X Server 1.6.0 Has Been Released

R00KIE · 2009-03-03 15:01:49

imho leaving Zap enabled is a security risk. if you lock your screen and someone zaps it you remain logged in.

Can you elaborate on that? I have tried it and the only thing left running owned by me was gamserver after a C-A-B.

And there is no vc/1 and ctrl+c for me. I use a console login _but_ I have it set so it logs out as soon as X starts so if I go to vc/1 there is just a plain login screen. I don't see any security risk there.

moljac024 · 2009-03-03 15:53:18

R00KIE wrote:

imho leaving Zap enabled is a security risk. if you lock your screen and someone zaps it you remain logged in.
Can you elaborate on that? I have tried it and the only thing left running owned by me was gamserver after a C-A-B.
And there is no vc/1 and ctrl+c for me. I use a console login _but_ I have it set so it logs out as soon as X starts so if I go to vc/1 there is just a plain login screen. I don't see any security risk there.

How did you set that up ?

R00KIE · 2009-03-03 16:18:32

My .bash.profile looks like this:

. $HOME/.bashrc
if [[ -z "$DISPLAY" ]] && [[ $(tty) = /dev/vc/1 ]]; then
#startx &> xlog.log &
startx &> /dev/null &
logout
fi

I have discovered the console login on this thread http://bbs.archlinux.org/viewtopic.php?id=57677 and I have made the proper changes so it logs out after X starts _and_ everything still works as it should. deluge would fail to start if I didn't use startx &> /dev/null & (I guess some other programs might fail too).

Edit:
This was mostly a copy/paste from the thread above and some trial and error. I have looked into this just now and it doesn't seem quite right. First the DISPLAY variable will not be set so -z "$DISPLAY" will always be true as it will be always empty, the original poster said this would prevent trying to start X when it was already running. Second .... I guess it should be $(tty) == /dev/vc/1 and not just = as something is being compared there (correct me if I'm wrong). This will try to start X if you login in vc/1 only (this part works ). Because there is the ampersand at the end even if X is already running it will complain (to the linux black hole or log file) and move on. I guess I should also add some kind of unique timestamp to the log file (when and if I need to use it), any ideas are welcome.

Well with this new zapp option default behaviour and this trick the good thing I can see is that I guess it will make it very hard for some joker to kill your X session.

Last edited by R00KIE (2009-03-03 16:47:47)

Arkane · 2009-03-03 21:24:37

Dieter@be wrote:

imho leaving Zap enabled is a security risk. if you lock your screen and someone zaps it you remain logged in.

xscreensaver prevents zapping (and all other special Xorg key sequences as well).

R00KIE · 2009-03-03 23:50:18

Arkane wrote:

Dieter@be wrote:
imho leaving Zap enabled is a security risk. if you lock your screen and someone zaps it you remain logged in.
xscreensaver prevents zapping (and all other special Xorg key sequences as well).

No it doesn't, not for me at least (with X Server 1.5), tried it before posting (a few posts ago).

Arkane · 2009-03-04 07:09:28

Oh s***, you're right...

andre.ramaciotti · 2009-03-04 12:44:46

Ah, so there is a reason for zapping these keys! I guess if somebody explained it before, the number of rants and complaints like my own would decrease.

Dieter@be · 2009-03-04 13:04:44

R00KIE wrote:

imho leaving Zap enabled is a security risk. if you lock your screen and someone zaps it you remain logged in.
Can you elaborate on that? I have tried it and the only thing left running owned by me was gamserver after a C-A-B.
And there is no vc/1 and ctrl+c for me. I use a console login _but_ I have it set so it logs out as soon as X starts so if I go to vc/1 there is just a plain login screen. I don't see any security risk there.

Nice trick. Without such trick you're just back at vc/1 (logged in) after killing X.
Can you explain how you do it? post the code? Then I can start using Zap again

R00KIE · 2009-03-04 13:53:48

@Dieter@be

After fiddling around with what I've posted before I have reached the conclusion that the next piece of code will do that. Will start X with what is configured on .xinitrc if and only if you login from vc/1, if you login from any other vc it will not try to start X.

In your ~/.bash_profile put

. $HOME/.bashrc
if [[ $(tty) == /dev/vc/1 ]]; then
#startx &> xlog.log &
startx &> /dev/null &
logout
fi

Then logout and make sure GDM, KDM, Slim or any other graphical login manager is not running. Then just login in vc/1 and watch X starting. Then check what happened in vc/1.

In my .xinitrc I have just

#!/bin/sh

#
# ~/.xinitrc
#
# Executed by startx (run your window manager from here)
#

exec startxfce4

Happy zapping

andre.ramaciotti · 2009-03-04 15:30:09

I use zsh and startx &; logout doesn't work, so I did this:

if [[ -z "$DISPLAY" ]] && [[ $(tty) = /dev/vc/1 ]]; then
  screen -d -m startx
  logout
fi

Maybe there are better option flags to pass to screen, but I wrote it quickly.

