-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathBUGS
272 lines (234 loc) · 11.1 KB
/
BUGS
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
some VALIDATE_WIN_COPY calls should be VALIDATE_WINVALID_COPY, but
haven't exhaustively looked for these. --10/02/99 gjb
-------
Out of color errors need to be reported better.
-------
closing the Fvwm pager causes a deadlock that makes Scwm hang;
kill -1 twice wakes it up, but need real fix.
--tlau
-------------
Race conditions cause scwm to hang when talking to fvwm2 modules when, e.g.,
recapturing.
May have just been forgetting the O_NONBLOCK flag... --09/17/98 gjb
-------
dialog popups that grab the keyboard focus should return focus better
after they are unmap (e.g., with tlau's fugu's search dialog, and escape
aborting out of it).
-> how get key focus
-> why keybindings not work after the Escape is pressed
-> focus-window-with-pointer should work...
-------
using (gtk gtk) before (app scwm gtk) causes a premature exit.
How can we force app scwm gtk to be called when (gtk gtk) module is loaded?
--10/02/99 gjb
-------
XEmacs pixmap icon doesn't survive after a restart (James Achtung Crippe) --09/15/99 gjb
-------
> btw, you might check for dl support before building utilities/pidproperty/,
> at least obsd didn't recognize the RTLD_GLOBAL flag on dlopen, had to
> replace it with DL_LAZY, at least it compiled, not sure if it works. had
> to remove -ldl from the Makefile too. (aren't there some dlopen tests
> done in configure already ?)
Bob Pepin [email protected]
--------
kept-on-top windows cannot be restacked relative to each other. I
tried finding some quick fixes but they cause segfaults, thus this is
something to fix by reconsidering all this code and implementing true
layers functionality.
-------------
[email protected] reports problems using ApplixWare's windows
and Scwm (perhaps related to bitmaps that Applixware puts on the windows)
-------------
Squashed titlebar feature:
it'd be nice if they could be squashed to the right, or centered,
instead of always only left-justified.
-------------
emacs lisp files get installed under / if no emacs (from forcer --09/23/98 gjb)
-------------
1. I pop up an xterm
2. I do a horizontal-toggle-maximize
3. I change to a smaller font.
4. The window automatically resizes due to 3, preserving the number of rows & columns.
5. I do a horizontal-toggle-maximize again to bring it back to normal size.
6. The window resizes to many more rows and columns than it originally
had. It ends up about the same size in pixels, though.
Evidentally toggle-maximize isn't differentiating between windows
sized in characters & windows sized in pixels.
Shouldn't toggling back to the original size go back to the orignal
number of rows & columns, not the original size in pixels? If I
recall correctly, that's what fvwm does.
hstein - 7/16/98
-------------
Build bugs --04/23/99 gjb
-lXpm was not on my link line when I linked on Hydra.cs.mu.oz.au -- why?
gcc didn't find -lstdc++ on hydra, either, and I needed to add
-L/usr/lib/gcc-lib/i486-linux/egcs-2.91.60 to get it. (Are these debian issues?)
-------------
Test with bad image path-- lots of things could be done better
when images are missing.
-------------
Visual bugs in boundaries of windows with some configurations of options.
-------------
icon-box not honoured correctly?
-------------
popup-menus attached to decorations come up centered over the click, instead of
flush with the window edge -- use of popup-menu-from-decoration is a bit hackish
---------------
xdvorak's change mapping should cause all keybindings to be reset and
new keycodes to be gotten (waiting for event rewrite)
---------------
Windows still aren't positioning correctly for me at startup. They're
starting too low (when positioned with -0, as in:
xterm -j -sb -sl 128 -si -geometry 82x48-0-0 -T $XTERMNAME &
xterm -j -sb -sl 128 -si -geometry 80x24+0-0 -T $XTERMNAME &
The 1st xterm also ends up a little to far to the right (looks like
it's off by the window border).
This is with "guile 1.3a" (snap 980630), & a cvs update this morning
of scwm.
--hjstein
-------------
I've noticed this off-and-on with 0.6, and just noticed again with
0.7: My xterms have #:focus 'sloppy set; sometimes (probably about
one time in 25), when I have a pair of partially overlapping
windows and I move the cursor quickly from the root window,
crossing one of the xterms (which held the sloppy focus), and
landing in the other xterm, the new window does not gain focus.
I have to take the cursor off of the window in which I desire
focus and bring it back in, crossing the border slowly.
Ken Pizzini
-------------
- Another window shade problem: clients that resize themsleves while
shaded make the wm lose. [ reported by Oleg Tihnov ]
------------
- Click to focus bugs
1a. At startup usually a window, which I don't want given focus,
gets focus.
1b. The same is true on a restart. Except sometimes no window gets
focus at all, which makes life difficult for `next'.
2. When I quit a X app, the focus is usually given to a window I don't
want to have focus.
Jens
-------
callbacks should run from a new dynamic root.
[ this would prevent them from calling continuations that escape
outside the callback handler - the code would lose big on this -ms]
GONE??? (but didn't note resolving them here)
-------
--01/09/00 gjb
By-window-resource group menus should auto-select the first window in
the sub-menus (they currently require a motion event for the menu item to
get highlighted).
-------------
--01/09/00 gjb
Incorrect usage of keyword arguments by a scheme function fails silently.
Both wrong keyword names and missing arguments should be reported.
