;; This is really starting to annoy me.. This workaround doesn't work either. ;; https://www.google.com/search?q=%22timed+out+waiting+for+reply+from+selection+owner%22&ie=utf-8&oe=utf-8 ;; https://lists.gnu.org/archive/html/bug-gnu-emacs/2015-01/msg00922.html New approach: - "emacs24 -nw" under screen - "emacsclient24 -c" under X This should make it possible to restart the server. Main emacs instance is then the one running in screen, which is otherwise not used. Also maybe use lucid version, as it doesn't have the bug hanging on remote X from disconnected hosts (e.g. laptop sleep). EDIT: just restart for processes, also make it easy to restart those. Maybe ther is really no way around this Erlang OTP idea of just restarting when one of those intermittent hard-to-reproduce bugs occurs. They are part of life. Deal with it!