Re: Announce: docker-buildpackage
Simon McVittie writes ("Re: Announce: docker-buildpackage"):
> On Wed, 02 May 2018 at 08:21:41 +0200, Johannes Schauer wrote:
> > Unfortunately, according to Martin  it is out of scope for
> > autopkgtest to also add support for making persistent changes to
> > the underlying backend. This in turn means, that an operation
> > like:
> > $ sbuild-update -udcar unstable
> > will never work for the autopkgtest backend.
As the original designer and author of the adt virt protocol, I
disagree with Martin.
IMO the right thing to do is to have a new adt-virt-server protocol
verb for "please start working with the `golden image'". Which might
or might not be supported by any particular virt server, of course,
but should be supported by ones that do some kind of snapshots.
I have disagreed with Martin before on the intended scope and
usefulness of the adt-virt protocols. Yes, this is in some sense
becoming a "virtualisation manager" but it is not competing with
things like libvirt or openstack because it's much simpler due to not
having to know how to configure or create a virt environment.
And it is for those same reasons that libvirt and openstack are not
good alternatives. No-one wants `sbuild-update -udcar unstable' to
have to involve a libvirt driver for sbuild chroots.
> This has the same predictability issues as upgrading a system
Of course I agree with this but this is what many of us do and it
should be as smooth as possible.