- xmonad - Haskell.org

0.3 coming soon.
by Spencer Janssen 24 Aug '07

24 Aug '07

It is that time again -- xmonad 0.3 is right around the corner. We're putting xmonad, XMonadContrib, and X11-extras under a feature freeze. If you've got patches that we haven't applied yet, don't worry, we haven't forgotten them! We'll apply them right after 0.3 is tagged and released Cheers, Spencer Janssen

1 0

Re: [Xmonad] xmonad-utils: small x apps for xmonad
by Gwern Branwen 23 Aug '07

23 Aug '07

On 0, Andrea Rossato <mailing_list(a)istitutocolli.org> scribbled: > On Wed, Aug 22, 2007 at 10:02:40AM -0400, Gwern Branwen wrote: > > Hey: would it be possible to have a version of hxsel which is a function you could call in Xmonad's Config.hs? > > well, you've got that version: it's just the code below...;-) > > I didn't test it but it looks fine to me. > > > else do destroyWindow dpy win >> return "" -- Is destroyWindow really needed? > > yes, as far as I remember, otherwise we leak memory. Ah, I suspected as much - I would've been surprised if you could create windows and not have to manually free them. > You can send it as a contrib module if you feel like. > > Generally speaking I prefer to keep this stuff outside XMonad: less > memory, less code, less bugs, but this is a matter of personal taste I > think. > > Cheers, > > Andrea Well, my reasoning is that given that Xmonad is already linking in so much X-related stuff, that calling the shell to call another program (with all the overhead involved in those indirections and standalone binaries,, like the 500KB RTS system GHC inserts) is less of a win than just putting in some stripped down code. Plus, shells are often sources of problems, and it's tidier to have all the logic in one place. Anyway, I don't think I'll send in a contrib module just yet. hxsel should probably be factored out into two functions, a main that prints stuff to stdout and the actual selection-fetching code; that way it doesn't have to be duplicated. Further, I'm having problems integrating it into Config.hs: I finally got it to compile like this: , ((modMask .|. shiftMask, xK_g ), spawn $ "google" ++ unsafePerformIO(getSelection)) But notice I used unsafePerformIO. The reason is that if you try to simply concatenate "google" and getSelection, there's a type mismatch. You can then try using liftM to do it, like ("google"++) `liftM` getSelection, and that will work well. Unfortunately, spawn :: String -> X (), so spawn needs to be lifted. (The internal /bin/sh stuff apparently requires a String too, so it's not spawn's fault.) But if you go 'spawn (("google"++) `liftM` getSelection' - which on its own would work - the resulting type is something like IO ( X ()). And that just can't be allowed inside 'keys'. At least, the only solution I could find was to just slice the Gordian knot and avoid introducing the IO signature in the first place. Monads may be the greatest thing since sliced bread, but most of the time they're just a pain :( -- gwern extraction Iris MITM Fortezza FBIS Chelsea Love Coderpunks 64 stakeout

4 4

xmonad-utils: small x apps for xmonad
by Andrea Rossato 23 Aug '07

23 Aug '07

Hi, I'm putting together some small x applications I find useful to run with xmonad. The idea *was* to make some nice small Haskell examples of X applications, but I don't think I'm really so willing to rewrite them in a nice style... I prefer invest some time in learning _new_ cool Haskell stuff...;-) Still I find them useful and I'm going to share them. Since one requires the darcs version of X11-Extras if I'll release them I'll do after xmonad-0.3. You can grab them here: darcs get http://gorgias.mine.nu/repos/xmonad-utils/ What's inside: 1. hslock: a CHaskell rewrite of slock/alock whatever. To lock the screen.[1] 2. hxsel: a CHaskell rewrite of sselp[2]. It returns the string currently in the X selection. I use it for opening urls, looking up words in the dictionary: I select a word with the mouse and then I bind a xmonad key to execute a small bash script like: #!/bin/sh URL=`hxsel` /usr/lib/firefox/mozilla-xremote-client -a firefox "openurl(${URL},new-tab)" 3. hmanage: to toggle override_redirect. Makes a window unmanaged. The window id you need can be retrieved with xwininfo. [1] http://www.suckless.org/download/slock-0.6.tar.gz [2] http://www.suckless.org/download/sselp-0.1.tar.gz Now that I have reimplemented dmenu, slock and sselp only lsw and swarp are left...;-) andrea ps: I will be off line till Sunday, starting from tonight (~22.00 UTC).

