I can confirm this issue, with version 1.2.5, ruby 1.9.3 and the error message "invalid byte squence in UTF-8"

A possible, imperfect workaround would be to replace in lib/git/lib.rb, line 672..

    def command_lines(cmd, opts = [], chdir = true, redirect = '')
       command(cmd, opts, chdir).split("\n")
    end

...with...

    def command_lines(cmd, opts = [], chdir = true, redirect = '')
      line = command(cmd, opts, chdir)
      if !line.valid_encoding?
        line = line.unpack('C*').pack('U*')
      end
      line.split("\n")
  end

+1 Having the same issue

The solution proposed by @ut works.

+1 too, this works fine for me.

Thanks for the feedback, I made a pull request out of this workaround.

If someone wants to use the gem with this fix right now, just use my fork

Seems abandoned 🗑

Looking at the queue of the waiting PR, that's a pity.

Myself at @perlun are in the process of resurrecting the project. I don't want this issue to get lost so I'm re-openong. We've not discussed priorities yet, however internationalization is likely to be high.

It seems that this issue has been solved in the meantime. I succesfully checked it with the example provided by @fwal

That's awesome @ut, then I'll close this for real 🎉