Hi everyone, I've released XQuartz 2.7.2_rc2. The main reason for this release is to address package signing. 2.7.2_rc1 (and I believe the beta that preceded it) were signed in a way that some users were unable to install the package without first installing the appropriate intermediate root CA into their system keychain. With this release, I changed how the package was signed to bundle the intermediate with the package, so that should alleviate such issues. If you have difficulty installing this package, please let me know. Other than that, there aren't many changes since rc1. I pulled in an updated version of libXi and some additional mesa patches for better logging and what I hope may be a fix for the GLX regression reportedly happening with Oblivion running in wine (http://xquartz.macosforge.org/trac/ticket/575). If you see any issues, please bring them to my attention. I think this is very close to what the final 2.7.2 will be. Thanks, Jeremy http://xquartz.macosforge.org/trac/wiki/X112.7.2
Am 10.05.2012 um 05:37 schrieb Jeremy Huddleston:
Hi everyone,
I've released XQuartz 2.7.2_rc2.
I have interesting effects... The cursor can disappear. For example when I'writing in Mail no cursor is visible. When I now leave this window and pick an X client it can happen that the cursor does no reappear. But it can affect the red-yellow-green button of windows or make Dock appear. But what puzzles me more is that XQuartz 2.7.2_rc2 (xorg-server 1.12.1.901) has these preferences set and pressing alt-7 produces |: With XQuartz 2.7.2 (xorg-server 1.12.1.901) from MacPorts these preferences are set to produce | by pressing alt-7: -- 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
Do you have a ~/.Xmodmap? Those settings will override anything set in your preferences. As far as the cursor appearing and disappearing, /shrug ... that's bizarre. --Jeremy On May 13, 2012, at 06:55, Peter Dyballa <Peter_Dyballa@Web.DE> wrote:
Am 10.05.2012 um 05:37 schrieb Jeremy Huddleston:
Hi everyone,
I've released XQuartz 2.7.2_rc2.
I have interesting effects...
The cursor can disappear. For example when I'writing in Mail no cursor is visible. When I now leave this window and pick an X client it can happen that the cursor does no reappear. But it can affect the red-yellow-green button of windows or make Dock appear.
But what puzzles me more is that XQuartz 2.7.2_rc2 (xorg-server 1.12.1.901) has these preferences set and pressing alt-7 produces |:
<XQuartz Prefs.png>
With XQuartz 2.7.2 (xorg-server 1.12.1.901) from MacPorts these preferences are set to produce | by pressing alt-7:
<XQuartz Ports Prefs.png>
-- 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
_______________________________________________ Xquartz-dev mailing list Xquartz-dev@lists.macosforge.org http://lists.macosforge.org/mailman/listinfo.cgi/xquartz-dev
On 13 May 2012, at 21:49, Jeremy Huddleston wrote:
As far as the cursor appearing and disappearing, /shrug ... that's bizarre.
I also experience that occasionally, but I never realised it might be related to XQuartz. I'll try to keep an eye on it in the future to see whether it also happens if XQuartz is not running. Jonas
Am 13.05.2012 um 21:49 schrieb Jeremy Huddleston:
Do you have a ~/.Xmodmap? Those settings will override anything set in your preferences.
Yes, I have. IS it working only in one X server?
As far as the cursor appearing and disappearing, /shrug ... that's bizarre.
I applied the latest security update and installed the latest Safari. After that I had a first bizarre encounter: the screen was used completely, other than after the Thunderbolt Software Update 1.1 which reduced it to one half or less, but it's size in pixels was just one fourth (or sixteenth) and text and (login) tools had four times the normal size. This was even left in Finder preferences: the windows were shifted into a small area of the screen when it was reset to the real physical size again. After that I first launched the updated XQuartz 2.7.2_rc2 and had that cursor effect. Now I'm using MacPorts' XQuartz... -- Greetings Pete A morning without coffee is like something without something else.
On May 13, 2012, at 13:36, Peter Dyballa <Peter_Dyballa@Web.DE> wrote:
Am 13.05.2012 um 21:49 schrieb Jeremy Huddleston:
Do you have a ~/.Xmodmap? Those settings will override anything set in your preferences.
