On Tue, 27 Mar 2018 00:16:06 +0200 René J.V. Bertin <[email protected]> wrote:
> > inherited or as set by the process. Their values are independent. > > For a program to take some action when a variable is read it would > > have to use some convention for variable reading, and that could be > > by-passed. > > Ouch, sorry, I keep making that shortcut despite what Jeremy pointed > out the other day. It's not the reading of $DISPLAY, it's accessing > the socket that is named by its value (when that's the value set by > launchd). I find that hypothesis implausible. $ echo $DISPLAY /private/tmp/com.apple.launchd.dnvCf0XLGk/org.macosforge.xquartz:0 Chrome is not an X client. It interacts with the screen via the API supplied by the OS. Why in the world would Chrome write to or read from that socket? How would it even know it exists? ISTM the questions to ask are: 1. What is the "bring to front" API call that XQuartz makes? (Because we don't think the OS is actively selecting XQuartz to be brought to the front). 2. Under what conditions is that call made? In response to some signal, perhaps? 3. What could Chrome do to provoke that condition? BTW, one more data point: the "active" application doesn't change when the XQuartz windows are brought to front. I don't know where the focus goes. I do know if I press command-tab to switch, and then -- without releasing the command key -- press the left arrow (to switch back), I land back in Chrome. --jkl _______________________________________________ Do not post admin requests to the list. They will be ignored. X11-users mailing list ([email protected]) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/x11-users/archive%40mail-archive.com This email sent to [email protected]
