I have sent it before, but it did not show up on the mailing list; maybe because I was not subscribed then ... 1 patch for repository http://code.haskell.org/xmonad: Tue Nov 12 00:25:51 CET 2013 Radim Krcmar * Use window border width instead of borderWidth configureWindow border width of floating windows is set to borderWidth regardless of the actual value. This produces an interesting effect when borderWidth < "window border set from layout", where window is repeatedly shrinked, because border makes it bigger than it should be. This patch should not be noticeable unless layout is setting border to different values than borderWidth (we already set border to borderWidth when enslaving the window) or third person is setting up window attributes. (then it is a good thing to leave them that way) Setting border width to 0 is required for sane fullscreen mplayer, because it decides to expand by the border width. (pretty bad behaviour) New patches: [Use window border width instead of borderWidth Radim Krcmar **20131111232551 Ignore-this: ffbaf5f6c085ee13a7b6338863a9693c configureWindow border width of floating windows is set to borderWidth regardless of the actual value. This produces an interesting effect when borderWidth < "window border set from layout", where window is repeatedly shrinked, because border makes it bigger than it should be. This patch should not be noticeable unless layout is setting border to different values than borderWidth (we already set border to borderWidth when enslaving the window) or third person is setting up window attributes. (then it is a good thing to leave them that way) Setting border width to 0 is required for sane fullscreen mplayer, because it decides to expand by the border width. (pretty bad behaviour) ] { hunk ./XMonad/Main.hsc 279 ws <- gets windowset wa <- io $ getWindowAttributes dpy w - bw <- asks (borderWidth . config) - if M.member w (floating ws) || not (member w ws) then do io $ configureWindow dpy w (ev_value_mask e) $ WindowChanges hunk ./XMonad/Main.hsc 286 , wc_y = ev_y e , wc_width = ev_width e , wc_height = ev_height e - , wc_border_width = fromIntegral bw + , wc_border_width = ev_border_width e , wc_sibling = ev_above e , wc_stack_mode = ev_detail e } when (member w ws) (float w) } Context: [Bump version to 0.12 Adam Vogt **20130720192124 Ignore-this: d73483c2dd30f7eea8f40daff7ef436f This is to make contrib build failures resulting from by data-default lead people to do the right thing: update xmonad-core. ] [warning police: name userCodeDef's argument defValue instead of def Daniel Wagner **20130528164401 Ignore-this: f32b218e2097a52c7ecccae28c23754b ] [depend on data-default, and deprecate the monomorphic name defaultConfig Daniel Wagner **20130528003531 Ignore-this: 1e746731695df3b6f684d5463a3da6a4 ] [use "modm" instead of "modMask" in the sample config Daniel Wagner **20130106174414 Ignore-this: 75ad65ae7a708bce69f811c28912cc4b ] [define the "help" string in the sample configuration bundled with xmonad Daniel Wagner **20130106174254 Ignore-this: 1db666ea921046378921356e14cc828f ] [configurableEventMasks mwlochbaum@gmail.com**20130205182858 Ignore-this: 3848de0f8f5ad5995e87a2a01e7752f ] [Grab all keycodes linked to each keysym, not just one Daniel Wagner **20130118225446 Ignore-this: 1a6c001560f68f99d75d5f550e7e83 This patch is based heavily on the one contributed by svein.ove@aas.no, but updated to avoid causing a conflict and to work with the newest X11 bindings. The name of the patch (and comment below) are copied verbatim from his patch. XKeysymToKeycode only gives the first code bound to a given symbol. To handle the case where multiple keys are bound to the same symbol, XKeycodeToKeysym is used instead, searching through all possible keycodes for each sym. ] [Issue 135 use wa_border_width for floating windows (neoraider) Adam Vogt **20130115170715 Ignore-this: c8ed6ceaf9483e31771ac25d86532f6c ] [Add flags for call to ghc closing issue 240 Adam Vogt **20130101035034 Ignore-this: 42a6a8599b615884c95626f74e3ba4a The -main-is flag goes back to at least ghc 6.10, and maybe the warning that this otherwise redundant flag enables (when xmonad.hs isn't a module Main) also dates back that far. ] [TAG 0.11 actual upload Adam Vogt **20130101014128 Ignore-this: 2c2a85caeed30cd23f02a7caf229fe7d ] Patch bundle hash: 90a0f22fb90471396fd95bb677d45da8ef116602