Yes, I have. IS it working only in one X server?
No. Both use it, but it can override that preference.
Am 13.05.2012 um 23:14 schrieb Jeremy Huddleston:
Yes, I have. IS it working only in one X server?
No. Both use it, but it can override that preference.
This overriding should then produce the same effect in both servers, because ~/.Xmodmap has since years: clear Mod1 keycode 66 = Mode_switch In XQuartz 2.7.2_rc2 I have to *activate* "Option keys send Alt_L and Alt_R" in order to have Alt-7 (Shift-7 is / on my German keyboard) produce | (and to have Alt-Shift-7 produce \). In XQuartz 2.7.2 from MacPorts I have to *deactivate* "Option keys send Alt_L and Alt_R" in order to have Alt-7 produce | (and to have Alt-Shift-7 produce \). I need to have contradicting settings with both identical documentation/explanation to have the same effect. Something must be wrong – and I presume my Mac is a piece of hardware and not something living. Which would explain the effect. -- Greetings Pete Encryption, n.: A powerful algorithmic encoding technique employed in the creation of computer manuals.
On May 13, 2012, at 14:40, Peter Dyballa <Peter_Dyballa@Web.DE> wrote:
Am 13.05.2012 um 23:14 schrieb Jeremy Huddleston:
Yes, I have. IS it working only in one X server?
No. Both use it, but it can override that preference.
This overriding should then produce the same effect in both servers, because ~/.Xmodmap has since years:
clear Mod1 keycode 66 = Mode_switch
In XQuartz 2.7.2_rc2 I have to *activate* "Option keys send Alt_L and Alt_R" in order to have Alt-7 (Shift-7 is / on my German keyboard) produce | (and to have Alt-Shift-7 produce \).
In XQuartz 2.7.2 from MacPorts I have to *deactivate* "Option keys send Alt_L and Alt_R" in order to have Alt-7 produce | (and to have Alt-Shift-7 produce \).
I need to have contradicting settings with both identical documentation/explanation to have the same effect. Something must be wrong – and I presume my Mac is a piece of hardware and not something living. Which would explain the effect.
Weird, I dunno. Move aside ~/.Xmodmap, setup both servers identically, and compare the output of 'xmodmap -pke' on both.
Am 13.05.2012 um 23:46 schrieb Jeremy Huddleston:
Move aside ~/.Xmodmap, setup both servers identically, and compare the output of 'xmodmap -pke' on both.
It seems as if the XQuartz 2.7.2_rc2's preferences list was defective. Meanwhile it's repaired. It seems as if my sparse ~/.Xmodmap is not needed. XQuartz 2.7.2_rc2 seems to produce a font path setting with the elements /opt/X11/share/fonts/OTF, /Library/Fonts, /System/Library/Fonts not ending with a slash and one element, "built-ins", that causes this report in Console: May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: xset: bad font path element (#17), possible causes are: May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Directory does not exist or has wrong permissions May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Directory missing fonts.dir May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Incorrect font server address or syntax MacPorts' XQuartz 2.7.2 has only /Library/Fonts and /System/Library/Fonts not ending with a slash, "built-ins" causes no report. But then I have this: May 14 00:42:48 sumac org.macports.X11.stub[14409]: launch_msg("CheckIn") IPC failure: Operation not permitted With both X servers I get May 14 00:43:07 sumac /usr/local/bin/emacs-23.3[14449]: BUG in libdispatch: 10K549 - 1986 - 0x4 Is it something I should report? Mac OS X 10.6.8, intel x86_64. -- Greetings Pete One person with a belief is a social power equal to ninety-nine who have only interests. – John Stuart Mill
On May 13, 2012, at 15:53, Peter Dyballa <Peter_Dyballa@Web.DE> wrote:
Am 13.05.2012 um 23:46 schrieb Jeremy Huddleston:
Move aside ~/.Xmodmap, setup both servers identically, and compare the output of 'xmodmap -pke' on both.
