Bug#832875: Avoid "utility" in package description

[Ulrich Eckhardt]

Package: python-json-schema-validator
Version: 2.3.1-3

Greetings!

The package description is "JSON schema validation utility", which is a bit
misleading. The problem is that it doesn't provide a utility with which you
can validate JSON agains a schema. Instead, it provides a "Python module"
or a "Python library" for integrating validation.

Also, minor typo, there's a "the" missing here:
"This package provides a JSON Schema validator conforming to a subset of
 second draft of the specification."

Further, some questions here:
"JSON files can be verified against a schema expressed in the python source
code, producing either a SchemaError exception or ValidationError
exception."
Can you only hard-code the schema in Python or also read the schema (which
itself is JSON) from a file? Also, can you only validate files or can you
validate in-memory data, too?

Thanks to you and the Debian team for your effort!

Uli

Bug#831554: cyrus-nntpd-2.4: nntp not working, regression of #734648?

[Ulrich Eckhardt]

Package: cyrus-nntpd-2.4
Version: 2.4.17+nocaldav-0~deb8u1
Severity: grave
Justification: renders package unusable

Dear Maintainer,

I have upgraded my server to from Debian 7 to Debian 8.
After this upgrade it is not possible anymore to feed news into
my imap spool. In the logs I get the following output from fetchnews:
cyrus/fetchnews[6474]: fetchnews: nntp.aioe.org offered 4; localhost rejected 
0, accepted 0, failed 4

This problem was registered as bugreport on cyrus and fixed in 2.4.18:
https://bugzilla.cyrusimap.org/show_bug.cgi?id=3833 . A detailed description of
the problem can be found there. Also I had a similar problem during
upgrading from debian 6 to debian 7: 
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=734648

I compiled the cyrus nntpd from cyrus 2.4.18. With this nntpd I could 
successfully deliver news into my imap spool.

-- System Information:
Debian Release: 8.5
  APT prefers stable
  APT policy: (500, 'stable')
Architecture: i386 (i686)

Kernel: Linux 3.2.41-042stab111.12 (SMP w/2 CPU cores)
Locale: LANG=de_DE.utf8, LC_CTYPE=de_DE.utf8 (charmap=UTF-8) (ignored: LC_ALL 
set to de_DE.utf8)
Shell: /bin/sh linked to /bin/dash
Init: systemd (via /run/systemd/system)

Versions of packages cyrus-nntpd-2.4 depends on:
ii  cyrus-nntpd  2.4.17+nocaldav-0~deb8u1
ii  dpkg 1.17.27

cyrus-nntpd-2.4 recommends no packages.

cyrus-nntpd-2.4 suggests no packages.

-- no debconf information

Bug#576829: dependency on KDE?

[Ulrich Eckhardt]

Per Olofsson  wrote:
> xdg-utils does not depend on KDE, actually. It is supposed to work
> with any desktop environment.
> 
> Which desktop environment (if any) are you using?

Hmmm, question is rather which DE I was using (on my old computer
at that time ...), but I think I used plain fluxbox at that time, which
is also what I'm running now, although on a different machine which was
setup from zero. I do have kde-runtime installed though (I don't
know why), which provides kreadconfig.


At the moment, I can't reproduce the issue in the original bug report
any more, although there still seems to be some kind of dependency on
KDE:

uli@zox ~> xdg-settings get default-web-browser
/usr/bin/xdg-settings: 1: /usr/bin/xdg-settings: kde-config: not found
chrome.desktop


The first line of output was actually to stderr, the second, expected
line was to stdout. However, after that, I tried what the bug reporter
initially complained about, and lo and behold, the error went away! By
pure luck, I happened to run the suggested "sh -x /usr/bin/xdg-settings
get default-web-browser" before that though:

uli@zox ~> sh -x /usr/bin/xdg-settings get default-web-browser
+ check_common_commands get default-web-browser
+ [ 2 -gt 0 ]
+ parm=get
+ shift
+ [ 1 -gt 0 ]
+ parm=default-web-browser
+ shift
+ [ 0 -gt 0 ]
+ [ -z  ]
+ unset XDG_UTILS_DEBUG_LEVEL
+ [ 0 -lt 1 ]
+ xdg_redirect_output= > /dev/null 2> /dev/null
+ [ xget = x--list ]
+ [ xget != x ]
+ [ xdefault-web-browser != x ]
+ [ xget = xget -o x != x ]
+ op=get
+ parm=default-web-browser
+ shift 2
+ [ xget != xget -a xget != xcheck -a xget != xset ]
+ detectDE
+ unset GREP_OPTIONS
+ [ -n  ]
+ [ x = x ]
+ [ xtrue = xtrue ]
+ DE=kde
+ [ xkde = x ]
+ [ xkde = x ]
+ [ xkde = xgnome ]
+ dispatch_specific
+ [ xget = xget ]
+ get_browser_kde
+ read_kde_browser
+ read_kde_config kdeglobals General BrowserApplication
+ configfile=kdeglobals
+ configsection=General
+ configkey=BrowserApplication
+ kreadconfig --file kdeglobals --group General --key BrowserApplication
+ application=
+ [ x != x ]
+ kde-config --path config
+ cut/usr/bin/xdg-settings: 1: /usr/bin/xdg-settings: kde-config: not
found -d : -f 1
+ configfile_dir=
+ configfile_path=/kdeglobals
+ [ ! -f /kdeglobals ]
+ return
+ browser=
+ [ x = x ]
+ get_browser_mime
+ test -z 
+ MIME=text/html
+ xdg_mime_fixup
+ [ kde = kde -a -z  ]
+ ktradertest text/html Application
+ XDG_MIME_FIXED=yes
+ xdg-mime query default text/html
chrome.desktop
+ [ 0 -eq 0 ]
+ exit_success
+ [ 0 -gt 0 ]
+ exit 0


I then changed the default-web-browser to test the original bug report,
but couldn't reproduce the issue. However, after that, even the error
message from xdg-settings is gone, I haven't rebooted or restarted the
session though! Anyhow, here's how the trace looks afterwards:

uli@zox ~> sh -x /usr/bin/xdg-settings get default-web-browser
+ check_common_commands get default-web-browser
+ [ 2 -gt 0 ]
+ parm=get
+ shift
+ [ 1 -gt 0 ]
+ parm=default-web-browser
+ shift
+ [ 0 -gt 0 ]
+ [ -z  ]
+ unset XDG_UTILS_DEBUG_LEVEL
+ [ 0 -lt 1 ]
+ xdg_redirect_output= > /dev/null 2> /dev/null
+ [ xget = x--list ]
+ [ xget != x ]
+ [ xdefault-web-browser != x ]
+ [ xget = xget -o x != x ]
+ op=get
+ parm=default-web-browser
+ shift 2
+ [ xget != xget -a xget != xcheck -a xget != xset ]
+ detectDE
+ unset GREP_OPTIONS
+ [ -n  ]
+ [ x = x ]
+ [ xtrue = xtrue ]
+ DE=kde
+ [ xkde = x ]
+ [ xkde = x ]
+ [ xkde = xgnome ]
+ dispatch_specific
+ [ xget = xget ]
+ get_browser_kde
+ read_kde_browser
+ read_kde_config kdeglobals General BrowserApplication
+ configfile=kdeglobals
+ configsection=General
+ configkey=BrowserApplication
+ kreadconfig --file kdeglobals --group General --key BrowserApplication
+ application=chrome.desktop
+ [ xchrome.desktop != x ]
+ echo chrome.desktop
+ browser=chrome.desktop
+ [ xchrome.desktop = x ]
+ resolve_kde_browser_desktop
+ [ -z chrome.desktop ]
+ echo chrome.desktop
chrome.desktop
+ [ 0 -eq 0 ]
+ exit_success
+ [ 0 -gt 0 ]
+ exit 0


The difference starts after the call to kreadconfig, which seems to
yield the selected browser after setting it once. In both cases, there
is a line "DE=kde", which is bogus. Looking at "xprop -root", I see
KDE_FULL_SESSION is set. This comes from my fluxbox startup script
where it was used to work around some bug in kded (used to be part of
KDE3, I used to run some KDE applications back then).


I just deactivated the workaround in fluxbox startup, and now at least
the DE is detected correctly:

uli@zox ~> sh -x /usr/bin/xdg-settings get default-web-browser
+ check_common_commands get default-web-browser
+ [ 2 -gt 0 ]
+ parm=get
+ shift
+ [ 1 -gt 0 ]
+ parm=default-web-browser
+ shift
+ [ 0 -gt 0 ]
+ [ -z  ]
+ unset XDG_UTILS_DEBUG_LEVEL
+ [ 0 -lt 1 ]
+ xdg_redirect_output= > /dev/null 2> /dev/null
+ [ xget = x--list ]
+ [ xget != x ]
+ [ xdefault-web-browser != x ]
+ [ xget = xget -o x != x ]
+ op=get
+ parm=default-web-browser
+ shift 2
+ [ xget != xget -a xget != xcheck 

Bug#755724: i3lock: segfault on start

[Ulrich Eckhardt]

Package: i3lock
Version: 2.6-1
Severity: important

Dear Maintainer,

I just tried to try out i3lock. When I run it, I expect it to lock my
X11 session until I unlock it with a password.
Instead, it just dies with a segfault.

I'm not using i3 (which seems to be the primary target) but fluxbox.
I also had xscreensaver running before. Stopping it doesn't make any
difference, but I don't know what persistant settings it leaves in X.


GDB backtrace:

#0  0x76a6d198 in ?? () from 
/usr/lib/x86_64-linux-gnu/libxkbcommon-x11.so.0
#1  0x76a6d3ae in ?? () from 
/usr/lib/x86_64-linux-gnu/libxkbcommon-x11.so.0
#2  0x76a6ce0c in ?? () from 
/usr/lib/x86_64-linux-gnu/libxkbcommon-x11.so.0
#3  0x76a6ac1c in xkb_x11_keymap_new_from_device ()
   from /usr/lib/x86_64-linux-gnu/libxkbcommon-x11.so.0
#4  0x00404079 in ?? ()
#5  0x00403288 in ?? ()
#6  0x75d88b45 in __libc_start_main (main=0x402fa0, argc=1, 
argv=0x7fffe3d8, 
init=optimized out, fini=optimized out, rtld_fini=optimized out, 
stack_end=0x7fffe3c8)
at libc-start.c:287
#7  0x00403819 in ?? ()


Tail of strace output:

socket(PF_FILE, SOCK_STREAM|SOCK_CLOEXEC, 0) = 3
connect(3, {sa_family=AF_FILE, path=@/tmp/.X11-unix/X0}, 20) = 0
getpeername(3, {sa_family=AF_FILE, path=@/tmp/.X11-unix/X0}, [20]) = 0
[...]
access(/usr/share/X11/xkb, R_OK|X_OK) = 0
stat(/home/uli/.xkb, 0x7fffa12da830)  = -1 ENOENT (No such file or directory)
poll([{fd=3, events=POLLIN|POLLOUT}], 1, 4294967295) = 1 ([{fd=3, 
revents=POLLOUT}])
writev(3, 
[{\207\1\4\0\3\0\7\0\0\0\7\0\337\0\337\0\207\30\4\0\0\1\0\0\0\0\0\0\0\0\0\0, 
32}], 1) = 32
poll([{fd=3, events=POLLIN}], 1, 4294967295) = 1 ([{fd=3, revents=POLLIN}])
recvmsg(3, {msg_name(0)=NULL, 
msg_iov(1)=[{\1\3\5\0\6\0\0\0\0\0\36\0\0\0\0\0\0\0\0\0\0\1\0\0\0\377\0\0\0\0\0\0...,
 4096}], msg_controllen=0, msg_flags=0}, 0) = 56
--- SIGSEGV (Segmentation fault) @ 0 (0) ---
+++ killed by SIGSEGV +++
Segmentation fault


Looking at strace, it seems as if XCB was reading into a buffer of insufficient 
size,
so it might be a bug there. Let me know if there's anything else I can do.


Cheers!

Uli


-- System Information:
Debian Release: jessie/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (500, 'testing'), (500, 'stable')
Architecture: amd64 (x86_64)

Kernel: Linux 3.13-1-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages i3lock depends on:
ii  libc6   2.19-5
ii  libcairo2   1.12.16-2
ii  libev4  1:4.15-3
ii  libpam0g1.1.8-3
ii  libxcb-dpms01.10-3
ii  libxcb-image0   0.3.9-1
ii  libxcb-shm0 1.10-3
ii  libxcb-util00.3.8-2
ii  libxcb-xinerama01.10-3
ii  libxcb-xkb1 1.10-3
ii  libxcb1 1.10-3
ii  libxkbcommon-x11-0  0.4.1-1
ii  libxkbcommon0   0.4.0-1

i3lock recommends no packages.

i3lock suggests no packages.

-- no debconf information


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#755724: i3lock: segfault on start

[Ulrich Eckhardt]

Michael Stapelberg stapelb...@debian.org wrote:
 Ulrich Eckhardt ulrich.eckha...@base-42.de writes:
  ii  libxkbcommon-x11-0  0.4.1-1
  ii  libxkbcommon0   0.4.0-1
 Note that you have version 0.4.1-1 of libxkbcommon-x11-0 but 0.4.0-1
 of libxkbcommon0. That _should_ not be a problem, but you can you
 please try using 0.4.1-1 of both and (if that doesn’t help) 0.4.0-1
 of both?

Good catch, Michael! I upgraded libxkbcommon0 to 0.4.1-1 and i3lock now
works.

Thanks!

Uli


--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#747860: Can't switch CRTC between different outputs

[Ulrich Eckhardt]

Package: arandr
Version: 0.1.7.1-1
Severity: wishlist

Hi!

I have a setup with a laptop, a docking station and two external monitors
connected to the docking station. Since the laptop can only drive two
displays at a time (it only has two CRTCs), I use xrandr to turn off the
internal LVDS display and activate the two external displays. This works,
but is cumbersome and requires three steps.

Using arandr, I was hoping that it could turn off the builtin display and
activate the external displays in one step and vice versa. However, this
doesn't work, since activating the second external display fails because
the required CRTC is already in use. It would be nice if arandr could
activate the displays in the required number of steps automatically.

Just for the record, a similar issue exists for lxrandr, the other GUI for
xrandr in Debian.

Thanks!

Uli

Bug#747861: Can't switch CRTC between different outputs

[Ulrich Eckhardt]

Package: lxrandr
Version: 0.1.2-3
Severity: wishlist

Hi!

I have a setup with a laptop, a docking station and two external monitors
connected to the docking station. Since the laptop can only drive two
displays at a time (it only has two CRTCs), I use xrandr to turn off the
internal LVDS display and activate the two external displays. This works,
but is cumbersome and requires three steps.

Using lxrandr, I was hoping that it could turn off the builtin display and
activate the external displays in one step and vice versa. However, this
doesn't work, since activating the second external display fails because
the required CRTC is already in use. It would be nice if lxrandr could
activate the displays in the required number of steps automatically.

Just for the record, a similar issue exists for arandr, the other GUI for
xrandr in Debian.

Thanks!

Uli

Bug#746462: Password entry via keyboard shortcut messed up

[Ulrich Eckhardt]

package: keepass2
version: 2.25+dfsg-1

The keyboard shortcut to auto-type a password (control+V) is causing me
problems, sometimes it works, sometimes not. I have the feeling that
keepass starts typing the password while I still have the control key held,
which changes the meaning of the generated keypresses. keepass2 should wait
until the user released the control key.

I think that an earlier version (the one in testing from ~6 months ago)
didn't show these problems. That version had the problem that the control
key remained stuck, i.e. X thought that the key was still held and only
pressing the according control key worked to get X in sync with reality
again.

As a workaround, using the auto-type function from the context menu seems
to work without issues.

Cheers!

Uli

Bug#734648: cyrus-nntpd-2.4: nntp not working after upgrade to wheezy

[Ulrich Eckhardt]

Package: cyrus-nntpd-2.4
Version: 2.4.16-4+deb7u1
Severity: grave
Tags: upstream
Justification: renders package unusable

Dear Maintainer,

I have upgraded my server to from Debian 6 (with cyrus 2.2) to Debian 7 (with
cyrus 2.4.16). After this upgrade it is not possible anymore to feed news into
my imap-spool.

This problem was registered as bugreport on cyrus:
https://bugzilla.cyrusimap.org/show_bug.cgi?id=3833 . A detailed description of
the problem can be found there.

This bug was resolved quickly by the cyrus team but planned for 2.4.18. I was
able to incorporate the two commits of the solution into the source package
cyrus-imapd-2.4_2.4.16-4+deb7u1 and build a new working nntpd for cyrus 2.4.16.
The following patch can be applied to fix this problem:

---

--- nntpd.c.deb 2014-01-07 20:33:00.827064145 +0100
+++ nntpd.c 2014-01-08 19:10:03.242389940 +0100
@@ -1065,7 +1065,7 @@
be = backend_current;
if (arg1.len 
(!is_newsgroup(arg1.s) ||
-(r = open_group(arg1.s, 1, be, NULL goto nogroup;
+(r = open_group(arg1.s, 0, be, NULL goto nogroup;
else if (be) {
prot_printf(be-out, %s, cmd.s);
if (arg1.len) {
@@ -1780,6 +1780,8 @@
 if (!has_prefix) {
snprintf(mailboxname, sizeof(mailboxname), %s%s, newsprefix, name);
name = mailboxname;
+
+   if (!is_newsgroup(name)) return IMAP_MAILBOX_NONEXISTENT;
 }

 if (!r) r = mlookup(name, newserver, acl, NULL);
@@ -2544,7 +2546,7 @@
}
 }
 else {
-   prot_printf(nntp_out, %s %u %u %c\r\n, name,
+   prot_printf(nntp_out, %s %u %u %c\r\n, name+strlen(newsprefix),
group_state-exists ? index_getuid(group_state,
group_state-exists) :
group_state-mailbox-i.last_uid,
group_state-exists ? index_getuid(group_state, 1) :
@@ -3000,7 +3002,7 @@
if (!rcpt) return -1;

/* construct the mailbox name */
-   sprintf(rcpt, %.*s, (int) n, p);
+   sprintf(rcpt, %s%.*s, newsprefix, (int) n, p);

/* skip mailboxes that we don't serve as newsgroups */
if (!is_newsgroup(rcpt)) continue;






-- System Information:
Debian Release: 7.3
  APT prefers stable-updates
  APT policy: (500, 'stable-updates'), (500, 'stable')
Architecture: i386 (i686)

Kernel: Linux 2.6.32-042stab078.27 (SMP w/1 CPU core)
Locale: LANG=de_DE.utf8, LC_CTYPE=de_DE.utf8 (charmap=UTF-8) (ignored: LC_ALL 
set to de_DE.utf8)
Shell: /bin/sh linked to /bin/bash

Versions of packages cyrus-nntpd-2.4 depends on:
ii  cyrus-common-2.4  2.4.16-4+deb7u1
ii  libc6 2.13-38
ii  libcomerr21.42.5-1.1
ii  libdb5.1  5.1.29-5
ii  libsasl2-22.1.25.dfsg1-6+deb7u1
ii  libssl1.0.0   1.0.1e-2+deb7u3
ii  libwrap0  7.6.q-24
ii  zlib1g1:1.2.7.dfsg-13

cyrus-nntpd-2.4 recommends no packages.

cyrus-nntpd-2.4 suggests no packages.

