December 2008 - Xquartz-dev - lists.macosforge.org

2.3.2-rc4 full screen oddities
by Viv Kendon 29 Dec '08

29 Dec '08

I just updated to 10.5.6 and installed 2.3.2-rc4. I've encountered some oddities with full screen mode. The first two aren't show stoppers, but might be quick and easy to fix: * if you enable full screen in the X11 prefs, then switch to it using cmd-opt-a, the bottom strip is still the mac desktop, to a height the size of the top menu bar. The missing strip really is mac desktop, clicking on it switches back to aqua/focus on finder. Another cmd-opt-a switch back and forth and it comes out the right full screen size. * at some points during the cmd-opt-a switching, the character å gets deposited into whichever X11 window gets focus. First find a window in which cmd-opt-a produces this character (or some effect) when typed directly in rootless mode. My ordinary xterms don't produce anything, but composing email in alpine does, for example. Then, in full-screen mode, with the mouse over that window, try a few switches. It gets all of them, so two appear after a switch out and back. (There are some other ffm related oddities with focus, but I'm assuming that's not high enough up the list to be worth reporting for now.) * now...full screen mode does NOT play nicely with spaces! I'm currently in full screen mode with the two X11 windows I had open on Space 3, an Aqua background, no menu bar, and with the full-screen desktop background and the X11 app I had open in Space 6 "lost" in Space 6 with no way to switch to it! I can also get the full screen desktop background interleaved with the X11 windows on first switching, I think others have reported that. And I just got the X11 desktop by itself, where in virtual spaces I don't know, but maybe on top of the X11 app that was in space 6. (How did I get into this state in the first place? well, I was trying to test the cmd-opt-a thing in another app, and RetroOffice (X11 build of NeoOffice) was open on Space 6...it's an app bundle so it is natural to start it in a different space. and yes, it does react to cmd-opt-a, by doing "select all".) OK, I know that's not how full screen is designed to be used :-) I think I'll have to find your instructions for running two copies of X11, one for full-screen and one rootless... Finally, something also appears to have changed with the cut and paste in relation to RetroOffice, I can't find a combination of pasteboard options that allows the ctr-c and ctl-v that should work within the app to function properly. There's a thread around the end of October where it was first discussed, that work around no longer appears to work (unselect both "update Pasteboard when CLIPBOARD" and "Update CLIPBOARD when Pastboard"). But I installed a RetroOffice update at the same time, so I need to do considerably more checking and come back with details. I think it works OK with fvwm as the window manager, but not with quartz-wm. White rectangle ghosts appear to be banished though, well exorcised! I still get a bazillion of these: DarwinPressModifierKey bad keycode: key=5 DarwinPressModifierKey bad keycode: key=7 and are these harmless? 19/12/2008 23:58:20 org.x.startx[5047] /usr/X11/bin/font_cache: line 207: syntax error near unexpected token `(' 19/12/2008 23:58:20 org.x.startx[5047] /usr/X11/bin/font_cache: line 207: ` echo " (-n just pertains to XFont cache, not fontconfig)"' 19/12/2008 23:58:20 org.x.privileged_startx[5056] /usr/X11/bin/font_cache: line 207: syntax error near unexpected token `(' 19/12/2008 23:58:20 org.x.privileged_startx[5056] /usr/X11/bin/font_cache: line 207: ` echo " (-n just pertains to XFont cache, not fontconfig)"' many many thanks for all the hard work!! -- Viv ________________________________________________ Dr Viv Kendon http://quantum.leeds.ac.uk/~viv tel: +44 113 343 4864 Physics and Astronomy Quantum Information Group University of Leeds

2 5

Re: [Xquartz-dev] xorg-server on Macports (works with Tiger)
by Peter Dyballa 28 Dec '08

