hi, Am Dienstag, den 26.09.2006, 12:11 +0000 schrieb Václav Šmilauer: > Thanks for your reply. But the patch you mention is only partial, not > sufficient, for 2 reasons: > > 1. relies on the LTSP_CLIENT variable to be set, which is not set by > default on XDMCP connections (and would not make sense to set it, since > it is not true) and is not a clean solution in that respect; do you > expect the admin to investigate such idiosyncratic defaults (user > killing sessions of all other users) and have him prevent them on case- > by-case basis?
well, we disable xdmcp on purpose simply because its insecure and evil (its also neither enabled nor encuraged in our ltsp unless you explicitly do so via an lts.conf file). but i agree the startx script in ltsp should probably set and export this variable. > 2. (maybe this should be filed as another bug?) XFCE calls shutdown > directly, not (AFAIK) by g-p-m or pmi. you didnt talk about xfce in your initial report ... i'm not sure if xfce handles power management through g-p-m (i dounbt it though, they seem to have everything hardcoded in xfce-session) if thats the case, that makes this bug a duplicate of bug 49503 -- XDMCP user is allowed to hibernate the terminal server https://launchpad.net/bugs/62144 -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
