(configurable?) CI resilience

If building your app on *one* platform fails, do you no longer care about the other platforms, or do you still want the other ones to try to go through so your draft Github Release has as much as possible?

I think "as much as possible" is nicer, although on balance it will probably just be wasteful?

There are some Github CI fields for configuring the fail-fast/keep-going behaviour.

Oh also we should probably make sure the CI dies immediately if a cargo-dist run fails. In the past I've seen a failed run keep going and slam into the gh commands. Not sure if that's still a thing though.

Start by locating the GitHub CI workflow that builds each platform and the cargo-dist invocation. Check the workflow's fail-fast/keep-going settings and whether a failed cargo-dist run still reaches gh commands. Done means the intended platform continuation behavior is configured and cargo-dist failures stop subsequent release commands.

Written by the indexing model from the issue text.