please enable dontzap by default

Bug #90434 reported by James Troup
20
Affects Status Importance Assigned to Milestone
xorg-server (Ubuntu)
Fix Released
Wishlist
Unassigned

Bug Description

Binary package hint: xorg

Please enable dontzap by default. It's really frustrating to lose a bunch of work because you hit the wrong key and killed X and all it's children.

Tags: bitesize
Revision history for this message
Brian Murray (brian-murray) wrote :

Thanks for your suggestion. However, the changes you are requesting aren't really a bug and require more discussion, which should be done on an appropriate mailing list or forum. http://www.ubuntu.com/community/forums/ might be a good start.

Changed in xorg:
status: Unconfirmed → Rejected
Changed in xorg:
importance: Undecided → Wishlist
status: Rejected → Confirmed
Revision history for this message
Bryce Harrington (bryce) wrote :

This has been accepted as a spec: https://wiki.ubuntu.com/XorgCtrlAltBackspace

Revision history for this message
Bryce Harrington (bryce) wrote :

We won't be implementing this for Gutsy because the spec needs to gain a stronger consensus.

Revision history for this message
Johan Dahlin (jdahlin-deactivatedaccount) wrote :

I agree with the reporter, I think DontZap should be enabled by default.
I'm using Ubuntu with the default settings and I press Ctrl-Alt-Backspace by accident quite often. I'm sure heavy users of workspaces will do the same, since if you switch workspace you hold down Ctrl and Alt. If you then want to erase text just after you switched workspace the whole X server goes down.

Revision history for this message
Adam Lebsack (alebsack) wrote :

xorg (1:7.3+8ubuntu3) hardy; urgency=low

  * dexconf:
    - Added DontZap "true" option by default

 -- Adam Lebsack <email address hidden> Mon, 31 Dec 2007 19:52:39 +0000

Changed in xorg:
assignee: nobody → alebsack
status: Confirmed → In Progress
Adam Lebsack (alebsack)
Changed in xorg:
status: In Progress → Confirmed
Revision history for this message
Timo Aaltonen (tjaalton) wrote :

If we end up using dontzap by default, it'll happen by patching the server, not by changing dexconf.

Revision history for this message
Daniel Holbach (dholbach) wrote :

Bryce: what's your take on this?

Revision history for this message
Daniel Holbach (dholbach) wrote :

Unsubscribing Sponsors team, there seems to be some disagreement about the patch.

Revision history for this message
Adam Lebsack (alebsack) wrote :

I personally don't have a problem enabling dontzap by default, but there should be a way to disable it. A patch to the actual server should just change the default value of the option from false to true, leaving the option usable in xorg.conf.

Once a decision is made, I'd be happy to attempt a patch for it.

Revision history for this message
Henrik Nilsen Omma (henrik) wrote :

Raising the priority of this since it can cause serious data loss.

Bryce, could we have your view on a simple implementation of this that changes the default for dontzap?

Changed in xorg-server:
assignee: alebsack → bryceharrington
importance: Wishlist → Medium
Revision history for this message
Bryce Harrington (bryce) wrote :

I'd also done a similar implementation of this earlier for a spec: https://wiki.ubuntu.com/XorgCtrlAltBackspace

However, after talking with the spec reviewers, it was not accepted, because there was lack of a consensus about the change, and while there were several favoring it strongly, there were others who opposed it just as strongly.

Adam, where your help could really be used here would be to help in resolving a few issues in the spec. In the comments section I listed 5 changes that should be made to better document the justification and to account for concerns that have been raised. Once that's done, then a patch can be put together; the approach I had been thinking of was to change xserver to DontZap by default and add a new option "DoZap" or whatever, to let people enable it.

Changed in xorg-server:
importance: Medium → Wishlist
Revision history for this message
Jan Niklas Hasse (jhasse) wrote :

I think CTRL+ALT+BACKSPACE should still reset the XServer as there are many tutorials out there which say something like "and now hit CTRL+ALT+BACKSPACE to restart your XServer". If this wouldn't work, many user would be helpless because they wouldn't now that this has been deactivated.

