maybe better to commit the file after each release in a github action?

@lorenzo i think it will not solve the issue. pre-commit checkouts the repository at the rev and then reads the pre-commit-hooks.yaml, which is where the version of the docker image is specified.

@lorenzo any other thoughts?

I’m not familiar with pre-commit hooks. What do you mean by “it checkouts the commit at the rev”? Do you have to keep updating the version locally?

What do you mean by “it checkouts the commit at the rev”?

from pre-commit docs

rev: the revision or tag to clone at.

A sample repository:

repos:
-   repo: https://github.com/pre-commit/pre-commit-hooks
    rev: v1.2.3
    hooks:
    -   ...

@lorenzo did i answer your question?

@foolioo Thanks, I understand what you mean. If this is merged, how do you usually tell docker to update to the latest tag when it is released? Once an image is downloaded to the cache, it will not change anymore

@lorenzo there are 3 ways:

1. you pin and bump the version in the hook with each release (of curse, this can be automated)
2. use pre-commit with the docker image directly, watch for changes, and update accordingly
3. expect the dependency in hadolint to be met

alternatives are enumerated by order of best practice.

Ok, I understand. I think it makes sense to change this. Thanks for your patience