-- no debconf information


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#418630: Similar issues for a Samsung printer

[Ulrich Eckhardt]

Hi!

I'm using 1.0.23-3 here and I'm facing similar issues. The device
is detected as

| uli@zox:~$ scanimage -L
| Failed cupsGetDevices
| device `smfp:usb;04E8;344F;Z72TB8KD4F00NFW' is a SAMSUNG SCX-3400 Series on 
USB Scanner

(I'm not sure what the CUPS-output is doing there...). Any --mode
argument I tried is rejected, both with and without the connecting =
and with arguments Gray, Grayscale and Color. I couldn't get the
--preview option to work either.

BTW: Have you already forwarded this bug to upstream? I sense that this
mixture of different argument styles isn't something that can be
resolved here in Debian but rather something that requires a decision
from upstream first.

Cheers!

Uli


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#711447: clicking on tabs doesn't focus windows

[Ulrich Eckhardt]

Hi!

I have just discovered what seems to be another aspect of this bug. If
you switch to click tab focus (Configuration/Focus Model/Click Tab
Focus), the window doesn't get focused when clicking on a tab.

Note that this requires external tabs, with internal tabs you can click
to change focus. Also, this bug is not present in stable (1.3.2-4) but
only in testing (1.3.5-1).

Cheers!

Uli


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#576829: dependency on KDE?

[Ulrich Eckhardt]

Hi!

I just stumbled across the same problem using 1.1.0~rc1+git20111210-7.
I'm not sure if that is the cause, but kfmclient is the KDE File
Manager client, as far as I know, so it seems that xdg-utils actually
depend on KDE. Could that be the reason?

Uli


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#711447: clicking on tabs doesn't raise windows

[Ulrich Eckhardt]

Package: fluxbox
Version: 1.3.5-1
Severity: wishlist

Hi!

I believe that since a recent update of fluxbox in testing, you can't
click on window tabs any more in order to raise the window. Similarly,
though not as grave, you can't drag a tab onto a group in order to add
the window to the group or to reorder the tabs.

I'm now wondering if this was by accident or intentional...

Greetings from Hamburg!

Uli



-- System Information:
Debian Release: jessie/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable')
Architecture: amd64 (x86_64)

Kernel: Linux 3.2.0-4-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages fluxbox depends on:
ii  libc6   2.17-3
ii  libfontconfig1  2.9.0-7.1
ii  libfreetype62.4.9-1.1
ii  libfribidi0 0.19.5-2
ii  libgcc1 1:4.8.0-7
ii  libice6 2:1.0.8-2
ii  libimlib2   1.4.5-1
ii  libsm6  2:1.2.1-2
ii  libstdc++6  4.8.0-7
ii  libx11-62:1.5.0-1+deb7u1
ii  libxext62:1.3.1-2+deb7u1
ii  libxft2 2.3.1-1
ii  libxinerama12:1.1.2-1+deb7u1
ii  libxpm4 1:3.5.10-1
ii  libxrandr2  2:1.3.2-2+deb7u1
ii  libxrender1 1:0.9.7-1+deb7u1
ii  menu2.1.46
ii  zlib1g  1:1.2.8.dfsg-1

Versions of packages fluxbox recommends:
ii  feh  2.3-2
ii  xfonts-terminus  4.35-1

Versions of packages fluxbox suggests:
pn  fbautostart  none
pn  fbdesk   none
pn  fbpager  none

-- no debconf information


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#708853: support browsing remote filesystems via SSH

[Ulrich Eckhardt]

Package: thunar
Version: 1.2.3-4+b1
Severity: wishlist

I'm trying to access a filesystem on a different system using SSH. I
know that thunar supports this, but I can't get it to work or even
extract any meaningful error message. Just to make that clear: I'm not
running XFCE, just thunar inside a fluxbox session!

I've been doing some research, and it seems that thunar requires some
external packages in order to use SSH, but I don't know which and they
are not hard dependencies. In any case, all recommends and suggests are
installed, apart from xfce4-panel. I have also installed gvfs-bin and
gvfs-fuse, but neither helped. I guess that it's just a matter of
documenting these dependencies (README.Debian perhaps?), if these are
the problem at all.

That said, the error messages are not really helpful. If I put
ssh://192.168.1.2/ into thunar's location field, it just displays a
red icon next to it, without any further error message. If I start it
from a terminal, I get a popup and the same text echoed on the console:

  ~$ thunar ssh://192.168.1.2/
  Thunar: Failed to open ssh://192.168.1.2/: The specified location
  is not supported

This is very similar to gvfs:

  ~$ gvfs-mount -m  ssh://192.168.1.2/
  Error mounting location: The specified location is not supported


I'm confident I could find out what's wrong myself, but I think a user
shouldn't be required to read the sources and search the web in order
to get this to work, hence the wishlist bug.


Greetings from Hamburg!

Uli


-- System Information:
Debian Release: jessie/sid
  APT prefers testing
  APT policy: (500, 'testing')
Architecture: amd64 (x86_64)

Kernel: Linux 3.2.0-4-amd64 (SMP w/4 CPU cores)
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages thunar depends on:
ii  desktop-file-utils  0.20-0.1
ii  exo-utils   0.6.2-5
ii  libatk1.0-0 2.4.0-2
ii  libc6   2.13-38
ii  libcairo2   1.12.2-3
ii  libdbus-1-3 1.6.8-1
ii  libdbus-glib-1-20.100.2-1
ii  libexo-1-0  0.6.2-5
ii  libgdk-pixbuf2.0-0  2.26.1-1
ii  libglib2.0-02.33.12+really2.32.4-5
ii  libgtk2.0-0 2.24.10-2
ii  libgudev-1.0-0  175-7.2
ii  libice6 2:1.0.8-2
ii  libnotify4  0.7.5-2
ii  libpango1.0-0   1.30.0-1
ii  libsm6  2:1.2.1-2
ii  libthunarx-2-0  1.2.3-4+b1
ii  libxfce4ui-1-0  4.8.1-1
ii  libxfce4util4   4.8.2-1
ii  shared-mime-info1.0-1+b1
ii  thunar-data 1.2.3-4

Versions of packages thunar recommends:
ii  dbus-x111.6.8-1
ii  gvfs1.12.3-4
ii  libfontconfig1  2.9.0-7.1
ii  libfreetype62.4.9-1.1
ii  thunar-volman   0.6.1-1
ii  tumbler 0.1.25-1+b1
ii  xdg-user-dirs   0.14-1
ii  xfce4-panel 4.8.6-4

Versions of packages thunar suggests:
ii  thunar-archive-plugin 0.3.0-4
ii  thunar-media-tags-plugin  0.2.0-1

-- no debconf information


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#652228: Works with Chrome, fails with Iceweasel

[Ulrich Eckhardt]

I just tried dragging an image from thunar (1.2.3-4+b1, running without
XFCE) onto the demo field in both Iceweasel (10.0.12esr-1+nmu1) and
Chromium (26.0.1410.43-1). It works only in Chromium.

Just my 2cc.

Uli


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#708853: [Pkg-xfce-devel] Bug#708853: support browsing remote filesystems via SSH

[Ulrich Eckhardt]

Yves-Alexis Perez wrote:
 Actually you miss one suggests (from gvfs): gvfs-backends.
 
 Can you install it and report back?

Ah yes, that one was indeed missing, and it works now! Can you please
add a note concerning this to /usr/share/doc/README.Debian? That would
have helped immensely! Other than that, it's still a useless error
message, though I'm not sure if that is from thunar or gvfs.

Thank you very much!

Uli


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#696221: Patch to provide more detailed info on errors

[Ulrich Eckhardt]

Package: cutecom
Version: 0.22.0-2

When opening one of the serial ports fails, you only get the error that it
failed but not why, which is frustrating and not helpful. The typical faults
are that a port isn't there or that you don't have access permissions. The
following piece of code changes the error handling after open() so that the
actual error becomes available to the user:

   m_fd=open(dev.toLatin1(), flags | O_NDELAY);
   if (m_fd0)
   {
  int e = errno;
  char* es = strerror(e);
  std::cerropening failed with errno=  e  (  es  ) 
std::endl;
  m_fd=-1;
  QMessageBox::information(this, tr(Error),
   tr(Could not open %1:\n error #%2 
(%3)).arg(dev).arg(e).arg(es));
  return;
   }


Cheers!

Uli


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#691693: Tabs in Python sourcefiles

[Ulrich Eckhardt]

Package: geany
Version: 1.22+dfsg-2

With the default configuration, Geany uses tabs to indent Python sources. 
According to Python's coding guidelines (PEP 8), sources should not use any 
tabs but exactly four spaces. As a workaround, you can manually switch to 
spaces under Document  Indent Type  Spaces, which already uses the correct 
width.

Two notes:
1. I'm not sure if the Python template file is generated on the fly or loaded, 
but the template also uses tabs.
2. A similar issue might be Makefiles, which actually rely on being indented 
with tabs. I haven't tested those though.

Thank you!

Uli


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#687696: Improper handling of non-ASCII paths

[Ulrich Eckhardt]

Package: svn-load
Version: 1.3-1

I'm trying to load a directory containing non-ASCII filenames, which fails. I'm 
pretty sure that the encoding is proper UTF-8 and that the locale is set up 
correctly. I have manually verified the encoding and tools like e.g. ls and 
thunar also handle it. If you want to verify yourself, the files are part of 
the ca-certificates package and stored in /etc/ssl.


What I did was this:

uli@zox:~$ svnadmin create tmp-repo

uli@zox:~$ svn-load file:///home/uli/tmp-repo ssl /etc/ssl
Adding /tmp/svn-load2yNV6x/working/openssl.cnf
Adding /tmp/svn-load2yNV6x/working/certs
Adding /tmp/svn-load2yNV6x/working/certs/Entrust.net_Secure_Server_CA.pem
Adding /tmp/svn-
load2yNV6x/working/certs/WellsSecure_Public_Root_Certificate_Authority.pem
Adding /tmp/svn-load2yNV6x/working/certs/5ad8a5d6.0
Adding /tmp/svn-load2yNV6x/working/certs/SwissSign_Platinum_CA_-_G2.pem
Traceback (most recent call last):
  File /usr/bin/svn-load, line 649, in module
overlay_files(client, workingdir, d)
  File /usr/bin/svn-load, line 438, in overlay_files
client.add(counterpath, ignore=False)
pysvn._pysvn_2_7.ClientError: Error during add of '/tmp/svn-
load2yNV6x/working/certs'
Can't convert string from native encoding to 'UTF-8':
AC_Ra?\195?\173z_Certic?\195?\161mara_S.A..pem

uli@zox:~$ svn-load file:///home/uli/tmp-repo ssl /etc/ssl/certs
Adding /tmp/svn-loadyTTNvb/working/Entrust.net_Secure_Server_CA.pem
Adding /tmp/svn-
loadyTTNvb/working/WellsSecure_Public_Root_Certificate_Authority.pem
Adding /tmp/svn-loadyTTNvb/working/5ad8a5d6.0
Adding /tmp/svn-loadyTTNvb/working/SwissSign_Platinum_CA_-_G2.pem
Traceback (most recent call last):
  File /usr/bin/svn-load, line 649, in module
overlay_files(client, workingdir, d)
  File /usr/bin/svn-load, line 420, in overlay_files
client.add(counterpath, ignore=False)
UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 33: 
ordinal not in range(128)


As you see, the actual error differs a bit depending on whether I load /etc/ssl 
or /etc/ssl/certs. Source browsing a bit seems to indicate that in the first 
case the file is added due to a recursive add while in the second case it is 
added directly. In both cases though, it seems that the passed parameter is 
treated as a string, while it is actually an array of bytes already encoded as 
UTF-8.

As a workaround, using svn_load_dirs from the subversion-tools packages seems 
to do the job.

Greetings!

Uli


-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#679035: Useless warning about disconnected USB device

[Ulrich Eckhardt]

Package: smartmontools
Version: 5.41+svn3365-1

I have a portable hard drive, a WD my passport. Recently, after 
disconnecting, smart-notifier started giving me regular warnings about not 
being able to access the disconnected device. Since the device is 
disconnected, this is little surprising and little helpful.

I haven't tried whether this is reliably reproducible. One thing making this 
case possibly special is that the harddisk actually has bad blocks that would 
be a reason for warning.

A single warning about a disconnected device is probably useless, but doesn't 
hurt. Repeatedly warning about a device that is disconnected is not only 
useless but annoying. In partictular since everytime I get a popup, I first 
have to unfold the part that contains the actual error. If I get too many, 
I'll start not looking at the errors. The net result is that important errors 
are shadowed by the useless warnings.

Please disable these notifications!

Thank you!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#671021: broken link in docs

[Ulrich Eckhardt]

Package: python-pyinotify
Version: 0.9.3-1

If you import the pyinotify module and invoke help(pyinotify), you get the 
following output:

| MODULE DOCS
| http://docs.python.org/library/pyinotify

This link doesn't exist, pyinotify is not (yet?) part of upstream's standard 
library.

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#665700: thunar scrolls erratically when hover-selecting files

[Ulrich Eckhardt]

Package: thunar
Version: 1.2.3-3

There are two observations which are related, I guess. I'm using the detailed 
list view (Control-2). The columns displayed are Name, Date Modified, 
Size. I have checked Automatically expand columns as needed in the 
configuration.

Case I: Here, the size of the Name column is less than the width of the 
window, but the other two columns exceed the width (just wondering, doesn't 
that already contradict the settings?), so that a scroll bar is visible at the 
bottom. When I scroll right, and hover-select a file (different or same as 
before), the scrollbar is reset to the leftmost position.


Case II: Here, the size of the Name column already exceeds the width of the 
window. If I change the selected file by hovering the mouse over it, the 
position is scrolled horizontally, alternating between the leftmost and a 
middle position. The middle position is so that the rightmost part of the 
Name column is aligned to the right border of the window.


How Thunar should behave: It shouldn't scroll horizontally at all. If I look 
at a column and select files there, it shouldn't scroll away from the column.

Further notes:
 - I'm not sure if the auto-expand code behaves correctly when there simply 
isn't enough space.
 - I'm not sure if the auto-expand code works at all. I have seen it change 
layout sometimes, but I couldn't find a rule to it.
 - I'm using Fluxbox as window manager, in case that matters (parenting, 
resize notifications...).


Thanks for maintaining this software, btw!

Uli


-- System Information:
Debian Release: wheezy/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (1, 'experimental')
Architecture: powerpc (ppc)

Kernel: Linux 3.2.0-1-powerpc
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages thunar depends on:
ii  desktop-file-utils  0.18-1
ii  exo-utils   0.6.2-4
ii  libatk1.0-0 2.2.0-2
ii  libc6   2.13-27
ii  libcairo2   1.10.2-7
ii  libdbus-1-3 1.4.18-1
ii  libdbus-glib-1-20.98-1
ii  libexo-1-0  0.6.2-4
ii  libgdk-pixbuf2.0-0  2.24.1-1
ii  libglib2.0-02.30.2-6
ii  libgtk2.0-0 2.24.10-1
ii  libgudev-1.0-0  175-3.1
ii  libice6 2:1.0.7-2
ii  libnotify4  0.7.4-1
ii  libpango1.0-0   1.29.4-3
ii  libsm6  2:1.2.0-2
ii  libthunarx-2-0  1.2.3-3
ii  libxfce4ui-1-0  4.8.1-1
ii  libxfce4util4   4.8.2-1
ii  shared-mime-info0.90-1
ii  thunar-data 1.2.3-3

Versions of packages thunar recommends:
pn  dbus-x111.4.18-1
pn  gvfs1.10.1-2+b1
pn  libfontconfig1  2.8.0-3.1
pn  libfreetype62.4.8-1
pn  thunar-volman   0.6.0-4+b1
pn  tumbler 0.1.23-1
pn  xdg-user-dirs   0.14-1
pn  xfce4-panel none

Versions of packages thunar suggests:
ii  thunar-archive-plugin 0.3.0-4
ii  thunar-media-tags-plugin  0.2.0-1

-- no debconf information



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#665700: [Pkg-xfce-devel] Bug#665700: thunar scrolls erratically when hover-selecting files

[Ulrich Eckhardt]

On Sunday 25 March 2012 14:05:28 Yves-Alexis Perez wrote:
 forwarded 665700 https://bugzilla.xfce.org/show_bug.cgi?id=8597
 thanks
 
 On dim., 2012-03-25 at 11:40 +0200, Ulrich Eckhardt wrote:
  Package: thunar
  Version: 1.2.3-3
  
  There are two observations which are related, I guess. I'm using the
  detailed list view (Control-2). The columns displayed are Name, Date
  Modified, Size. I have checked Automatically expand columns as
  needed in the configuration.
  
  
  Case I: Here, the size of the Name column is less than the width of the
  window, but the other two columns exceed the width (just wondering,
  doesn't that already contradict the settings?),
 
 Exactly the opposite, if you read the help text. That means the columns
 are expanded so the text is visible. So yes, that might means
 overflowing the window.

Yes, that's the fault in my expectations. I was expecting it to automatically 
resize the columns but within the limits set by the window. As a programmer, I 
guess that this would be very hard to program in a way that it works correctly 
under all corner cases.


  Further notes:
   - I'm not sure if the auto-expand code behaves correctly when there
   simply isn't enough space.
 
 Why?

Wrong expectations, see above.

   - I'm not sure if the auto-expand code works at all. I have seen it
   change layout sometimes, but I couldn't find a rule to it.
 
 That's pretty not clear at all. Care to explain?

