Nick Miyake

Happy to submit a PR to fix, but would need to decide on an approach first. Broadly, I think there are 2 approaches that make sense: 1. If "quiet" is...

My error handling philosophy generally aligns with this as well, except for point (4) -- I typically model logging as separate from control flow, so don't use a "fatal" level...

I've been thinking about this problem, and in the abstract I think it could be solved pretty cleanly if inputs were considered in the form of packages rather than directories....

Thanks for the feedback! I think the amount of work largely depends on the design of how the packages logic interacts with the other pieces of mockery and the current...

FWIW, I'm a big fan of https://github.com/google/truth -- very fluent/legible, broad compatibility, and find that the method-based completion for matchers makes it one of the easiest to write.

@iamdanfox @carterkozak would like to revive discussion about this and tagging you folks as core Conjure members (if I'm missing others who should be looped in let me know/feel free...

@iamdanfox I think that enforcing unique package names across all of a definition would be a much larger break -- the only real upside I see is that it would...

## Updated proposal: clarify/codify that the "package" value exists only for namespacing/disambiguation, and that the Conjure must not use the package to distinguish any semantic meaning for things like access...

@iamdanfox @carterkozak do you have any objections to closing this issue? Based on our discussions in #131, it sounds like we're pretty comfortable with the current state of the world,...

@natefinch or @jaczhao any plans on picking this up? I'm running into this now as well. If I don't hear an ACK, will post a new PR that picks up...