch~] object in them which has the DSP turned off for that canvas. Why is this happening even now that Python does the HTTP requests and just sends the results via OSC to Pd?
Thanks for the help so far!