It seems as if the XQuartz 2.7.2_rc2's preferences list was defective. Meanwhile it's repaired. It seems as if my sparse ~/.Xmodmap is not needed.
Ok, so the issue is resolved? Can you give more specifics to cure my curiosity?
XQuartz 2.7.2_rc2 seems to produce a font path setting with the elements /opt/X11/share/fonts/OTF, /Library/Fonts, /System/Library/Fonts not ending with a slash and one element, "built-ins", that causes this report in Console:
May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: xset: bad font path element (#17), possible causes are: May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Directory does not exist or has wrong permissions May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Directory missing fonts.dir May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Incorrect font server address or syntax
Which is element 17 in the list? Is that directory missing a fonts.dir?
MacPorts' XQuartz 2.7.2 has only /Library/Fonts and /System/Library/Fonts not ending with a slash, "built-ins" causes no report.
Ok.
But then I have this:
May 14 00:42:48 sumac org.macports.X11.stub[14409]: launch_msg("CheckIn") IPC failure: Operation not permitted
Expected.
May 14 00:43:07 sumac /usr/local/bin/emacs-23.3[14449]: BUG in libdispatch: 10K549 - 1986 - 0x4
Is it something I should report?
That message just indicates that kevent() returned EINTR. This is harmless; just ignore it. I believe Lion's libdispatch no longer prints that message.
Am 14.05.2012 um 01:09 schrieb Jeremy Huddleston:
Ok, so the issue is resolved? Can you give more specifics to cure my curiosity?
I just had to deactivate a few times that option to make it find its way reliably into the PLIST file. I think there was some interaction with that sticky active state and the disappearing cursor. Reenabling ~/.Xmodmap does not change anything, so it must have been an effect in the X server itself. And the cursor has no tendency to disappear again... I edited some X11 related PLIST files with Xcode's Plist editor to make menu entries equal...
XQuartz 2.7.2_rc2 seems to produce a font path setting with the elements /opt/X11/share/fonts/OTF, /Library/Fonts, /System/Library/Fonts not ending with a slash and one element, "built-ins", that causes this report in Console:
May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: xset: bad font path element (#17), possible causes are: May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Directory does not exist or has wrong permissions May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Directory missing fonts.dir May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Incorrect font server address or syntax
Which is element 17 in the list?
It's that "built-ins" element.
Is that directory missing a fonts.dir?
No, it's just an obviously unrelated name like "built-ins". -- Greetings Pete I love deadlines. I love the whooshing noise they make as they go by. – Douglas Adams
On May 13, 2012, at 16:45, Peter Dyballa <Peter_Dyballa@Web.DE> wrote:
XQuartz 2.7.2_rc2 seems to produce a font path setting with the elements /opt/X11/share/fonts/OTF, /Library/Fonts, /System/Library/Fonts not ending with a slash and one element, "built-ins", that causes this report in Console:
May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: xset: bad font path element (#17), possible causes are: May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Directory does not exist or has wrong permissions May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Directory missing fonts.dir May 14 00:19:19 sumac org.macosforge.xquartz.startx[11811]: Incorrect font server address or syntax
Which is element 17 in the list?
It's that "built-ins" element.
