headers
merlin | 1 Dec 2002 13:48

display-window/warp-to-window-enabled bugs


In recent sawfishen, warp-to-window-enabled defaults to
nil, so display-window (from cycling or the menu) will
focus windows without moving the pointer. This is pretty
unintuitive for focus-mode 'enter-exit and 'enter-only.

It would seem to me that warp-to-window-enabled should
perhaps go, and its value should be a property of the
focus mode?

Obviously I can configure this to work properly from
lisp, however focus mode is still one of the available
GUI options, and this incongruity makes it confusing.

Merlin
Permalink | Reply | 4 comments |
headers
John Harper | 1 Dec 2002 22:42

Re: display-window/warp-to-window-enabled bugs

On Sunday, Dec 1, 2002, at 04:48 US/Pacific, merlin wrote:
> In recent sawfishen, warp-to-window-enabled defaults to
> nil, so display-window (from cycling or the menu) will
> focus windows without moving the pointer. This is pretty
> unintuitive for focus-mode 'enter-exit and 'enter-only.

My reasoning was that warping the pointer is always a bad idea, so I 
thought it should be disabled by default. There are already cases where 
the constraint that in focus follows mouse focus modes the pointer is 
always inside the focused window is broken, so adding another instance 
of this seemed better than forcing everyone using ffm modes to put up 
with the pointer being warped

>
> It would seem to me that warp-to-window-enabled should
> perhaps go, and its value should be a property of the
> focus mode?

this is already true, the function warp-pointer-if-necessary calls 
through the current focus mode. Only ffm modes will ever warp the 
pointer

>
> Obviously I can configure this to work properly from
> lisp, however focus mode is still one of the available
> GUI options, and this incongruity makes it confusing.

I think the default behaviour - never warping the pointer - is the best 
solution.
(Continue reading)

Permalink | Reply | 3 comments |
headers
Daniel Pfeiffer | 2 Dec 2002 21:32
X-Face
Picon

Re: display-window/warp-to-window-enabled bugs

John Harper <jsh <at> unfactored.org>  skribis:
> On Sunday, Dec 1, 2002, at 04:48 US/Pacific, merlin wrote:
> > In recent sawfishen, warp-to-window-enabled defaults to
> > nil, so display-window (from cycling or the menu) will
> > focus windows without moving the pointer. This is pretty
> > unintuitive for focus-mode 'enter-exit and 'enter-only.

> My reasoning was that warping the pointer is always a bad idea, so I 
> thought it should be disabled by default. There are already cases where 
> the constraint that in focus follows mouse focus modes the pointer is 
> always inside the focused window is broken, so adding another instance 
> of this seemed better than forcing everyone using ffm modes to put up 
> with the pointer being warped

Well, I for one had meant to report this as a bug, but was never quite bothered enough to sit down and write it.  I
have ffm (focus-mode enter-only) and focus goes to new windows (focus-windows-when-mapped t).  But even with

(setq warp-to-window-enabled t)

I get the startling focus-doesn't-follow-mouse situation where the pointer stays where it was, and if
that isn't where the new window pops up, that gets the focus but not the mouse.

While you seem aware of this, I've-done-it-wrong-once is no excuse for doing it wrong in another situation
rather than fixing the first.  Ffm (that's short for Frankfurt am Main, btw. :-) sometimes should imply
warping when combined with certain other options, and there's nothing wrong with that.

Now if you want to take warping away from the Kindergarden lot, then you must also take away everything that
then becomes inconsistent!

coralament / best Grötens / liebe Grüße / best regards / elkorajn salutojn
(Continue reading)

Permalink | Reply | 2 comments |
headers
David J. Bakeman | 2 Dec 2002 21:24
Picon

How NOT to raise window with mouse click

I don't want to have the window I click the mouse in raised.  How do I 
do this?  I've looked through the docs and the preferences editor and 
can't find a way. In fact there is a question in the FAQ that is the 
exact opposite of want I want (#8 How do I make clicking on a window 
raise it).

Thanks
Permalink | Reply | 0 comments |
headers
Ezequiel Martín Cámara | 3 Dec 2002 21:51
Picon
Favicon

doc discrepancy

Hello.

This is to report a bug, either in the documentation or the code of
Sawfish 1.0.

The documentation tells of a "check-frame-availability" function
(node: frame styles), but you can't call it from "sawfish-client". I
think it is because it is not in the (export ...) list of the
sawdish.wm.frames structure.

I have downloaded the cvs source, changed the code by adding to the
export clause, and run "cvs diff -u", and the result is:

****** Start "cvs diff -u"
--- lisp/sawfish/wm/frames.jl	20 Oct 2002 08:07:50 -0000	1.91
+++ lisp/sawfish/wm/frames.jl	3 Dec 2002 20:30:18 -0000
 <at>  <at>  -26,6 +26,7  <at>  <at> 
      (export after-setting-frame-option
 	     define-frame-type-mapper
 	     add-frame-style
+	     check-frame-availability
 	     reload-frame-style
 	     reframe-window
 	     reframe-all-windows
****** End "cvs diff -u"

I don't know how to test the thing properly (this is my first open
source collaboration...). If I do a "./configure && make && make
install" in my Mandrake 8.1 installation, ¿will it break any
Mandrake-specific changes, or are those all in config files that
(Continue reading)

Permalink | Reply | 1 comment |
headers
Kristian G. Kvilekval | 4 Dec 2002 00:42
Picon
Gravatar

PROBLEM: dual head window placement


I am using sawfish 1.2 with "first-fit" placement  on 
a dual head setup that works well.  However, new windows 
always appear the second (rightmost) screen (upper left corner).   

