I have a python script with import RPi.GPIO as GPIO. When I execute in my Raspberry Pi3 Python3 IDLE, it runs just fine with no issues.
If I launch the script with shell pi@my_user_name, it says "no module named RPi". I am trying to add this script to be executed in my /etc/rc.local file so it can start on startup.
I have other scripts that startup just fine in Python 3 with "/usr/bin/env python3 ...myscript.py".
I have installed python 3.6 correctly many times but in the shell if I do python --version, it is 2.7.9. So I believe my shell is trying to execute in python 2 regardless if I do this: python3 /home/pi/Desktop/myscript.py
I have been going at this for hours, I also did sudo -H -u pi /user/bin/python3 /home/pi/Desktop/myscript.py in the rc.local file and it does execute, however the counts are not correct from GPIO ports when compared to python3 IDLE execute.
Is this something to do with a virtualenv of where I have python install or my RPi.GPIO? Any help is appreciated. Again the script runs perfectly when I launch through python3 IDLE.
1 Answer 1
You are barking up the wrong tree.
python IS python2 - this is in accordance with python.org standards.
If you want Python3 the correct command is python3.
There is no such file/directory as /user/bin/python3, /usr/bin/python3 is the included version /usr/bin/python3 -> python3.5
I don't know why you would install Python3.6; Debian includes Python3.5 so this would already be installed, and there is little difference - See https://docs.python.org/3/whatsnew/3.6.html
The other part of your question is unanswerable without detail of your script, BUT using /etc/rc.local is not the most reliable method. This is a hangover from SysV and systemd does its best to convert/run SysV scripts.