IMHO the best solution would be to just restart the XServer after holding CTRL+ALT+BACKSPACE for several seconds. This way it's unlikely you would restart your XServer by accident. Also if a newbie reads a tutorial with the advice to press CTRL+ALT+BACKSPACE, he might have the idea to hold it for several seconds.

What do you think?

Revision history for this message
Bryce Harrington (bryce) wrote :

Perhaps... but I don't think there is an easy solution to this that will make everyone satisfied. I could imagine those who support keeping the shortcut would dislike it no longer working immediately, whereas people who don't want to have X exit from inadvertent keystrokes may prefer to see the function disabled entirely, not just delayed. (But I could be wrong.)

I suspect the best solution will be involve including a gui checkbox control someplace, and a script to turn it on/off in xorg.conf, perhaps a Hardy+1/+2 sort of thing. But lets see what ideas develop consensus in the specification first.

Bryce Harrington (bryce)
Changed in xorg-server:
assignee: bryceharrington → nobody
Revision history for this message
Rthaduthd Anthnhkrc (nthnuekeu-deactivatedaccount) wrote :

this thing is a huge fail for ubuntu and linux in general .. screw backwards-compatibility wrt. blog-posts mentioning how to restart X; we're supposed to be moving forward ...... make it ctrl-alt-shift-backspace or something that requires some tiiny significant amount of effort from the user to really show that he "means it" ....... x)

i've added dontzap to xorg.conf .. bought new hardware or switched between screens, reconfigured xorg.conf, then had this crap hit me again and again as my edited xorg.conf got overwritten ..

i've tried to do what's described here:
http://www.foldr.org/~michaelw/log/programming/lisp/dontzap-emacs

..ie; "rewiring" the backspace key using a configfile in my ~ directory ... but it doesn't work, X seems to ignore this; things keep failing badly and shit keeps on hitting the fan as i manage to hit this key-combo time and time again .. ctrl-backspace is a very common key when deleting words etc. .. it's just too close ...

Revision history for this message
Jan Niklas Hasse (jhasse) wrote :

@Lars
That's why my idea is to make it necessary to hold the keys for several seconds. If a user reads "Press STRG+ALT+BACKSPACE" to restart your XServer" and it doesn't work immediately he will very likely try to hold the keys for several seconds. He wouldn't try to add SHIFT to the combination, would he?

Revision history for this message
Rthaduthd Anthnhkrc (nthnuekeu-deactivatedaccount) wrote :

to add some real life context; practically all programs dealing with text have support for ctrl-backspace:

i'm editing text in Firefox (while i'm typing this)? .. it's there
i'm editing text in OpenOffice? .. it's there
i'm editing code in Emacs? .. it's there
i'm chatting with someone on MSN (Pidgin)? .. it's there
i'm chatting with someone on IRC (X-Chat)? .. it's there

"this 'linux thing' sucks! .. it crashes(?) for no reason at all, and i wasn't even doing anything exciting; just editing text .. rofl!"

..and i _know_ this isn't just me .. i've been using this 'linux thing' since 1998 and this has _always_ been a pesky annoying issue that never ever seems to go away permanently

.. i've added "dontzap" to that xorg.conf file (or whatever it used to have as name before xorg) so many times i've lost count .. and i've seen and read so many others get hit by this nonsense "feature" too

so please, please, fix already! .. it's easy to fix manually, but it gets tiresome "re-fixing" time and time again

at least..while you're arguing for something "permanent", make the default keycombo something not so close to the commonly used ctrl-backspace combo .... :(

Revision history for this message
Bryce Harrington (bryce) wrote :

Lars, ranting about it is only going to sour people who can fix this for you and turn them off from wishing to help.

