Closed olleolleolle closed 3 years ago
@pitr-ch 👋 I think I made one of the matrix elements pass! At what cost? We'll learn.
Update I took a look around, and could not find any obviously-broken details connected to Timecop changes.
Can you please rebase?
@chrisseaton 🟢 Rebased!
Thanks for your fix.
Does this make any of the experimentals less experimental?
Which experimental Ruby versions?
I need to try to get a grip on what Ruby versions we officially claim to support and document it and reflect it in CI.
@chrisseaton There are 2 workflows in CI, one being scheduled to run at midnight, and targets [head, truffleruby, truffleruby-head].
The regular ci.yml one includes 2.4 and up, including latest ruby: [2.4, 2.5, 2.6, 2.7, 3.0, jruby, jruby-head]
.
The gemspec requires a Ruby 1.9.3+...
(Sorry, these are just random facts, which do not contribute to making an understanding of what you mentioned.)
Counter-note: I was running around in the codebase looking for opportunities to use __dir__
, but that's 2.0. Now, there are hits for __dir__
in the code-base, in the tests. It has been backported: so, still 1.9.3
Rakefile
8: def __dir__
The "Supported Ruby Versions" section claims 2.0+ - so we could go to use __dir__
, I guess.
https://github.com/ruby-concurrency/concurrent-ruby#supported-ruby-versions
This PR asks for a Timecop with a non-exact version,
0.9
, which resolves to 0.9.4. That version supports Ruby 3.1 aka ruby-head. Timecop changelog.Background
Last commit to that line in the Gemfile was 6 years ago, with exact versions everywhere. https://github.com/ruby-concurrency/concurrent-ruby/commit/4cfd9c75ec2ee092796edfa7e4fba842c43f316b
Result
With this version of timecop, concurrent-ruby is green 🟢 on ruby-head.