|
|
That's a messy commit, but it required a lot of changes to get
everything right.
* Docker: create user jekyll with UID/GID that match the user that built
the image (for seamless writes to /project).
* Docker: run the container by the current user for the same purpose.
* Docker: add an ENTRYPOINT to drop root privileges & check if the
running user is the same as the one who built the image.
* Jekyll: use --drafts.
* Makefile: add docker/logs.
As a side note, Docker + non-root users + bind mounts are a pain, I even
wrote a blog post to make sense of it all:
https://egor-tensin.github.io/blog/2020/05/06/docker-bind-mounts.html
|
