On Wed, Feb 19, 2014 at 2:39 PM, Gary Gapinski <gapinski@nasa.gov> wrote:

Our intent is to score a subset of all checklist items but have all items evaluated on all systems. This requires allč <Rule>s to be selected.

A <Profile> cannot alter the roleČ attribute of a <Rule> (to unscored rather than full).


This is an interesting use case I relate to: wanting to know results of all tests, but only really caring about a subset.

For example, the current/default XCCDF report format via OSCAP does not summarize pass/fail by severity.  In the scan I am using there are only 17 critical controls.  Don't I want my reports to guide where I put my attention first?  It's on my ist to change that for my needs.

Being able to categorize system state and then apply varying states against varying contexts is highly useful for prioritizing and analysis, IMO. Do I want to take automatic actions according to some algorithm of tests and criticality?  Is pass/fail really a binary or even a scalar operation, or is it more flexible and context driven?

I well know the anxiousness from someone forking my carefully-crafted software to use in unintended ways. I also know the joy of finding open source code that I can bend to my use case.

Openness means never having to say your sorry for forking around.  Experiments are the driving force of evolution.

Greg Elin
personal cell: 917-304-3488
personal email: greg@fotonotes.net
email: gregelin@gitmachines.com