Wrong expectations, I was assuming that Thunar auto-adjusted when resizing the 
window, but as you said it only adjusts to the required size of the columns.


Thanks for the quick feedback!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#663329: has_capture_channel() returns wrong value

[Ulrich Eckhardt]

Package: libasound2
Version: 1.0.25-2

snd_mixer_selem_has_capture_channel() should return 0 or 1, depending on 
whether the element has a capture channel. However, it also regularly returns 
-22 (-EINVAL) here for mixers that are output-only. This not only differs from 
the documentation, but also from the complementary 
snd_mixer_selem_has_playback_channel(), which never does this, even for input-
only channels.

Looking at the sources (slightly reformatted to fit into a mail), the problem 
seems obvious:

int
snd_mixer_selem_has_capture_channel(
snd_mixer_elem_t *elem,
snd_mixer_selem_channel_id_t channel)
{
  CHECK_BASIC(elem);
  CHECK_DIR(elem, SM_CAP_CVOLUME|SM_CAP_CSWITCH);
  return sm_selem_ops(elem)-is(
elem, SM_CAPT, SM_OPS_IS_CHANNEL, channel);
}

int
snd_mixer_selem_has_playback_channel(
snd_mixer_elem_t *elem,
snd_mixer_selem_channel_id_t channel)
{
  CHECK_BASIC(elem);
  return sm_selem_ops(elem)-is(
elem, SM_PLAY, SM_OPS_IS_CHANNEL, (int)channel);
}

The difference is that has_capture_channel() has an additional CHECK_DIR()
which contains a return -EINVAL when the element doesn't have either the 
CVOLUME or the CSWITCH capability. My gut feeling is that the CHECK_DIR() 
needs to be removed, but I haven't tried that and leave the to people that 
fully understand the API.


Cheers!

Uli




-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#660461: persistent app settings applied to wrong window

[Ulrich Eckhardt]

Package: fluxbox
Version: 1.3.2-2

Following steps to reproduce the problem:
1. install fspanel (0.7-10) and alsamixergui (0.9.0rc2-1-9.1)
2. start alsamixergui
3. right-click on its titlebar - remember - position
4. start fspanel

Initially, fspanel is now positioned on top of the ALSA mixer. The problem is 
also visible in the apps file after making the position persistent:

[app] (name=) (class=)
  [Position](UPPERLEFT) {1438 17}
[end]

Of course the actual bug is in both fspanel and alsamixergui, i.e. that they 
both don't set the mandatory WM_CLASS[1]. However, here is a point where the 
behaviour of fluxbox could be improved. In the presence of a broken client 
like those two, it should simply refuse to store any resource information, 
because the stored information would match every such client.

Cheers!

Uli


[1] http://tronche.com/gui/x/icccm/sec-4.html says This property must be 
present when the window leaves the Withdrawn state



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#660333: failure to switch desktops and hidden by other windows

[Ulrich Eckhardt]

Here's a patch against the Debian sources. The modifications go beyond just 
this bug, I discovered a few others on the way. Changes:

1. Set WM_CLASS property. See #660461 for a reason why.
2. Fix some indention mistakes and typos.
3. Handle sticky windows (desktop=-1) as visible everywhere
4. EWMH compliance:
 a) test _NET_SUPPORTED to detect compliant WM
 b) use _NET_NUMBER_OF_DESKTOPS and _NET_CURRENT_DESKTOP for switching
 c) use _NET_WM_DESKTOP to locate client windows
 d) prefer _NET_WM_NAME, fallback to WM_NAME for client name
 e) use _NET_WM_STATE to position fspanel above all other windows
 f) use _NET_WM_WINDOW_TYPE to mark fspanel as dektop part
 g) use _NET_CLIENT_LIST instead of _WIN_CLIENT_LIST


Note that I neither use fspanel regularly nor have I tested those changes 
extensively.


BTW: I found out that when clicking on a task in fspanel, the window is not 
raised to the top. Only if the window was previously minimized/iconified it is 
restored and raised. This might be a fluxbox quirk or configuration, but if 
you can reproduce this with any other window managers I would file a bug about 
it, so it isn't forgotten. As a workaround, clicking three times helps, the 
first focuses, the second minimizes and the third restores and raises the 
window.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#660333: Patch...

[Ulrich Eckhardt]

--- fspanel-0.7/fspanel.c	2012-02-19 13:32:54.0 +0100
+++ fspanel.c	2012-02-19 13:59:12.0 +0100
@@ -38,6 +38,14 @@
 
 Display *dd;
 Window root_win;
+/* WM supports EWMH
+This flag is set if the window manager supports the EWMH protocol for e.g.
+switching workspaces. The fallback if this is not supported is to use the
+Gnome variant. This is determined by looking for the presence of the
+_NET_SUPPORTED property of the root window. Note that this is only used
+for communication with the WM, whether each client supports this protocol
+is up to the individual client. */
+int wm_use_ewmh;
 Pixmap generic_icon;
 Pixmap generic_mask;
 GC fore_gc;
@@ -72,7 +80,17 @@ char *atom_names[] = {
 	_WIN_CLIENT_LIST,
 	_WIN_WORKSPACE_COUNT,
 	_WIN_STATE,
-	WM_STATE
+	WM_STATE,
+	_NET_NUMBER_OF_DESKTOPS,
+	_NET_CURRENT_DESKTOP,
+	_NET_WM_STATE,
+	_NET_WM_STATE_ABOVE,
+	_NET_SUPPORTED,
+	_NET_WM_WINDOW_TYPE,
+	_NET_WM_WINDOW_TYPE_DESKTOP,
+	_NET_WM_DESKTOP,
+	_NET_WM_NAME,
+	UTF8_STRING,
 };
 
 #define ATOM_COUNT (sizeof (atom_names) / sizeof (atom_names[0]))
@@ -89,6 +107,16 @@ Atom atoms[ATOM_COUNT];
 #define atom__WIN_WORKSPACE_COUNT atoms[7]
 #define atom__WIN_STATE atoms[8]
 #define atom_WM_STATE atoms[9]
+#define atom__NET_NUMBER_OF_DESKTOPS atoms[10]
+#define atom__NET_CURRENT_DESKTOP atoms[11]
+#define atom__NET_WM_STATE atoms[12]
+#define atom__NET_WM_STATE_ABOVE atoms[13]
+#define atom__NET_SUPPORTED atoms[14]
+#define atom__NET_WM_WINDOW_TYPE atoms[15]
+#define atom__NET_WM_WINDOW_TYPE_DESKTOP atoms[16]
+#define atom__NET_WM_DESKTOP atoms[17]
+#define atom__NET_WM_NAME atoms[18]
+#define atom_UTF8_STRING atoms[19]
 
 
 void *
@@ -132,8 +160,8 @@ fill_rect (taskbar *tb, int x, int y, in
 void
 scale_icon (task *tk)
 {
-int xx, yy, x, y;
-unsigned int w, h, d, bw;
+	int xx, yy, x, y;
+	unsigned int w, h, d, bw;
 	Pixmap pix, mk = None;
 	XGCValues gcv;
 	GC mgc = None;
@@ -223,13 +251,40 @@ get_task_kdeicon (task *tk)
 	}
 }
 
+/* returns whether the window is visible on the desktop */
 int
-find_desktop (Window win)
+is_visible_on_desktop (Window win, int desk)
+{
+	int client_desk = -1;
+	unsigned long *data;
+
+	if (wm_use_ewmh)
+		data = get_prop_data (win, atom__NET_WM_DESKTOP, XA_CARDINAL, 0);
+	else
+		data = get_prop_data (win, atom__WIN_WORKSPACE, XA_CARDINAL, 0);
+
+	if (data)
+	{
+		client_desk = *data;
+		XFree (data);
+	}
+
+	/* If the client_desk is -1, it is visible on all desktops */
+	return (client_desk == -1) || (client_desk == desk);
+}
+
+/* index of the currently displayed desktop */
+int
+get_current_desktop ()
 {
 	int desk = 0;
 	unsigned long *data;
 
-	data = get_prop_data (win, atom__WIN_WORKSPACE, XA_CARDINAL, 0);
+	if (wm_use_ewmh)
+		data = get_prop_data (root_win, atom__NET_CURRENT_DESKTOP, XA_CARDINAL, 0);
+	else
+		data = get_prop_data (root_win, atom__WIN_WORKSPACE, XA_CARDINAL, 0);
+
 	if (data)
 	{
 		desk = *data;
@@ -271,19 +326,36 @@ is_iconified (Window win)
 	return ret;
 }
 
+/* window name
+The returned pointer must be freed using XFree() after use.
+
+TODO: The encoding for WM_NAME can be STRING or COMPOUND_TEXT. In any case
+this encoding should be normalized before returning from this function. */
+char*
+get_window_name (Window win)
+{
+	char* res = NULL;
+	/* try EWMH's _NET_WM_NAME first */
+	res = get_prop_data (win, atom__NET_WM_NAME, atom_UTF8_STRING, 0);
+	if (!res)
+		/* fallback to WM_NAME */
+		res = get_prop_data (win, XA_WM_NAME, XA_STRING, 0);
+	return res;
+}
+
 void
 add_task (taskbar * tb, Window win, int focus)
 {
 	task *tk, *list;
 
-	/* is this window on a different desktop? */
-	if (tb-my_desktop != find_desktop (win) || is_hidden (win))
+	/* is this window on a different desktop or hidden? */
+	if (!is_visible_on_desktop (win, tb-my_desktop) || is_hidden (win))
 		return;
 
 	tk = calloc (1, sizeof (task));
 	tk-win = win;
 	tk-focused = focus;
-	tk-name = get_prop_data (win, XA_WM_NAME, XA_STRING, 0);
+	tk-name = get_window_name (win);
 	tk-iconified = is_iconified (win);
 
 	get_task_kdeicon (tk);
@@ -331,6 +403,7 @@ gui_create_taskbar (void)
 {
 	taskbar *tb;
 	Window win;
+	XClassHint wm_class;
 	MWMHints mwm;
 	XSizeHints size_hints;
 	XWMHints wmhints;
@@ -353,11 +426,30 @@ gui_create_taskbar (void)
   /*value mask*/ CWBackPixel | CWEventMask,
   /* attribs */ att);
 
-	/* don't let any windows cover fspanel */
-	set_prop (win, atom__WIN_LAYER, 10);	/* WIN_LAYER_ABOVE_DOCK */
+	/* set name and class */
+	wm_class.res_name = fspanel;
+	wm_class.res_class = fspanel;
+	XSetClassHint (dd, win, wm_class);
 
-	set_prop (win, atom__WIN_STATE, WIN_STATE_STICKY |
- WIN_STATE_FIXED_POSITION);
+	/* don't let any windows cover fspanel */
+	if (wm_use_ewmh)
+	{
+		XChangeProperty (dd, win,
+		 atom__NET_WM_STATE, XA_ATOM, 32,
+		 PropModeReplace,
+		 (unsigned char *) atom__NET_WM_STATE_ABOVE, 1);
+
+		XChangeProperty (dd, win,
+		 

Bug#660333: failure to switch desktops and hidden by other windows

[Ulrich Eckhardt]

Package: fspanel
Version: 0.7-10

When using a contemporary fluxbox (1.3.2) as window manager, fspanel fails in 
at least two ways:
1. It gets hidden by other windows, although it should be on top of everything 
but fullscreen apps.
2. The two up/down buttons to switch desktops don't work.

Note that this still worked with fluxbox 1.1. However, I'm not filing this 
against fluxbox, because they simply Removed (obsolete) Gnome window manager 
hints (quoting their changelog). fspanel should instead use EWMH[1] to set 
itself on top and to switch desktops. Also, one by one, other WMs will also 
drop support for this, so this isn't only a problem with fluxbox.

I'm currently facing similar bugs in a different program (cobweb panel, not 
yet packaged in Debian). I've already fixed issue 1 above, maybe I'll backport 
the patch from XCB to Xlib, unless someone beats me to it.

Cheers!

Uli


[1] http://standards.freedesktop.org/wm-spec/wm-spec-latest.html



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#636632: Still present...

[Ulrich Eckhardt]

...in the new 1.3.2-2, verified on one PowerPC system and another x86 system. 
I'm wondering, is there a problem with either the description of the problem 
or the patch?

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#405702: sdl-net1.2 bug #405702 fixed long ago

[Ulrich Eckhardt]

On Thursday 29 December 2011 01:20:43 you wrote:
 Sorry for the big delay, there are new efforts to try to maintain SDL
 packages in better shape.
 
 This bug seems to have been fixed long ago (Thu Apr 20 16:48:25 2006
 +, about 9 months before this Debian bug report was submitted), in
 this commit upstream:
 http://hg.libsdl.org/SDL_net/rev/8551dd6a2235

There is one thing that strikes me though, and that is that I only consider 
the fix above as a workaround, not as a fix. There are problems with it:
 1. the ignored signal is ignored process-wide
This means that other code which relies on the signal being raised will fail 
when libSDL switches the process to ignoring it.
 2. the signal is not ignored if a handler is in place
If other areas actually registered a handler, the signal will not be ignored, 
the code in SDL checks this. This could mean that SDL then again starts to 
fail (I don't think so though).
 3. the wrong handler is restored if the signal was ignored
If the calling code ignored the signal process-wide, SDL will not ignore it 
again, but when uninitialized it will again switch it to the default, while 
the remaining code might still rely on it being ignored.

What I don't get is why send() isn't simply called with MSG_NOSIGNAL as I 
suggested. I switched to dropping SDL_net and creating my own network wrapper 
doing just that and I never had problems again. I can't provide proper tests 
though, since I have abandoned that code altogether.


 Anyway, this bug report was fixed long ago althought it was never
 properly explained, documented or closed... so doing that right now :)

Thank you for this work, Manuel! This bug was nagging me everytime I looked at 
the list of bugs I filed, because I never understood how it could linger for 
so long without more people complaining. The fact that a functioning 
workaround is in place explains it. Note that I don't object to closing this 
bug, as obviously the fix is good enough, but I wanted to raise the concerns 
so they are at least documented.

Cheers! :)

Uli




-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#651207: whitespace handling in commandline parsing

[Ulrich Eckhardt]

Package: aptsh
Version: 0.0.7+nmu2

Entering remove foo tries to remove foo. Entering  remove foo complains 
about not being able to find e,   remove foo complains about not being 
able to finde ve. It seems like after detecting the remove token, it 
simply skips ahead six chars from the beginning before looking for the next 
token.

Cheers and thanks for this tool!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#631127: very strange KMail error

[Ulrich Eckhardt]

Hi!

I guess the very strange KMail error mentioned in [1] is actually
described in [2], right? In any case, I can also confirm that in
testing, current kmail (4:4.4.11.1+l10n-1) with libsasl
( 2.1.24~rc1.dfsg1+cvs2011) is still broken. I've tried version
2.1.25 from [3], too, but it doesn't fix those two bugs either.

What I'm still puzzled by is the issue with the domain in /etc/resolv.
What exactly do I have to do to get this to work again? Also, is that
a workaround for the libsasl2 bug or is it a bug in my /etc/resolf
that just happened to work with older libsasl versions?


Cheers!

Uli


[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=636534
[2] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=631127
[3] 
ftp.de.debian.org:/debian/pool/main/c/cyrus-sasl2/libsasl2-2_2.1.25.dfsg1-2_powerpc.deb
 



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#639837: UI hangs for 30s while loading pages

[Ulrich Eckhardt]

On Tuesday 30 August 2011 22:32:31 Mike Hommey wrote:
 How big is the ~/.mozilla/firefox/*/sessionstore.js file ?

 
uli@fruitbox:~$ ll ~/.mozilla/firefox/*/sessionstore.js
-rw--- 1 uli uli 305k Aug 30 21:43 
/home/uli/.mozilla/firefox/xzqx6a0f.default/sessionstore.js


305k on one line .. is that much/little/normal?

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#639837: UI hangs for 30s while loading pages

[Ulrich Eckhardt]

On Wednesday 31 August 2011 08:18:02 Mike Hommey wrote:
 Another thing I can think of that could explain your symptoms is if you
 have ipv6 enabled, but ipv6 doesn't actually work. I'd suggest you try
 to disable ipv6 (about:config, network.dns.disableIPv6) and see how it
 goes from there.

Tried that, no change. Maybe a tiny bit less often, but still regularly 
present.

In any case, it would not explain why the UI freezes (I'd hope that Iceweasel 
doesn't do blocking networking requests in the UI loop) or the presence of 
mentioned error messages about FD abuse in the logfile.

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#639837: UI hangs for 30s while loading pages

[Ulrich Eckhardt]

Package: iceweasel
Version: 5.0-6
Severity: important

Iceweasel regularly hangs while loading pages. Many pages show the
symptoms, though never regularly. I haven't determined if some do and 
some don't. It could also be that some advertisments that are being 
randomly served influence the issue.

Following symptoms:
1. Complete UI freeze
No switching to other tabs, aborting loading of the website, not even 
redrawing the window.

2. There is always some pending IO
The lower-left status popup (where e.g. the URL you're hovering over is 
displayed) shows the state of the loading operation. While the UI hangs, 
this is typically in the state Waiting for www.example.com

3. In the output (~/.xsession-errors), I get these four lines:

 ** (firefox-bin:29217): WARNING **: Serious fd usage error 14

 ** (firefox-bin:29217): WARNING **: Serious fd usage error 10
Yes, two empty ones and two with these warnings! These always (or usually) 
come in such pairs. I have seen errors 10, 13, 14 (according to errno 
those are ECHILD, EACCES, EFAULT) there.

4. Neither high CPU (htop) nor IO (iostat) load occurs.

5. After 30s, loading and UI operation resumes, without any faults. The 
interesting thing here is that these 30s are reproducible, not randomly. 
I watched the clock after a few such hangs to find that this is systematic 
indeed.


Good luck!

