@mcphail we’re aware of the problem, @niemeyer thought of a workaround, then with @jdstrand and @zyga-snapd we thought of a better one, and it’s on @zyga-snapd’s plate for working on asap.
For the record, the plan is, as I recall it, that as a snap refresh is attempted, if there is a process from that snap running, the refresh is postponed (within reason). At the same time we’d make SNAP_DATA and SNAP_USER_DATA point to the current symlink instead of the current revisioned directory.
We might look into alerting the user of this if possible, but I don’t think that’s part of the first iteration of the fix.