Breaking change for fastlane in 0.16.1

[janpio]

(Note: Read all the comments first before thinking about the actual problems, possible this here in the issue itself is not the real problem or error)

Basic Info

• Faraday Version: 0.16.1
• Ruby Version: 2.3

Issue description

fastlane is using faraday (and after breaking on error classes in general with 0.16.0 which fortunately were quickly fixed by 0.16.1) we are encountering new problems with 0.16.1 when running our tests:

  1) Spaceship::Client retry re-raises when retry limit reached throwing Faraday::TimeoutError
     Failure/Error:
       stub_request(:get, test_uri).
         to_raise(error).times(times).then.
         to_return(status: status, body: body)

     NoMethodError:
       undefined method `to_ary' for Faraday::TimeoutError:Class
     # ./.bundle/ruby/2.3.0/gems/faraday-0.16.1/lib/faraday/deprecated_constant.rb:42:in `method_missing'
     # ./.bundle/ruby/2.3.0/gems/webmock-2.3.2/lib/webmock/request_stub.rb:32:in `flatten'
     # ./.bundle/ruby/2.3.0/gems/webmock-2.3.2/lib/webmock/request_stub.rb:32:in `to_raise'
     # ./spaceship/spec/client_spec.rb:34:in `stub_client_request'
     # ./spaceship/spec/client_spec.rb:108:in `block (4 levels) in <top (required)>'

https://circleci.com/gh/fastlane/fastlane/35752

I am not a ruby developer, so I might be missing something really obvious here - please let me know if this is the case. From my naive understanding, it might be that #1030 did not at back all the previous behavior (that fastlane unfortunately relies on).

(This is not happening with Ruby 2.5 which we are also testing.)

[janpio]

To add on the fun, we are also testing on Windows with ruby 2.4.6p354 (2019-04-01 revision 67394) [x64-mingw32]. There the same tests are failing, but with different error messages:

  1) Spaceship::Client retry re-raises when retry limit reached throwing Faraday::TimeoutError
     Failure/Error:
       expect do
         subject.req_home
       end.to raise_error(thrown)
       expected Faraday::TimeoutError, got #<RuntimeError: can't modify frozen String> with backtrace:
         # ./spaceship/lib/spaceship/client.rb:811:in `force_encoding'
         # ./spaceship/lib/spaceship/client.rb:811:in `log_response'
         # ./spaceship/lib/spaceship/client.rb:846:in `block in send_request'
         # ./spaceship/lib/spaceship/client.rb:620:in `with_retry'
         # ./spaceship/lib/spaceship/client.rb:844:in `send_request'
         # ./spaceship/lib/spaceship/client.rb:688:in `request'
         # ./spaceship/spec/client_spec.rb:8:in `req_home'
         # ./spaceship/spec/client_spec.rb:111:in `block (5 levels) in <top (required)>'
         # ./spaceship/spec/client_spec.rb:110:in `block (4 levels) in <top (required)>'
     # ./spaceship/spec/client_spec.rb:110:in `block (4 levels) in <top (required)>'
  2) Spaceship::Client retry retries when Faraday::TimeoutError error raised
     Failure/Error: request(:get, TestClient.hostname)
     RuntimeError:
       can't modify frozen String
     # ./spaceship/lib/spaceship/client.rb:811:in `force_encoding'
     # ./spaceship/lib/spaceship/client.rb:811:in `log_response'
     # ./spaceship/lib/spaceship/client.rb:846:in `block in send_request'
     # ./spaceship/lib/spaceship/client.rb:620:in `with_retry'
     # ./spaceship/lib/spaceship/client.rb:844:in `send_request'
     # ./spaceship/lib/spaceship/client.rb:688:in `request'
     # ./spaceship/spec/client_spec.rb:8:in `req_home'
     # ./spaceship/spec/client_spec.rb:118:in `block (4 levels) in <top (required)>'

https://ci.appveyor.com/project/KrauseFx/fastlane/builds/27733501

As it's in the same area of the code and also mentions Faraday::TimeoutError (at least the first time) I assume this is caused by a similar underlying problem.

[janpio]

Ahem, we are now also seeing test failures with Ruby 2.5 (and not on Ruby 2.3 any more):

  1) Spaceship::Client retry re-raises when retry limit reached throwing Faraday::TimeoutError
     Failure/Error:
       expect do
         subject.req_home
       end.to raise_error(thrown)

       expected Faraday::TimeoutError, got #<FrozenError: can't modify frozen String> with backtrace:
         # ./spaceship/lib/spaceship/client.rb:811:in `force_encoding'
         # ./spaceship/lib/spaceship/client.rb:811:in `log_response'
         # ./spaceship/lib/spaceship/client.rb:846:in `block in send_request'
         # ./spaceship/lib/spaceship/client.rb:620:in `with_retry'
         # ./spaceship/lib/spaceship/client.rb:844:in `send_request'
         # ./spaceship/lib/spaceship/client.rb:688:in `request'
         # ./spaceship/spec/client_spec.rb:8:in `req_home'
         # ./spaceship/spec/client_spec.rb:111:in `block (5 levels) in <top (required)>'
         # ./spaceship/spec/client_spec.rb:110:in `block (4 levels) in <top (required)>'
     # ./spaceship/spec/client_spec.rb:110:in `block (4 levels) in <top (required)>'

  2) Spaceship::Client retry retries when Faraday::TimeoutError error raised
     Failure/Error: request(:get, TestClient.hostname)

     FrozenError:
       can't modify frozen String
     # ./spaceship/lib/spaceship/client.rb:811:in `force_encoding'
     # ./spaceship/lib/spaceship/client.rb:811:in `log_response'
     # ./spaceship/lib/spaceship/client.rb:846:in `block in send_request'
     # ./spaceship/lib/spaceship/client.rb:620:in `with_retry'
     # ./spaceship/lib/spaceship/client.rb:844:in `send_request'
     # ./spaceship/lib/spaceship/client.rb:688:in `request'
     # ./spaceship/spec/client_spec.rb:8:in `req_home'
     # ./spaceship/spec/client_spec.rb:118:in `block (4 levels) in <top (required)>'

https://circleci.com/gh/fastlane/fastlane/35783

Totally not clear to me why this is happening, doesn't make sense to me. But the error messages here are consitent to the ones we are seeing on Windows - so maybe the first one was a fluke or... I have no idea seriously 🤷‍♀

[janpio]

This also breaks fastlane in production for users, e.g. fastlane/fastlane#15397

[janpio]

For now we decided to pin < 0.16.0 until this is figured out: fastlane/fastlane@596a388

[BobbyMcWho]

Same issue as #1033, I have a pending PR to fix it

[janpio]

Thanks @BobbyMcWho, I saw #1033 but didn't make the connection as the error message(s) was so different. Hope your PR #1035 will make it to 0.16.2 soon 🚀

[janpio]

Replaced by #1038