28 Dec '08

Am 27.12.2008 um 16:00 schrieb Peter Dyballa: > There is something buggy left in the new version: when I switch off > full-screen mode from the menu, then the X clients appear on screen ? > but the ? menu bar goes away! When I then switch to a Quartz > application the X clients go away and the ? menu bar returns. The ``?´´ stands for an apple. Is there no improvement for neither the list software nor Mail to handle UTF-8 or similiar encodings? The white rectangles did not reappear, and either the apple menu bar. When the X11 clients are visible, achieved by clicking at the X11 icon in dock, the apple menu bar is invisible. When I pick a Quartz application the X clients become invisible and the apple menu bar appears. In this state only Desktop Manager "feels" the mouse cursor and appears. Dock is immune. When I press ⌘-, the preference pane of X11 appears – and the X11 clients vanish. When I switch from Mail to the X11 preference pane for a moment the X clients appear and the Quartz applications seem to be invisible. The Quartz again is visible und usable. When I switch off full-screen mode in the preference pane, blackbox WM's "dock" never appears, i.e., neither when the focus is owned by a Quartz application nor when it's owned by an X client. It's nice that I can un-bury X clients by clicking at the X11 icon in dock, but it's not nice in full-screen mode that X11 takes reliably over when I switch from X11 to a Quartz application that I then quit. Although it seems to be regular Quartz or Aqua behaviour ... -- Greetings Pete America believes in education: the average professor earns more money in a year than a professional athlete earns in a whole week. – Evan Esar

2 1

Re: [Xquartz-dev] xorg-server on Macports (works with Tiger)
by Peter Dyballa 28 Dec '08

28 Dec '08

Am 28.12.2008 um 16:00 schrieb Jeremy Huddleston: > What do you mean? Do you mean when you disable fullscreen from the > Preferences or when you toggle fullscreen from the X11 menu? Toggle. -- Greetings Pete Bake pizza not war!

1 0

Re: [Xquartz-dev] xorg-server on Macports (works with Tiger)
by Peter Dyballa 28 Dec '08

28 Dec '08

Am 27.12.2008 um 16:00 schrieb Peter Dyballa: > Using XQuartz 2.3.2 (xorg-server 1.4.2-apple29) the rectangles bug > seems to have vanished. My GNU Emacs 23.0.60, without being updated, > re-compiled, and re-installed, also does not show the prompt bug. The prompt bug still exists. And it's spreading out. In *one* Terminal window (small, reduced font size) it also exists. In Terminal's tcsh prompt is defined as: set prompt = "%{\033[43;34m%}%B%n ! /\\ %b" inside GNU Emacs' *shell* buffer it's defined as set prompt = "%n ! /\\ " and GNU Emacs takes care of colourisation. For root it is defined as set prompt = "`echo $user` ! /\\ " Another effect, which is particularly visible in root's GNU Emacs 23.0.60, is that when a process (compilation, recursive ls or find) produces lots of output and the cursor is put at the (momentary) end of output, new output does not push the prompt forward, instead the prompt is left where it landed. -- Greetings Pete Don't just do something, sit there.

2 1

Re: [Xquartz-dev] xorg-server on Macports (works with Tiger)
by Peter Dyballa 27 Dec '08

27 Dec '08

Am 26.12.2008 um 16:00 schrieb Peter Dyballa: > Right now I am upgrading to to xorg-server1.4.2_apple29_x and updates > to xinit and xextproto. There is something buggy left in the new version: when I switch off full-screen mode from the menu, then the X clients appear on screen – but the  menu bar goes away! When I then switch to a Quartz application the X clients go away and the  menu bar returns. This morning, after System Wake, Console reported: 2008-12-27 11:04:56.126 DashboardClient[647] IP Change 2008-12-27 11:04:56.177 DashboardClient[647] IP Change 2008-12-27 11:05:00.520 DashboardClient[647] IP Change 2008-12-27 11:05:00.913 DashboardClient[647] IP Change DarwinPressModifierKey bad keycode: key=7 DarwinPressModifierKey bad keycode: key=7 2008-12-27 11:05:30.449 Mail[27816] Loaded GPGMail 1.1.2 (Tiger) So still some key events are sent to X11 ... -- Greetings Pete Basic, n.: A programming language. Related to certain social diseases in that those who have it will not admit it in polite company.

2 1

Re: [Xquartz-dev] xorg-server on Macports (works with Tiger)
by Peter Dyballa 27 Dec '08

27 Dec '08

Am 26.12.2008 um 16:00 schrieb Peter Dyballa: > Right now I am upgrading to to xorg-server1.4.2_apple29_x and updates > to xinit and xextproto. Using XQuartz 2.3.2 (xorg-server 1.4.2-apple29) the rectangles bug seems to have vanished. My GNU Emacs 23.0.60, without being updated, re-compiled, and re-installed, also does not show the prompt bug. -- Greetings Pete Specifications are for the weak and timid!

