I've loaded XFree86 and XDarwin 1.0a3 using the installers from
osxgnu.org. When I try to run XDarwin on my Pismo 400 I get the
following error:
******
XFree86 Version 4.1.99.1 / X Window System
(protocol Version 11, revision 0, vendor release 6510)
Release Date: xx September 2001
If the server is older than 6-12 months, or if your hardware is
newer than the above date, look for a newer version before
reporting problems. (See http://www.XFree86.Org/FAQ\)
Operating System: Darwin
Using keymapping provided in /System/Library/Keyboards/
USA.keymapping.
Warning: no access to tty (Inappropriate ioctl for device).
Thus no job control in this shell.
Display mode: Rootless Quartz
Screen 0 added: 1024x747 @ (0,21)
Screen 0 placed at X11 coordinate (0,0).
Keymapping file did not contain appropriate keyboard interface.
Reverting to kernel keymapping.
Fatal server error:
Could not get kernel keymapping! Load keymapping from file instead.
OsVendorFatalError
AbortDDX
Quitting XDarwin...
****
This happens regardless of whether I have installed the "Pismo patch" from
osxgnu.org or checked the "Load from file" box in preferences. Any
thoughts?
Thanks!
-Chris
Logged In: YES
user_id=37183
Hmm, that's odd. I have a Pismo at work, and building from source (approximately version 1.0a3) works fine on it. I'll try the osxgnu.org installer on Monday.
The "kernel keymapping" is known to fail on Powerbooks with 10.0 and all machines with 10.1. The "keymapping file" should work. Does /System/Library/Keyboards/USA.keymapping exist? You might try one of the other .keymapping files in /System/Library/Keyboards (`startx -- -rootless -keymap Deutsch.keymapping`, for example).
Logged In: YES
user_id=343014
Tried reinstalling directly from the XDarwin tarball and switching a few keymaps but to no avail. I have a hunch it's the XFree86 pkg that's the culprit but haven't yet installed XF86 from binaries. Do I need to install the XF86 Xquartz module if I'm installing XDarwin as well?
Thanks!
-C
Logged In: YES
user_id=37183
You probably should install Xquartz.tgz. I don't know whether it's completely replaced by the XDarwin updates.
Logged In: YES
user_id=37183
Someone else saw this problem and reported it to xdarwin.org. He says that XDarwin works fine until he runs Classic. After Classic has started, XDarwin will refuse to start until he reboots, and it fails with precisely the error above.
Does this match your experience? If you reboot, don't run Classic, and do run XDarwin, does it work?
Logged In: YES
user_id=343014
Yes, that's it! I had been starting Classic at login; when I turn that off XDarwin
works ok. Interestingly it continues to work if I start Classic after XDarwin launches,
but then won't reliably re-launch until I reboot.
Nice to know I'm not the only one. Only extraneous component I'm aware of is
URL Manager Pro's shared menus, if that helps.
Logged In: YES
user_id=133579
I'm upping this ones priority and transferring it to Bugs.
This looks like some weird interaction between Classic and
XDarwin on Pismos with Mac OS X 10.1. Has this been seen on
any other type of machine? Unfortunately I don't have a Pismo
to test on.
Logged In: YES
user_id=133579
Classic is likely changing what the kernel thinks its
keyboard interface ID is. We could work around this problem
by finding out what the bad ID is and substituting the proper
ID whenever the kernel returns the bad ID. If someone has a
Pismo and a debugger, it would be very useful to know what
the keyboard interface ID is when XDarwin launches correctly
and when it fails to launch.
Logged In: YES
user_id=37183
I haven't been able to duplicate this on my Pismo. I'm
running 10.1, Classic 9.1, and default extensions (I
think). Does XDarwin still have trouble if you start
Classic with extensions off? Are you running Classic
9.1 or 9.2?
Logged In: YES
user_id=353356
I originally posted about my problem in the "Help"
forum under "Can't start XDarwin twice in 10.1".
I have the same problem where as soon as I start
Classic I can't launch XDarwin again, but I have an
iBook, not a Pismo. I've tried starting with other
keymapping files, but that didn't work. I installed
XFree86 with the binaries from xfree86.org. I'm running
10.1 and Classic 9.2.1. How would I find out the
keyboard interface ID, if that would help track down the
bug? Thanks.
