| $ screen -v | Screen version 4.01.00devel (GNUa6eea7b) 2-May-06 Well I tried with that version too and I had the same results. Also there were no messages in the buffer when I scrolled up. I have checked the permissions of the pseudo terminals - it seems that when I create a new terminal in screen it appears in the /dev/pts/? where ? is a number (0,1,2,3...). The default permissions are crw-------. I changed them to have read-write permission for all users on all files in /dev/pts/ but the messages are still not shown.
| Since I understand that you have several users (physical users, | I assume) logged on to the same machine, would it make sense to test | this with two ‘physical terminals’ or even with two separate X11 | sessions (on the same machine) on vt7/vt8..? I haven't thought about having two separate X sessions. I guess that it will work as long as the terminals are not running inside screen - but if they are two separate xterm instances. Thanks, Vladimir On Sun, Jan 1, 2012 at 11:06 PM, Chris Jones <cjns1...@gmail.com> wrote: > On Sun, Jan 01, 2012 at 10:41:36AM EST, Vladimir Todorov wrote: > > > | Were you trying to tell your users they should go home & celebrate the > > | New Year...? > > > Haha. I was just trying to find a way to broadcast a message stating > > that something has happened (like an event). The 'wall' command works > > fine for the terminals outside screen. But if we say I have two > > regions in screen with two terminals. And if I execute the 'wall' > > linux command (not :wall in screen) in one of the terminals then the > > message is not shown in the second terminal in the second region (in > > the same screen session). But the message is shown on the terminals > > that are not running in screen. I tried to execute 'wall' in > > a terminal outside screen and the messages is again visible on all > > terminals not running in screen and not visible on all terminals > > running in screen. It's strange. It doesn't matter if I execute 'wall' > > with root or with my regular user - the message is not shown in either > > case. > > > I am using Debian 6 Squeeze with 2.6.39-bpo.2-686-pae kernel from > > http://backports.debian.org/. > > > Hmm.. that's odd. Since quite obviously (at least that's the way I see > it..) the superuser should be able to send a message to all users at any > time regardless of whatever options they have set that might block > messages from unprivileged users, I vagueley thought that what might be > causing this was that you were logged in as an unprivileged user when > you experienced the problem. Seems to be more to it than that. > > This is the version of GNU screen I am running (more recent than the one > that comes with ‘lenny’): > > | $ screen -v > | Screen version 4.01.00devel (GNUa6eea7b) 2-May-06 > > In the test I ran, I su'd to root in a bash session running within > screen and issued the ‘wall’ command. Then I typed ‘hello’ followed by > <Enter> and then CTRL-D to exit back to the shell & actually send the > message. > > As expected, and regardless of what apps they were running, all the > screen ‘terminals’ within my screen session (including the root shell > I had sent the broadcast from), a couple of xterms running outside > screen, and the linux console from which I had started my X11 session > showed: > > | Broadcast Message from user@host > | (/dev/pts/1) at 15:26 ... > | > | hello > > Might be worth checking the permissions on the pseudo terminals: > > | $ ls -al /dev/pts/ > > Of course, I can't see what you are actually doing, and maybe there is > an outside chance that the message does get written to what you call the > ‘second screen region’ but that switching from region #1 (where you > issued the wall command) to region #2 causes the message to disappear > before you get a chance to see it. Something that wouldn't happen if you > were sending the message to s/o else. > > You could try hitting ‘CTRL-A [’ to switch screen to ‘copy mode’ and > scroll up via, e.g. CTRL-U, see if you find your message? > > Since I understand that you have several users (physical users, > I assume) logged on to the same machine, would it make sense to test > this with two ‘physical terminals’ or even with two separate X11 > sessions (on the same machine) on vt7/vt8..? > > Otherwise, I'm not sure what's going on, but I thought I'd give > a detailed account of the way I ran my test anyway, just in case.. > > Maybe someone more familiar with quirks relative to IPC and such on > linux will be able to hazard a guess? > > CJ > > > > _______________________________________________ > screen-users mailing list > screen-users@gnu.org > https://lists.gnu.org/mailman/listinfo/screen-users >
_______________________________________________ screen-users mailing list screen-users@gnu.org https://lists.gnu.org/mailman/listinfo/screen-users