2 1

darcs patch: XPrompt: just code formatting
by Andrea Rossato 22 Aug '07

22 Aug '07

Hi, this is just code formatting and other cosmetic clean up for the upcoming release. ciao andrea Wed Aug 22 21:32:20 CEST 2007 Andrea Rossato <andrea.rossato(a)unibz.it> * XPrompt: just code formatting

1 0

haddock documentation
by Andrea Rossato 22 Aug '07

22 Aug '07

Hi, since I'm going offline and I hope to come back and find a new XMonad release, I'm sending the stuff I use to generate the documentation, so that, if you release before my coming back, you can generate the documentation by yourself. Attached you'll find: 1. a patch to haddock (needed for the deriving clause of the X newtype) 2. a modified Setup.lhs with a haddock hook that will use the package description on the first page of the documentation 3. a cabal file with all the modules: the one I'm sending is updated to the last darcs version. I use the metamodule to get the list of modules. That's it: runhaskell Setup.haddockHook.lhs haddock Ciao andrea

1 0

slock issue
by Alexander Tsamutali 22 Aug '07

22 Aug '07

Hi, all. I have strange issue with slock [0]. Here is how it's bound in my Config.hs: ((modMask, xK_F12 ), spawn "slock") -- @@ Lock screen Dead simple. But it doesn't work. When i launch slock from urxvt or dmenu everything works. When i launch it using modMask + xK_F12 screen goes black and then after few seconds everything comes back and no locking occurs. I noticed that if for ex. before launching slock my urxvt is in focus and i type something when screen is black, urxvt catches my input. Seems like slock fails to lock screen and catch all input events. But this happens only when i launch slock with a keybinding. Can someone help? [0] http://www.suckless.org/wiki/tools/xlib -- Alexander Tsamutali

2 8

Show name of focused window in status bar
by Kai Grossjohann 21 Aug '07

21 Aug '07

Do people think that showing the name of the currently focused window in the status bar (in addition to the workspaces) would be a worthwhile addition? If so, could you advise how to improve the quality of the code I've concocted? Given that I don't know Haskell, I'm happy that it runs. Here's what I did. To NamedWindows.hs I added the following function (and exported it): unWindow :: NamedWindow -> String unWindow (NW n _) = n In Config.hs I wrote: dynamicLogWName :: X () dynamicLogWName = do withWindowSet $ io . putStr . pprWindowSet io (putStr " // ") withNamedWindow $ io . putStr . unWindow io (putStrLn "") logHook = dynamicLogWName tia, Kai

2 3

darcs patch: CycleWS: a couple of simple functions to cycle between...
by Andrea Rossato 21 Aug '07

21 Aug '07

Hi, this is a module that exports a couple of simple functions to cycle between workspaces. It'll keep working when we will switch the type WorkspaceID = String Andrea Tue Aug 21 07:45:18 CEST 2007 Andrea Rossato <andrea.rossato(a)unibz.it> * CycleWS: a couple of simple functions to cycle between workspaces

1 1

xmonad on one head display
by jason 20 Aug '07

20 Aug '07

Hi all, I am satisfied user xmonad on PC with two head display configuration (with XINERAMA). I would like use xmonad on my notebook too. But I haven't found any way how to configure it. Xmonad always try to use XINERAMA and fails on one display notebook. Is there any howto or decription for this simple case? Thanks Jason

4 4

I just noticed...
by Isaac Dupree 20 Aug '07

20 Aug '07

XMonad layouts make the menu-bars of a window often be at the top of the screen. Well, one pixel away from the top of the screen. It would be better for mouse movement not to go off the window, like in Mac OS menu-bars: if the cursor wouldn't go onto those pixels, it is ergonomically easier (Fitts's law) to reach the menubars, because there is no limit on how much vertical movement you make. If there are buttons or something on the side or bottom of windows, this applies to them too. Is it a disadvantage not to be able to remove the cursor from within the bounds of the window? The ability to have full-screen windows automatically when there is only one window displayed at once seems like a really nice property (for VNC, or qemu, ... perhaps). I am then concerned about the small possibility of there being a way to remove the cursor from any window (assuming the current workspace contains ANY windows). If the border highlighting doesn't appear at the edges of the monitor, how to distinguish which of two windows is active? A two pixel (or four pixel possibly? depending on screen resolution...) buffer between windows, could indicate that with a color difference or gradient, and it could be as if the cursor were in the *nearest* window, on the edge of it (I wonder if that would confuse any apps, or if it could be done right...) P.S. I also have xmodmap my command/apple keys to Control modifier Here is what I currently do to my MacBook keyboard in case anyone's interested #Turn off caps-lock. #Make left and right command (apple) keys function as control keys # as well as the actual control keys. #keycode 108 (keysym 0xff8d, KP_ENTER) #I guess I'll turn "enter" into a "right alt/option" for now #rightAlt-shift-Return doesn't seem to work, but others # (rightAlt-shift-c) seem to (even with right shift) xmodmap \ -e 'remove Lock = Caps_Lock' \ -e 'clear Control' \ -e 'keycode 115 = Control_L' \ -e 'keycode 116 = Control_R' \ -e 'add Control = Control_L' \ -e 'add Control = Control_R' \ -e 'clear Mod1' \ -e 'keycode 108 = Alt_R' \ -e 'add Mod1 = Alt_L' \ -e 'add Mod1 = Alt_R' (for middle- and right- click I use F12 and [eject], which could be improved) Isaac

2 2