1 0

Re: [Xquartz-dev] xorg-server on Macports (works with Tiger)
by Peter Dyballa 26 Dec '08

26 Dec '08

Am 10.12.2008 um 16:00 schrieb Jeremy Huddleston: > port install xorg-server # if you don't get version 1.4.2_apple27_2, > resync in a bit and try again This version has one more kind of bug: when I leave X11 in full- screen, work in some already running Quartz application, and quit it when work is done, X11 takes over and I am pushed away from Quartz. The prompt bug in GNU Emacs 23.0.60 is getting more interesting: I had running also one as root (to update in Fink after I had removed Apple's defective security update 8 from 2008) and here the cursor finds the end of the prompt. Settings are a bit different. Right now I am upgrading to to xorg-server1.4.2_apple29_x and updates to xinit and xextproto. -- Greetings Pete Perl—the only language that looks the same before and after RSA encryption. (Keith Bostic)

2 1

Re: [Xquartz-dev] 2.3.3_rc4 - cpu goes up to 100% - sampling process
by Christof Wolf 26 Dec '08

26 Dec '08

Hi Jeremy, > Is it X11.bin that is using the CPU? The sample provided in the > previous email shows it waiting almost 99% of the time... is it > another process that's using up the CPU? Or am I missing something > here... yes it is X11.bin. please read my answer to Gene :-) ciao christof -- public key www.hfph.mwn.de/~chwolf/chwolf.asc Please respect my privacy and do not make my contact information available to third parties.

2 2

Re: [Xquartz-dev] 2.3.3_rc4 - cpu goes up to 100% - sampling process
by Christof Wolf 26 Dec '08

26 Dec '08

On Thursday 25 December 2008, xquartz-dev-request(a)lists.macosforge.org wrote: > Does this condition arise randomly, or do you know what to do to > provoke it? I am still unable to reproduce it, but I swear I saw the > same thing 3-4 times after I installed rc3. It was before I knew how > to sample it, so I cannot claim it was the same, but I saw the 100% > cpu load by the idle process -- by idle I mean undisturbed by user > events. It is interesting that it did go down substantially while I > used it, and returned to 100% as soon as I left it alone. > > Also, I suspect that it is caused by something outside X11, because > whenever this condition arose, restarting the server did not help; it > went away only after system restart. Hi Gene, it goes randomly - sometimes it takes up to 3 minutes, sometimes shorter - if you do nothing on your macbook than the fan goes up - its heating up quite a bit - I did some testing today installed 2.3.2_beta1 - where the CPU does not go crazy at all www.lp-muc.com/console_msg_beta1.txt www.lp-muc.com/kcminit_beta1.crash www.lp-muc.com/ksplash_beta1.crash installed 2.3.2_rc4 - where the CPU does go crazy www.lp-muc.com/console_msg_rc4.txt www.lp-muc.com/kcminit_rc4.crash www.lp-muc.com/ksplash_rc4.crash in both cases the pastboard is disabled and no klipper is running, in both cases I got two crash reports kcminit.crash and ksplash.crash, do those have anything to do with it? ciao christof -- public key www.hfph.mwn.de/~chwolf/chwolf.asc Please respect my privacy and do not make my contact information available to third parties.

