Paolo Bonzini <pbonz...@redhat.com> writes:
> On 10/07/19 10:39, Marc-André Lureau wrote: >>> Why not do this in docker.py (either as part of patch 1 or separately)? >>> Also, can you document in a comment why this is not needed with docker? >> >> Doing it in docker.py would probably mean parsing and tweaking >> arguments given to Docker.run(). Since it's a "temporary" work around, >> I would rather have it at the top-level caller, in the Makefile. > > On the other hand that splits the choice of docker vs. podman in two > places, and Python is a better place to implement workarounds. Yeah I agree we should move this trickery away from the Makefiles. -- Alex Bennée
