Peter Solnica

If you're just seeding data for specs, then I'd recommend using `Relation#insert`, this way your setup will be much simpler and faster. We're still discussing how to approach test data...

Does it work when you define the schema explicitly? I think this is the only way to make it work now. Schemas are inferred from canonical db schemas, and if...

> Is that the correct usage? (required(:hash_field).hash) For now, yes, but in 2.0.0 it won't work. `value(:hash)` should work, so this is a bug. I'm assigning this to 2.0.0 milestone,...

@mereghost wow that's some major investigation right there! 🙇🏻

This is expected since it's not a regular word that includes an acronym that you need to register in your inflector. I'm not sure if registering just the acronym or...

@deepj yeah but the path to files does not match the constant name, `lib/cloudnats/http.rb` should be `lib/cloud_nats/http.rb`. Whatever you do, you need to make sure that the inflector correctly converts...

> Suggestion, dry-rails may would expose Zeitwerk inflection API yeah I've thought about it. In general I would like to make dry-rails zeitwerk-only eventually, because it's a much better solution...

> And this is dangerous to set in Rails Inflector because it can affect 3rd-party gems which can use HTTP acronym in their names. It's safer to use Zeitwerk inflection...

Thank you for reporting this and especially for checking it in a plain dry-system setup. I'll look into this, it's very likely that test mode in Rails will require some...

@stefanrendevski you definitely want to set up stubs in a before hook and the rspec integration (that we don't have yet) should make sure that container stubbing is setup/teared down...