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

2.3.2_rc2
by Jeremy Huddleston 01 Dec '08

01 Dec '08

Happy Turkey Day. I just rolled together another release candidate for 2.3.2. http://static.macosforge.org/xquartz/downloads/X11-2.3.2_rc2.pkg I think we're pretty much set for release, but I wanted to check for any last show-stoppers. Please give this a thorough beating. The full release notes can be seen in our 2.3.2 draft release notes: http://xquartz.macosforge.org/trac/wiki/X112.3.2 The changes from 2.3.2_rc1 were minor server updates: # Improved support for running X11 without launchd # Improved support for running multiple X11 servers # Fixed dead-acute with the Czech and Greek layouts and updated packages from x.org: # xedit 1.1.2 # xproto 7.0.14 # cairo 1.8.4 # libpng 1.2.33 # libXaw 1.0.5 o libXaw.8.dylib is still provided from libXaw-1.0.4 for compatability # libXi 1.1.4 # xcb-util 0.3.1 Plus an update to libxcb which should make it play more nicely with the launchd DISPLAY. Thanks, Jeremy

7 18

GLX updates on 2.3.2_rc2
by ryan woodsmall 30 Nov '08

30 Nov '08

Jeremy, George, et al: Thanks so much for your work on X11 under Leopard! I do a lot of SSH- tunneled X forwarding from Linux servers to my MacBook at my job, and your work has made X not only more stable, but a treat to use. Kudos. As far as "play" goes, I'm attempting to make open source Wine work with straight X11, including OpenGL/Direct3D support. Wine itself works great for running the few Windows apps I need that have no Mac/ Linux counterparts, but I'm still running into some problems - likely known - with Wine running games. This issue was marked fixed with the recent 2.3.2_rc2 release: http://xquartz.macosforge.org/trac/ticket/122 I can indeed see the GLX_SGIX_fbconfig extension, but it fails to initialize automatically in Wine. When forcing Wine to use the GLX_SGIX_fbconfig path by #ifdef'ing a section of source in the X11 OpenGL driver, there's still a crash when initializing a Wine client like, say, Steam to run a Direct3D game. I'd bet this is related to the following open issues, but I could be completely wrong: http://xquartz.macosforge.org/trac/ticket/135 http://xquartz.macosforge.org/trac/ticket/140 http://xquartz.macosforge.org/trac/ticket/196 Using a Wine-ified version of the realtech Windows OpenGL Extensions viewer doesn't cause a full crash of the process, but the program warn of attempting to read from an unknown part of memory address. This looks to be an attempt to use a function pointer to a GL/GLX function that's not present. As always, debugging Wine is something of a nightmare - there are so many jumps to different areas that it's hard to even track exactly where in the code the crash is occurring. I have some experience with C and OpenGL, but little with direct X or GLX coding, so I'm grasping a bit to understand everything here. Before digging too deep for my own good in Wine, should I expect most GLX stuff to "just work" at this point, or is this an exercise in futility until the OpenGL.framework DRI work is done? As this is mostly for games, and nothing in my day-to-day work requires any OpenGL/Direct3D Windows apps, even a recommendation of just wait for the DRI framework would suffice, though any advice would be greatly appreciated. Thanks again! ryan woodsmall rwoodsmall(a)mac.com "Be well, do good work, and keep in touch." - Garrison Keillor

2 1

Re: [Xquartz-dev] Tiger fixes in 1.4.2-apple24 and 1.4.2-apple25
by Peter Dyballa 29 Nov '08

29 Nov '08

Am 29.11.2008 um 16:00 schrieb Jeremy Huddleston: > If I get Tiger setup locally, ... which shouldn't be so complicated since modern Macs (and also portables) can boot from external disks ... > I'll look into it more, but you may need to do > some digging on your own for now... I have no idea of what *I* can do further. To me it looks as if the components don't talk to each other via Mac OS X channels. Is it possible to build an X11 that works like in any other UNIX software? I don't want an half-initialised X server start without window manager when I make a mistake and invoke an X client without X actually running. The time to wait until I can kill this nuisance costs too much ... -- Greetings Pete "No man was ever taken to hell by a woman unless he already had a ticket in his pocket, or at least had been fooling around with timetables." – Archie Goodwin

