aha! And I have discovered that if I give xnest geom settings that are small enough so that the xnest window is smaller than full-screen to start with, then indeed everything works as expected. If you give it geom settings that are bigger than full-screen, then it must pass the bigger ones one to xmonad, but meanwhile it displays its own window full-screen. Hence truncation. Simon | -----Original Message----- | From: Spencer Janssen [mailto:sjanssen@cse.unl.edu] | Sent: 05 July 2007 15:32 | To: Simon Peyton-Jones | Cc: xmonad@haskell.org | Subject: Re: [Xmonad] Re: Emailing: A taste of Haskell.pdf | | On Thu, 5 Jul 2007 14:01:58 +0100 | Simon Peyton-Jones <simonpj@microsoft.com> wrote: | | > What I'm still missing is: why doesn't xmonad get told the correct | > window size by xnest? The -geom thing tells xnest how big a window | > it should throw up, but regardless of its window size it should | > report it faithfully to xmonad. | > | > Similarly on resizing, which comprehensively does not work -- ie | > resizing the xnest window, making it smaller, is not reflected in the | > xmonad layout. Instead xnest just truncates even more of it. | > | > S | | Xnest is just not very nice in this respect -- it chooses a geometry at | start up and sticks with it. It continues to report the same geometry | to clients, no matter how you resize the containing Xnest window. | | | Cheers, | Spencer Janssen