Make sure entrypoint is chown'ed as well

[stavros-k]

Was looking something unrelated, and noticed this. With the . you only chown the WORKDIR

Proposed Changes

Please check that the PR fulfills these requirements

• Tests for the changes have been added (for bug fixes / features)
• Added yourself to AUTHORS.rst

Other questions

• Do users need to run some commmands in their local instances due to this PR
  (e.g. database migration)?

[rolandgeider]

Thanks for the PR.

I'm not sure if there was a reason why we only chown the source folder. Alternatively, couldn't we just update the chmod line above?

[stavros-k]

The chmod does not do chown actions. Or am I missreading something?

[rolandgeider]

no, but I meant that we could change the chmod +x line to chmod a+x (unless +x and a+x is already the same, I can never remember)

[stavros-k]

Aha, well yes a+x will set the execute bits for everyone (user, group, others).
But it will still be owned by root.
Tbh I don't know how it worked so far :P As the wger user isn't root.
I haven't looked deep into to check, so I can't really tell.

But I believe it's best to have it owned by the user it will be executed.
To avoid headaches in the future. With user inconsistencies.

[rolandgeider]

ahh, I see what you mean. I'll try to do some tests later today and check that everything still works

[stavros-k]

Cool! I noticed this in our CI (TrueCharts Catalog For TrueNAS Scale)
exec /home/wger/entrypoint.sh: exec format error
And thought this might be it, but this was running like this for months.
And I don't see anything changed. Anyway. If I get some time I might try to dive into it myself too

(Not sure if you can see full logs https://github.com/truecharts/charts/actions/runs/3681327880/jobs/6228221329)

[rolandgeider]

we had some problems a couple of days ago so that the arm64/amd64 images weren't being built correctly (or only one of them was so that you'd sometimes get the wrong arch), but that is fixed already. As far as I could reproduce it, you'd get the same "format error" when using the wrong image

[stavros-k]

we had some problems a couple of days ago so that the arm64/amd64 images weren't being built correctly (or only one of them was so that you'd sometimes get the wrong arch), but that is fixed already. As far as I could reproduce it, you'd get the same "format error" when using the wrong image

I see, I'll trigger a CI run to pull latest images

[stavros-k]

we had some problems a couple of days ago so that the arm64/amd64 images weren't being built correctly (or only one of them was so that you'd sometimes get the wrong arch), but that is fixed already. As far as I could reproduce it, you'd get the same "format error" when using the wrong image

I see, I'll trigger a CI run to pull latest images

This fixed it, thanks for the heads up