1 0

Re: [Xquartz-dev] Tiger fixes in 1.4.2-apple24
by Peter Dyballa 29 Nov '08

29 Nov '08

Am 28.11.2008 um 20:01 schrieb Jeremy Huddleston: > diff --git a/privileged_startx/server.c b/privileged_startx/server.c Your extended patch made it – almost! /usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: warning prebinding disabled because of undefined symbols /usr/libexec/gcc/powerpc-apple-darwin8/4.0.1/ld: Undefined symbols: _bootstrap_strerror I patched the files app/xinit/privileged_startx/client.c:45: fprintf(stderr, "bootstrap_look_up(): %s\n", bootstrap_strerror(kr)); app/xinit-1.1.0/privileged_startx/client.c:45: fprintf (stderr, "bootstrap_look_up(): %s\n", bootstrap_strerror(kr)); as you did with xserver/hw/xquartz/mach-startup/bundle-main.c:380: fprintf (stderr, "bootstrap_look_up(): %s\n", bootstrap_strerror(kr)); xserver/hw/xquartz/mach-startup/stub.c:280: fprintf (stderr, "Xquartz: bootstrap_look_up(): %s\n", bootstrap_strerror(kr)); and xinit built! I installed with sudo. Before I had enough time to run 'sudo font_cache -s -v -f', but now one more change had happened: /usr/X11/lib/X11/xinit/privileged_startx.d contains two shell scripts! And I also have now: -rwxr-xr-x 1 pete admin 2595856 27. Nov 23:46 /Applications/ Utilities/X11.app/Contents/MacOS/X11 -rwxr-xr-x 1 pete admin 28812 27. Nov 23:46 /usr/X11/bin/Xquartz -rwxr-xr-x 1 root admin 6918 28. Nov 22:36 /usr/X11/bin/startx -rwxr-xr-x 1 root admin 24664 28. Nov 22:36 /usr/X11/bin/xinit /Library/LaunchAgents: insgesamt 4 -rw-r--r-- 1 root wheel 638 28. Nov 22:36 org.x.startx.plist /Library/LaunchDaemons: insgesamt 8 -rwxr-xr-x 1 root wheel 619 21. Mai 2008 com.distortedvista.istatmenusprocessserver.plist -rw-r--r-- 1 root wheel 479 28. Nov 22:36 org.x.privileged_startx.plist X11's behaviour is still the same: pete 391 /\ /Applications/Utilities/X11.app/Contents/MacOS/X11 X11.app: main(): argc=1 argv[0] = /Applications/Utilities/X11.app/Contents/MacOS/X11 X11.app: Could not connect to server (DISPLAY=":0", unsetting). Starting X server. X11.app: Launching /usr/X11/bin/startx: argv[0] = /usr/bin/login argv[1] = -fp argv[2] = pete argv[3] = /bin/sh argv[4] = -c argv[5] = /usr/X11/bin/startx Waiting for startup parameters via Mach IPC. Last login: Fri Nov 28 22:22:05 on tty?? Welcome to Darwin! You have new mail. pete 124 /\ while Xquartz's has changed (fixed font is found!): Xquartz :0 Xquartz: Starting X server: /Applications/Utilities/X11.app/Contents/ MacOS/X11 --listenonly X11.app: main(): argc=2 argv[0] = /Applications/Utilities/X11.app/Contents/MacOS/X11 argv[1] = --listenonly Waiting for startup parameters via Mach IPC. X11.app: do_start_x11_server(): argc=2 argv[0] = Xquartz argv[1] = :0 2008-11-28 22:53:01.606 X11[11439] Could not connect the action apps_table_cancel: to target of class X11Controller X11.app: Debug Info: kKLuchrData failed, trying kKLKCHRData. If you are using a 3rd party keyboard layout, please see http:// xquartz.macosforge.org/trac/ticket/154 Xquartz starting: X.Org X Server 1.4.2-apple25 Build Date: 20081127 _XSERVTransSocketUNIXCreateListener: ...SocketCreateListener() failed _XSERVTransMakeAllCOTSServerListeners: server already running Fatal server error: Cannot establish any listening sockets - Make sure an X server isn't already running AbortDDX Quitting Xquartz... Xquartz: start_x11_server: (ipc/mig) server died Exit 1 Again, there is no X server running. Last time it did is some days ago ... When I double-click X11's icon in Dock it leaps a few times (launches as can be seen by ps) and finally stops being responsive. There must be an alternative to a reboot ... libGL*'s install names in Mesa 7.0.4: the errors are not in the Makefile's, they have the proper lines: -id $(INSTALL_LIB_DIR)/lib$(xxx_LIB).$(xxx_MAJOR).dylib the problem is that neither configs/darwin nor configs/default define INSTALL_LIB_DIR ... LibOSMesa is not needed in /usr/X11/lib? -- Greetings Pete The box said "Use Windows 95 or better," so I got a Macintosh.

