Fwd: Is disabling ctrl-alt-backspace really such a good idea? - no.

[Thomas Jaeger]

Martin Pitt wrote:
> Thomas Jaeger [2009-02-12 17:16 -0500]:
>> This is not a healthy discussion.  We have people claiming that they
>> can't live without C-A-B
> 
> Nobody stops them from re-enabling it (to the contrary, there's a new
> tool "dontzap" which makes this very easy). Please keep in mind that
> we aren't discussing taking this away entirely, just changing the
> default.
This misses the point, though.  What has surfaced here, and I think
needs addressing is that C-A-B is not just used as a convenient way of
restarting the X server after recompiling it, but it's actually used as
a general-purpose something's-not-working-so-let's-kill-X key (this
might have been clear to you, but it actually surprised me).  This might
work for some people, but I find this a completely unacceptable solution
and I think we need to get to the bottom of those issues.

One thing that has been mentioned is how a process that requests an
infinite amount of memory can completely trash the system.  This is
definitely something we can get out of this discussion.  I'm not
familiar enough with what the kernel does, but I'm sure we could find
some more desktop-oriented kernel options that alleviate the problem
(deny requests for new memory earlier when we're low on memory, set a
hard limit on how much memory a process can request, maybe even a key
combination that kills the application that is currently using the most
memory).  I'm sure these kinds of issues have been discussed upstream
before.

> Also, please be aware of the fact that this is a *developer* list,
> who naturally has a differnent and more expert-oriented approach to
> using a system.
Yeah, this is why I thought I could get good descriptions of what
problems people are running into, but I guess they just don't care as
long as there's a quick way to zap X.