nanogui: nanox+flnx problem


Previous by date: 17 Jul 2002 15:26:34 -0000 Re: nanox+flnx problem, Florian Berger
Next by date: 17 Jul 2002 15:26:34 -0000 lcd test pattern ..?, ashish anand
Previous in thread: 17 Jul 2002 15:26:34 -0000 Re: nanox+flnx problem, Florian Berger
Next in thread:

Subject: Re: [nanogui] nanox+flnx problem
From: Jordan Crouse ####@####.####
Date: 17 Jul 2002 15:26:34 -0000
Message-Id: <1026918673.17785.1610.camel@cosmic>

Ok - I understand what you are trying to do, and it seems that your code
works correctly, but I am concerned that we are masking a more serious
problem with Microwindows with this fix, which is a practice that I am
trying to stop. 

The problem seems to be that during creation the window manager
reparents the window and moves it to a position of (0,5) relative to the
parent window, and the position of the window is now reported as such
with a GR_UPDATE_MOVE event, which screws up the internal system (which
your patch sets straight).  What I have been unable to verify is the X
equivalent of these messages - Does X11 also report a move to (0, 5), or
does it report the absolute position of the reparented window?  If it
reports the absolute position of the reparented window, then Nano-X
reports an incorrect value which we should fix (or apply your patch).

But if X11 reports the same values as Nano-X, then there are problems
elsewhere that we need to address.  I am especially concerned with
hacking the FLNX code to fix Microwindows problems - it should be the
other way around.

I'm not going to apply your patch to CVS right away, but I urge others
to test it and report their results.  I am especially concerned how well
it works with lots of child windows.

Jordan




Previous by date: 17 Jul 2002 15:26:34 -0000 Re: nanox+flnx problem, Florian Berger
Next by date: 17 Jul 2002 15:26:34 -0000 lcd test pattern ..?, ashish anand
Previous in thread: 17 Jul 2002 15:26:34 -0000 Re: nanox+flnx problem, Florian Berger
Next in thread:


Powered by ezmlm-browse 0.20.