XMonad.Layout.ResizableTile fails if floating window present on same screen.

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Import XMonad.Layout.ResizableTile
2. Bind the MirrorExpand / MirrorShrink commands to keys as in:

, ((modm,  xK_a), sendMessage MirrorExpand)
, ((modm, xK_z), sendMessage MirrorShrink)

3. Now place a floating window on the same screen as some tiled windows;
this is irrlevant of of Xinerama, although that's what I am using.

What is the expected output? What do you see instead?

When focusing a window already tiled with a floating window present on the
screen, the keys bounf in (2) above should shrink/expand that window, but
it does not.

Instead nothing happens, or something ""weird"" happens, where ""weird""
could be nothing at all, or the focused tiled window is only resizable in
one direction,

If I move the floating window to another monitor (assuming Xinerama), or
close it, the the Mirror{Expand,Shrink} commands work as expected,  Clearly
something odd is happening with the presence of a floating window present,
although no visible events are sent to the floating window when trying to
shrink/expand a tiled window which overlaps the floating window.

What version of the product are you using? On what operating system?

XMonad darcs (post 0.5 release).  Running on Debian Linux SID, compiled myself.

Please provide any additional information below.

N/A

Original issue reported on code.google.com by thomas.a...@gmail.com on 12 Dec 2007 at 10:24

[GoogleCodeExporter]

Hmm, I think I see the problem.  The doLayout function in ResizableTile's 
LayoutClass
instance applies its tiling to the results of W.integrate -- but W.integrate 
results
in the entire contents of the stack, which includes any floating windows.  I'll 
try
to push a fix later today.

Original comment by byor...@gmail.com on 13 Dec 2007 at 12:37

[GoogleCodeExporter]

Original comment by byor...@gmail.com on 24 Dec 2007 at 2:10

• Changed state: Accepted

[GoogleCodeExporter]

I just pushed a patch to darcs.  It seems to work, but independent
verification/testing is welcome.

Original comment by byor...@gmail.com on 25 Dec 2007 at 2:03

• Changed state: Fixed