When saving a ruby-mode buffer, if the content is not plain ASCII, then the default behaviour is to add a specially-formatted comment that indicates the encoding to the Ruby interpreter.  E.g.,

# coding: utf-8

However, since Ruby 2.0 released in 2013, the default encoding for Ruby has been UTF-8.  Consequently, users of other editors tend not to include this comment when using UTF-8.  When you edit such a file with Emacs, you end up with a messy diff.

Two patches are attached to address this:

- The first patch adds a new choice to ruby-insert-encoding-magic-comment, unless-utf8, which causes the magic comment not to be inserted if the encoding is UTF-8.

- The second patch, perhaps more controversially, makes this the default.

-- 
Peter Oliver