Difference between PUID PGID and the docker-compose user and groupadd

[dimitur2204]

Hey,

I am unsure this is the right place to ask that question, but since it is the case for most (maybe all) lscr images I figured it does make sense to ask here in the organization's discussions.

Basically the question is in the title, I struggle to understand what the difference is between PGID and PUID and docker-compose properties, from my research what I figured is that user in docker compose runs that specific container in an environment where that user is the owner, even though my docker could be running as root, it sort of isolates it. If that is the case then I struggle to understand what is the point of PGID and PUID?

Edit: I have read the page for PGID and PUID before writing the post, but it did not answer my question - https://docs.linuxserver.io/general/understanding-puid-and-pgid/?h=pgid

[thespad]

TL;DR. PUID & PGID predate user being a viable option for running containers reliably as a non-root user and are our implementation allowing root init with a switch to a non-root user for running the application itself. The user directive starts the container init as that user, rather than root, but the docker daemon is still running as root.

Some of our images do now support running as a non-root user but not all of them.