2 1

Re: [Xquartz-dev] 2.3.2_rc2
by Jeremy Huddleston 28 Nov '08

28 Nov '08

>>> I can confirm that this has something to do with changing fonts. >>> If I remove all font-related business from my xemacs startup >>> files, it opens in the right spot. If I now change the font to >>> anything else than it is, or if I change the font size, the window >>> moves to somewhere top left (top left corner of window just >>> outside the screen). This has nothing to do with xemacs startup, >>> since I can now move the window around, but changing the font (or >>> size) again moves it back to the top left of the screen. >> >> What is the output of 'quartz-wm --version'? > > $ quartz-wm --version > 1.0.0 > > Details: > - ~/.xemacs/: empty > - ~/.Xdefaults: empty > - XEmacs 21.4 (patch 21) "Educational Television" [Lucid] (i386- > apple-darwin9.2.1, Mule) > - X11: 2.3.2_rc2, quartz-wm version 1.0.0 > - OS X: 10.5.5 on an Intel MacBook > > What _does_ work: > - The initial positioning: > - xemacs -geometry -0+0 --> correctly opens window in the top right > corner > - xemacs -geometry 80x40-0+0 --> idem, different size window > - xemacs -geometry +0+0 --> correctly opens window in the top left > corner Yeah, this working was what I meant by it's been fixed... so this is a different bug now. > What _does not_ work: > - When changing either the font or the font size after moving the > window, this change makes the window jump back to its initial > position. > - xemacs -geometry +0+0: opens window correctly in top left corner > - now manually move (drag) the window out of that corner > - change the font (menu: options->font->... and pick one) > - watch the window move back into the top left corner Thanks, I can see this and know what the problem is... but it's gonna take a bit to get the fix. > Hmmm, my quartz-wm is in /usr/bin/quartz-wm. From the date, it looks > like the one from the rc2 installer (and I don't see another one > anywhere). But I replaced that with the debug version. Does this > different location of quartz-wm tell you anything? As in: did I muck > up my X11 at some point? No, it just means that I sent the email late at night and typed the wrong location. /usr/bin/quartz-wm is correct. > This quartz-wm also tells me it's version 1.0.0, is that correct? Yes. They're both reporting 1.0.0. It'll only get bumped in version on "real" releases.

1 1

Re: [Xquartz-dev] Tiger fixes in 1.4.2-apple24
by Peter Dyballa 28 Nov '08

28 Nov '08

Am 28.11.2008 um 10:19 schrieb Jeremy Huddleston: > diff --git a/privileged_startx/server.c b/privileged_startx/server.c > index d0356b5..86bfbf5 100644 > --- a/privileged_startx/server.c > +++ b/privileged_startx/server.c > @@ -153,7 +153,7 @@ int server_main(const char *dir) { > > mp = launch_data_get_machport(svc); > #else > - kr = bootstrap_check_in(bootstrap_port, bname, &mp); > + kr = bootstrap_check_in(bootstrap_port, "org.x.X11", &mp); > if (kr == KERN_SUCCESS) > return mp; This cures one problem. There is still this: server.c: In function ‘do_privileged_startx’: server.c:213: error: ‘idle_globals’ undeclared (first use in this function) server.c:213: error: (Each undeclared identifier is reported only once server.c:213: error: for each function it appears in.) server.c:216: warning: passing argument 1 of ‘fts_open’ from incompatible pointer type -- Greetings Pete <] o __o |__ o HPV, the real ___o /I -\<, |o \ -\),-% high speed! ___/\ /\___./ \___...O/ O____.....`-O-'-()--o_________________

