darcs patch: Combo.hs: tweak usage instructions.

Wed Jun 20 13:25:55 CEST 2007 joachim.fasting@gmail.com * Combo.hs: tweak usage instructions. tabbed needs an additional argument.

On Wed, Jun 20, 2007 at 01:27:13PM +0200, joachim.fasting@gmail.com wrote:
Wed Jun 20 13:25:55 CEST 2007 joachim.fasting@gmail.com * Combo.hs: tweak usage instructions. tabbed needs an additional argument.
[Combo.hs: tweak usage instructions. joachim.fasting@gmail.com**20070620112555 tabbed needs an additional argument. ] { hunk ./Combo.hs 33 --- > combo [(full,1),(tabbed,1)] (twoPane 0.03 0.5) +-- > combo [(full,1),(tabbed shrinkText,1)] (twoPane 0.03 0.5)
I suppose it might also be worth adding a note that using tabbed with combo is broken, now that xmonad has decided that stacking order and focus order are the same thing. Someday I hope we'll be able to fix this, but it'll require changes in core (or reimplementation of core, which I'm not going to do) that seem likely to be unacceptable to core developers. -- David Roundy http://www.darcs.net

On Wed, 20 Jun 2007 06:35:06 -0700
David Roundy
On Wed, Jun 20, 2007 at 01:27:13PM +0200, joachim.fasting@gmail.com wrote:
Wed Jun 20 13:25:55 CEST 2007 joachim.fasting@gmail.com * Combo.hs: tweak usage instructions. tabbed needs an additional argument.
[Combo.hs: tweak usage instructions. joachim.fasting@gmail.com**20070620112555 tabbed needs an additional argument. ] { hunk ./Combo.hs 33 --- > combo [(full,1),(tabbed,1)] (twoPane 0.03 0.5) +-- > combo [(full,1),(tabbed shrinkText,1)] (twoPane 0.03 0.5)
I suppose it might also be worth adding a note that using tabbed with combo is broken, now that xmonad has decided that stacking order and focus order are the same thing. Someday I hope we'll be able to fix this, but it'll require changes in core (or reimplementation of core, which I'm not going to do) that seem likely to be unacceptable to core developers.
What exactly is broken with tabbed/combo? I've just pushed a change that stacks windows in the order they're returned by layouts, does this fix the problem? Cheers, Spencer Janssen

On Wed, Jun 20, 2007 at 10:21:16AM -0500, Spencer Janssen wrote:
On Wed, 20 Jun 2007 06:35:06 -0700 David Roundy wrote:
I suppose it might also be worth adding a note that using tabbed with combo is broken, now that xmonad has decided that stacking order and focus order are the same thing. Someday I hope we'll be able to fix this, but it'll require changes in core (or reimplementation of core, which I'm not going to do) that seem likely to be unacceptable to core developers.
What exactly is broken with tabbed/combo? I've just pushed a change that stacks windows in the order they're returned by layouts, does this fix the problem?
Ah, that would greatly help (and is a very good idea). But the trouble is that for Combo+Tabbed to work we need to track which window in each unfocussed Tabbed is visible. This used to "just work" because xmonad didn't modify the stacking order, so the window that was most recently raised was visible. Tracking "previously focussed" windows is a pain, particularly as layouts aren't notified when windows are added or removed. The "right" way to do this is to support hierarchical Stacks directly. This is the direction I was hoping to move after we'd made the current hierarchy more uniform, but you didn't like that idea. The last thing that I want to do is to add yet another distinct layer with yet another distinct API for finding out which objects are focussed. The truth is that Combo was an ugly hack that worked because of the undocumented/unintended stacking behavior of xmonad (which allowed X itself to keep track of information for us) and now it no longer works. Adding onto the ugliness is not the way to go. It's a beautiful layout, but if I'm going to rewrite it, I'll to it properly. -- David Roundy http://www.darcs.net

On Wed, 20 Jun 2007 13:27:13 +0200 joachim.fasting@gmail.com wrote:
Wed Jun 20 13:25:55 CEST 2007 joachim.fasting@gmail.com * Combo.hs: tweak usage instructions. tabbed needs an additional argument.
Applied, thanks.
participants (3)
-
David Roundy
-
joachim.fasting@gmail.com
-
Spencer Janssen