run identification is unclear
each time one executes a "scan" command, it creates a new run, and each time a name is required. Thing is, all runs are kept in the store, but when we query a run using the "store" command, we use a name, and if several runs have the same name, only the first created can be reached. This is confusing and shows that the interaction between filtering, caching and resuming a scan is unsound (see also #33 (closed)). What would be a better usage workflow?