Thank you for this wonderfully written, clear article. I got so much out of it.

I like how you elaborate on some of Hane and Atherton's concepts and methods, e.g. in "Expression", the use of block diagrams and priority guides. And in "Meaning", the way you identified core concepts and conceptual groupings in the domain model.

I've recently started with a government agency tasked with managing COVID-19 rules and guidelines. One challenge I've identified is that COVID-19 remains an 'unstable domain'.

Relationships between concepts change frequently. For example, months ago, a "rule" related to the vaccination status of the whole population – e.g. "we'll remove these measures when the state is over 80% double vaccinated". Before this, the "rule" related to case numbers of a specific area ("this local government area must do these things while their test positivity rate is over x%"). More recently, the rules are individual, relating to someone's status as a close contact or having contracted COVID-19, or their individual vaccination status. There are also other challenges endemic of government organisations, such as ambiguous content ownership.

My current strategy is to use a process similar to yours, even if the content may not be formally structured. In other words, we can identify the types of content we so frequently need to update and publish, and define attributes like "who this applies to", relationships etc., even if they're not discrete fields in the CMS.

I'm rambling lol, but keen to get your take on the above. Love the approach, but how do you think it'd fair in a government environment, managing an unstable domain like COVID-19?