Fixes #27202, [ci skip]

Fix arguments passing in testing isolation

The issue affects MRI 2.2.5, MRI 2.3.3, JRuby 9.1.6.0. It can be reproduced by:

```
$ cd activemodel
$ NO_FORK=1 bundle exec rake test
```

If we wrap original arguments in quotes, it will be considered as a one big single argument.
Later, [`rake/rake_test_loader.rb`](https://github.com/ruby/rake/blob/7863b97/lib/rake/rake_test_loader.rb#L15)
will iterate over ARGS and try to require that huge single "argument" (which is a list of multiple .rb files).
This leads to an exception:

```
/Users/kir/Project
s/opensource/rails/vendor/bundle/gems/rake-11.3.0/lib/rake/rake_test_loader.rb:15:in `require': cannot load such file -- /Users/kir/Projects/opensource/rails/activemodel/test/cases/
attribute_assignment_test.rb [stripped] /Users/kir/Projects/opensource/rails/activemodel/test/cases/validations/with_validation_test.rb /Users/kir/Projects/opensource/rails/activemodel/test/cases/validations_test
.rb (LoadError)

        from /Users/kir/Projects/opensource/rails/vendor/bundle/gems/rake-11.3.0/lib/rake/rake_test_loader.rb:15:in `block in <main>'
        from /Users/kir/Projects/opensource/rails/vendor/bundle/gems/rake-11.3.0/lib/rake/rake_test_loader.rb:4:in `select'
        from /Users/kir/Projects/opensource/rails/vendor/bundle/gems/rake-11.3.0/lib/rake/rake_test_loader.rb:4:in `<main>'
```

Originally quotes were introduced in https://github.com/rails/rails/pull/19819 to fix MRI 2.2.2.

The fix solves issue on all affected platforms: MRI 2.2.5, MRI 2.3.3, JRuby 9.1.6.0.

While working on updating Paper Trail for 5.1 compatibility, I noticed
that I was required to pass a second argument to `attribute`. I didn't
intend for this to be the case, as `attribute :foo` is totally
reasonable shorthand for "I want `attr_accessor :foo`, but also have it
work with things like `.attributes` and `ActiveRecord::Dirty`"

I had pointed the messages at the new behavior, not the old.

use `Gem.win_platform?` to check windows Ruby platforms

reactor_running? will be true just after the thread enters
EventMachine.run; reactor_thread only gets set after the internal
initialize_event_machine method has been called, the C extension is set
up, and it is entering its run loop.

`Gem.win_platform?` check if it is Windows more accurately.
Ref: https://github.com/ruby/ruby/blob/ruby_2_2/lib/rubygems.rb#L945..L952

Make sure the file doesn't exist until we've finished writing it.

removed `@current` as it is not used

fix typo in changelog

Fix typo in Rails 5.0 release notes – “when when”

[ci skip]

Explicitly document 'false' as value for 'action_on_unpermitted_parameters'

[ci skip]

Prior to this commit, `3.months - 3.months` would result in a duration
that has the "parts" of `[[:months, 3], [:months, -3]]`. This would mean
that it was subtly different than `2.months - 2.months`. When applied to
a time, the date might actually change if the resulting day doesn't
exist however many months in the future, even though in both cases we
just wanted to add `0`, which should always be an identity operation.

With this change, we now store the parts as a hash, so `3.months -
3.months` is simply stored as `{ months: 0 }`.

Fix that unsigned with zerofill is treated as signed

Add Yarn support in new apps using --yarn option

Add note about step needed to get tests to pass if you implemented basic auth [ci skip]

Add rubocop to Gemfile

use public Module#include instead of send :include

I found this issue and wanted to give the fix.

fixing update_all and delete_all when chained with left_joins

Mention the Accept header and how that figures into the request format.

Wrongly added when fixing the request path wrangling.