2 1

Re: [Xquartz-dev] Tiger fixes in 1.4.2-apple24
by Peter Dyballa 28 Nov '08

28 Nov '08

Am 28.11.2008 um 04:13 schrieb Jeremy Huddleston: > Uhm, then why wasn't that in the output of your 'grep fixed /usr/X11/ > lib/X11/fonts/misc/font*'? I think I was only looking in fonts.alias ... -- Greetings Pete To drink without thirst and to make love all the time, madam, it is only these which distinguish us from the other beasts. – Beaumarchais

1 0

Re: [Xquartz-dev] Tiger fixes in 1.4.2-apple24
by Peter Dyballa 28 Nov '08

28 Nov '08

Am 23.11.2008 um 19:32 schrieb Jeremy Huddleston: > This is triggered by privileged_startx on Leopard, but that doesn't > build on Tiger... Try building xinit (you don't need to rebuild > everything) with the '--with-launchd' configure option... then let me > know how it fails. Jeremy, I'm sorry! I read and forgot this paragraph! But since I started to rebuild X11R7.4 to instal directly in /usr/X11 some hours ago and it ran into problems in the updated app/xinit, I re-configured xinit for the first time with '--with-launchd' – but the failure is the same as when without this option: server.c: In function ‘server_main’: server.c:156: error: ‘bname’ undeclared (first use in this function) server.c:156: error: (Each undeclared identifier is reported only once server.c:156: error: for each function it appears in.) server.c: In function ‘do_privileged_startx’: server.c:213: error: ‘idle_globals’ undeclared (first use in this function) server.c:216: warning: passing argument 1 of ‘fts_open’ from incompatible pointer type make[3]: *** [server.o] Error 1 -- Greetings Pete We also sponsor National Invisible Chronic Illness Awareness Week annually in September. Join the millions

2 1

Re: [Xquartz-dev] Experiences with 2.3.2_rc1
by Richard Tobin 28 Nov '08

28 Nov '08

> > What should I do to get a crash log? Is it possible to attach to > > the server process with a debugger and see what it's doing? > could be you need to enable the Crash Reporter first with / > Developer/Applications/Utilities/CrashReporterPrefs.app (on Tiger). The problem is that this isn't a crash, it's a hang. I suppose I could kill it with "kill -SEGV" or similar to get it to be treated as a crash, but that might mess up the crash log somewhat. -- Richard -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.

2 1

Re: [Xquartz-dev] Tiger fixes in 1.4.2-apple24
by Peter Dyballa 28 Nov '08

28 Nov '08

Am 27.11.2008 um 19:48 schrieb Jeremy Huddleston: > What is in your /usr/X11R7.4/lib/X11/fonts/misc/fonts.dir? You should > have this line in it somewhere: > > /usr/X11/lib/X11/fonts/misc/fonts.dir:6x13-ISO8859-1.pcf.gz -misc- > fixed-medium-r-semicondensed--13-120-75-75-c-60-iso8859-1 grep -n 6x13-ISO8859-1 /usr/X11/lib/X11/fonts/misc/fonts.dir 98:6x13-ISO8859-1.pcf.gz -misc-fixed-medium-r- semicondensed--13-120-75-75-c-60-iso8859-1 99:6x13-ISO8859-10.pcf.gz -misc-fixed-medium-r- semicondensed--13-120-75-75-c-60-iso8859-10 100:6x13-ISO8859-11.pcf.gz -misc-fixed-medium-r- semicondensed--13-120-75-75-c-60-iso8859-11 101:6x13-ISO8859-13.pcf.gz -misc-fixed-medium-r- semicondensed--13-120-75-75-c-60-iso8859-13 102:6x13-ISO8859-14.pcf.gz -misc-fixed-medium-r- semicondensed--13-120-75-75-c-60-iso8859-14 103:6x13-ISO8859-15.pcf.gz -misc-fixed-medium-r- semicondensed--13-120-75-75-c-60-iso8859-15 104:6x13-ISO8859-16.pcf.gz -misc-fixed-medium-r- semicondensed--13-120-75-75-c-60-iso8859-16 I've never seen failures when the fonts were built. Could LC_CTYPE=de_DE.UTF-8 LANG=de_DE.UTF-8 cause problems? -- Greetings Pete "A designer knows he has arrived at perfection not when there is no longer anything to add, but when there is no longer anything to take away." – Antoine de Saint-Exupéry

2 1