to compare plans to reality and to maybe give others ideas on how much time is needed/used

because otherwise the output hasn't changed and it would waste energy on uberspace server

this is actually a bit sad since I don't want to specify in my repo which runner to use. It should be possible with docker and shell executor

because the master branch is the one people should comment on

more and more details will come when I notice them

copied from issue in gitlab

taken from Joshuas master thesis which is turn taken from the default from the Mathfaculty

more professional?

this way I always know which version someone commented on if they comment directly into the pdf or the printed version

This reverts commit c5a4a890.
Microtype caused errors for some weird reason.