thoughtbot / suspenders

A Rails template with our standard defaults.
https://thoughtbot.com
MIT License
4.01k stars 530 forks source link

Circle 2.0 config #865

Closed croaky closed 5 months ago

croaky commented 7 years ago

The current Circle config for Suspenders is Circle 1.0 config. 2.0 is a Docker-ized environment.

https://circleci.com/docs/2.0/faq/ https://circleci.com/docs/2.0/migrating-from-1-2/

derekprior commented 7 years ago

Also worth noting here that even under Circle 1.0, our configuration removes the benefit of automatic parallelization. If we leave the default test step, Circle will detect we want to use RSpec and use it's automatic parallelization strategy for RSpec, provided we have slots.

Fixing this or even trying to get the most out of Circle 2.0 likely means the end to "run exactly what CI runs just by running rake", right? We'd have to manually keep these in sync?

mike-burns commented 6 years ago

@croaky @derekprior do either of you want me to leave this issue open as a reminder? Or do you mind if I close it and we can continue the discussion on the PR?

toobulkeh commented 6 years ago

Despite the problematic use of the term "Sunset":

https://circleci.com/blog/sunsetting-1-0/

We’re sunsetting CircleCI 1.0: August 31, 2018 is the final day for 1.0 builds. Refer to CircleCI 2.0 Hello World document for users and developers to get started with CircleCI 2.0.

mike-burns commented 6 years ago

We would need to change this one file: https://github.com/thoughtbot/suspenders/blob/master/lib/suspenders/generators/ci_generator.rb

iwz commented 6 years ago

@mike-burns, I think this is the file that needs to be updated: https://github.com/thoughtbot/suspenders/blob/master/templates/circle.yml.erb

timup commented 6 years ago

CircleCI 2.0 looks for configuration in a different location. Will try to submit a PR with example config this weekend. I have couple of suspenders based apps passing through 2.0, but not entirely happy with the setup.

stevepolitodesign commented 5 months ago

Closing in favor of #1135 and #1172.