-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA512 On Thu, Mar 15, 2012 at 19:42:44 GMT, Stroller wrote: > You know, there probably is a case for DBus, at least. I don't know > about this stuff, but I know that I'm generally the exception. But > you're going about it in a very bad way to make your point.
I agree. > I would have thought that amongst tmux users there were fewer people > who needed DBus than in the wider Ubuntu-using population. It's nice for things such as upower and udisks so that I don't need to login as root to do things such as suspend or mounting USB sticks (technically, I currently use autofs for USB sticks and the like, but I need to either use root to umount or wait for the timeout to get them unmounted). So a system bus makes sense in many cases. It would be nice to be able to use the local DBus session bus when ssh'ing around, but until ssh forwards DBus session buses across the pipe, I don't think tmux would have much reason to deal with it. Hopefully whatever solutions for networking crop up with Wayland will automatically forward whatever is necessary across the pipe instead of leaving ssh pick up all the envvars IPC and such may need in the future[1]. As for fixing your issue for non-ssh use cases, I'd recommend running "eval $( dbus-launch )" within xinitrc so that the entire session is using one session bus. For one-off uses (typically KDE and GNOME apps), I just do dbus-launch within one window and leave it around as needed and then kill all dbus related things later. > My problem with starting X11 apps from within tmux is that I get a > screenful of shite - diagnostic output about GTK libraries being > accessed and stuff. So I just don't tend to do it, but use a separate > window on an otherwise-empty virtual desktop -I run `ssh host -X > firefox` then ignore that terminal window completely. I'd love it if > you'd help me use tmux more effectively, though. I don't have that issue... Starting things from tmux Just Works in pretty much all cases (except after attaching to a session over ssh which either clears DISPLAY or makes it point remotely, but that's not easily solvable). - --Ben [1]From discussion on LWN, it'd be nice if Qt/GTK took care of network transparency because then they could have things like DBus use the *local* bus instead of the remote one. They could also expose URI schemes such as wayland-local:/ or wayland-remote:/ to access the local FS as well as the remote one. I firmly believe that the "it *looks* local, but it really exists remotely" solution is not the best that we can do. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) iQIcBAEBCgAGBQJPYqYCAAoJEKaxavVX4C1XQPEQAJhuATPopJ6AzXMzWokhdarc hACJMXa0KDJKNOQgN0Moh4kWSjOVI6H2gsDPvukdnlU3ashxe598nUJV1cujxBkc eKRcy8DUlBQdGAFd1/hD5lHeXhnlTctxAbf41Q2YVUNu/6utijqj4RzZcqo+npTE pdgKt6J6S0LOLHr9dQfc+4vnufQMNF9tMt9HbgTEwKk6YncXeq7PtaPh7NdZIS61 6oeGl7hOhQ+9zzpsdmKbhI/Qxtw3SdJJWuGkEyM9TVLzrCfhtSG3Ch2xHm5ffn4u KpWwEInKE+D9HmLSczGKFkUSvcFetCctvxm2sWYTfvfHSOBPeL6N4p1qCgoF2KDy XHh/mD0epT40CudB2d6oMczfI6kXpdmohDSYU6sGJgNOKVira1+S5QDBoaTV4xTz Y28e6UiBJH8BV7X0EzuC387Jgy6Sd7t284zhnrOiIByk2i8lwW304fKZOTA5hpE5 pD8TNbgAimskuquS5P3QBr0NAOCNcKjvluRVDKozuK4ab63SBkTpSbDkyTfCrRXO r1ZFfwGy3Go0Z8ylo7x8U4qfX/AxuQVow2YXzab/BcZjlcWTvM+Ja0Yj7iU8ifUz o0EakiVVuzVR4jgbVZXQL9/DyNhTyspRTM0XRxL0FbeM49kAwqZcZcoR4MKZR3Bc gbLwGb9uCiHWCivTsW2B =z0dq -----END PGP SIGNATURE----- ------------------------------------------------------------------------------ This SF email is sponsosred by: Try Windows Azure free for 90 days Click Here http://p.sf.net/sfu/sfd2d-msazure _______________________________________________ tmux-users mailing list tmux-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/tmux-users
