problem building qt4-x11 on SL

[Ryan Schmidt]

Remember to Reply All.

On Dec 4, 2011, at 02:17, Frédéric Dubois wrote:

> I suspected something with launchd
> 
> Here is the log ...
> 
> vpn2:PythonOCC dubois$ python draw.py
> Traceback (most recent call last):
>  File "draw.py", line 226, in <module>
>    Window(display.Display()).loop()
>  File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/Xlib/display.py", line 85, in __init__
>    self.display = _BaseDisplay(display)
>  File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/Xlib/display.py", line 67, in __init__
>    apply(protocol.display.Display.__init__, (self, ) + args, keys)
>  File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/Xlib/protocol/display.py", line 45, in __init__
>    name, host, displayno, screenno = connect.get_display(display)
>  File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/Xlib/support/connect.py", line 67, in get_display
>    return mod.get_display(display)
>  File "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/Xlib/support/unix_connect.py", line 53, in get_display
>    raise error.DisplayNameError(display)
> Xlib.error.DisplayNameError: Bad display name "/tmp/launch-1hAg4j/org.x:0"
> 
> and /tmp/launch-1hAg4j/org.x:0 is the value of DISPLAY

I see this was reported on the pythonocc list November 13 two years ago:

https://mail.gna.org/public/pythonocc-users/2009-11/msg00040.html

The reply says this had already been fixed at that time in a newer version of python-xlib which was about to be released. Version 0.14, to which the port was just updated, is four years old, so doesn't include the fix. Maybe 0.15rc1, which was released two years ago, on November 14, the day after the question was asked, does. We usually only update ports to stable versions, but maybe in this case, since the developers of python-xlib are clearly being very relaxed about their release schedule, we should make an exception, if 0.15rc1 does resolve the problem. (Could you test whether it does?)