Re: Triggering gnome screen-saver timeout from console keypresses?



Mario Lang, le Wed 13 Aug 2008 20:57:34 +0200, a écrit :
> Samuel Thibault <samuel thibault ens-lyon org> writes:
> > Unfortunately there is no way for the X server to know that things are
> > being typed in the console, since it's the kernel that handles that.
> 
> It is true that the kernel handles that, but it isn't quite true that
> there is "no way" for the X server to still monitor keypresses.
> One technically doable solution would be for the X server
> to listen in on the linux input event interface.

Right...  And brltty would like to do that too, and another program too,
and the interface is cumbersome to achieve that, to say the least.  It
looks to me like a feature is missing here.

One problem is then to get the screen saver know that there is activity.
This has been raised several times on the Xorg mailing list and the
conclusion is that it's a mess: each screen saver has its own way to
detect idleness...

> > The X server knows when it loses focus however.
> 
> Well, thats of no use here since if the X server looses focus
> but no further keyboard interaction happens, the screen saver
> should still start to protect the users account...

That's why I said some people may not be happy with such a solution,
yes.

> I only see the prompt on my display and moving up/down does
> immediately give me the "end of screen" beep, as if there was just one
> line in the terminal.

Well, at the beginning the AT-SPI terminal has only one line yes.

> After a bit of fiddling back and forth, brltty seems to die again

Ah, so there are still bugs in the text to terminal translation... :/
We really need an implementation of the AT-SPI terminal interface to
simplify all of that code.

> and I see the Orca representation of that terminal...  Maybe you can
> tell me what I need to do to get this working?

Fix bugs :/

> >> The X11 screen saver has kicked in every time I switch back to my X11
> >> console (Alt+F7).
> >
> > It's not only that, but animated screen savers may also eat CPU...
> 
> Well, that might be true, but is also easy to fix by configuring
> an appropriate screen saver setting I guess?

What do you mean?

I'm not only thinking about blind people here.
Some people do like to see fancy things on their screen.  That eats CPU
but that's fine since the machine is supposed to be idle.  Now, if the
screen saver starts to eat CPU while they are working on text consoles,
that's a problem, which can not be solved through the configuration.

> > These screen savers have always been a problem, e.g. while playing a
> > movie in a webbrowser, etc. :/
> 
> >>  1.) A hack that I could use to fix this ASAP?
> >
> > Disable the screen saver :)
> 
> I am afraid my company security policy does not allow for that.

Mmm, do you really always remember to either close all your text
consoles (but then you could switch to VT7) or use vlock (in which
latter case there is no need for an X lock)?

> Besides, I think the timeout feature is a useful one, it just needs
> to work properly :-).

Sure.  The problem is that you are asking for it to work globally, while
it has always been meant to work only for a given xserver.

Samuel


[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]