Uli


-- Package-specific info:

-- Extensions information
Name: Default themefalse
Location: /usr/lib/iceweasel/extensions/{972ce4c6-7e08-4474-a285-3208198ce6fd}
Package: iceweasel
Status: enabled

Name: DownloadHelper extensionfalse
Location: ${PROFILE_EXTENSIONS}/{b9db16a4-6edc-47ec-a1f4-b86292ed211d}
Status: enabled

-- Plugins information
Name: Shockwave Flash
Location: /usr/lib/mozilla/plugins/libflash-mozplugin.so
Package: libflash-mozplugin
Status: enabled


-- Addons package information
ii  iceweasel  5.0-6  Web browser based on Firefox
ii  libflash-mozpl 0.4.13-9   GPL Flash (SWF) Library - Mozilla-compatible

-- System Information:
Debian Release: wheezy/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (1, 'experimental')
Architecture: powerpc (ppc)

Kernel: Linux 2.6.39-rc7-powerpc
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages iceweasel depends on:
ii  debianutils   4.0.2  Miscellaneous utilities specific t
ii  fontconfig2.8.0-3generic font configuration library
ii  libc6 2.13-16Embedded GNU C Library: Shared lib
ii  libgcc1   1:4.6.1-4  GCC support library
ii  libgdk-pixbuf2.0-02.23.5-3   GDK Pixbuf library
ii  libglib2.0-0  2.28.6-1   The GLib library of C routines
ii  libgtk2.0-0   2.24.4-3   The GTK+ graphical user interface 
ii  libnspr4-0d   4.8.9-1NetScape Portable Runtime Library
ii  libstdc++64.6.1-4GNU Standard C++ Library v3
ii  procps1:3.2.8-11 /proc file system utilities
ii  xulrunner-5.0 5.0-6  XUL + XPCOM application runner

iceweasel recommends no packages.

Versions of packages iceweasel suggests:
ii  libgssapi-krb5-21.9.1+dfsg-1 MIT Kerberos runtime libraries - k
pn  mozplugger  none   (no description available)
ii  ttf-lyx 2.0.0-1  TrueType versions of some TeX font
pn  ttf-mathematica4.1  none   (no description available)
pn  xfonts-mathml   none   (no description available)
pn  xprint  none   (no description available)

Versions of packages xulrunner-5.0 depends on:
ii  libasound2  1.0.24.1-2   shared library for ALSA applicatio
ii  libatk1.0-0 2.0.1-2  ATK accessibility toolkit
ii  libbz2-1.0  1.0.5-6  high-quality block-sorting file co
ii  libc6   2.13-16  Embedded GNU C Library: Shared lib
ii  libcairo2   1.10.2-6.1   The Cairo 2D vector graphics libra
ii  libdbus-1-3 1.4.14-1 simple interprocess messaging syst
ii  libevent-1.4-2  1.4.13-stable-1  An asynchronous event notification
ii  libfontconfig1  2.8.0-3  generic font configuration library
ii  libfreetype62.4.6-1  FreeType 2 font engine, shared lib
ii  libgcc1 1:4.6.1-4GCC support library
ii  libgdk-pixbuf2.0-0  2.23.5-3 GDK Pixbuf library
ii  libglib2.0-02.28.6-1 The GLib library of C routines
ii  libgtk2.0-0 2.24.4-3 The GTK+ graphical user interface 
ii  libhunspell-1.2-0   1.2.14-4 spell checker and morphological an
ii  libjpeg62   6b1-2Independent JPEG Group's JPEG runt
ii  libmozjs5d  5.0-6Mozilla SpiderMonkey JavaScript li
ii  libnspr4-0d

Bug#636632: fbrun history is broken

[Ulrich Eckhardt]

Package: fluxbox
Version: 1.3.1~dfsg1-2

The history management of fbrun is broken. The most obvious symptom is that 
after starting something with it, the next time fbrun is started the cursor up 
button doesn't reproduce that command. While one could argue that this is 
still subjective, there is another thing that isn't.

The part that is more of a proof is the smashed content of the history file. 
For that try the following:
0. remove ~/.fluxbox/fbrun_history (or back it up)
1. run fbrun and enter xterm as command
2. run fbrun and enter sed as command
3. run fbrun and enter grep as command
4. run fbrun and enter xterm as command
5. run fbrun and enter xterm as command

Don't use the history for entering xterm the second and third time, type them 
manually! The contents of fbrun_history on my machine are now:

| sexterm
|
| xterm

I've been looking around the sources a bit, and I'll try to provide a patch. I 
have already found one error and one thing that at least makes no sense to me.


Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#636632: Patch...

[Ulrich Eckhardt]

There are two changes in this patch:
1. Loading the history, the code used 

  while (!infile.eof()) {
  getline(infile, ...);
  ...
  }

That's simply wrong, because it will read one superfluous empty line at the 
end. You need to check if reading succeeded instead of checking if the stream 
already encountered EOF.

2. Writing back the history, the code tried to skip elements until it found 
the current one, then write the rest and then finally the current command. 
This avoids writing the beginning of the file, but firstly it doesn't work, 
and secondly it is IMHO a completely unnecessary optimization because a) it 
isn't called often enough and b) the file is 1kB here after years of use. I'm 
using the most simple implementation and just write the whole history every 
time, skipping all duplicates and finally append the current command.


There was one thing I'm not sure how the code was supposed to work and which 
wasn't documented either. The question is how editing of previously entered 
commands is reflected in the history. There are two approaches:
1. If I enter xter (if I pressed enter too early), going back in history and 
then appending the m, it replaces the xter with xterm in history.
2. If I enter xter, going back in history and appending the m yields two 
entries xter and xterm.

I chose approach 1 in the patch. However, that means that if a user wants to 
have two entries like xterm sh and xterm bash, they can't achieve that by 
just editing the history, they must enter them completely separately instead. 
If you don't like that, just skip the code leading to the second continue in 
the look that writes the history. ;)

Cheers!

Uli
Index: FbRun.cc
===
--- FbRun.cc	(revision 34)
+++ FbRun.cc	(revision 39)
@@ -137,43 +137,22 @@
 hide(); // hide gui
 
 // save command history to file
-if (text().size() != 0) { // no need to save empty command
+if (!command.empty()) { // no need to save empty command
 
-// don't allow duplicates into the history file, first
-// look for a duplicate
-if (m_current_history_item  m_history.size()
- text() == m_history[m_current_history_item]) {
-// m_current_history_item is the duplicate
-} else {
-m_current_history_item = 0;
-for (; m_current_history_item  m_history.size();
- ++m_current_history_item) {
-if (m_history[m_current_history_item] == text())
-break;
-}
-}
-
-fstream inoutfile(m_history_file.c_str(), ios::in|ios::out);
-if (inoutfile) {
-// now m_current_history_item points at the duplicate, or
-// at m_history.size() if no duplicate
-if (m_current_history_item != m_history.size()) {
-unsigned int i = 0;
-// read past history items before current
-for (; inoutfile.good()  i  m_current_history_item; i++)
-inoutfile.ignore(1, '\n');
-
-// write the history items that come after current
-for (i++; i  m_history.size(); i++)
-inoutfilem_history[i]endl;
-
-} else {
-// set put-pointer at end of file
-inoutfile.seekp(0, ios::end);
+ofstream outfile(m_history_file.c_str());
+if (outfile) {
+for (unsigned i = 0; i != m_history.size(); ++i) {
+// don't allow duplicates into the history file
+if (m_history[i] == command)
+continue;
+// skip currently selected item, this will be appended but is
+// not caught as duplicate when it was modified by the user
+if (i == m_current_history_item)
+continue;
+outfilem_history[i]endl;
 }
 // append current command to the file
-inoutfilecommandendl;
-
+outfilecommandendl;
 } else
 cerrFbRun Warning: Can't write command history to file: m_history_fileendl;
 }
@@ -197,8 +176,7 @@
 m_history.clear();
 // each line is a command
 string line;
-while (!infile.eof()) {
-getline(infile, line);
+while (getline(infile, line)) {
 if (line.size()) // don't add empty lines
 m_history.push_back(line);
 }

Bug#598456: Solved...

[Ulrich Eckhardt]

Hi!

It seems there is a setting that tells fluxbox to ignore or not ignore the 
tabs when maximizing, it is under the Tab Options and called Maximize 
Over. With that deactivated, windows are now maximized as expected.

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#633157: Using wrong configuration file leads to data loss

[Ulrich Eckhardt]

Package: fluxbox
Version: 1.3.1~dfsg1-2

I can tell fluxbox to use a different configuration file:

  $ fluxbox -help
  [...]
  -rc stringuse alternate resource file.
  [...]

This works to the extent that if I specify a different path, it will actually 
use it. However, and that is inacceptable, inside the file which is 
autogenerated if it doesn't exist, there are still references to my other 
configuration:

  session.styleOverlay:   ~/.fluxbox/overlay
  session.keyFile:/home/uli/.fluxbox/keys
  session.appsFile:   ~/.fluxbox/apps
  session.slitlistFile:   /home/uli/.fluxbox/slitlist
  session.styleFile:  /home/uli/.fluxbox/styles/Sn33z-mod

If I save anything in e.g. the apps file, which happens during normal WM 
interaction, the configuration there is lost, although I specified a different 
path exactly in order to avoid overwriting it.


Some more notes:
1. It looks weird that some paths above are absolute, while some are relative 
to the homedir.
2. Fluxbox doesn't have a configuration file, but a configuration dir. The 
commandline should honour that, i.e. not expect just a single init file but a 
directory.
3. If I specify a path to a file that doesn't exist, I'd expect to either get 
an error or that the file is created. This works to some extent, but if I 
specify a file where the parent dir doesn't exist, I get errors on stdout 
(most users won't see that) and fluxbox uses the default configuration from my 
home dir, which again can lead to data loss.


Cheers!

Uli




-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#632688: ConfigureWindow not handled correctly

[Ulrich Eckhardt]

Package: fluxbox
Version: 1.3.1~dfsg1-2

The attached file is an excerpt from a program. It creates a single window, 
undecorated, on top of other windows and in the upper left corner of the 
screen. At least it should do that, and it used to. At the moment, it only 
achieves parts of this, the resulting position is wrong.

The program uses XCB, but I have tried an earlier version of the program that 
uses Xlib, and it suddenly has the same problems, so that is not the cause for 
problems.

I have tried openbox, sawfish and icewm, and all three the WMs work as 
expected. I have also tried fluxbox 1.1.1+dfsg2-1 and it shows the same wrong 
behaviour as the current version. If that had always been the case, I'm pretty 
sure that I would have noticed before, because I've been running fluxbox for 
ages. That again makes me wonder if this is really just a problem with 
fluxbox, at least I suspect other influences, like possibly in the X server.


Cheers!

Uli
// g++ -g -O0 -W test.cpp $(pkg-config xcb xcb-icccm xcb-shape xcb-image --libs --cflags)

#include xcb/xcb.h
#include xcb/xcb_icccm.h
#include stdint.h
#include stdexcept
#include iostream
#include string.h
#include stdlib.h
#include iomanip

xcb_connection_t* connection = 0;

xcb_atom_t
resolve_atom_name(char const* name)
{
size_t const len = strlen(name);
if(static_castuint16_t(len) != len)
throw std::runtime_error(xatom: name too long);
xcb_intern_atom_cookie_t cookie = xcb_intern_atom(connection,
  0, // only_if_exists
  static_castuint16_t(len),
  name);
xcb_generic_error_t* error = 0;
xcb_intern_atom_reply_t* reply = xcb_intern_atom_reply(connection, cookie, error);
if(error)
throw std::runtime_error(xcb_intern_atom_reply failed with error);
if(!reply)
throw std::runtime_error(xcb_intern_atom_reply failed without reply);
xcb_atom_t const res = reply-atom;
free(reply);
return res;
}

int
main()
{
try
{
/* open connection with the server */
connection = xcb_connect(NULL,NULL);
if (xcb_connection_has_error(connection))
throw std::runtime_error(xcb_connect() failed);

/* retrieve the setup */
xcb_setup_t const* setup = xcb_get_setup(connection);
if(!setup)
throw std::runtime_error(xcb_get_setup() failed);

/* get the first screen */
xcb_screen_t* screen = xcb_setup_roots_iterator(setup).data;
if(!screen)
throw std::runtime_error(missing root screen);

/* create window */
uint32_t const win_values[] =
{
screen-white_pixel,
( XCB_EVENT_MASK_EXPOSURE
 |XCB_EVENT_MASK_BUTTON_PRESS
 |XCB_EVENT_MASK_BUTTON_RELEASE)
};
xcb_window_t window_id = xcb_generate_id(connection);
xcb_create_window(connection,
  XCB_COPY_FROM_PARENT, // depth
  window_id,
  screen-root, // parent
  0, 0, // position
  100, 100, // size
  0, // border width
  XCB_WINDOW_CLASS_INPUT_OUTPUT, // class
  XCB_COPY_FROM_PARENT, // visual
  XCB_CW_BACK_PIXEL | XCB_CW_EVENT_MASK, win_values);

// resolve properties, (IDs, types and values)
xcb_atom_t const wm_name_atom = resolve_atom_name(_NET_WM_NAME);
xcb_atom_t const wm_window_type_atom = resolve_atom_name(_NET_WM_WINDOW_TYPE);
xcb_atom_t const win_layer_atom = resolve_atom_name(_WIN_LAYER);
xcb_atom_t const utf8_string_atom = resolve_atom_name(UTF8_STRING);
xcb_atom_t const cardinal_atom = resolve_atom_name(CARDINAL);
xcb_atom_t const atom_atom = resolve_atom_name(ATOM);
xcb_atom_t const wm_window_type_desktop_atom = resolve_atom_name(_NET_WM_WINDOW_TYPE_DESKTOP);

// set window name
xcb_change_property(connection,
XCB_PROP_MODE_REPLACE,
window_id,
wm_name_atom,
utf8_string_atom,
8,
strlen(test), test);

// set layer to 10 (above dock)
uint32_t layer = 10;
xcb_change_property(connection,
XCB_PROP_MODE_REPLACE,
window_id,
win_layer_atom,
cardinal_atom,
32,
1, layer);

// set window type to desktop element to avoid decoration
xcb_change_property(connection,
XCB_PROP_MODE_REPLACE,
window_id,
  

Bug#630148: Failure to clean up resources in temporary dir

[Ulrich Eckhardt]

Package: konqueror
Version: 4:4.4.5-3

Note up front: This is probably not only a problem with konqueror but also 
with other KDE parts, but it hit me most with konqueror.

The problem is that konqueror caches files in /var/tmp/ from webbrowsing, 
which are never cleaned up. I recently removed a folder /var/tmp/kdecache-
uli/http which took up 1.5GiB(!) of space there. Note that rebooting or 
anything like that doesn't help, these files are never cleaned up AFAICT.

Two wishes here:
1. If applications cache things, they also need to clean up their cache.
2. If it is not the applications, the surrounding system itself should clean 
up regularly.

Thank you!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#630150: lnusertemp fails to honour TEMPDIR

[Ulrich Eckhardt]

Package: kdelibs4c2a
Version: 4:3.5.10.dfsg.1-5

TEMPDIR is the canonical Unix way to specify the location for temporary files. 
lnusertemp doesn't honour this environment variable and simply creates these 
files in /tmp or /var/tmp.

As a workaround, you can set KDETMP and KDEVARTMP, as per manpage.


Thanks!

Uli




-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#629581: Zombie windows after closing terminal

[Ulrich Eckhardt]

Package: fluxbox
Version: 1.3.1~dfsg1-2

Problem:

The simple procedure to reproduce the problem is to run

  xterm -e false

in a shell (xterm -e false works, too). This will create an xterm that is 
immediately closed again. Sometimes, this window will get stuck. Checking the 
output of ps ux|grep xterm, I see that the xterm process was already 
terminated, so the window is left without the according X client.


Workaround:

In order to remove that window, just repeat the command that actually caused 
it. In rare cases, you will end up with one more zombie, but now and then the 
zombies are removed.


Notes:
 - What makes this special is that the client only lives for a very short 
time. Due to the asynchronous nature of X, it is possible that the client 
terminates even before the window manager has fully mapped the window that was 
created. At least that's what I would look at, but I don't really know.
 - This might be related to[1], but there it mentions tab groups.


Cheers!

Uli

[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=543387



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#598456: Still present...

[Ulrich Eckhardt]

...using version 1.3.1~dfsg1-2.

Also, just wondering, was this forwarded upstream yet?

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#465301: Partially fixed...

[Ulrich Eckhardt]

Hi!

I'm running icewm 1.3.7-1 here and blast 1.1-22 here and windows don't close 
their holes again when they are moved. When they are resized, they will close 
the holes though.

That said, I'm not sure I would call this behaviour a bug though.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#628203: System load tooltip in panel displays wrong data

[Ulrich Eckhardt]

Package: icewm
Version: 1.3.7-1

In the panel/taskbar/system tray is a small applet displaying the CPU load. If 
you hover the mouse over it, you get a tooltip window displaying various 
information textually:

1. CPU Load: 0.58 0.97 1.04, 270
I can recognize the first three and know where to look up their meaning 
(Though I really doubt Joe average User can!), but the last value is a 
mystery to me.

2. Ram: 1007.03/80.86M
Firstly, it should probably be RAM, not Ram. Secondly, I guess it should be 
80.86/1007.03M instead, i.e. the free RAM. Or maybe it should be 
926.17/1007.03M, which is the used RAM just as the CPU and swap usage display. 
As optional feature, I would display the usage in either percent, like the CPU 
load and/or in MiB instead of MB, in order to avoid fractional values.

3. Swap: 1953.15/1868.25M
Again, I would swap these two values like with the RAM above or even make this 
a percentage.

4. ACPI Temp:
I'm on a PowerPC system, but ACPI is a x86-ism, AFAIK. Hence also no 
temperature here, but I'd rather say the whole entry should be removed when no 
temperature is available.

5. CPU Freq: 0.000GHz
This value is simply wrong, this is also the main issue that prompted me to 
write this bug report, before I actually really looked at the values there and 
started thinking. ;)


Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#628417: Missing menu entries to start application

[Ulrich Eckhardt]

Package: kmail
Version: 4:4.4.7-3

Hi!

Since quite some time (roughly since the transition of testing to KDE 4, I 
believe) there haven't been any menu entries for most KDE applications, like 
dolphin, konqueror, knode and kmail. Interestingly k3b and ktorrent have menu 
entries, but I believe those are not parts of the KDE core.

It would be great to have those menu entries back! ;)


Thanks!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#442908: Seems fixed....

[Ulrich Eckhardt]

I tried versions 1:3.6.1-4 (unstable) and 1:3.6.99svn11761-1 (experimental). 
Neither of these three bugs were reproducible with these two versions.

Good job, thank you! :)

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#257641: Update?

[Ulrich Eckhardt]

On Tuesday 10 May 2011 20:45:32 Pierre Habouzit wrote:
 I'm unable to reproduce. valgrind is actually a shell script that does
 what it has to wrt LD_LIBRARY_PATH, are you sure you can reproduce? what
 do you set in your LD_LIBRARY_PATH, please give steps to reproduce.