Logged In: NO
I also have this exact same bug on an iBook Dual USB
(500Mhz w/Combo Drive), and it only occurrs after
some 'unknown' happens in my system... what I mean
is that it will launch fine when I first boot up, but after
loading an application or two (not just Classic) I get the
same keymapping error if I choose to try to relaunch
XDarwin.
I'm on about my 4th reinstall of everything, and the
problem still crops up. Could something be happening
that's corrupting the information read from the
keymapping files for Laptop Macs? This occurs no
matter which keymapping file I use, or even if I copy the
keymapping file to my own ~/Library/Keyboards/ folder.
Please look into this, as it's very frustrating... especially
because it always crops up when I'm trying to
demonstrate how cool my iBook is running OSX with
rootless XFree86 (and I end up eating crow instead ;) ).
Otherwise, thanks for a great project.
Here's the terminal output, even though it's verbatim of
the post above:
[localhost:~] david%startx -- -rootless
Running rootless inside Mac OS X window server.
This is a pre-release version of XFree86, and is not
supported in any
way. Bugs may be reported to XFree86@XFree86.Org
and patches submitted
to fixes@XFree86.Org. Before reporting bugs in pre-
release versions,
please check the latest version in the XFree86 CVS
repository
(http://www.XFree86.Org/cvs)
XFree86 Version 4.1.99.1 / X Window System
(protocol Version 11, revision 0, vendor release 6510)
Release Date: xx September 2001
If the server is older than 6-12 months, or if your
hardware is
newer than the above date, look for a newer
version before
reporting problems. (See http://www.XFree86.Org/
FAQ)
Operating System: Darwin
Using keymapping provided in /Users/david/Library/
Keyboards/USA.keymapping.
Display mode: Rootless Quartz
Screen 0 added: 1024x747 @ (0,21)
Screen 0 placed at X11 coordinate (0,0).
Keymapping file did not contain appropriate keyboard
interface.
Reverting to kernel keymapping.
Fatal server error:
Could not get kernel keymapping! Load keymapping
from file instead.
// Is this an intentional irony? ;) [my comment]
OsVendorFatalError
AbortDDX
Quitting XDarwin...
X connection to :0.0 broken (explicit kill or server
shutdown).
[localhost:~] david%
Logged In: NO
When starting XDarwin (a3) on a freshly booted machine (TiBook), everything works fine.
If I start it after the machine has _slept_ I get the following 'keymapping issue'
XFree86 Version 4.1.99.1 / X Window System
(protocol Version 11, revision 0, vendor release 6510)
Release Date: xx September 2001
If the server is older than 6-12 months, or if your hardware is
newer than the above date, look for a newer version before
reporting problems. (See http://www.XFree86.Org/FAQ\)
Operating System: Darwin
Using keymapping provided in /System/Library/Keyboards/USA.keymapping.
_XSERVTransmkdir: Owner of /tmp/.X11-unix should be set to root
Display mode: Rootless Quartz
Screen 0 added: 1152x747 @ (0,21)
Screen 0 placed at X11 coordinate (0,0).
Keymapping file did not contain appropriate keyboard interface.
Reverting to kernel keymapping.
Fatal server error:
Could not get kernel keymapping! Load keymapping from file instead.
OsVendorFatalError
AbortDDX
Quitting XDarwin...
Note that classic has nothing to do with it, I don't even have OS9 installed on my machine. Sleeping is the only thing.
Niv
Logged In: NO
I have a Lombard running 10.1, with XFree86 4.1.0 and
XDarwin 1.0a3 installed, and have the identical keymapping
problem as posted by Niv when trying to start XDarwin after
my laptop has slept. Have not tried running Classic to see
whether that also produces the same symptoms.
-Eric
Logged In: NO
What about the case where this happens without classic being
involved at all? Anything known about that?
I'm having this problem on a TiBook after the machine has slept ... It is 100% reproducible. Boot and everything is fine, put the machine to sleep and after that no X11.
Niv
Logged In: NO
Niv, that is just what I was talking about! I'm having the
identical symptoms as you (but on a Lombard), i.e. without
Classic being involved at all. Sorry if I was unclear.
-Eric
Logged In: YES
user_id=37183
Summary: this has only been seen on laptops, and there are three proposed causes.
