Ben Hoyt

We're considering this solved by the new collect-status events released in [ops 2.5.0](https://github.com/canonical/operator/releases/tag/2.5.0)

I would be opposed to adding this complexity to `ParseDuration`. It already has units (a scale, kind of like an exponent, so kind of doubles up that functionality) and this...

@jameinel I know this has been a year (time flies!), but did you just end up working around this in your charm by catching `socket.timeout`? I agree this should be...

Assigned to @IronCore864 to do together with the related doc fix here: https://github.com/canonical/operator/issues/1131

Per Madrid discussion, this seems reasonable (to add the `__setitem__` assignment ability). The raw dict would be validated and converted to a proper `RelationDataContent` object. Let's do a proof of...

Charm Tech discussed, and we would rather not implement this. It's a simple shortcut that you could easily factor out in your tests, if it's something that you're repeating over...

I don't quite understand the need for this: how does `framework.abort()` differ from `sys.exit(0)` (or `sys.exit(nonzero)` if you want an error signalled)? @PietroPasotti @beliaev-maksim

What about doing this in your charm class's `__init__`: ```python def __init__(self, *args): super().__init__(*args) if self.config_is_bork(): self.unit.status = BlockedStatus('config is bork') self.framework.commit() # if you want to save snapshots, commit,...

Yep, I just checked, and -- unlike relation data and so on -- `status-set` updates go through immediately (by design), whether or not the hook exits early, or even whether...

@PietroPasotti Good explanation in your [doc](https://discourse.charmhub.io/t/talking-to-a-workload-control-flow-from-a-to-z/6161) about why status-set updates are immediate -- nice doc BTW! I'm still not sure this is worth it. If we did want to add...