I'm not using LD_LIBRARY_PATH and it still segfaults, but I'm on a PowerPC 
system, see #603961. Maybe Martin is on a PowerPC system, too, and 
LD_LIBRARY_PATH is actually irrelevant?

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#626009: Should suggest evas-engine

[Ulrich Eckhardt]

Package: libevas1
Version: 1.0.0-1

Hi!

Firstly, thanks for packaging this, I'm thrilled to try out what the 
Enlightenment team has created. For that reason, I installed python-evas and 
started toying with it. However, none of the examples I tried worked, and that 
for the simple reason that I didn't have any rendering engines installed. I'd 
Suggests: libevas1-engines-core|libevas1-engines-x or something like that. 
Alternatively, I'd maybe add a Provides: libevas1-engine to these two 
packages and a Depends: libevas1-engine to libevas1, but I guess you will 
find the best way to assure that at least some engine can be loaded.

BTW: There is a bug in libevas1 that it reports all possible engines in 
evas_render_method_list(), even those that are not installed and that can't be 
loaded. Only when trying to load one it returns zero, but the Python wrapper 
doesn't interpret that as error. These two separate issues contributed to my 
problems getting things to run. Forcing me to install at least one engine 
would have helped me a lot!

Thank you!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#567793: your mail

[Ulrich Eckhardt]

On Saturday 26 February 2011 12:29:37 you wrote:
 is it going better in squeeze or sid?

I just switched to 24BPP and EXA, and the artifacts are still there. Switching 
to XAA fixes the issues again. The version of xserver-xorg-video-radeon is
1:6.13.1-2+squeeze1.


Uli





-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#610269: Could not write to file newsserver when sending posting

[Ulrich Eckhardt]

Package: knode
Version: 4:4.4.7-2

Since a few days, knode fails to send postings to at least some newsgroup. The
group in question that doesn't work is comp.lang.c++. A group that does work
is de.test. When sending, I get Could not write to file news.hansenet.de,
which is my newsserver!?

Further notes:
 - Around the time that happened first, I ran out of space on the root
   partition (Just wondering, why does KDE store stuff there even though I
   have set $TMPDIRCo to something else?). I did not get any disk full
   errors though and this is also no longer the case.
 - I have quit and restarted knode.
 - I have unsubscribed from that group and resubscribed again.
 - I subscribed to de.test just for testing.
 - I have not yet tried any other newsgroups.
 - I have started knode from the commandline, no errors visible there expect:
   Connecting to deprecated signal 
QDBusConnectionInterface::serviceOwnerChanged(QString,QString,QString)
   knode(16839)/kdecore (KConfigSkeleton) KCoreConfigSkeleton::writeConfig:


Any ideas how to fix this or what else to try?

Thanks!

Uli


-- System Information:
Debian Release: 6.0
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (1, 'experimental')
Architecture: powerpc (ppc)

Kernel: Linux 2.6.37-rc5-powerpc
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages knode depends on:
ii  kdebase-runtime   4:4.4.5-1  runtime components from the offici
ii  kdepim-runtime4:4.4.7-1  Runtime components for akonadi-kde
ii  kdepimlibs-kio-plugins4:4.4.5-2  kio slaves used by KDE PIM applica
ii  libc6 2.11.2-7   Embedded GNU C Library: Shared lib
ii  libgcc1   1:4.4.5-8  GCC support library
ii  libkabc4  4:4.4.5-2  library for handling address book 
ii  libkde3support4   4:4.4.5-2  the KDE 3 Support Library for the 
ii  libkdecore5   4:4.4.5-2  the KDE Platform Core Library
ii  libkdepim44:4.4.7-2  KDE PIM library
ii  libkdeui5 4:4.4.5-2  the KDE Platform User Interface Li
ii  libkhtml5 4:4.4.5-2  the KHTML Web Content Rendering En
ii  libkio5   4:4.4.5-2  the Network-enabled File Managemen
ii  libkmime4 4:4.4.5-2  library for handling MIME data
ii  libkontactinterface4  4:4.4.5-2  Kontact interface library
ii  libkparts44:4.4.5-2  the Framework for the KDE Platform
ii  libkpgp4  4:4.4.7-2  gpg based crypto library
ii  libkpimidentities44:4.4.5-2  library for managing user identiti
ii  libkpimtextedit4  4:4.4.5-2  library that provides a textedit w
ii  libkpimutils4 4:4.4.5-2  library for dealing with email add
ii  libkutils44:4.4.5-2  various utility classes for the KD
ii  libmailtransport4 4:4.4.5-2  mail transport service library
ii  libqt4-dbus   4:4.6.3-4  Qt 4 D-Bus module
ii  libqt4-qt3support 4:4.6.3-4  Qt 3 compatibility library for Qt 
ii  libqt4-xml4:4.6.3-4  Qt 4 XML module
ii  libqtcore44:4.6.3-4  Qt 4 core module
ii  libqtgui4 4:4.6.3-4  Qt 4 GUI module
ii  libstdc++64.4.5-8The GNU Standard C++ Library v3

knode recommends no packages.

knode suggests no packages.

-- no debconf information



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#608034: Useless error message on startup failure

[Ulrich Eckhardt]

Package: pavucontrol
Version: 0.9.9-1

When trying to start pavucontrol, it gave me the error Connection failed: 
connection refused. This is rather unhelpful, since the problem is that the 
pulseaudio package is not installed (shouldn't it depend on it?) and that it 
therefore can't start the pulseaudio daemon, I guess.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#541538: Sound volume jumps on play

[Ulrich Eckhardt]

On Saturday 25 December 2010 19:26:44 you wrote:
 It looks like the symptom that you describe is likely a sound driver
 issue.

Depending on what you call a driver, I agree. The point is that plain ALSA 
output without pulseaudio works, so that part of the driver stack is not the 
problem. However, adding pulseaudio as intermediate driver layer causes 
problems.

 Are you still able to reproduce it in current testing?

I'm not experiencing the jumps in the output level, though I haven't tested it 
extensively. The output sliders twitch when starting, but neither is the 
effect permanent nor audible. However, I also mentioned that the problem was 
not 100% reproducible even at that time (see point 5).

Considering point one, there's still a similar effect in place. If I run 
alsamixergui in parallel, changing the PCM volume there affects both the 
sliders for Amarok on the Playback tab and those on the Output Devices tab 
of pavucontrol. Using pavucontrol, I can change both pairs of sliders 
separately though. The behaviour of alsamixergui is to follow changes on both 
to some extent.

Considering point 6, this effect is still present. However, the absolute 
limite of 43% percent actually depends on the the settings of both the sliders 
on the Playback and Output Devices tabs of pavucontrol. In any case, the 
static begins when turning the PCM output to close to zero on alsamixergui or 
when it reaches this level through the combined pavucontrol slider levels.

That and the occasional cracks and plops made stop using pulseaudio, so I 
can't really say whether the main part of the bug report is still applicable 
with current versions.

 https://wiki.ubuntu.com/Audio/AlsaInfo

See attachment, hope that helps!


Cheers and thanks!

Uli
upload=truescript=truecardinfo=
!!
!!ALSA Information Script v 0.4.59
!!

!!Script ran on: Sun Dec 26 11:03:13 UTC 2010


!!Linux Distribution
!!--

Debian GNU/Linux squeeze/sid \n \l


!!DMI Information
!!---

Manufacturer:  
Product Name:  


!!Kernel Information
!!--

Kernel release:2.6.37-rc5-powerpc
Operating System:  GNU/Linux
Architecture:  ppc
Processor: unknown
SMP Enabled:   No


!!ALSA Version
!!

Driver version: 1.0.23
Library version:1.0.23
Utilities version:  1.0.23


!!Loaded ALSA modules
!!---

snd_aoa_fabric_layout


!!Sound Servers on this system
!!

Pulseaudio:
  Installed - Yes (/usr/bin/pulseaudio)
  Running - Yes

ESound Daemon:
  Installed - Yes (/usr/bin/esd)
  Running - No

aRts:
  Installed - Yes (/usr/bin/artsd)
  Running - No


!!Soundcards recognised by ALSA
!!-

 0 [SoundByLayout  ]: AppleOnbdAudio - SoundByLayout
  SoundByLayout


!!PCI Soundcards installed in the system
!!--



!!Advanced information - PCI Vendor/Device/Susbsystem ID's
!!



!!Modprobe options (Sound related)
!!

snd-atiixp-modem: index=-2
snd-intel8x0m: index=-2
snd-via82xx-modem: index=-2
snd-pcsp: index=-2
snd-usb-audio: index=-2
snd-powermac: index=0


!!Loaded sound module options
!!--

!!Module: snd_aoa_fabric_layout
* : 


!!ALSA Device nodes
!!-

crw-rw+ 1 root audio 116, 4 Dec 24 18:58 /dev/snd/controlC0
crw-rw+ 1 root audio 116, 6 Dec 24 18:58 /dev/snd/pcmC0D0c
crw-rw+ 1 root audio 116, 5 Dec 26 11:21 /dev/snd/pcmC0D0p
crw-rw+ 1 root audio 116, 3 Dec 24 18:58 /dev/snd/seq
crw-rw+ 1 root audio 116, 2 Dec 24 18:58 /dev/snd/timer

/dev/snd/by-path:
total 0
drwxr-xr-x 2 root root  60 Dec 24 18:58 .
drwxr-xr-x 3 root root 160 Dec 24 18:58 ..
lrwxrwxrwx 1 root root  12 Dec 24 18:58 pci-0001:10:17.0 - ../controlC0


!!Aplay/Arecord output
!!

APLAY

 List of PLAYBACK Hardware Devices 
card 0: SoundByLayout [SoundByLayout], device 0: Master []
  Subdevices: 0/1
  Subdevice #0: subdevice #0

ARECORD

 List of CAPTURE Hardware Devices 
card 0: SoundByLayout [SoundByLayout], device 0: Master []
  Subdevices: 1/1
  Subdevice #0: subdevice #0

!!Amixer output
!!-

!!---Mixer controls for card 0 [SoundByLayout]

Card hw:0 'SoundByLayout'/'SoundByLayout'
  Mixer name: 'SoundByLayout'
  Components: ''
  Controls  : 5
  Simple ctrls  : 5
Simple mixer control 'Headphone',0
  Capabilities: pswitch pswitch-joined penum
  Playback channels: Mono
  Mono: Playback [on]
Simple mixer control 'Headphone Detect Autoswitch',0
  Capabilities: pswitch pswitch-joined penum
  Playback channels: Mono
  Mono: Playback [on]
Simple mixer control 'Headphone Detected',0
  Capabilities: pswitch pswitch-joined penum
  Playback channels: Mono
  Mono: Playback [on]
Simple mixer control 'Speakers',0
  

Bug#544112: Fixed?

[Ulrich Eckhardt]

Hi!

I recently noticed that the problem seems to be gone.

I then downloaded and installed fluxbox_1.1.1+dfsg2-1_powerpc.deb, restarted 
fluxbox, and the problem reappeared, so that version is affected.

Installing fluxbox_1.1.1+git20100908.df2f51b-2_powerpc.deb again, the effect 
is gone again, so something in between must have fixed the issue. Note that I 
picked this package directly from the FTP server, APT doesn't seem to know 
about it.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#605988: Can't undo all changes without first selecting a specific one

[Ulrich Eckhardt]

Package: kompare
Version: 4:4.4.5-1

I'm looking at a kompare window opened by kdesvn when selecting diff local 
changes there. 'ps aux' says the commandline is kompare -on -. The button 
to revert all changes to the file (Apply All) is deactivated until I select 
a specific change in the file, it should always be activated if there is any 
change to revert.


Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#605998: Text View behaves weirdly

[Ulrich Eckhardt]

Package: kompare
Version: 4:4.4.5-1

Under the Settings menu, there is an option Show Text View. There are 
several things surprising with this text view:

1. The option is not saved anywhere, so if I select the text view, close 
kompare and reopen it, the text view is gone and I have to activate it again. 
I would have expected that kompare remembers things selected under Settings.

2. The position in the text view is the last difference, regardless of which 
position in the source/target views I selected previously. The position also 
doesn't move, neither the position in the source/target view when moving 
through the text view nor the one in the text view when moving through the 
source/target views.

3. I can edit text in the text view. I have looked around a bit, but I 
couldn't find any lasting effect of these edits. They neither show up in 
source/target view nor after saving the files.

4. I can not really edit text in the text view. For example, trying to type a 
space triggers a hotkey that is bound to Apply Difference instead.

5. I can't detach the text view from the main window. If I do, it looses all 
its content and decoration. Further, I get a stream of errors on stdout:

  X Error: BadMatch (invalid parameter attributes) 8
Major opcode: 62 (X_CopyArea)
Resource id:  0x58053fc

These come every second or so while the mouse is inside the detached window 
which I can still move and resize or reattach, which restores the content and 
stops the errors. I tried with dolphin, which uses similar detachable windows 
and there I don't have any such problems. I tried with both fluxbox and icewm 
in order to rule out the WM as the problem for this.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#606014: various errors printed to console

[Ulrich Eckhardt]

Package: kompare
Version: 4:4.4.5-1

When I just start kompare without any arguments, it outputs the following
two errors on the terminal window:

  kompare(2276)/kdecore (KLibrary) findLibraryInternal: plugins should not have 
a 'lib' prefix: libkomparepart.so
  kompare(2276)/kdecore (KLibrary) findLibraryInternal: plugins should not have 
a 'lib' prefix: libkomparenavtreepart.so

When running it with two files to compare, it gives me this:

  kompare(2466)/kdecore (KLibrary) findLibraryInternal: plugins should not have 
a 'lib' prefix: libkomparepart.so
  kompare(2466)/kdecore (KLibrary) findLibraryInternal: plugins should not have 
a 'lib' prefix: libkomparenavtreepart.so
  kompare(2466)/kdecore (services) KMimeTypeFactory::parseMagic: Now parsing  
/usr/share/mime/magic
  load glyph failed err=6 face=0x108a32e0, glyph=512
  QFontEngine: Glyph neither outline nor bitmap format=0

The last two lines are then repeated over and over while the program is
running. Just in case it matters, Unicode codepoint 512 (u200) is Latin
capital letter A with double grave. Since I'm comparing text files that
consist of a subset of ASCII, their content shouldn't be the reason for
these errors.

Greetings!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#606026: handle_event.c:565: handle_breakpoint: Assertion `sbp' failed.

[Ulrich Eckhardt]

Package: ltrace
Version: 0.5.3-2.1
Severity: important

Consider this three-liner:

  #include iostream
  int main()
  {}

Compiled with g++ x.cpp and run with ltrace ./a.out, I get the
following output:

  _start(1, 0xbfb13834, 0xbfb1383c, 0xbfb138d4, 0x4800f930 unfinished ...
  __libc_start_main(1, 0xbfb13834, 0xbfb1383c, 0xbfb138d4, 0x4800f930 
unfinished ...
  _ZNSt8ios_base4InitC1Ev(0x10010b50, 65535, 0xbfb1383c, 0xbfb138d4, 0ltrace: 
handle_event.c:565: handle_breakpoint: Assertion `sbp' failed.
  Aborted

If I remove the first line, no such crash happens. g++ is at version
4:4.4.5-1, libstdc++6 is version 4.4.5-8, just in case that matters.
I also tried downgrading to ltrace to 0.5-3.1 from stable, with that
version I don't get any crash either.


Cheers!

Uli


-- System Information:
Debian Release: squeeze/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (1, 'experimental')
Architecture: powerpc (ppc)

Kernel: Linux 2.6.33-2-powerpc
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages ltrace depends on:
ii  libc6 2.11.2-7   Embedded GNU C Library: Shared lib
ii  libelfg0  0.8.13-1   an ELF object file access library

ltrace recommends no packages.

ltrace suggests no packages.

-- no debconf information



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#603961: valgrind: segmentation fault on PowerPC

[Ulrich Eckhardt]

Package: valgrind
Version: 1:3.6.0~svn11254+nmu1
Severity: important

Hi!

Valgrind is broken for PowerPC, just starting it without any further
command to run already causes a segmentation fault. Running it on any
executable has the same problem.

There is a patch available, the simple recipe mentioned in
https://bugs.launchpad.net/ubuntu/+source/valgrind/+bug/484048 or
http://bugs.kde.org/show_bug.cgi?id=214223 works. Or maybe one should
say it works around the issue, which could be nonportable valgrind code
or a GCC bug.

Note: This is not related to http://bugs.debian.org/cgi-
bin/bugreport.cgi?bug=529016, which also reports breakage on PowerPC,
but the problem is a different one. With the change mentioned above,
valgrind is again usable on PowerPC.

Thanks!

Uli




-- System Information:
Debian Release: squeeze/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (1, 'experimental')
Architecture: powerpc (ppc)

Kernel: Linux 2.6.33-2-powerpc
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages valgrind depends on:
ii  libc6 2.11.2-7   Embedded GNU C Library: Shared lib
ii  libc6-dbg 2.11.2-7   Embedded GNU C Library: detached d

Versions of packages valgrind recommends:
ii  gdb   7.0.1-2+b1 The GNU Debugger

Versions of packages valgrind suggests:
pn  alleyoop  none (no description available)
pn  kcachegrind   none (no description available)
pn  valkyrie  none (no description available)




-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#257641: Update?

[Ulrich Eckhardt]

Greetings!

I'm facing the same problem, but:
1. valgrind's version is 1:3.6.0~svn11254+nmu1
2. libc6-dbg's version is 2.11.2-7
3. valgrind actually depends on libc6-dbg, so uninstalling is not a viable 
workaround.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#529016: Still broken...

[Ulrich Eckhardt]

FWIW, libc6-dbg's version is 2.11.2-7 here, valgrind's is 
1:3.6.0~svn11254+nmu1.

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#533468: Workaround

[Ulrich Eckhardt]

Since there obviously is no activity on Splix towards fixing this bug, I'd 
like to point out that the Foomatic driver now supports the ML-1610, so 
installing that instead of Splix gets you a working printer.

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#458018: This doesn't only happen to short files

[Ulrich Eckhardt]

Hi!

I'm just evaluating VLC as alternative to xine or mplayer, and I stumbled 
across the same bug. However, in my case the file wasn't called a.avi but 
داخلحماممغربي.flv, and picks up a bunch of files called *.txt or *.utf8.

Cheers!

