Luca Bruno

Nice, code so far looks fine to me. In case you want to externalize part of this (now or at some later point), I started collecting other non-FFI systemd things...

For clarity, the fetching step (`cosa fetch` above) is not supposed to be performed by rpm-ostree in this scenario, so we should take care of noting down all the required...

I realized there are some packages (e.g. `clevis`) which are calling `usermod`, so we should also intercept that and translate it to a `m` entry. Overall progress: - [x] intercept...

@jorhett this is a bug report and not a PR, so it can't really be "merged". Are you maybe looking at some specific PR? Which top-level missing directory concerns you?...

I agree that having a lossless pkgcache would be the key ingredient here. After that, I'd go for `perform these filesystem changes at checkout time` because maintaining another slightly-different cache...

Yes I got that, I'd be ok either paths. The /var cleanup isn't on a roadmap hot-path, so we aren't really in a hurry to have it. If there are...

The regression at https://github.com/coreos/rpm-ostree/issues/3182 actually happened because of the split logic for `--editor` which is described here. Thus a bugfix will intersect this too. The current method already takes a...

@martinpitt https://github.com/coreos/rpm-ostree/pull/3909

@cgwalters I'm fine to **also** have an umbrella issue in fcos-tracker to tie all the components and parts together. But I currently don't have all the required knowledge to see...

I started looking into this, and then realized it has further implications on other parts of the codebase. As a transitory measure, perhaps we can try starting to hardlink the...