Allen Madsen

I'm not aware of any tools to test these things. Not sure about it being at a lower level either. I assumed since this lib implemented the server side of...

Yea, I can do that. The slightly complicated part is that another change removed a deprecation, so I need to remove the rest so this can go 1.0.0.

This is my workaround right now: ``` elixir defp maybe_validate_password(changeset, params, config) do new_user? = !Changeset.get_field(changeset, :id) has_password_change? = Map.get(params, "password") if new_user? || has_password_change? do Pow.Ecto.Schema.Changeset.new_password_changeset(changeset, params, config) else...

Don't feel any need to rush on this for my behalf. The workaround I have works fine. What you're proposing sounds interesting.

I have the same problem on SauceLabs. I believe this is the actual bit that fails: https://github.com/HashNuke/hound/blob/master/lib/hound/session.ex#L29 Is there a reason that there are no retries for this command?

Maybe what people are experiencing here is a hackney timeout rather than the GenServer call. You can configure the hackney timeouts with: `config :hound, http: [recv_timeout: 60000, connect_timeout: 60000]` These...

For the Phoenix part, it's actually pretty simple, because they already instrument parts of your app including the web requests, web sockets, and view rendering. So, you just provide a...

I agree, this library should be usable outside a phoenix context. It makes sense to be able to instrument any type of application.

I looked and it looks like you implemented it as a new relic plugin API instead of using the APM API. That's workable, but from my perspective inferior, because new...

Hey @ShaneKilkelly, added some tests for the various cases