If you've read the blueprint discussion you'll see that the idea of changing the keys, adding a dialog, or turning it off entirely have all been rejected by those who favor keeping the traditional behavior. However, they appear to be *somewhat* open to the idea of a 1-2 sec delay (many I've spoken to are still uncertain, and have not given the go-ahead for implementation). However, if those who are advocating for disabling the shortcut, don't express support for the compromise and simply continue with rants or pushing for already rejected options, then I anticipate the already limited support for the compromise is going to disappear, and we'll be back to staying with the status quo.

Revision history for this message
Rthaduthd Anthnhkrc (nthnuekeu-deactivatedaccount) wrote : Re: [Bug 90434] Re: please enable dontzap by default

On Fri, Jun 27, 2008 at 10:55 PM, Bryce Harrington
<email address hidden> wrote:
> Lars, ranting about it is only going to sour people who can fix this for
> you and turn them off from wishing to help.
>
> If you've read the blueprint discussion you'll see that the idea of
> changing the keys, adding a dialog, or turning it off entirely have all
> been rejected by those who favor keeping the traditional behavior.
> However, they appear to be *somewhat* open to the idea of a 1-2 sec
> delay (many I've spoken to are still uncertain, and have not given the
> go-ahead for implementation). However, if those who are advocating for
> disabling the shortcut, don't express support for the compromise and
> simply continue with rants or pushing for already rejected options, then
> I anticipate the already limited support for the compromise is going to
> disappear, and we'll be back to staying with the status quo.

Yeah, going to add that I think the ones who want to vote for this fix
(note; in any form, it doesn't matter) mostly doesn't know the vote
exists in the first place.

..and the ones voting against it are totally clueless wrt. the "real
world". Yeah, there; I said it: you people (voting against), whoever
you are, are wrong, and I/we are right! x)

See; Ubuntu is marketed as a "Desktop OS"(
https://launchpad.net/ubuntu/+bug/1 ) and a "Desktop OS" doesn't
_instantly kill all the friggin applications_ when one press a wrong
key-combo. It's really as simple as that; there is absolutely no ..
discussion, or vote .. needed .. with regards to that.

So where does one vote for this stuff anyway? Who gets to vote? ..
What type of people know about this vote in the first place; and what
is their percentage compared to the ones who get hit most often by
this?

..anyway, I'll stop ranting. Who knows; I might get kicked/banned from
this CoC tainted web-site thingy also for having a shockingly honest
opinion. haha x)

--
Lars Rune Nøstdal
http://nostdal.org/

Revision history for this message
Rthaduthd Anthnhkrc (nthnuekeu-deactivatedaccount) wrote :

.. and it happened again .. just lost a long message i was working on ..

ok, enough .. where do i send a patch? .. this needs to go away for good ..

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

dontzap is enabled by default in upstream, so in jaunty this will be fixed.

Changed in xorg-server:
status: Confirmed → In Progress
Revision history for this message
Rthaduthd Anthnhkrc (nthnuekeu-deactivatedaccount) wrote :

excellent news! :)

On Fri, Oct 31, 2008 at 12:13 PM, Timo Aaltonen <email address hidden>wrote:

> dontzap is enabled by default in upstream, so in jaunty this will be
> fixed.
>
> ** Changed in: xorg-server (Ubuntu)
> Status: Confirmed => In Progress
>
> --
> please enable dontzap by default
> https://bugs.launchpad.net/bugs/90434
> You received this bug notification because you are a direct subscriber
> of the bug.
>

--
Lars Rune Nøstdal || AJAX/Comet GUI type stuff for Common Lisp
http://nostdal.org/ || http://groups.google.com/group/symbolicweb

Revision history for this message
Timo Aaltonen (tjaalton) wrote :

fixed in jaunty.

Changed in xorg-server:
status: In Progress → Fix Released
Revision history for this message
Raptor Ramjet (ramjet) wrote :

Disabling Ctrl+Alt+Delete by default is a dumb idea. Dumb. Dumb. Dumb.

The correct way to handle this would have been for the desktop state to be saved when the three keys are pressed to end the X session. Then when X is restarted you could be prompted as to whether you wish to recover the session. Then the people "inadvertently" using the key combination could LEARN how to use their computer.