-------------
----------------------------------------------------------------------
RESOLVED:
----------------------------------------------------------------------
Sticky windows do not de-iconify properly. --01/09/00 gjb
-------
--01/09/00 gjb
There seems to be a bug in window size (or allowable size)
communication between scwm & applications.
For example, bring up a netscape window. Maximize its height. It
very nicely stretches from the top of the window to the bottom. Now
start giving it lots of Alt-Ns to pop up new netscape windows.
Netscape always starts up new windows the same size as the netscape
window last created or adjusted in size. So, all the windows being
poped up should be the same size. However, each window poped up ends
up one pixel larger in height.
--Harvey Stein
-------
Changing desktops on elements GWS machines (as tzoompy) crashes Scwm; I can't
reproduce as gjb
NLR (No longer reproducible)--09/29/99 gjb
-------
start-x2x-ita causes scwm to grow wildly out of control until that
new x2x process is killed.
ConfigNotify fighting + guile leak? --09/29/99 gjb
-------
*preferences-use-notebook* set to #f causes a seg fault in the gc
from guile-gtk code, it seems. --09/13/99 gjb
Res'd -- Vollmer's patch to guile-gtk --09/29/99 gjb
-------
xload windows (among others, probably) move and resize on restarts --
gravity related?
Res'd --07/17/99 gjb
-------------
wildcard-matcher in wininfo doesn't permit distinguishing based on title,
class, and resource name separately. This is essential for window styles
to do the right thing (a misfeature compared to fvwm2)
----------------------------------------------------------------------
Really smart placement (i.e., clever placement) has an off my one bug.
Try starting a couple of xterms in the viewport one below the home viewport,
then move back to the home viewport -- you can see one row of pixels at the
bottom of the screen. --04/14/99 gjb
Res'd--07/17/99 gjb
----------------------------------------------------------------------
make-color can return #f, we should have make-closest-color that is
guaranteed to work and return the closest color to the requested one.
-------------
Hooks which take wrong number of params cause seg faults!
--07/04/98 gjb
----------------------------------------------------------------------
windows having a different border-width than the default (6) are
placed incorrectly if they have SE gravity. They are several pixels
[(6-border_width)*2] off the screen-border.
This is due to the fact that `PlaceWindow' is called before the
evaluation of window-styles. `PlaceWindow' needs to be called if the
border-width changed (from `set_border_width_x', or `SetupFrame', or
whatever). It would be nice if the call to `PlaceWindow' before the
window is properly styled could then be removed, but I don't know what
this could break.
Robert Bihlmeyer
Robert verified this has been fixed -- 09/26/98 jtl
------------
Changing title-height in the decor (e.g., by changing the font) doesn't
affect new windows using that style
-------------
xterm-s started from scwm die when scwm dies.... why?
see scwm-system in base.scm -- the terminal process group was getting
the SIGINT from the controlling tty --09/17/98 gjb
-------------
gjb gets this on recapture
[Scwm][ScwmErrorHandler]: <<ERROR>> Request 54, Error 4, EventType: 2
XFreePixmap
This could be due to the recapture code losing track of the pixmap due
to some change I made in the pixmap code --11/08/97 gjb
Not reproducible any longer --08/02/98 gjb
- Animated window-shade bug, reported by Sam Steingold <[email protected]>,
exploit confirmed by MS, but I have no clue how to fix it right now:
Open 2 windows, one (1) partially obscuring the other's (2) title bar.
Double click on the (1)'s title bar, shading it. You can see that the
portion of the (2)'s title bar with was obscured before but it not
obscured now is not repainted. Also, if you move the mouse towards (2)
while (1) is being (animatedly ==> with delay) shaded, so that the (1)
is "pulled from under the mouse pointer", leaving it on (2), (2) will
*not* have the focus, although it should as per sloppy focus.
--03/10/98 gjb; XSync was discarding events, and we needed a CoerceEnterNotifyOnCurrentWindow() to restore proper focus
- Jim Blandy's patch to load boot-9 for newer guile versions is
problematic; it can't tell wether or not we are running under an older
guile which has already loaded boot-9, resulting in a segfault. - MS
[fixed for now - MS]
- buggy icon handling everywhere -- needs thorough testing
-- 2-20-97 mstachow
should initially focus/hilight some window (also a bug in fvwm2)
-- Fixed by adding CoerceEnterNotifyOnCurrentWindow()
Memory allocation errors can occur if guile was built with Rx support;
autoconf sets this option if it finds a librx.{a,so} in the paths it
checks. --11/08/97 gjb
Passing a bad window from guile can cause scwm to dump core in
ensure_valid(); MS says fixed ~ 11/1/97
functions should be renamed to end with _x
-- This looks done... is it not? --11/08/97 gjb
(get-window) acts wierdly when a window needs to be selected, or
when executed via scwmsend's property mechanism -- was a problem with
DeferExecution(); not sure if it's completely right now, but seems to
work in the cases I've tried -- 10/25/97, gjb
there needs to be a way to remove mouse bindings; unbind_key added 10/24/97, gjb
start interactive moves and resizes right, so that initial mouse motion is not
lost. ~ 10/15/97, MS
Oleg Tihonov <[email protected]>'s bug re: bind-mouse
and unbind-mouse. bind-mouse grabbed for all windows unnecessarily, and
unbind-mouse did not permit `1' to mean moust button 1 --gjb 11/27/97