Require Mint 1.6 #287
Require Mint 1.6 #287
Conversation
| @@ -235,9 +235,6 @@ defmodule Finch.HTTP2.PoolTest do | |||
| ]) | |||
|
|
|||
| assert_receive {:resp, {:ok, _}} | |||
|
|
|||
| assert_recv_frames([rst_stream(stream_id: ^stream_id, error_code: :no_error)]) | |||
|
|
|||
There was a problem hiding this comment.
@sneako tbh I'm not sure what this whole test is/was testing, it's not asserting on:
assert_receive {:resp, {:error, %Finch.Error{reason: :request_timeout}}}like other similar tests do.
There was a problem hiding this comment.
@zachallaun blast from the past! Could you help changing this test? Or if it outlived its usefulness, we let it go?
There was a problem hiding this comment.
Nevermind, you showed up in git blame but I think you were just moving code around, sorry about that!
There was a problem hiding this comment.
I think its just testing the happy path? Seems fine to me to delete this assertion, let's do it
There was a problem hiding this comment.
@sneako per:
test "request timeout with timeout > 0 that fires after request is done"it makes sense we don't assert on any returned errors because there aren't any. Perhaps this test was asserting the side effects (a frame with no_error from Mint) but since there aren't any, perhaps it's not so useful anymore. Probably OK to to keep it though!
|
Thanks again Wojtek! |
This is in preparation for #274.
As mentioned in #274 (comment), the test change is caused by upstream change in Mint.