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