* Running Classic.
* Putting the machine to sleep.
* Using the Kensington mouse driver.
There are already reports of failure without running Classic. Please post here if
* you see this problem WITHOUT putting the machine to sleep, or
* you see this problem WITHOUT any mouse drivers installed, or
* you see this problem on any non-laptop machine.
I am still unable to reproduce this on my Pismo using any of the above methods (although I haven't yet tried the Kensington driver with any Kensington hardware plugged in).
Logged In: NO
If this helps, I posted previously about having this
problem on the iBook 500 and I 'am' using the
Kensington Mouseworks drivers (with an Optical Pro
mouse). Also, Classic does not appear to be the
culprit in my case, but I have not tried to see if putting
the iBook to sleep is the cause. Possibly it's a
combination of using both the Kensington Drivers and
putting the machine to sleep.
Anyway, thought I'd add that if it helps steer you in the
right direction.
good luck, and thanks again.
Logged In: YES
user_id=366209
I have this exact problem too, on my 1999 G3 Lombard 400,
XDarwin 1.0a3, XFree86 4.1.99.1.
Using Kensington Mouseworks, Classic and sleeping. I haven't
isolated it to one of the above. I tried the
Schweizer.keymapping file and had the same result as with
USA.keymapping:
XFree86 Version 4.1.99.1 / X Window System
(protocol Version 11, revision 0, vendor release 6510)
Release Date: xx September 2001
If the server is older than 6-12 months, or if your hardware is
newer than the above date, look for a newer version before
reporting problems. (See http://www.XFree86.Org/FAQ\)
Operating System: Darwin
Using keymapping provided in
/System/Library/Keyboards/Schweizer.keymapping.
_XSERVTransmkdir: Owner of /tmp/.X11-unix should be set to root
Warning: no access to tty (Inappropriate ioctl for device).
Thus no job control in this shell.
Display mode: Rootless Quartz
Screen 0 added: 1024x747 @ (0,21)
Screen 0 placed at X11 coordinate (0,0).
Keymapping file did not contain appropriate keyboard interface.
Reverting to kernel keymapping.
Fatal server error:
Could not get kernel keymapping! Load keymapping from file
instead.
OsVendorFatalError
AbortDDX
Quitting XDarwin...
Logged In: YES
user_id=366209
Additional input on this problem, i.e. "keymapping file did
not contain appropriate keyboard interface", as manifested
on my PB Lombard 400:
1) Does not occur after restart, no Classic & no sleep.
2) Does not occur after Classic startup without sleep.
3) Does occur after sleep (Classic running).
Logged In: YES
user_id=37183
Don reported that updating to Kensington's new MouseWorks (version 1.1, released November 1) appears to fix the problem.
If you see this problem and you do not have ANY MouseWorks installed, please post here.
If you see this problem with the old MouseWorks installed, please upgrade MouseWorks, reboot, and report here whether it worked or not. Try running Classic and putting the machine to sleep, too, just to make sure it's really fixed.
I have not tried the old MouseWorks on my Pismo.
Logged In: YES
user_id=37183
I finally saw the problem on my Pismo. I have
MouseWorks 1.1b1 installed, and the problem didn't
show itself until after sleep. Don says that 1.1b2 is the
one that works for him. Make sure you upgrade to 1.1b2
and not 1.1b1. (In the Finder, Get Info on /Applications/
Utilities/MouseWorks should say "Version: 1.1b2".)
If you see this problem with MouseWorks 1.1b2 or
without any MouseWorks at all, please post here!
Logged In: YES
user_id=226250
This is clearly not just a Pismo problem. I see the
exactly the same error message on my G4/500 Ti
laptop when trying to restart XDarwin after the machine
has been to sleep. On my installation of 10.1 the
problem is definately tied to Kensington Mouse Works.
I've tried the older version and the latest release. If
either version of Mouseworks is installed, I see the
problem, it I uninstall Mouseworks, the problem goes
away. I tried the install/uninstall cycle several times.
Logged In: YES
user_id=226250
I should add that the new version I tried and still have
trouble with is 1.1b2
Logged In: YES
user_id=37183
I think I've fixed it (patch #477901). There's no
workaround except to uninstall MouseWorks, but we
should get a binary update out soon with this fix.