Closed botimer closed 4 years ago
https://circleci.com/gh/samvera/hydra-editor/54?utm_campaign=vcs-integration-link&utm_medium=referral&utm_source=github-build-link leads me to believe that bundle install
has to be run either from inside of .internal_test_app
, or outside of it after engine_cart:generate
has finished.
Thanks, James. I mangled the clean-up in multiple ways... I've just restored the second bundle_for_gem
(after engine_cart and before rubocop). Let's see if it's happy this time.
Hmmm... I'm not sure what's wrong, here. I modeled the simplification on hydra-derivatives. Maybe there is something weird with the caching, how I tried to use the orb, or fold the jobs together. Do you see anything obvious, @jrgriffiniii?
Looking at https://github.com/samvera/hydra-editor/pull/168/files, I don't think that Rubocop was a part of the Travis CI build (https://github.com/samvera/hydra-editor/pull/168/files#diff-354f30a63fb0907d4ad57269548329e3L22 would just execute bundle rake
). I'm not certain what exactly is failing for the environment that leads to this.
The latest commits to https://github.com/samvera/hydra-editor/compare/jul19-ruby-versions...jrgriffiniii:jul19-ruby-versions get it passing. I'm going to experiment with another branch to try and get rubocop
working.
https://github.com/samvera/hydra-editor/pull/179 restores the build with Rubocop. I think that the redundant bundle install
invocations are necessary, but I can verify that this is the case tomorrow or Sunday.
This gem had a particularly verbose config. This attempts to trim it down and normalize with other gems that use engine_cart.
Re: samvera/maintenance#20 Re: samvera/maintenance#21