Jesse Glick

> Move the master bug fix to a seperate pr Well, desirable practice generally, so that the obvious bug fix can get to users while we are discussing the main...

> Can you provide more details on how specifically I can provide manual verification steps? Well, something I can run locally (perhaps inside VirtualBox if host permissions matter) that would...

> Will this PR break your environment somehow? If so I may be able to add in some checks to make that less likely If there is any concern about...

> you just need to specify the user a la `docker.image('rpm-with-systemd-tester').inside('--user root:root')` Interesting, though it would be nicer to have this be an advertised option. [JENKINS-38438](https://issues.jenkins-ci.org/browse/JENKINS-38438) was recently filed, which...

I am not soliciting any new DSL functions. It is impossible to maintain the ones we already have.

I am inclined to deprecate the code touched here as being intrinsically unreliable and hard to understand, test, or fix.

CI build seems to be complaining about a SpotBugs violation I cannot reproduce locally, and which only happens on Windows: ``` launchEnv must be non-null but is marked as nullable...

ping @rsandell

So there is no way to reproduce the problem programmatically?

> the evidence suggests that the docker commands return the exit code from the child This smells like a Docker bug to me.