On Fri, 29 May 2015 12:48:17 -0700 Marc MERLIN <[email protected]> said:

> On Fri, May 29, 2015 at 08:42:29AM +0900, Carsten Haitzler wrote:
> > On Thu, 28 May 2015 22:15:18 +0200 Massimo Maiurana <[email protected]>
> > said:
> > 
> > > Marc MERLIN ha scritto il 28/05/2015 alle 22:00:
> > > 
> > > > Actually, just to be clear: I have 3 virtual screens, and I just got the
> > > > bug again. All my windows of all 3 virtual screens got centered in their
> > > > respective virtual screen.
> > > > Does this match the problem you're describing?
> > > > 
> > > > Is there any workaround? This is starting to be problematic for me :-/
> > > 
> > > There is one in current git as Raster said.
> > > I had a similar problem but in my case windows were changed in size
> > > instead of being centered, always after a resume from suspend.
> > > 
> > > In current git there is now two new options in screen setup dialog,
> > > "monitor hotplug" and "lid events"; unchecking the latter did the trick
> > > for me ;)
> > 
> > yup - so i am listening to feedback on the new randr code and as people
> > point out issues am fixing them (in some cases the only sane fix is either
> > to totally drop a feature - eg auto reconfig of screens on plug/unplug or
> > like the above - add an option if you happen to suffer from it, as drivers
> > all behave differently - for example nvidia doesn't go indicate monitors
> > are disconnected just because they are dpms turned off, but it would seem
> > intel does (which imho is a bit broken on the part of intel)).
> 
> So I found out some more. It's worse than I thought.
> 
> When this happens, I resume from sleep, xscreensaver is running and the
> screen is locked, but I don't see the unlock window.
> Instead I see all my X windows in the cneterg of their virtual screen,
> and my mouse is bound by the corners of the invisible xscreensaver
> password window.
> 
> Given that, it sounds like this new intel xorg driver is causing
> problems not just to E, but also to xscreensaver and that upgrading E
> will not fix the remaining problems like xscreensaver being confused and
> not doing its job of hiding my locked windows underneath.
> Correct?

i don't know - i dont use xscreensaver. i let e's own locker do it's thing. :)
i have no idea what xscreensaver may or may not be doing. :( but if what is
happening is what i think is (that the intel drivers are literally reporting
screen unplug events when dpms kicks in), then this explains a lot. imho it's
pretty broken to report screen unplugs unless a screen was actually unplugged.
but that is a guess. the new randr code in e has a workaround for this that at
least stops you from having ZERO screens - and as massimo said - options to
ignore the monitor hotplug events if your drivers do this.

so to some extent, your solution atm is to move to a newer e (git master), or
downgrade your intel drivers back to where they used to work.xscreensaver -
dunno.

-- 
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler)    [email protected]


------------------------------------------------------------------------------
_______________________________________________
enlightenment-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/enlightenment-users

Reply via email to