- Sep 13, 2023
-
-
Tim Trense authored
- Sep 04, 2023
-
-
Tim Trense authored
-
Tim Trense authored
and some other common spellings of common files
-
- Jun 05, 2023
-
-
Tim Trense authored
-
- Jun 01, 2023
-
-
Tim Trense authored
-
Tim Trense authored
-
Tim Trense authored
-
Tim Trense authored
-
- May 31, 2023
-
-
Tim Trense authored
-
Tim Trense authored
-
Tim Trense authored
And another round of "we dont do this here". Since the `pwd` in the pages build script gives "/" (the root of the dind container), but `docker run` uses this as the host path (not the dind filesystems path), this is effectively broken. So we have to do quite some workaround by first creating a temporary container, than using `docker cp` (since that copies from the current filesystem of the shell that executes the `docker` binary), than running the container, than copying the results back out and ultimately destroying the container (even on failed builds).
-
Tim Trense authored
The "tee" step produced a file within the filesystem of the gitlab runner docker:dind container. But when executing `docker run`, the left side of the --volume option is the **host** path. Not the path from the filesystem that is valid, inside dind. Thus the `README.meta.yaml` was never actually created in the correct place to be picked up by mdcompile. With this commit, the file is not part of the CI pipeline anymore, but actual code now. This also is more logical since it is code metadata and should not be "magically" created.
-
Tim Trense authored
-
Tim Trense authored
-
Tim Trense authored
-
Tim Trense authored
-
Tim Trense authored
It also contains links to some resources for better detailing
-
Tim Trense authored
-
Tim Trense authored
-
Tim Trense authored
It is generally a good idea to version control the projects code repository configuration to make it the same for all team members.
-
Tim Trense authored
-
Tim Trense authored
-
Tim Trense authored
-