Hi Charming folks 
I’m planning to write recommendations on using the reconciler pattern in charms.
While I’m studying your code that I found on GitHub and our analysis batch (we track 149 charms or all ages at the time of writing), I also wanted to connect to some of you in person to get your opinions: what’s better, what worked well, what needs extra care, what should in fact be featured in our recommendations.
Here’s my rough list of things to investigate:
- reconciler pattern itself (e.g. how it’s different from the holistic umbrella term)
- charm library APIs for reconciler charms
- where to draw the line between unit tests and integration tests
- performance considerations (workload, juju, maybe caching)
I’m happy to interact here, or on Matrix or to schedule a few calls.
Thank you!