Dieter@be · 2009-03-04 19:26:41

andre.ramaciotti wrote:

I use zsh and startx &; logout doesn't work, so I did this:
if [[ -z "$DISPLAY" ]] && [[ $(tty) = /dev/vc/1 ]]; then
  screen -d -m startx
  logout
fi
Maybe there are better option flags to pass to screen, but I wrote it quickly.

That's because you have 2 command delimiters after each other.

I just did

alias startx='startx & logout'

and it works fine No auto-starting of X but I don't want that anyway

andre.ramaciotti · 2009-03-04 22:16:26

Actually it was

startx &
logout

The problem here is that zsh won't quit while there's a running proccess.

ataraxia · 2009-03-04 22:52:47

andre.ramaciotti wrote:

Actually it was
startx &
logout
The problem here is that zsh won't quit while there's a running proccess.

So maybe do

startx &
disown %%
logout

ebano · 2009-03-04 23:07:50

wow.....

x.org 1.6.0 + new xf86-intel-2.6.3 rulez!!!!!!!

just upgraded here via pacman and performace almost doubled with uxa+dri2.
no more transparency bug.

nice work arch developers !

ConnorBehan · 2009-03-05 04:18:38

... and Xorg developers.

I just added Option "DontZap" "false" to my Serverflags because you people have converted me to the zapping doctrine. However I will probably never use Ctrl+Alt+Backspace.

R00KIE · 2009-03-05 08:43:11

I have seen the solution with screen and it does seem to work, I just didn't feel like using it
I am using bash and what I have posted does work ... for bash, but as it seems it will not work so well for other shells.
As for startx &; logout ..... be careful because some programs may stop working without any apparent reason. When I did this change I noticed deluge stopped working, at that time it seemed to be quite verbose about what it was doing so I _really_ needed to redirect all output somewhere (the linux black hole /dev/null was a good place) otherwise I guess it would not have anywhere to spam with what it was doing and fail to start without any error message. I guess that may be an issue with other solutions too, be careful about that, when in doubt start a X session normally without the auto logout trick and make sure that it is not causing trouble.

Lexion · 2009-03-06 00:59:51

I only zap to stop xorg.... I will have to use xbindkeys.... bugger....

muflax · 2009-03-07 20:03:30

Since I've been using Gentoo 2 years ago I dread X.org updates. They are the only ones that always cause pain and don't seem to do any good. But, this is actually the first one I'm optimistic about.

No zapping is great. Having a default shortcut that kills the X server, including all apps you've been using, that is easy to reach and can come up accidentally (I killed several games that way...), is just stupid. I really appreciate the change.
While I hate XML with a passion and find the new way to configure keyboards retarded, at least it gets rid of the old system. The old keyboard and layout handling is so udderly broken that any change is welcome. At least somebody is working on it.
And more performance never hurts. ^^

So, yay, for the first time?

Arch Linux

#51 2009-03-03 15:01:49

Re: X Server 1.6.0 Has Been Released

#52 2009-03-03 15:53:18

Re: X Server 1.6.0 Has Been Released

#53 2009-03-03 16:18:32

Re: X Server 1.6.0 Has Been Released

#54 2009-03-03 21:24:37

Re: X Server 1.6.0 Has Been Released

#55 2009-03-03 23:50:18

Re: X Server 1.6.0 Has Been Released

#56 2009-03-04 07:09:28

Re: X Server 1.6.0 Has Been Released

#57 2009-03-04 12:44:46

Re: X Server 1.6.0 Has Been Released

#58 2009-03-04 13:04:44

Re: X Server 1.6.0 Has Been Released

#59 2009-03-04 13:53:48

Re: X Server 1.6.0 Has Been Released

#60 2009-03-04 15:30:09

Re: X Server 1.6.0 Has Been Released

#61 2009-03-04 19:26:41

Re: X Server 1.6.0 Has Been Released

#62 2009-03-04 22:16:26

Re: X Server 1.6.0 Has Been Released

#63 2009-03-04 22:52:47

Re: X Server 1.6.0 Has Been Released

#64 2009-03-04 23:07:50

Re: X Server 1.6.0 Has Been Released

#65 2009-03-05 04:18:38

Re: X Server 1.6.0 Has Been Released

#66 2009-03-05 08:43:11

Re: X Server 1.6.0 Has Been Released

#67 2009-03-06 00:59:51

Re: X Server 1.6.0 Has Been Released

#68 2009-03-07 20:03:30

Re: X Server 1.6.0 Has Been Released

Board footer