I'm not suggesting moby-engine as an alternative to podman, but as an alternative to +1.5 year old docker package
Why? because in OpenSooq.com (where I work) we have mixed environment (centos and fedora),
> Was there a
specific use case that Podman was not meeting?
my concern is that centos version of podman is too old (0.12, latest is 1.12)
and setup of user containers seems not supported (as opposed to fedora), for example
$ podman ps
ERRO[0000] cannot find newuidmap: exec: "newuidmap": executable file not found in $PATH
# yum provides '*/newuidmap'
No matches found
I've already make podman-compose (to replace docker-compose).
But I have two docker-specific projects
one of them give our developers shell access to the container or its logs (using nothing but good old ssh)
and of course this can be ported to podman easily, or even better with podman users can run containers as their own users (but in that case centos needs more recent podman stack too)
the other project listen on docker socket for events and triggers custom shell scripts or python classes or even update haproxy config from a template by reading labels
something like
docker run -d --name wp1 -l glue_http_80_host='
wp1.example.com' mywordpress/wordpress
would be added automatically to nginx or haproxy rules using templates
this can also be ported to podman as of v1.2.0 which introduced "podman events",
but the haproxy case is still tricky because there is no way to listen to events on multiple users (can be solved using systemd @ magic, let's say podman-glue myuser)
but non of those are blocker for the adoption of podman,
because we use containers on dev machines
and we don't rely on the above projects
> Is there some particular
reason why you still want/need the Moby Engine package?
no, I just want to have that option (even if in a non default repo) for convenient and easy migration and switch between old scripts/ansible playbooks, new fedora ...etc..