Uli



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#483836: Can't reproduce this

[Ulrich Eckhardt]

Command line is

  vlc http://dump.com/wp-content/uploads/golf.flv

There's a bunch of spurious warnings and errors and the video output is broken 
(but it's like that for all videos, which is a different bug), but VLC works 
without crashing.

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#602708: Video output is X-compressed to 25% and colors are off

[Ulrich Eckhardt]

Package: vlc
Version: 1.1.3-1

Greetings!

Any video I try to play is reduced in size and only on the left quarter of the
video area, the rest of the video area is black. Further, the colors in the
video are off, you can still recognize the content though. Lastly, there are
horizontal stripes, somehow like interlacing gone wrong.

When I just run vlc, I get this:
| u...@fruitbox:~$ vlc
| VLC media player 1.1.3 The Luggage (revision exported)
| Blocked: call to unsetenv(DBUS_ACTIVATION_ADDRESS)
| Blocked: call to unsetenv(DBUS_ACTIVATION_BUS_TYPE)
| Warning: call to signal(13, 0x1)
| [0x10103918] main libvlc: Running vlc with the default interface. Use 'cvlc' 
to use vlc without interface.
| Blocked: call to setlocale(6, )
| Blocked: call to sigaction(17, 0x484dda48, 0x484ddad4)
| Warning: call to rand()
| Warning: call to rand()
| Warning: call to rand()
| Warning: call to rand()
| Warning: call to rand()

When I drag'n'drop a video onto the window, I get this:
| swScaler: pal8 is not supported as output pixel format
| [0x103aaf58] swscale scale error: could not init SwScaler and/or allocate 
memory
| [0x103bb5d8] xcb_xv generic error: no available XVideo adaptor
| object not accessible
| [swscaler @ 0x1042cf30] No accelerated colorspace conversion found from 
yuv420p to rgb4.
| object not accessible

The first three lines of this are in red, I'm not sure what this means.
In any any case, VLC itself doesn't display any error, this is only on
the commandline.

About the system, it's an old MacMini G4, a big-endian CPU, using a
Radeon graphic chip. I've also attached the output of xvidinfo in case
this gives some insight.

Cheers!

Uli


-- System Information:
Debian Release: squeeze/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (1, 'experimental')
Architecture: powerpc (ppc)

Kernel: Linux 2.6.33-2-powerpc
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages vlc depends on:
ii  libaa1  1.4p5-38 ascii art library
ii  libc6   2.11.2-7 Embedded GNU C Library: Shared lib
ii  libfreetype62.4.2-1  FreeType 2 font engine, shared lib
ii  libfribidi0 0.19.2-1 Free Implementation of the Unicode
ii  libgcc1 1:4.4.5-4GCC support library
ii  libgl1-mesa-glx [libgl1 7.7.1-4  A free implementation of the OpenG
ii  libqtcore4  4:4.6.3-4Qt 4 core module
ii  libqtgui4   4:4.6.3-4Qt 4 GUI module
ii  libsdl-image1.2 1.2.10-2+b1  image loading library for Simple D
ii  libsdl1.2debian 1.2.14-6.1   Simple DirectMedia Layer
ii  libstdc++6  4.4.5-4  The GNU Standard C++ Library v3
ii  libtar  1.2.11-6 C library for manipulating tar arc
ii  libvlccore4 1.1.3-1  base library for VLC and its modul
ii  libx11-62:1.3.3-3X11 client-side library
ii  libx11-xcb1 2:1.3.3-3Xlib/XCB interface library
ii  libxcb-keysyms1 0.3.6-1  utility libraries for X C Binding 
ii  libxcb-randr0   1.6-1X C Binding, randr extension
ii  libxcb-shm0 1.6-1X C Binding, shm extension
ii  libxcb-xv0  1.6-1X C Binding, xv extension
ii  libxcb1 1.6-1X C Binding
ii  libxext62:1.1.2-1X11 miscellaneous extension librar
ii  ttf-freefont20090104-7   Freefont Serif, Sans and Mono True
ii  vlc-nox 1.1.3-1  multimedia player and streamer (wi
ii  zlib1g  1:1.2.3.4.dfsg-3 compression library - runtime

Versions of packages vlc recommends:
ii  vlc-plugin-notify 1.1.3-1LibNotify plugin for VLC
ii  vlc-plugin-pulse  1.1.3-1PulseAudio plugin for VLC

Versions of packages vlc suggests:
pn  mozilla-plugin-vlcnone (no description available)
pn  videolan-doc  none (no description available)

Versions of packages vlc-nox depends on:
ii  liba52-0.7.4   0.7.4-14  library for decoding ATSC A/52 str
ii  libasound2 1.0.23-2.1shared library for ALSA applicatio
ii  libass40.9.9-1   library for SSA/ASS subtitles rend
ii  libavahi-client3   0.6.27-2  Avahi client library
ii  libavahi-common3   0.6.27-2  Avahi common library
ii  libavc1394-0   0.5.3-1+b2control IEEE 1394 audio/video devi
ii  libavcodec52   5:0.6~svn20100726-0.1 library to encode decode multimedi
ii  libavformat52  5:0.6~svn20100726-0.1 ffmpeg file format library
ii  libavutil494:0.5.2-6 ffmpeg utility library
ii  libc6  2.11.2-7  Embedded GNU C Library: Shared lib
ii  libcaca0   0.99.beta17-1 colour ASCII art library
ii  

Bug#521275: Close bug?

[Ulrich Eckhardt]

Hi!

I'm just looking through bugs I filed, and this one is IMHO fixed or invalid. 
As far as I'm concerned it was a user error, not a bug in the package.

Also sorry about the Reply-To in the initial mail, I had the wrong personality 
selected in my mail client.

Thanks!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#598456: Maximizing windows doesn't account for tabs, pushes them offscreen

[Ulrich Eckhardt]

Package: fluxbox
Version: 1.1.1+git20100908.df2f51b-2


Hi!

The description of this is really that short, if you have tabs activated and
maximize a window, the window frame is resized to the screen's size (title
bar touching the top), but the tabs are sitting on top of this title bar, so
they are pushed off the top of the screen. I would expect the tabs to still
be visible inside the screen.

Thanks!

Uli



-- System Information:
Debian Release: squeeze/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (1, 'experimental')
Architecture: powerpc (ppc)

Kernel: Linux 2.6.32-5-powerpc
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages fluxbox depends on:
ii  libc6   2.11.2-6 Embedded GNU C Library: Shared lib
ii  libfontconfig1  2.8.0-2.1generic font configuration library
ii  libfreetype62.4.2-1  FreeType 2 font engine, shared lib
ii  libfribidi0 0.19.2-1 Free Implementation of the Unicode
ii  libgcc1 1:4.4.4-8GCC support library
ii  libice6 2:1.0.6-1X11 Inter-Client Exchange library
ii  libimlib2   1.4.2-8+b2   powerful image loading and renderi
ii  libsm6  2:1.1.1-1X11 Session Management library
ii  libstdc++6  4.4.4-8  The GNU Standard C++ Library v3
ii  libx11-62:1.3.3-3X11 client-side library
ii  libxext62:1.1.2-1X11 miscellaneous extension librar
ii  libxft2 2.1.14-2 FreeType-based font drawing librar
ii  libxinerama12:1.1-3  X11 Xinerama extension library
ii  libxpm4 1:3.5.8-1X11 pixmap library
ii  libxrandr2  2:1.3.0-3X11 RandR extension library
ii  libxrender1 1:0.9.6-1X Rendering Extension client libra
ii  menu2.1.43   generates programs menu for all me
ii  zlib1g  1:1.2.3.4.dfsg-3 compression library - runtime

Versions of packages fluxbox recommends:
ii  xfonts-terminus   4.30-2 Fixed-width fonts for fast reading

Versions of packages fluxbox suggests:
pn  fbdesknone (no description available)
pn  fbpager   none (no description available)

-- no debconf information



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#593364: X11 SHAPE extension not updating windows correctly

[Ulrich Eckhardt]

Package: xcompmgr
Version: 1.1.5-1

Hi!

I'm writing a program that uses the SHAPE extensions to give a window a
non-rectangular shape. This works up to the point where I try to modify
the shape at runtime. The attached example program creates a simple
window of 50x50 pixels. It then creates a moving square of 10x10 pixels
inside that window which is not part of the window, using the SHAPE
extension, so that you can watch the background or underlying window
through this hole.

While xcompmgr is running, the window isn't updated correctly. I tried
this using fluxbox (which itself has some problems with X11 SHAPE), but
it also happens with openbox, icewm or windowmaker, it even happens when
not using any windowmanager at all. I tried using compiz as compositing
and window manager, but failed getting it to run even halfway. 


The system is an old Mac Mini, using an ATI Radeon videochip and running
a current Debian/testing system.

Give me some feedback if you can't reproduce this, I'll try to circle
in the other factors then.

Cheers!

Uli


-- System Information:
Debian Release: squeeze/sid
  APT prefers testing
  APT policy: (500, 'testing'), (500, 'stable'), (1, 'experimental')
Architecture: powerpc (ppc)

Kernel: Linux 2.6.32-5-powerpc
Locale: LANG=en_GB.UTF-8, LC_CTYPE=en_GB.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages xcompmgr depends on:
ii  libc6 2.11.2-2   Embedded GNU C Library: Shared lib
ii  libx11-6  2:1.3.3-3  X11 client-side library
ii  libxcomposite11:0.4.2-1  X11 Composite extension library
ii  libxdamage1   1:1.1.3-1  X11 damaged region extension libra
ii  libxfixes31:4.0.5-1  X11 miscellaneous 'fixes' extensio
ii  libxrender1   1:0.9.6-1  X Rendering Extension client libra

// Example program for bug in xcompmgr
//
// This should display a window with a moving hole in it, but with xcompmgr
// running, it isn't updated correctly.
//
// g++ -Wall `pkg-config xcb xcb-shape --cflags --libs` bug-shape-changer.cpp -o bug-shape-changer
//

#include stdlib.h
#include string.h
#include xcb/xcb.h
#include xcb/shape.h
#include iostream
#include stdexcept
#include assert.h

/* global objects, the connection to the X server and the window
created there. */
xcb_connection_t* connection = 0;
xcb_window_t win;
unsigned const window_size = 50;

unsigned const hole_size = 10;
int hole_position = 0;

void
init_connection()
{
/* open connection to the server */
connection = xcb_connect(NULL,NULL);
if (xcb_connection_has_error(connection))
throw std::runtime_error(xcb_connect() failed);
}


void
check_shape_extension()
{
// query for availability
xcb_query_extension_reply_t const* reply = xcb_get_extension_data(
connection, xcb_shape_id);
if(!reply)
throw std::runtime_error(xcb_get_extension_data failed);
if(!reply-present)
throw std::runtime_error(X11 shape extension not present);

// query version
xcb_shape_query_version_cookie_t cookie = xcb_shape_query_version_unchecked(connection);
xcb_shape_query_version_reply_t* version = xcb_shape_query_version_reply(connection,
cookie, 0);
if(!version)
throw std::runtime_error(xcb_shape_query_version failed);
}


void
create_window(int16_t x, int16_t y, uint16_t width, uint16_t height)
{
/* retrieve the first screen from the setup */
xcb_setup_t const* setup = xcb_get_setup(connection);
if(!setup)
throw std::runtime_error(xcb_get_setup() failed);
xcb_screen_t* screen = xcb_setup_roots_iterator(setup).data;
if(!screen)
throw std::runtime_error(missing root screen);

// create window
xcb_window_t id = xcb_generate_id(connection);
uint32_t const win_values[] =
{
screen-black_pixel,
XCB_EVENT_MASK_EXPOSURE
};
xcb_create_window(connection,
XCB_COPY_FROM_PARENT, // depth
id,
screen-root, // parent
x, y, width, height, // position, size
0, // border width
XCB_WINDOW_CLASS_INPUT_OUTPUT, // class
XCB_COPY_FROM_PARENT, // visual
XCB_CW_BACK_PIXEL | XCB_CW_EVENT_MASK, win_values);
win = id;
}

void
set_window_mask()
{
// set rectangualar base outline first
xcb_rectangle_t const rect = {0, 0, window_size, window_size};
xcb_shape_rectangles(connection,
XCB_SHAPE_SO_SET,
XCB_SHAPE_SK_BOUNDING,
0,
win,
0, 0, // offset
1, rect);

// cut out moving rectangle
xcb_rectangle_t const hole = {hole_position, hole_position, hole_size, hole_size};
xcb_shape_rectangles(connection,
XCB_SHAPE_SO_SUBTRACT,
XCB_SHAPE_SK_BOUNDING,
0,
win,
0, 0, // offset
1, hole);
xcb_flush(connection);
}

int
main(int argc, char** argv)
{
try
{
init_connection();

create_window(0, 0, 

Bug#544112: Example program

[Ulrich Eckhardt]

Hi!

I noticed in a different context that fluxbow was sending superfluous expose 
events when losing and gaining focus. These events are normally sent when 
parts of a window were exposed and thus need redrawing. I guess these could be 
responsible for the anomalous flickering.

The attached example program just creates a simple window using the X11 SHAPE 
extension. You can see that when losing focus, the window receives two expose 
events and when the window gains focus it receives four expose events. All 
these events are seemingly unnecessary, since nothing in the window was 
actually exposed.

Notes:
 - I also tried the same program under icewm, where these events are not sent.
 - I have noticed that both gmplayer and xine suffer from the flickering, but 
I don't actually know if it is caused by the expose events.
 - I'm not sure if this is related, but #593061 is also about focus handling. 
I noticed here that when entering the window created by the example program, 
the window doesn't always receive focus. In particular when I cross the border 
quickly it seems to sometimes miss this.

I hope this helps a bit.

Uli
// Example program for Debian bug #544112
//
// g++ -Wall `pkg-config xcb xcb-shape --cflags --libs`
//

#include stdlib.h
#include string.h
#include xcb/xcb.h
#include xcb/shape.h
#include iostream
#include stdexcept
#include assert.h

/* global objects, the connection to the X server and the window
created there. */
xcb_connection_t* connection = 0;
xcb_window_t win;


void
init_connection()
{
/* open connection to the server */
connection = xcb_connect(NULL,NULL);
if (xcb_connection_has_error(connection))
throw std::runtime_error(xcb_connect() failed);
}


void
check_shape_extension()
{
// query for availability
xcb_query_extension_reply_t const* reply = xcb_get_extension_data(
connection, xcb_shape_id);
if(!reply)
throw std::runtime_error(xcb_get_extension_data failed);
if(!reply-present)
throw std::runtime_error(X11 shape extension not present);

// query version
xcb_shape_query_version_cookie_t cookie = xcb_shape_query_version_unchecked(connection);
xcb_shape_query_version_reply_t* version = xcb_shape_query_version_reply(connection,
cookie, 0);
if(!version)
throw std::runtime_error(xcb_shape_query_version failed);
}


void
create_window(int16_t x, int16_t y, uint16_t width, uint16_t height)
{
/* retrieve the first screen from the setup */
xcb_setup_t const* setup = xcb_get_setup(connection);
if(!setup)
throw std::runtime_error(xcb_get_setup() failed);
xcb_screen_t* screen = xcb_setup_roots_iterator(setup).data;
if(!screen)
throw std::runtime_error(missing root screen);

// create window
xcb_window_t id = xcb_generate_id(connection);
uint32_t const win_values[] =
{
screen-black_pixel,
XCB_EVENT_MASK_EXPOSURE
};
xcb_create_window(connection,
XCB_COPY_FROM_PARENT, // depth
id,
screen-root, // parent
x, y, width, height, // position, size
0, // border width
XCB_WINDOW_CLASS_INPUT_OUTPUT, // class
XCB_COPY_FROM_PARENT, // visual
XCB_CW_BACK_PIXEL | XCB_CW_EVENT_MASK, win_values);
win = id;
}


int
main(int argc, char** argv)
{
try
{
init_connection();

create_window(0, 0, 50, 50);

// set rectangualar base outline
xcb_rectangle_t const rect = {0, 0, 50, 50};
xcb_shape_rectangles(connection,
XCB_SHAPE_SO_SET,
XCB_SHAPE_SK_BOUNDING,
0,
win,
0, 0, // offset
1, rect);

// actually display window
xcb_map_window(connection, win);

int const e = xcb_flush(connection);
if(e=0)
// TODO: provide more detailed diagnostics if possible
throw std::runtime_error(xcb_flush failed);

// poll connection's file descriptor for events
int const ss = xcb_get_file_descriptor(connection);
while(true)
{
// wait on the connection socket
fd_set fd;
FD_ZERO(fd);
FD_SET( ss, fd);
timeval timeout;
timeout.tv_usec = 0;
timeout.tv_sec =  1;
switch(select( ss+1, fd, 0, 0, timeout))
{
case -1:
throw std::runtime_error(select failed);
case 0:
std::cout  select timed out  std::endl;
break;
case 1:
break;
default:
std::cout  select return unexpected value  std::endl;
break;
}

// handle all available events
xcb_generic_event_t* e = 0;
while((e = xcb_poll_for_event(connection)))
{
switch (e-response_type  ~0x80)
{
case XCB_EXPOSE:

Bug#593061: Pointer moving between windows not detected correctly

[Ulrich Eckhardt]

Package: fluxbox
Version: 1.1.1+dfsg2-1

When moving between two windows and at the same time between groups, the focus 
is not updated correctly. Following steps illustrate the setup:

 * Delete your .fluxbox, I'm assuming a default configuration.
 * Start fluxbox.
 * In the Tab Options (Right-click on the background) turn off Tabs in 
Titlebar to get external tabs.
 * In the Focus Model, select Mouse Focus and Mouse Tab Focus.

You now have a setup where you change the focus with just the mouse movement, 
no clicks involved. It allows you to switch between unrelated windows and, by 
moving over the tabs, to switch within one group of windows.

Now, in order to reproduce the buggy behavior, use following steps:
 * Create three windows A, B, C (I used three uxterms, but it doesn't matter).
 * Combine two of the three to a group, by middle-dragging the tab of one onto 
the other, so you get A, [B, C]
 * Top-align all three windows so that they overlap partially:
   [A*]  [B|C]
   ,-+,
   [ ]]
In the picture, A has the focus and is on top. The important part is that you 
can move seamlessly between these windows, hence the alignment. You can also 
position B and C below, but the above setup is easiest to explain and 
reproduce.
 * Now, move the pointer from the A window onto either of the B or C tabs.

The last step should have the effect of switching focus away from A to the 
according window in the B/C group. However, it only switches the focus to the 
B/C group, it does not switch to the tab's window within that group.