Really? That's odd. I don't think the built-ins element shouldn't trigger a message like that, and interestingly, built-ins is *forced* into the default font path if it isn't already. Take a look at http://cgit.freedesktop.org/xorg/xserver/tree/dix/dixfonts.c Furthermore, I really don't see how "built-ins" is landing in the list we hand to xset. Take a look at /opt/X11/lib/X11/xinit/xinitrc.d/10-fontdir.sh(*). built-ins is not in the list that we hand to xset, and I don't see why there should be 17 elements. Are you doing something special with your font path in a custom file in ~/.xinitrc.d or /opt/X11/lib/X11/xinit/xinitrc.d? I tried adding built-ins in a call to xset and didn't see a problem: $ xset fp= "/opt/X11/lib/X11/fonts/misc/,/opt/X11/lib/X11/fonts/TTF/,/opt/X11/lib/X11/fonts/OTF,/opt/X11/lib/X11/fonts/Type1/,/opt/X11/lib/X11/fonts/75dpi/:unscaled,/opt/X11/lib/X11/fonts/100dpi/:unscaled,/opt/X11/lib/X11/fonts/75dpi/,/opt/X11/lib/X11/fonts/100dpi/,built-ins" *: /opt/X11/lib/X11/xinit/xinitrc.d $ cat 10-fontdir.sh if [ -x /opt/X11/bin/xset ] ; then fontpath="/opt/X11/lib/X11/fonts/misc/,/opt/X11/lib/X11/fonts/TTF/,/opt/X11/lib/X11/fonts/OTF,/opt/X11/lib/X11/fonts/Type1/,/opt/X11/lib/X11/fonts/75dpi/:unscaled,/opt/X11/lib/X11/fonts/100dpi/:unscaled,/opt/X11/lib/X11/fonts/75dpi/,/opt/X11/lib/X11/fonts/100dpi/" [ -e "$HOME"/.fonts/fonts.dir ] && fontpath="$fontpath,$HOME/.fonts" [ -e "$HOME"/Library/Fonts/fonts.dir ] && fontpath="$fontpath,$HOME/Library/Fonts" [ -e /Library/Fonts/fonts.dir ] && fontpath="$fontpath,/Library/Fonts" [ -e /System/Library/Fonts/fonts.dir ] && fontpath="$fontpath,/System/Library/Fonts" /opt/X11/bin/xset fp= "$fontpath" unset fontpath fi
Am 14.05.2012 um 04:30 schrieb Jeremy Huddleston:
Are you doing something special with your font path in a custom file in ~/.xinitrc.d or /opt/X11/lib/X11/xinit/xinitrc.d?
With XQuartz 2.7.2_rc2 installed as /Applications/Utilities/XQuartz.app obviously ~/.xinitrc is used. I can deduce this from the fact that GNU Emacs 23.3 is launched and not GNU Emacs 23.4. In Console I can see: May 14 10:52:44 sumac org.macosforge.xquartz.startx[26427]: xrdb: xset: bad font path element (#11), possible causes are: May 14 10:52:44 sumac org.macosforge.xquartz.startx[26427]: Directory does not exist or has wrong permissions May 14 10:52:44 sumac org.macosforge.xquartz.startx[26427]: Directory missing fonts.dir May 14 10:52:44 sumac org.macosforge.xquartz.startx[26427]: Incorrect font server address or syntax I have commented all xset commands in ~/.xinitrc. With a bit of UNIX I get: pete 145 /\ xset q | grep /opt | tr ',' '\012' | pr -n3 -t 1 /opt/X11/share/fonts/misc/ 2 /opt/X11/share/fonts/TTF/ 3 /opt/X11/share/fonts/OTF 4 /opt/X11/share/fonts/Type1/ 5 /opt/X11/share/fonts/75dpi/:unscaled 6 /opt/X11/share/fonts/100dpi/:unscaled 7 /opt/X11/share/fonts/75dpi/ 8 /opt/X11/share/fonts/100dpi/ 9 /Library/Fonts 10 /System/Library/Fonts 11 built-ins You may count yourself to eleven... In ~/.xinitrc I do: xset +fp /opt/local/share/fonts/misc/,/opt/local/share/fonts/TTF/,/opt/local/share/fonts/OTF/,/opt/local/share/fonts/Type1/,/opt/local/share/fonts/100dpi/,/opt/local/share/fonts/75dpi/ xset fp+ /sw/lib/X11/fonts/naga10/ ... xset fp rehash (Needs an update, obviously.) Mac OS X 10.6.8, x86_64 – 'locate libdispatch' still does not find anything... -- Greetings Pete "Once you've heard Beefheart, it's hard to wash him out of your clothes." – Tom Waits
On May 14, 2012, at 2:13 AM, Peter Dyballa <Peter_Dyballa@web.de> wrote:
Am 14.05.2012 um 04:30 schrieb Jeremy Huddleston:
Are you doing something special with your font path in a custom file in ~/.xinitrc.d or /opt/X11/lib/X11/xinit/xinitrc.d?
