[Xmonad] Re: darcs patch: mod-wer for Xinerama was inadvertently changed

[Spencer Janssen-2]

Alec sent me a patch:
> Thu Apr 12 09:20:33 EDT 2007  Alec Berryman <[hidden email]>
>   * mod-wer for Xinerama was inadvertently changed

I believe this change was intentional, but I can't remember why.  Does
anyone have comments on this?


Cheers,
Spencer Janssen


PS - I've copied this mail to the new [hidden email] mailing list,
perhaps you could join?

[Alec Berryman]

Spencer Janssen on 2007-04-12 15:57:02 -0500:

> Alec sent me a patch:
> > Thu Apr 12 09:20:33 EDT 2007  Alec Berryman <[hidden email]>
> >   * mod-wer for Xinerama was inadvertently changed
>
> I believe this change was intentional, but I can't remember why.  Does
> anyone have comments on this?

darcs-annotate says that the change was introduced in:

Jason Creighton <[hidden email]>, 04/10/07 06:27:31
 * moved screen <-> workspace mapping from XMonad to StackSet

The diff is:

diff -rN -u old-xmonad-upstream/Config.hs new-xmonad-upstream/Config.hs
--- old-xmonad-upstream/Config.hs       2007-04-12 17:51:56.006088500 -0400
+++ new-xmonad-upstream/Config.hs       2007-04-12 17:51:56.050091250 -0400
@@ -136,7 +136,7 @@
     -- Keybindings to each screen :
     -- mod-wer (underneath 123) switches to physical/Xinerama screens 1 2 and 3
     ++
-    [((m .|. modMask, key), screenWS sc >>= f)
-        | (key, sc) <- zip [xK_w, xK_e, xK_r] [1..]
+    [((m .|. modMask, key), gets workspace >>= f . (+1) . fromMaybe 0 . W.workspace sc)
+        | (key, sc) <- zip [xK_s, xK_d, xK_f] [0..]
         , (f, m) <- [(view, 0), (tag, shiftMask)]]

The first added line is definitely a change related to the
screen/workspace mapping move, but it looked to me like the second line
was a personal configuration change that snuck in since none of the
other documentation in the file was updated and it wasn't related to the
subject of the patch.

I'm a fan of using wer, but either the documentation should be updated
or that line reverted for consistency.

[Donald Bruce Stewart]

sjanssen:

> Alec sent me a patch:
> > Thu Apr 12 09:20:33 EDT 2007  Alec Berryman <[hidden email]>
> >   * mod-wer for Xinerama was inadvertently changed
>
> I believe this change was intentional, but I can't remember why.  Does
> anyone have comments on this?
>
>
> Cheers,
> Spencer Janssen
>
>
> PS - I've copied this mail to the new [hidden email] mailing list,
> perhaps you could join?
> _______________________________________________

Did Jason's private 'sdf' patch get in?

-- Don

[Jason Creighton]

On Thu, Apr 12, 2007 at 03:57:02PM -0500, Spencer Janssen wrote:
> Alec sent me a patch:
> > Thu Apr 12 09:20:33 EDT 2007  Alec Berryman <[hidden email]>
> >   * mod-wer for Xinerama was inadvertently changed
>
> I believe this change was intentional, but I can't remember why.  Does
> anyone have comments on this?

What change are you referring to? Back when I wrote the screen-switching
code, I picked mod-wer to be the bindings. No huge reason for it; just
makes sense to me.

Then, somebody (I don't know who [hidden email] is) got this patch into
the repo:

[Config.hs avoids conflict with essential M-w Emacs shortcut.
[hidden email]**20070401015135] {
hunk ./Config.hs 98
-        | (key, sc) <- zip [xK_w, xK_e, xK_r] [1..]
+        | (key, sc) <- zip [xK_e, xK_r, xK_t] [1..]
}

...because they wanted alt-w to work in Emacs. I take a dim view of this
rationale, because in my opinion, the answer to key clobbering is to change
your modMask to the windows key or similar approach.

Later, dons was cleaning up the hodgepodge of keybindings, and asked me
about that in IRC. At that point, dons changed it back to mod-wer. IRC
log: http://tunes.org/~nef/logs/haskell/07.03.31 discussion starting at
20:09:22

Jason Creighton

[Jason Creighton]

On Thu, Apr 12, 2007 at 05:55:23PM -0400, Alec Berryman wrote:

> Spencer Janssen on 2007-04-12 15:57:02 -0500:
>
> > Alec sent me a patch:
> > > Thu Apr 12 09:20:33 EDT 2007  Alec Berryman <[hidden email]>
> > >   * mod-wer for Xinerama was inadvertently changed
> >
> > I believe this change was intentional, but I can't remember why.  Does
> > anyone have comments on this?
>
> darcs-annotate says that the change was introduced in:
>
> Jason Creighton <[hidden email]>, 04/10/07 06:27:31
>  * moved screen <-> workspace mapping from XMonad to StackSet
>
> The diff is:
>
> diff -rN -u old-xmonad-upstream/Config.hs new-xmonad-upstream/Config.hs
> --- old-xmonad-upstream/Config.hs       2007-04-12 17:51:56.006088500 -0400
> +++ new-xmonad-upstream/Config.hs       2007-04-12 17:51:56.050091250 -0400
> @@ -136,7 +136,7 @@
>      -- Keybindings to each screen :
>      -- mod-wer (underneath 123) switches to physical/Xinerama screens 1 2 and 3
>      ++
> -    [((m .|. modMask, key), screenWS sc >>= f)
> -        | (key, sc) <- zip [xK_w, xK_e, xK_r] [1..]
> +    [((m .|. modMask, key), gets workspace >>= f . (+1) . fromMaybe 0 . W.workspace sc)
> +        | (key, sc) <- zip [xK_s, xK_d, xK_f] [0..]
>          , (f, m) <- [(view, 0), (tag, shiftMask)]]
>
> The first added line is definitely a change related to the
> screen/workspace mapping move, but it looked to me like the second line
> was a personal configuration change that snuck in since none of the
> other documentation in the file was updated and it wasn't related to the
> subject of the patch.

Yes, exactly. Sorry, totally my fault. Ignore my other reply.

I was playing around with mod-sdf to see if I liked it better. (for
avoiding accidental mod-shift-q).

> I'm a fan of using wer, but either the documentation should be updated
> or that line reverted for consistency.

Okay. I think we should just change it back to wer in the repo then.

Jason Creighton

[Spencer Janssen-2]

Applied, thanks.


Spencer Janssen