Note that this is a rather old bug, it didn't come up with the version that 
recently entered testing.

Cheers!

Uli




-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#551102: Can't reproduce

[Ulrich Eckhardt]

Hi!

I just tried to reproduce this, but failed. The installed programs are a 
current Debian testing:

fluxbox 1.1.1+dfsg2-1
tilda 0.09.6-2
conky-all 1.8.0-1

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#592520: missing keybord mapping example file

[Ulrich Eckhardt]

Package: stormbaancoureur
Version: 2.1.6-1

Hi!

The README mentions a .stormbaancoureur.keys.example file, but that is 
apparently not part of the package. It would probably also be nice for Debian 
beginners if it was not hidden, even if it must be stored as hidden file 
eventually.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#592075: patch: add mixer bindings for mousewheel

[Ulrich Eckhardt]

Package: alsamixergui
Version: 0.9.0rc2-1-9
Severity: whishlist

The attached patch allows using the mousewheel to increase/decrease the 
volume.

Cheers!

Uli
--- alsamixergui-0.9.0rc2-1.orig/src/Fl_AM.cxx	2002-05-09 17:41:35.0 +0200
+++ src/Fl_AM.cxx	2010-08-06 20:15:06.0 +0200
@@ -385,6 +384,28 @@ public:
 	}
 	int handle(int e)
 	{
+switch(e) {
+case FL_MOUSEWHEEL:
+{
+if(!Fl::event_inside(this))
+break;
+Fl_AMGroup* g = (Fl_AMGroup*) this-parent();
+mixer_focus_elem = g-elem_id;
+double new_value = increment(value(), Fl::event_dy());
+if (this==g-volumeL || g-lock-value()) {
+g-volumeL-value(new_value);
+mixer_volume_hard[MIXER_CHN_LEFT] = new_value;
+}
+if (this==g-volumeR || g-lock-value()) {
+g-volumeR-value(new_value);
+mixer_volume_hard[MIXER_CHN_RIGHT] = new_value;
+}
+gui_update();
+return 1;
+}
+default:
+break;
+}
 		return Fl_Slider::handle(e);
 	}
 };

Bug#587999: turning off radeonfb crashes system on console switch

[Ulrich Eckhardt]

Hi!

I got around to testing with video=radeonfb:off on the kernel commandline. 
What you then get is a completely hanging system when switching terminale 
(chvt), when trying to stop the login manager KDM or when logging out of an X 
session. Note that this means that rebooting the system becomes impossible, 
similarly partitions aren't unmounted properly either.

I also tried activating KMS explicitly then, but that crashes even earlier 
during boot, even before init is started. The kernel seems to detect that my 
video card has 256MiB of VRAM, while it only has 32. Some other memory layouts 
it detects are similarly off.

If I find a free hour I'll try to disable KDM and then try without radeonfb, 
if that is broken, too, or if it is only in combination with X. I'll also try 
to copy the kernel output when using KMS.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#587999: further info

[Ulrich Eckhardt]

On Sunday 11 July 2010 10:05:08 Brice Goglin wrote:
 What about the original problem (pixel columns switched when using XV,
 with 16 or 24 bits color depth), is it gone now?

No, not really, but it doesn't affect me now.

If I understand correctly, what happened was that KMS broke DRI for X, so it 
switched back to non-accelerated rendering. Non-accelerated rendering gets the 
pixel columns wrong. In addition, it seems that DRI with 24BPP requires more 
memory now that I have (can someone confirm/explain that?), which made 
diagnosis more complicated.

After turning off KMS and upgrading, I was to switch back to 16BPP, which 
works accelerated now, so the bug doesn't affect me. However, this is rather a 
workaround than a fix.


Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#567793: EXA causes arifacts

[Ulrich Eckhardt]

On Sunday 11 July 2010 10:12:23 Brice.Goglin wrote:
 On Sun, Jan 31, 2010 at 02:14:33PM +0100, Ulrich Eckhardt wrote:
  I have artifacts with colors being off and sometimes areas being filled
  with black/white noise. 
[...]
 Assuming this is the problem you referred to at the end of
   http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=587999#66
 could you give an update to the status of this exact bug report too?

No, the artifacts that prompted me to file this bug were different. However, 
I'll give EXA another try with the new driver and report the current state.

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#587999: further info

[Ulrich Eckhardt]

On Friday 09 July 2010 17:49:57 Michel Dänzer wrote:
   http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=565313
   http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=567616
   http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=567793
 
 The depth 16 issue could be fixed in xserver-xorg-video-radeon
 1:6.13.1-1, not sure about the other EXA issue.

Thanks, that seems to do the job. I downloaded 6.13.1-1 from unstable and now 
I can again use 16 BPP without the greenish tint. Further, if I deactivate KMS 
(modprobe radeon modeset=0), X also initialises DRI (even though it complains 
about this and that) and bug 587999 doesn't appear, so it's indeed only the 
non-DRI code it seems.

Just for the record, I had the effect that areas were not recognized as newly 
exposed with EXA and the new radeon driver, though only with a KDE app. 
Switching back to XAA seems to work around that. I'll try to reproduce this 
and file a separate report if I get to it.


Thank you for your help, I owe you a beer!

Uli



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#587999: further info

[Ulrich Eckhardt]

On Friday 09 July 2010 14:56:39 Michel Dänzer wrote:
 This is the reason for the DRI being disabled now:
  (II) RADEON(0): Memory manager initialized to (0,0) (1728,4854)
  (II) RADEON(0): Reserved area from (0,1680) to (1728,1682)
  (II) RADEON(0): Largest offscreen area available: 1728 x 3172
  (II) RADEON(0): Will use front buffer at offset 0x0
  (II) RADEON(0): Will use back buffer at offset 0x9cf000
  (II) RADEON(0): Will use depth buffer at offset 0x14e2000
  (II) RADEON(0): Will use 0 kb for textures at offset 0x1ff5000
  (EE) RADEON(0): Static buffer allocation failed.  Disabling DRI.
  (EE) RADEON(0): At least 34020 kB of video memory needed at this
  resolution and depth.
 
 Maybe it'll work without Option AccelMethod XAA

I'll try that.

 (or with explicit EXA instead), otherwise you'll need to
 reduce the maximum desktop size via the Virtual directive or
 run in depth 16.

EXA and 16 BPP are out, except if using neither of EXA and XAA changes 
anything there:

 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=565313
 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=567616
 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=567793

Hopefully...

Uli





--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#587999: further info

[Ulrich Eckhardt]

On Monday 05 July 2010 09:03:53 you wrote:
 On Sam, 2010-07-03 at 23:58 +0200, Ulrich Eckhardt wrote:
  (EE) RADEON(0): [dri] RADEONDRIGetVersion failed to open the DRM
  [dri] Disabling DRI.
 
 It should work better with the DRI enabled.

Good catch, but shouldn't it work even without DRI?

 Do you have the firmware-linux package installed?

Yes.

 If so, please post the output of running
 
 sudo modprobe radeon
 
 as well as the messages appearing in the dmesg output after it.

The module is already loaded. I can unload it though, even with a running X11 
session, lsmod shows that its refcount is zero. When loading it, it outputs 
[drm] radeon kernel modesetting enabled. in dmesg. However, when starting 
the X server, it outputs radeonfb :00:10.0: Invalid ROM contents there.

Note that this error is nothing new. X11 has been complaining about a missing 
ROM since I got this machine (MacMini), but has always worked nonetheless. If 
I understand correctly, the graphics adapter doesn't have a ROM and doesn't 
need one, because it is initialized by the firmware.

Thanks for your effort!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#587999: Xv switches pixel columns

[Ulrich Eckhardt]

Package: xserver-xorg
Version: 1:7.5+6

Using mplayer and xine (didn't try others) with Xv as video output driver, I 
have the effect that the order of pixel columns is switched. If I use 24 bits 
color depth, the order is 1,0,3,2 etc. Using 16 bits color depth, the order is 
3,2,1,0,7,6,5,4.

Notes:
 * I'm using the radeon driver backend.
 * I'm on a big-endian machine (PPC).

As a workaround, I can use xshm as video output driver.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#587999: further info

[Ulrich Eckhardt]

-- Package-specific info:
/var/lib/x11/X.roster does not exist.

/var/lib/x11/X.md5sum does not exist.

X server symlink status:
lrwxrwxrwx 1 root root 13 Dec 10  2007 /etc/X11/X - /usr/bin/Xorg
-rwxr-xr-x 1 root root 1940408 Jun  3 18:28 /usr/bin/Xorg

/var/lib/x11/xorg.conf.roster does not exist.

VGA-compatible devices on PCI bus:
:00:10.0 VGA compatible controller: ATI Technologies Inc RV280 [Radeon 
9200] (rev 
01)

/var/lib/x11/xorg.conf.md5sum does not exist.

Xorg X server configuration file status:
-rw-r--r-- 1 root root 6148 Jul  3 16:47 /etc/X11/xorg.conf

Contents of /etc/X11/xorg.conf:
# This configfile demonstrates the broken behaviour of xorg
# The problem is that this file as configuration is capable of starting X in
# the configured layout, but once shut down, this does _not_ work a second
# time.
#
# Note:
#   Option  UseFBDev  true
#  will enable startup.
#
# Configuration:
# xserver-xorg-video-ati 6.6.3-2
# Monitor:
# SyncMaster 225BW on VGA
# Machine:
# Mac Mini PPC

Section Files
FontPath/usr/share/fonts/X11/misc
FontPath/usr/X11R6/lib/X11/fonts/misc
FontPath/usr/share/fonts/X11/cyrillic
FontPath/usr/X11R6/lib/X11/fonts/cyrillic
FontPath/usr/share/fonts/X11/100dpi/:unscaled
FontPath/usr/X11R6/lib/X11/fonts/100dpi/:unscaled
FontPath/usr/share/fonts/X11/75dpi/:unscaled
FontPath/usr/X11R6/lib/X11/fonts/75dpi/:unscaled
FontPath/usr/share/fonts/X11/Type1
FontPath/usr/X11R6/lib/X11/fonts/Type1
FontPath/usr/share/fonts/X11/100dpi
FontPath/usr/X11R6/lib/X11/fonts/100dpi
FontPath/usr/share/fonts/X11/75dpi
FontPath/usr/X11R6/lib/X11/fonts/75dpi
# path to defoma fonts
FontPath/var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType
EndSection

Section Module
Loadbitmap
Loaddbe
Loadddc
Loaddri
Loadextmod
#Load   freetype
Loadglx
Loadint10
Loadvbe
EndSection

#Section Extensions
#Option Composite enable
#EndSection

Section InputDevice
Identifier  Generic Keyboard
Driver  kbd
Option  CoreKeyboard
Option  XkbRules  xorg
Option  XkbModel  pc104
Option  XkbLayout dvorak
Option  XkbVariantdvorak
Option  XkbOptionslv3:ralt_switch
EndSection

Section InputDevice
Identifier  Configured Mouse
Driver  mouse
Option  CorePointer
Option  Device/dev/input/mice
EndSection

Section Device
Identifier  Builtin Radeon
Driver  radeon
# select model with integrated DVI adapter
Option  MacModel mini-internal
#   Option  MonitorLayout TMDS,NULL
#   Option  PanelSize 1680x1050
# ignore monitor mode lines received via DDC but use DDC to detect the 
monitor 
itself
Option  IgnoreEDID  true
# dynamically scale the chip clock to reduce energy consumption
#Option  DynamicClocks true
# TODO: try this on
# Option  DefaultTMDSPLL true
#   Option  AGPMode   4
#Option  XAANoOffscreenPixmaps true

#   Option  UseFBDev  false
#   Option  UseFBDev  true
#   Option  AGPFastWrite  off
#   Option  DDCMode   true
#   Option  AccelMethod   EXA
#   Option  AccelMethod   XAA
#   Option  ReverseDDCtrue
###
#   BusID   PCI:0:16:0

# All options supported by the driver (2008-08-24)
#
#Option SWcursor boolean
#Option SWcursor true
#Option NoAccel boolean
#Option Dac6Bit boolean
#Option VideoKey integer
#Option ScalerWidth integer
#Option AGPMode integer
#Option AGPFastWrite boolean
#Option BusType string
#Option DDCMode boolean
#Option DisplayPriority string
#Option ColorTiling boolean
#Option IgnoreEDID boolean
#Option PanelSize string
#Option EnablePageFlip boolean
#Option ForceMinDotClock frequency
#Option RenderAccel boolean
#Option AccelMethod string
# old X accel
Option AccelMethod XAA
# new X accel
#Option AccelMethod EXA
#Option AccelDFS boolean
#Option FBTexPercent integer
#Option DepthBits integer
#Option DMAForXv boolean
#Option SubPixelOrder string
#Option DynamicClocks boolean
#Option VGAAccess boolean
#Option ReverseDDC boolean
#Option LVDSProbePLL boolean
#Option TVDACLoadDetect boolean
#Option DefaultTMDSPLL boolean
#Option DefaultTVDACAdj boolean
#Option DRI boolean
#Option DefaultConnectorTable boolean
#Option MacModel string
#Option TVStandard string
#Option ForceTVOut boolean
#Option IgnoreLidStatus boolean
#Option Int10 boolean

Bug#586919: Missing manpage for apctest

[Ulrich Eckhardt]

Package: apcupsd
Version: 3.14.5-2

There is a manpage source file in package apcupsd-doc, installed as 
apctest.man, which should probably be compiled and installed as manpage 
instead. There is also a file called apcupsd.man in the same folder, which 
seems to be the source file for the supplied manpage and which (I'd say) 
doesn't have to be installed there.

Cheers!

Uli

-- 
Sator Laser GmbH, Fangdieckstraße 75a, 22547 Hamburg, Deutschland
Geschäftsführer: Thorsten Föcking, Amtsgericht Hamburg HR B62 932

**
Sator Laser GmbH, Fangdieckstraße 75a, 22547 Hamburg, Deutschland
Geschäftsführer: Thorsten Föcking, Amtsgericht Hamburg HR B62 932
**
   Visit our website at http://www.satorlaser.de/
**
Diese E-Mail einschließlich sämtlicher Anhänge ist nur für den Adressaten 
bestimmt und kann vertrauliche Informationen enthalten. Bitte benachrichtigen 
Sie den Absender umgehend, falls Sie nicht der beabsichtigte Empfänger sein 
sollten. Die E-Mail ist in diesem Fall zu löschen und darf weder gelesen, 
weitergeleitet, veröffentlicht oder anderweitig benutzt werden.
E-Mails können durch Dritte gelesen werden und Viren sowie nichtautorisierte 
Änderungen enthalten. Sator Laser GmbH ist für diese Folgen nicht 
verantwortlich.
**




--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#584850: Provide better diagnostic for directories [patch]

[Ulrich Eckhardt]

Package: mmv
Version: 1.01b-15

I just tried to rename a bunch of directories:

  $ mmv -v DSide-* DSide #1
  DSide-* - DSide #1 : no match.
  Nothing done.

This is completely unhelpful, as the error no match seems to indicate the 
match string was wrong. Only after many fruitless attempts and thoroughly 
searching the manpage I found the -r option that (for reasons I don't 
understand) is necessary to convince mmv to move directories. At the very 
least I would have expected the -v to tell me that the matching directory 
entries are not considered in the rename.

Looking at the sources, I'd do something like the following to function 
keepmatch():


 #ifdef IS_MSDOS
-   if ((ffrom-fi_attrib  FA_DIREC) ? !dirs : !fils)
+   if ((ffrom-fi_attrib  FA_DIREC) ? !dirs : !fils) {
 #else
getstat(pathbuf, ffrom);
-   if ((ffrom-fi_stflags  FI_ISDIR) ? !dirs : !fils)
+   if ((ffrom-fi_stflags  FI_ISDIR) ? !dirs : !fils) {
 #endif
+   if (verbose)
+   printf(ignoring directory %s\n, ffrom-fi_name);
return(0);
+   }


Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#581531: Can't disable text in image

[Ulrich Eckhardt]

Package: webcam
Version: 3.95.dfsg.1-8.1

'webcam' adds a text containing a timestamp by default. This text can not be 
turned off, if you specify an empty text in the configuration file, you only 
get a message telling you that there's a syntax error on that line and the 
value is reverted to the default.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#551124: Bug still present, overall condition even worse

[Ulrich Eckhardt]

Firstly, I still can't play consecutive Flash videos. Note that this only 
applies to Flash videos, not other types of videos, in case that wasn't clear.

Further, when while xine is idle (either after finishing one or on startup) 
and then I drag'n'drop another file onto xine, it takes around ten seconds 
until the video output starts, while the audio output comes much earlier. 
During that time, the UI is unresponsive, stopping, closing, volume all don't 
work. After that, audio and video are not out of sync, the video only starts 
being displayed later. This applies to all videos btw, not just flash, and it 
doesn't apply to videos started directly from the commandline. Also, when 
starting via the play button, the video starts correctly.

For the record, here are the now installed packages and versions:

xine-ui0.99.6-1
libc6  2.10.2-6
libcurl3-gnutls7.20.0-3
liblircclient0 0.8.3-5
libpng12-0 1.2.43-1
libreadline6   6.1-1
libx11-6   2:1.3.3-2
libxext6   2:1.1.1-3
libxft22.1.14-2
libxine1   1.1.18.1-1+b1
libxine1-ffmpeg1.1.18.1-1+b1
libxine1-x 1.1.18.1-1+b1
libxinerama1   2:1.1-3
libxtst6   2:1.1.0-2
libxv1 2:1.0.5-1
libxxf86vm11:1.1.0-2


Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#575583: Unusable on non-QWERT keyboards

[Ulrich Eckhardt]

Package: tint
Version: 0.03b

TINT is not usable e.g. on a Dvorak keyboard, it assumes that the keys JKL are 
arranged in a way similar to how the cursor keys are arranged. By default, it 
should use the cursor keys instead of JKL, their arrangement is more portable 
between different keyboards.

Thanks!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#570830: Patch to avoid invalid fstat calls when opening empty files

[Ulrich Eckhardt]

Package: libxine1-bin
Version: 1.1.17-1+b1

Hi!

When opening empty files with xine, it reports:
 This is xine (X11 gui) - a free video player v0.99.6cvs.
 (c) 2000-2007 The xine Team.
 system call fstat: Bad file descriptor
 system call fstat: Bad file descriptor
 call failed
 object not accessible
 object not accessible
 call failed

The attached patch fixes the two error messages from fstat() calls and a typo. 
The problem was that when detecting an empty file, it is closed, but the file 
descriptor in struct file_input_plugin_t is still left at the previous value, 
causing the errors when it is used later.