With XQuartz 2.7.2_rc2 installed as /Applications/Utilities/XQuartz.app obviously ~/.xinitrc is used.
...
In ~/.xinitrc I do:
xset +fp /opt/local/share/fonts/misc/,/opt/local/share/fonts/TTF/,/opt/local/share/fonts/OTF/,/opt/local/share/fonts/Type1/,/opt/local/share/fonts/100dpi/,/opt/local/share/fonts/75dpi/ xset fp+ /sw/lib/X11/fonts/naga10/ ... xset fp rehash
Please move aside your ~/.xinitrc and use the system default. Reproduce the issue using that sanitized environment. Tell me what xset commands to execute to cause the error message. I don't see it with: ~ $ xset fp+ built-ins ~ $ xset fp rehash --Jeremy
Am 14.05.2012 um 20:41 schrieb Jeremy Huddleston:
Please move aside your ~/.xinitrc and use the system default.
Now I get: May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: xrdb: xset: warning, no entries deleted from font path. May 14 23:16:44: --- last message repeated 1 time --- May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: xrdb: xset: bad font path element (#11), possible causes are: May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: Directory does not exist or has wrong permissions May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: Directory missing fonts.dir May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: Incorrect font server address or syntax May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: xrdb: xset: bad font path element (#11), possible causes are: May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: Directory does not exist or has wrong permissions May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: Directory missing fonts.dir May 14 23:16:44 sumac org.macosforge.xquartz.startx[57118]: Incorrect font server address or syntax May 14 23:16:45 sumac org.macosforge.xquartz.startx[57118]: xrdb: Warning: Cannot convert string "Libris ADF Std:autohint=true:antialias=true:style=Bold:size=9" to type FontStruct
Reproduce the issue using that sanitized environment.
This seems to fail. When I also mv ~/.xinitrc.d out of the way this message vanishes.
Tell me what xset commands to execute to cause the error message.
It might help to have more than one version of xset: -rwxr-xr-x 1 root wheel 88544 10. Mai 04:54 /opt/X11/bin/xset -rwxr-xr-x 1 root admin 38200 20. Apr 21:01 /opt/local/bin/xset -rwxr-xr-x 1 root wheel 88624 17. Dez 2009 /usr/X11/bin/xset -- Greetings Pete "Evolution" o __o _o _ °\___o /0~ -\<, ^\___ /=\\_/-% oo~_______ /\ /\______/ \_________O/ O_______________o===>-->O--o____
On May 14, 2012, at 2:53 PM, Peter Dyballa <Peter_Dyballa@web.de> wrote:
Am 14.05.2012 um 20:41 schrieb Jeremy Huddleston:
Please move aside your ~/.xinitrc and use the system default.
Now I get:
<snip>
Reproduce the issue using that sanitized environment.
This seems to fail. When I also mv ~/.xinitrc.d out of the way this message vanishes.
I assumed you didn't have a ~/.xinitrc.d since you had a ~/.xinitrc (which would mean ~/.xinitrc.d wouldn't be used...), so yeah ... move both out of the way to sanitize your environment...
Tell me what xset commands to execute to cause the error message.
It might help to have more than one version of xset:
-rwxr-xr-x 1 root wheel 88544 10. Mai 04:54 /opt/X11/bin/xset -rwxr-xr-x 1 root admin 38200 20. Apr 21:01 /opt/local/bin/xset -rwxr-xr-x 1 root wheel 88624 17. Dez 2009 /usr/X11/bin/xset
That is irrelevant. They are all the same. With your sanitized environment (no ~/.xinitrc, no ~/.xinitrc.d), the issue doesn't occur, so obviously it's something about your custom xinitrc that is causing the issue. You need to figure out what xset commands you are running in your xinitrc and come up with a reduced test case. That will help you figure out the problem. --Jeremy
participants (3)
-
Jeremy Huddleston
-
Jonas Maebe
-
Peter Dyballa