RE: [openc2] Action/Target Pairings

The LS is necessary but not sufficient to define compliance.
We will need to define compliance in context of what is being called compliant - eg an actuator interface.
You then test compliance with an actuator profile which defines the subset of the language that actuator must implement. Similarly, it you are testing the orchestrator interface to a particular actuator, it also use the actuator profile.
However there are still some obvious non-compliant tests  that could be done even without the actuator profile since all the  actions and all the targets are in the LS. Ie you can define non-compliance with  just the LS. For example is you used 'block' (not in openc2) instead of  'deny' (the proper openc2 action) than it's clear you are non-compliant.
Once we get into defining compliance, and if there is concern on 'looking at two documents', maybe that will cause us to change the structure (eg maybe the current language spec and the profiles become 'parts' of one doc like Stix did it). But compliance and one/two docs are separate issues from action/target pairings (ie we already had actuator profiles).
Duncan Sparrell  sFractal Consulting LLC  iPhone, iTypo, iApologize