I have fiddled with XF86Config(below) and have had no 
luck.   "top-left" placement does work, but put the windows in a 
cascading list.  I haven't seen any options in sawfish that allow me
to set the first and second screens.

Has anybody else had this problem and is there anything to do about it?

Thx.

Section "ServerLayout"
        Identifier     "Dual monitor"
        Screen  0 "Dell panel" 0 0
        Screen  1 "KDS flat" RightOf "Dell panel"
#        Screen         "Dell panel" LeftOf "KDS flat"
#        Screen         "Dell panel" 0 0
        InputDevice    "Generic Keyboard"
        InputDevice    "Configured Mouse"
EndSection

Section "Device"
        Identifier      "ATI Radeon 7500:DVI"
        Driver          "radeon"
        BusID           "PCI:1:0:0:"
#       VideoRam        64000
        Option          "hw cursor" "off"
(Continue reading)

Permalink | Reply | 0 comments |
headers
John Harper | 4 Dec 2002 05:47

Re: display-window/warp-to-window-enabled bugs

Daniel Pfeiffer writes:
|I get the startling focus-doesn't-follow-mouse situation where the
|pointer stays where it was, and if that isn't where the new window
|pops up, that gets the focus but not the mouse.

You really want the pointer to jump to new windows?

|
|While you seem aware of this, I've-done-it-wrong-once is no excuse for
|doing it wrong in another situation rather than fixing the first.  Ffm
|(that's short for Frankfurt am Main, btw. :-) sometimes should imply
|warping when combined with certain other options, and there's nothing
|wrong with that.

Yes, if you take things to their logical conclusion, that is true. I
don't think what you would arrive at would be very usable, but I'd be
happy to apply a patch that called warp-pointer-if-necessary in more
situations (since it will have no effect for the vast majority of users)

|
|Now if you want to take warping away from the Kindergarden lot, then
|you must also take away everything that then becomes inconsistent!

Please don't insult people who you think are less technically competent
than yourself,

	John
Permalink | Reply | 1 comment |
headers
John Harper | 4 Dec 2002 05:52

Re: doc discrepancy

Ezequiel Martín Cámara writes:
|The documentation tells of a "check-frame-availability" function
|(node: frame styles), but you can't call it from "sawfish-client". I
|think it is because it is not in the (export ...) list of the
|sawdish.wm.frames structure.

thanks for pointing this out. The function doesn't do what the manual
describes anymore, so I think that's why it's not exported. I just
deleted the entry from the manual.

|
|I don't know how to test the thing properly (this is my first open
|source collaboration...). If I do a "./configure && make && make
|install" in my Mandrake 8.1 installation, ¿will it break any
|Mandrake-specific changes, or are those all in config files that
|survive? (¿will my new Sawfish know how to find those files?) 

it should be ok, since it will get installed in /usr/local by default

|
|I came to this doc discrepancy while working on a new theme. I put the
|theme.jl and *.png files in "~/.sawfish/theme/mencial-minimal/" and set
|theme "mencial-minimal" as a default in the Gnome control. But when I
|restart X, windows have no frames. Then I put the "theme.jl" file in
|"~/.sawfish/lisp/" and the "*.png" files in
|"/usr/share/sawfish/images/", start "sawfish-client" and do ",load-file
|"theme.jl"", and from them on all new windows get frames. ¿How can I
|check what's wrong with the theme?

try putting it in ~/.sawfish/themes ? It should be possible to load it
(Continue reading)

Permalink | Reply | 0 comments |
headers
Daniel Pfeiffer | 4 Dec 2002 22:10
X-Face
Picon

Re: display-window/warp-to-window-enabled bugs

John Harper <jsh <at> unfactored.org> skribis:
> Daniel Pfeiffer writes:
> |I get the startling focus-doesn't-follow-mouse situation where the
> |pointer stays where it was, and if that isn't where the new window
> |pops up, that gets the focus but not the mouse.

> You really want the pointer to jump to new windows?

Well I remember the first couple of times it happened, and I temporarily didn't want to take care of the new
window.  I was mousing around the old one, which was registering the mouse activities, but didn't get the
keyboard.  I thought I had done something wrong, and only the third time it happened I figured out what was
going on.

> |While you seem aware of this, I've-done-it-wrong-once is no excuse for
> |doing it wrong in another situation rather than fixing the first.  Ffm
> |(that's short for Frankfurt am Main, btw. :-) sometimes should imply
> |warping when combined with certain other options, and there's nothing
> |wrong with that.

> Yes, if you take things to their logical conclusion, that is true. I
> don't think what you would arrive at would be very usable, but I'd be
> happy to apply a patch that called warp-pointer-if-necessary in more
> situations (since it will have no effect for the vast majority of users)

That would be better :-)

> |Now if you want to take warping away from the Kindergarden lot, then
> |you must also take away everything that then becomes inconsistent!

> Please don't insult people who you think are less technically competent
(Continue reading)

Permalink | Reply | 0 comments |
headers
Christopher Pinon | 6 Dec 2002 19:43

changing font in menus and configurator?

Hi,

I'm using sawfish (v. 1) and was wondering how I can change the font in
the menus and the configurator? How is the default font for these
purposes chosen?

Thank you,
Christopher Pinon
Permalink | Reply | 1 comment |

Gmane