Remove version overwriting hacks

[deivid-rodriguez]

What was the end-user problem that led to this PR?

The problem was that we are doing some hacks in our version file that are... dangerous. At very least they cause "circular problems" like #6927, but I also have bad feelings about it.

What was your diagnosis of the problem?

My diagnosis was we're overwriting the version of the currently loaded bundler 😬. This hack is almost unnecessary (see the spec run here with the hack fully removed: https://travis-ci.org/bundler/bundler/builds/509497021. Only three jobs failed, and for old rubies/rubygems/bundler).

What is your fix for the problem, implemented in this PR?

My fix is to limit the hack to bundler's spec run, since it's only needed there.

Why did you choose this fix out of the possible options?

I chose this fix because fully getting the build green with the hack fully removed is a bit of work. I dedicated a bit of time to investigate one of the failures and it all comes down to the priority of loading default gems vs -I. So I think something like ruby/rubygems#1868 should fix it. But that's out of the scope of this PR.

Closes #6927.

[deivid-rodriguez]

Ok, so I looked a bit more into this, and it turns out that this "version hack" is only needed for rubygems versions the doesn't provide the BundlerVersionFinder. So I added a check for the rubygems version, and removed the hack.

While debugging this, I added a few other changes that are not directly related to this, so I have extracted them to separate PRs (#7136, #7137, #7138, #7139).

@segiddins What do you think about this? I really don't like changing the version of a loaded gem, if we can avoid it 😬...

[deivid-rodriguez]

Sorry @segiddins. This is not ready for review. A general idea of whether you think this is a good idea is appreciated though.

[deivid-rodriguez]

@segiddins This should be ready for a review now!

[deivid-rodriguez]

These version overwriting hacks bit me again in #7166, but I think, at least some of them can be fixed for all supported rubygems versions, so I'm back to working on this PR 😅. I'll re-request a review when I'm ready :)

[deivid-rodriguez]

I'm so happy I finally managed to keep this part of the diff

diff --git a/lib/bundler/version.rb b/lib/bundler/version.rb
index cfd630095..21eeeb0fd 100644
--- a/lib/bundler/version.rb
+++ b/lib/bundler/version.rb
@@ -1,23 +1,7 @@
 # frozen_string_literal: false
 
 module Bundler
-  # We're doing this because we might write tests that deal
-  # with other versions of bundler and we are unsure how to
-  # handle this better.
-  VERSION = "2.1.0.pre.1".freeze unless defined?(::Bundler::VERSION)
-
-  def self.overwrite_loaded_gem_version
-    begin
-      require "rubygems"
-    rescue LoadError
-      return
-    end
-    return unless bundler_spec = Gem.loaded_specs["bundler"]
-    return if bundler_spec.version == VERSION
-    bundler_spec.version = Bundler::VERSION
-  end
-  private_class_method :overwrite_loaded_gem_version
-  overwrite_loaded_gem_version
+  VERSION = "2.1.0.pre.1".freeze
 
   def self.bundler_major_version
     @bundler_major_version ||= VERSION.split(".").first.to_i

while also keeping CI green for all supported rubygems & rubies 🎉.

The price to pay is the admittedly ugly style of using absolute paths with autoloads, but I think it's definitely worth it. We could make it prettier if autoload_relative ever lands into core.

This should fix a bunch of problems with manually installed bundler versions coexisting with bundler versions as default gems 😃.

As per vendored dependencies, I created CocoaPods/Molinillo#130 for Molinillo, and fileutils has already been updated but changes have not yet been incorporated here (I created #7156 to do that separately). The more complicated one would be thor because it still supports 1.8.7. I requested that to be dropped in rails/thor#643 (comment). In any case, the activity in the thor repo is really low, so I think it's fine to incorporate these changes directly, and propose them upstream once they drop support.

I'd like others to chime in here 😃

[deivid-rodriguez]

Also, I think this might get green several specs that are currently skipped when running with the "ruby-core" flag.

[indirect]

This looks good to me. 👍🏻 Thanks!

[deivid-rodriguez]

Thanks @indirect!

@bundlerbot r+

[ghost]

Build succeeded

• continuous-integration/travis-ci/push