As for moving between workspaces then it should be required that all three keys are pressed down simulataneously to have any effect. e.g. pressing Ctrl+ALt then releasing the keys and pressing backspace should not work the same as pressing all three simultaneously.

On which note I use the Ctrl+ALt+Left and Ctrl+ALt+Left combinations all the time and I have never "inadvertently" managed to kill X by pressing backspace to delete text in an editor.

FInally I have to use this key combination every single day as Xorg does not recognise my monitor correctly and I have to keep giving X the "3 finger salute" (mark 2) to get a usable X (it gets there eventually)

If you can't manage not to type this key combination then you really should question whether you're able to use a computer at all.

Disabl;ing the combination by default was a dumb idea.

Revision history for this message
Rthaduthd Anthnhkrc (nthnuekeu-deactivatedaccount) wrote :

On Sat, May 16, 2009 at 10:21 AM, Raptor Ramjet <email address hidden>wrote:

>
> The correct way to handle this would have been for the desktop state to
> be saved when the three keys are pressed to end the X session. Then
> when X is restarted you could be prompted as to whether you wish to
> recover the session.

lol .. and this would be implemented how? "Zap" is pretty much an instant ..
hard .. process .. kill. It kills Xorg and takes all the child processes
down with it. Do you understand what this implies?

Stuff like Firefox has a "Would you like to restore the tabs of your
previously "crashed" session?" feature, but, uh ....

> On which note I use the Ctrl+ALt+Left and Ctrl+ALt+Left combinations all
> the time and I have never "inadvertently" managed to kill X by pressing
> backspace to delete text in an editor.

Right. There's a chance there are other people than you out there. People
with different keyboard layouts, who use different software which use the
CTRL-ALT-and-a-key-close-to-backspace keycombo often.

FInally I have to use this key combination every single day as Xorg does
> not recognise my monitor correctly and I have to keep giving X the "3
> finger salute" (mark 2) to get a usable X (it gets there eventually)

Maybe what's causing this should be fixed instead?

> If you can't manage not to type this key combination then you really
> should question whether you're able to use a computer at all.

lol. Yeah, it would perhaps have been better to have one require to hold the
keys down for a few seconds before it triggered, but I'm not going to make
an ass out of myself by _assuming_ this would be trivial to implement just
because I think it would be a better idea.

--
Lars Rune Nøstdal
http://nostdal.org/

Revision history for this message
Jan Niklas Hasse (jhasse) wrote :

@Raptor

Where's the problem in using the "new" key combo? (Alt + Druck + K on my German keyboard)

I think it's great they disabled Ctrl+Alt+Delete since it really is possible to hit this accidentally (e.g. gedit uses Ctrl+Alt+PageUp to switch tabs).

Revision history for this message
Thomas Zehetbauer (realborg) wrote :

could you please make "dontzap" a accessability option for those who are unable to control their fingers?

today I ended up having to pull the plug because I could not find another way to get out of xorg. my /home partition was not mounted so gnome would not come up and would also not load my "dontzap=no" config

dontzap belongs to the xorg config. keep it there. users use linux because they do not want to find new ways to do common tasks with every release. ubuntu started and was successful as a better maintained debian. now it seems the focus has shifted to "change and break everything that has reliably worked for 30 years"

Revision history for this message
Jan Niklas Hasse (jhasse) wrote :

Thomas why didn't you just use the new key combo? (Alt + Druck + K on my German keyboard, see above)

The advantage is that you can also use it when your XServer crashed completely AFAIK.

Revision history for this message
damian (damian-lang) wrote :

> If you can't manage not to type this key combination then you really
> should question whether you're able to use a computer at all.

He is right you know. Its absurd than anyone could press those buttons accidentally.

If you would be to press them "accidentally" theres a bigger problem at hand..

Revision history for this message
Jan Niklas Hasse (jhasse) wrote :

@damian

My backspace key is only 1 cm away from Page Up. So it can happen if you want to switch tabs (e.g. in gedit) using CTRL+ALT+PageUp.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Duplicates of this bug

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.