1 0

2.3.3_rc4 - cpu goes up to 100% - sampling process
by Christof Wolf 25 Dec '08

25 Dec '08

Hi Jeremy, I did disable all pasteboard options, killed klipper, no proxy in .xinitrc - but still the same pattern, here are the actual sampling process: :-( Mery Christmas :-) christof Sampling process 1748 for 3 seconds with 1 millisecond of run time between samples Sampling completed, processing symbols... Analysis of sampling X11.bin (pid 1748) every 1 millisecond Call graph: 1910 Thread_2503 1910 start 1910 main 1910 mach_msg_server 1910 mach_startup_server 1910 _Xstart_x11_server 1910 do_start_x11_server 1910 server_main 1910 X11ApplicationMain 1910 -[NSApplication run] 1910 -[NSApplication nextEventMatchingMask:untilDate:inMode:dequeue:] 1910 _DPSNextEvent 1910 BlockUntilNextEventMatchingListInMode 1910 ReceiveNextEventCommon 1910 RunCurrentEventLoopInMode 1910 CFRunLoopRunInMode 1910 CFRunLoopRunSpecific 1861 mach_msg 1861 mach_msg_trap 1861 mach_msg_trap 49 0xffffffff 49 _sigtramp 49 _sigtramp 1910 Thread_2603 1910 thread_start 1910 _pthread_start 1910 CAPThread::Entry(CAPThread*) 1910 HALRunLoop::OwnThread(void*) 1910 CFRunLoopRunInMode 1910 CFRunLoopRunSpecific 1910 mach_msg 1910 mach_msg_trap 1910 mach_msg_trap 1910 Thread_2703 1910 thread_start 1910 _pthread_start 1910 server_thread 1910 dix_main 1910 Dispatch 1763 select$DARWIN_EXTSN 1763 select$DARWIN_EXTSN 143 WaitForSomething 83 BlockHandler 59 QuartzBlockHandler 32 NSPopAutoreleasePool 18 NSPopAutoreleasePool 4 objc_assign_strongCast 4 objc_assign_strongCast 3 NSClassFromObject 2 -[NSObject class] 2 -[NSObject class] 1 object_getClass 1 object_getClass 3 objc_msgSend 3 objc_msgSend 2 objc_collecting_enabled 2 objc_collecting_enabled 1 _CFExecutableLinkedOnOrAfter 1 _CFExecutableLinkedOnOrAfter 1 object_getClass 1 object_getClass 9 objc_msgSend 9 objc_msgSend 6 -[NSAutoreleasePool init] 5 -[NSAutoreleasePool initWithCapacity:] 3 NSPushAutoreleasePool 3 NSPushAutoreleasePool 1 -[NSAutoreleasePool initWithCapacity:] 1 -[NSObject class] 1 -[NSObject class] 1 -[NSAutoreleasePool init] 4 objc_collecting_enabled 4 objc_collecting_enabled 3 QuartzBlockHandler 2 +[NSObject alloc] 2 +[NSObject alloc] 2 -[NSAutoreleasePool release] 2 -[NSAutoreleasePool release] 1 dyld_stub_objc_collecting_enabled 1 dyld_stub_objc_collecting_enabled 12 AnimCurScreenBlockHandler 10 AnimCurScreenBlockHandler 2 miSpriteBlockHandler 2 miSpriteBlockHandler 7 objc_msgSend 7 objc_msgSend 3 BlockHandler 1 RootlessBlockHandler 1 RootlessBlockHandler 1 dyld_stub_objc_msgSend 1 dyld_stub_objc_msgSend 19 WaitForSomething 18 GetTimeInMillis 17 gettimeofday 14 __gettimeofday 8 __nanotime 8 __nanotime 6 __gettimeofday 3 gettimeofday 1 GetTimeInMillis 11 WakeupHandler 8 WakeupHandler 2 QuartzWakeupHandler 2 QuartzWakeupHandler 1 NoopDDA 1 NoopDDA 6 select$UNIX2003 6 select$UNIX2003 5 __bzero 5 __bzero 1 select$DARWIN_EXTSN 1 select$DARWIN_EXTSN 2 Dispatch 1 __error 1 __error 1 memset 1 memset 1910 Thread_2803 1910 thread_start 1910 _pthread_start 1910 glvmDoWork 1910 pthread_cond_wait$UNIX2003 1910 __semwait_signal 1910 __semwait_signal 1910 Thread_2903 1910 thread_start 1910 _pthread_start 1910 DarwinProcessFDAdditionQueue_thread 1910 pthread_cond_wait$UNIX2003 1910 __semwait_signal 1910 __semwait_signal 1910 Thread_2a03 1910 thread_start 1910 _pthread_start 1910 xpbproxy_input_thread 1910 pthread_cond_wait$UNIX2003 1910 __semwait_signal 1910 __semwait_signal 1910 Thread_2b03 1910 thread_start 1910 _pthread_start 1910 select$DARWIN_EXTSN 1910 select$DARWIN_EXTSN 1910 Thread_2c03 1910 thread_start 1910 _pthread_start 1910 _xp_async_thread 1910 _xp_async_dequeue 1910 pthread_cond_wait$UNIX2003 1910 __semwait_signal 1910 __semwait_signal Total number in stack (recursive counted multiple, when >=5): 7 _pthread_start 7 thread_start Sort by top of stack, same collapsed (when >= 5): __semwait_signal 7640 mach_msg_trap 3771 select$DARWIN_EXTSN 3674 _sigtramp 49 WaitForSomething 19 objc_msgSend 19 NSPopAutoreleasePool 18 AnimCurScreenBlockHandler 10 WakeupHandler 8 __nanotime 8 __gettimeofday 6 objc_collecting_enabled 6 select$UNIX2003 6 __bzero 5 Sample analysis of process 1748 written to file /dev/stdout -- public key www.hfph.mwn.de/~chwolf/chwolf.asc Please respect my privacy and do not make my contact information available to third parties.

3 2