Enjoy!

Uli
--- xine-lib-1.1.17/src/input/input_file.c	2009-11-30 21:55:45.0 +0100
+++ ../xine-lib-1.1.17/src/input/input_file.c	2010-02-21 19:46:26.0 +0100
@@ -292,7 +292,7 @@ static uint32_t file_plugin_get_blocksiz
 }
 
 /*
- * Return 1 is filepathname is a directory, otherwise 0
+ * Return 1 if filepathname is a directory, otherwise 0
  */
 static int is_a_dir(char *filepathname) {
   struct stat  pstat;
@@ -409,6 +409,7 @@ static int file_plugin_open (input_plugi
   if (file_plugin_get_length (this_gen) == 0) {
   _x_message(this-stream, XINE_MSG_FILE_EMPTY, this-mrl, NULL);
   close (this-fh);
+  this-fh = -1;
   xine_log (this-stream-xine, XINE_LOG_MSG,
 		_(input_file: File empty: %s\n), this-mrl);
   return -1;

Bug#569591: patch

[Ulrich Eckhardt]

Greetings!

After some hacking, I think I found the culprit, it is the algorithm that 
linebreaks text in order to fit it into the given width of a dialog box. When 
a single word (e.g. a path) is so long it doesn't fit into a line on its own, 
it tries to step back to the last whitespace first, but that isn't reset after 
each line.

Cheers!

Uli
--- xine-ui-0.99.5/src/xitk/xine-toolkit/image.c	2007-03-30 01:07:56.0 +0200
+++ ../xine-ui-0.99.5/src/xitk/xine-toolkit/image.c	2010-02-21 18:52:34.0 +0100
@@ -515,6 +515,7 @@ xitk_image_t *xitk_image_create_image_wi
   
   xitk_font_string_extent(fs, bp, lbearing, rbearing, NULL, NULL, NULL); 
   linew = rbearing - lbearing;
+  lastws = 0;
 }
 
 p++;

Bug#561779: Bug fixed

[Ulrich Eckhardt]

On Thursday 18 February 2010 14:54:27 Ana Guerrero wrote:
 After a quick chat with Antonis by IRC he told me he has tried to reproduce
 the bug several times with a new version (4:4.3.4-1) and the bug was not
 reproducible for it.
 Ulrich, you reported it against 4:4.3.2-1, could you check with this
 newer version?

I have tested it here with 4:4.3.4-1, and it still exists and with the same 
steps for reproduction. What I have noticed is that the visible marker 
(highlighter) for the selected message sometimes disappears when changing 
sorting order, but I'm not sure when. Just wondering, did you reproduce the 
problem with an older version?

In any case, I have just taken a few screenshots to illustrate the problem, 
just in case my description was unclear. I'll clean these up and post them 
later.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#569591: Filename error handling

[Ulrich Eckhardt]

Package: xine-ui
Version: 0.99.5+cvs2007

Hi!

I was looking at bug #510606 and started trying to reproduce it.
However, I can't see exactly the behaviour the original reporter described, 
but I see that xine gets a bit picky with long filenames, empty files and 
invalid paths, so I decided to write that up separately.



 # test 1
 $ touch 1234567890
 $ xine 1234567890
 
 This is xine (X11 gui) - a free video player v0.99.6cvs.
 (c) 2000-2007 The xine Team.
 system call fstat: Bad file descriptor
 system call fstat: Bad file descriptor
 object not accessible
 call failed

The weird things there are the calls to fstat with bad file descriptors. While 
this seems to be caught, it is an indicator of something going very wrong. 
Note that the filename is really only ten characters long, it rather seems to 
be related to the fact that the file is empty. When I ran xine on a logfile 
(just for testing), it didn't produce those two lines.

Anyway, what xine does here is to produce two popups. The first one is

  *sight*, unknown error. (File is empty: 1234567890)

Well, obviously the error is known, so that doesn't make sense. (Also it 
probably should be *sigh* not *sight*.) The second popup is

 - xine engine error -
 Input plugin failed to open mrl '1234567890'

This error seems a) wrong, since it did manage to open the file and b) 
superfluous since the first popup already explains all. Another weird things 
is that these two popups don't come in a fixed order! Most of the time the 
first one is in front of the second one, but not always!



 # test 2
 $ touch 1234567890
 $ xine `pwd`1234567890
 
 This is xine (X11 gui) - a free video player v0.99.6cvs.
 (c) 2000-2007 The xine Team.
 object not accessible
 call failed

Note here that the resulting argument does not actually resolve to a file, 
since 'pwd' doesn't have a trailing slash in its output. Again, I get two 
error popups, while one is enough.



 # test 3
 # these are actually 200 characters below!
 $ touch 1234567890...1234567890
 $ xine `pwd`1234567890...1234567890
 This is xine (X11 gui) - a free video player v0.99.6cvs.
 (c) 2000-2007 The xine Team.
 xiTK received SIGSEGV signal, RIP.
 Aborted
 xdg-screensaver: Window 71303875 does not exist

Here, xine actually aborted with a segmentation violation. For a split second, 
it displayed the message

 - xine engine error -
 Input plugin failed to open mrl '/home/uli/tmp12345678901234567

Note here that the text really ends with a cut off path. Neither the full path 
is given nor even the closing single quote! Running the same in GDB produces 
the following output:

 Program received signal SIGSEGV, Segmentation fault.
 [Switching to Thread 0x4f0f1480 (LWP 4765)]
 _wordcopy_bwd_dest_aligned (dstp=1318006776,
 srcp=1318006768,
 len=1071651524) at wordcopy.c:391
 object not accessible
 391 wordcopy.c: No such file or directory.
 in wordcopy.c
 (gdb) bt
 #0  _wordcopy_bwd_dest_aligned (dstp=1318006776,
 srcp=1318006768,
 len=1071651524) at wordcopy.c:391
 #1  0x0faa89dc in *__GI_memmove (dest=0x4f0ec4f2,
  src=value optimized out,
  len=4294967282) at memmove.c:99
 #2  0xe010069b in ?? ()
 Cannot access memory at address 0xf44f0ee8

Things to note here are that xine is trying to copy one billion (hex 
0x3fe01ac4) words, which is more that a process in a 32-bit environment has as 
virtual address space, I think.
Further up the stack, it is doing a __GI_memmove() which has as an actual size 
the value 0xfff2. How that value converts to a length of 0x3fe01ac4 in the 
call to _wordcopy_bwd_dest_aligned() is beyond me, and possibly worth a bug 
for some other code.

Further testing with GDB turned up a few more things:
 1. (gdb) run /home/uli/tmp1234567890
This creates an error message (no surprise here) but otherwise works as 
expected.
 2. (gdb) run /home/uli/tmp12345678901234567890
This creates the same error messages, but on shutdown, xine receives a SIGSEGV 
in _int_free(). In the call stack is XCloseDisplay() though, so that might be 
a bug in Xlib.
 3. (gdb) run /home/uli/tmp123456789012345678901234567890
Same as case 2.
 4. (gdb) run /home/uli/tmp1234567890123456789012345678901234567890
At this size, it suddenly runs normally! I tried multiple times, I didn't see 
the same effect as with cases 2 and 3.
 5. (gdb) run /home/uli/tmp12345678901234567890123456789012345678901234567890
Here we get the effects described above, with the SIGSEGV in the call to 
_wordcopy_bwd_dest_aligned().



The effects are all easy to reproduce. I'm going to compile a version locally 
with debug symbols and see if I can create more info or even a patch tomorrow.

Cheers and good evening!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact 

Bug#567616: Some more info...

[Ulrich Eckhardt]

Hi!

Since an update yesterday (I'm running testing here), I'm seeing the same 
effect here. I can also work around this by switching to 24 bits color depth.

Anyway, some info on this system here:
 - Mac Mini with a G4 PowerPC CPU
 - Radeon graphics chip, 'lspci' says VGA compatible controller: ATI 
Technologies Inc RV280 [Radeon 9200] (rev 01)
 - Some installed packages:
  xserver-xorg1:7.5+2
  xserver-xorg-core   2:1.7.4-2
  xserver-xorg-video-ati  1:6.12.4-2
  xserver-xorg-video-mach64   6.8.2-2
  xserver-xorg-video-r128 6.8.1-2
  xserver-xorg-video-radeon   1:6.12.4-2

HTH

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#532616: Translation...

[Ulrich Eckhardt]

The title would translate to nv  nvidia driver: refresh rate from 60 to 50.

mein Problem: die Wiederholrate wird beim nv-Treiber auf 60Hz gesetzt und
bei dem nvidia-Treiber auf 50Hz gesetzt. das führt dazu, das die Zeichen
plötzlich verdoppelt werden.

My problem: The refresh rate is set to 60Hz with the 'nv' driver and to 50Hz 
with the 'nvidia' driver. This leads to the characters suddenly being doubled.


For the record, I can't tell what the reporter actually meant with characters 
being doubled, the description isn't clear in German either.

HTH

Uli



--
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#567616: Duplicate bug report?

[Ulrich Eckhardt]

Hi!

I think that

16 bit colour on M64 gives green tinted screen
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=565313

new update has problem using DefaultDepth 16 in xorgs config
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=567616

actually describe the same problem.

Uli




-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#566622: Wrong URL in manpage

[Ulrich Eckhardt]

Package: scrounge-ntfs
Version: 0.9-6

Simple patch:
 - http://memberwebs.com/swalter/software/scrounge/
 + http://memberwebs.com/stef/software/scrounge/

For the record, the URL on the package's page at Debian's is correct.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#563162: ownership problems give bogus error message

[Ulrich Eckhardt]

Package: cryptsetup
Version: 2:1.1.0~rc2-1

I tried activating an encrypted partition as normal user, and it gave
me this:

 u...@fruitbox:~$ cryptdisks_start hda5
 Starting crypto disk...hda5 (skipped, device /dev/hda5 does not 
 exist)...failed.
 u...@fruitbox:~$ ls /dev/hda
 hda   hda1  hda2  hda3  hda4  hda5  

As you see, the partition exists:

 u...@fruitbox:~$ cat /etc/crypttab 
 # target name source device key file  options
 hda5/dev/hda5   noneluks,noauto

And it works as root:

 r...@fruitbox:/home/uli# cryptdisks_start hda5
 Starting crypto disk...hda5 (starting)...
 Unlocking the disk /dev/hda5 (hda5)
 Enter passphrase: 
 Key slot 0 unlocked.
 hda5 (started)...done.

I guess the code is simply assuming ENOENT when open fails, but it
should also handle EACCESS and similar errors.

Cheers!

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#560126: Similar problems here...

[Ulrich Eckhardt]

Mine aren't as severe, but at least they are enough to crash my X server, see 
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=561331.

Further, also with a radeon card, I'm suffering a randomly corrupted cursor 
when moving between two windows ... I'll file a separate report unless already 
done by someone else.

Uli



-- 
To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Bug#506666: Can't reproduce here...

[Ulrich Eckhardt]

I tried the simple SDL_Init/SDL_Quit example from [1] and it doesn't crash. I 
also tried running tuxpaint, not crashing my machine either. The system is a 
current testing system, also on a Mac Mini G4.

Attached is the xorg.conf file, notable differences are:
 - The whole Module section.
 - In the Device section the MacModel option, I'm not sure if this is 
important though.
 - In the Device section the IgnoreEDID option. My monitor actually 
reports more (60Hz) than it is really capable of (58Hz), so I have to program 
the timing manually.
 - I'm running at a display depth of 16 bit instead of 8.

Actually, what depth are you running? I assumed it would use the first one, 
i.e. 8BPP in your case, is that true? However, I think this is mostly a mode 
for legacy code, as all others now use some kind of RGB encoding. I would 
definitely suggest trying 15/16/24 bits, too, that might already fix your 
problem.

Uli

[1] http://lazyfoo.net/SDL_tutorials/lesson01/linux/cli/index.php
Section Files
FontPath/usr/share/fonts/X11/misc
FontPath/usr/X11R6/lib/X11/fonts/misc
FontPath/usr/share/fonts/X11/cyrillic
FontPath/usr/X11R6/lib/X11/fonts/cyrillic
FontPath/usr/share/fonts/X11/100dpi/:unscaled
FontPath/usr/X11R6/lib/X11/fonts/100dpi/:unscaled
FontPath/usr/share/fonts/X11/75dpi/:unscaled
FontPath/usr/X11R6/lib/X11/fonts/75dpi/:unscaled
FontPath/usr/share/fonts/X11/Type1
FontPath/usr/X11R6/lib/X11/fonts/Type1
FontPath/usr/share/fonts/X11/100dpi
FontPath/usr/X11R6/lib/X11/fonts/100dpi
FontPath/usr/share/fonts/X11/75dpi
FontPath/usr/X11R6/lib/X11/fonts/75dpi
# path to defoma fonts
FontPath/var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType
EndSection

Section Module
Loadbitmap
Loaddbe
Loadddc
Loaddri
Loadextmod
Loadglx
Loadint10
Loadvbe
EndSection

Section InputDevice
Identifier  Generic Keyboard
Driver  kbd
Option  CoreKeyboard
Option  XkbRules  xorg
Option  XkbModel  pc104
Option  XkbLayout dvorak
Option  XkbVariantdvorak
Option  XkbOptionslv3:ralt_switch
EndSection

Section InputDevice
Identifier  Configured Mouse
Driver  mouse
Option  CorePointer
Option  Device/dev/input/mice
EndSection

Section Device
Identifier  Builtin Radeon
Driver  radeon
# select model with integrated DVI adapter
Option  MacModel mini-internal
# ignore monitor mode lines received via DDC but use DDC to detect the 
monitor itself
Option  IgnoreEDID  true
Option AccelMethod XAA
EndSection

Section Monitor
Identifier  SyncMaster 225BW on VGA
Option  DPMS
EndSection

Section Monitor
Identifier SyncMaster 225BW on DVI
ModelName SyncMaster

# 1680x1050 57.96 Hz (CVT) hsync: 63.06 kHz; pclk: 141.25 MHz
Modeline 1680x1...@58  141.25  1680 1784 1960 2240  1050 1053 1059 
1088 -hsync +vsync
Modeline 1680x1...@57  140.25  1680 1784 1960 2240  1050 1053 1059 
1088 -hsync +vsync

Option  PreferredMode 1680x1...@57
EndSection


Section Screen
Identifier  Default Screen
Device  Builtin Radeon
Monitor SyncMaster 225BW on DVI
DefaultDepth16
SubSection Display
Depth   1
Modes   1680x1050 1280x1024 1024x768 800x600 
640x480
EndSubSection
SubSection Display
Depth   4
Modes   1680x1050 1280x1024 1024x768 800x600 
640x480
EndSubSection
SubSection Display
Depth   8
Modes   1680x1050 1280x1024 1024x768 800x600 
640x480
EndSubSection
SubSection Display
Depth   15
Modes   1680x1050 1280x1024 1024x768 800x600 
640x480
EndSubSection
SubSection Display
Depth   16
Modes   1680x1050 1280x1024 1024x768 800x600 
640x480
EndSubSection
SubSection Display
Depth   24
Modes   1680x1050 1280x1024 1024x768 800x600 
640x480
EndSubSection
EndSection

Section ServerLayout
Identifier  Default Layout
Screen  Default Screen
InputDevice Generic Keyboard
InputDevice Configured Mouse
EndSection

Section DRI
Mode0666
EndSection

Bug#562971: Alternative patch with ls-like '--si' option

[Ulrich Eckhardt]

Well, I think all involved parties made clear what their position on that 
matter is. I hope no hard feelings remain.

That said, behave like GNU ls as an explicit goal is a good idea regardless. 
With that in mind, I just hacked together exactly that, see attached patch.

Uli
diff -upwr tree-1.5.3.orig/man/tree.1 tree-1.5.3/man/tree.1
--- tree-1.5.3.orig/man/tree.1	2009-12-31 15:54:48.0 +0100
+++ tree-1.5.3/man/tree.1	2009-12-31 16:27:24.0 +0100
@@ -115,7 +115,13 @@ Print the size of each file in bytes alo
 .B -h
 Print the size of each file but in a more human readable way, e.g. appending a
 size letter for kilobytes (K), megabytes (M), gigabytes (G), terrabytes (T),
-petabytes (P) and exabytes (E).
+petabytes (P) and exabytes (E). These are similar to how 'ls' works, i.e. using
+a base of 1024.
+.PP
+.TP
+.B --si
+Like '-h', print the size of each file in a more human readable way, but using
+SI prefixes and a base of 1000.
 .PP
 .TP
 .B -u
@@ -131,16 +137,16 @@ Print the date of the last modification 
 .PP
 .TP
 .B --inodes
-Prints the inode number of the file or directory
+Prints the inode number of the file or directory.
 .PP
 .TP
 .B --device
-Prints the device number to which the file or directory belongs
+Prints the device number to which the file or directory belongs.
 .PP
 .TP
 .B -F
 Append a `/' for directories, a `=' for socket files, a `*' for executable files
-and a `|' for FIFO's, as per ls -F
+and a `|' for FIFO's, as per ls -F.
 .PP
 .TP
 .B -q
Only in tree-1.5.3.orig: strverscmp.c.orig
Only in tree-1.5.3: tree
diff -upwr tree-1.5.3.orig/tree.c tree-1.5.3/tree.c
--- tree-1.5.3.orig/tree.c	2009-12-31 15:54:48.0 +0100
+++ tree-1.5.3/tree.c	2009-12-31 16:08:24.0 +0100
@@ -182,7 +182,7 @@ int strverscmp (const char *s1, const ch
 
 /* Globals */
 int dflag, lflag, pflag, sflag, Fflag, aflag, fflag, uflag, gflag;
-int qflag, Nflag, Dflag, inodeflag, devflag, hflag;
+int qflag, Nflag, Dflag, inodeflag, devflag, hflag, siflag;
 int noindent, force_color, nocolor, xdev, noreport, nolinks, flimit;
 char *pattern = NULL, *ipattern = NULL, *host = NULL, *title = Directory Tree, *sp =  ;
 const char *charset=NULL;
@@ -206,7 +206,7 @@ int main(int argc, char **argv)
 
   q = p = dtotal = ftotal = 0;
   aflag = dflag = fflag = lflag = pflag = sflag = Fflag = uflag = gflag = FALSE;
-  Dflag = qflag = Nflag = Hflag = Rflag = hflag = FALSE;
+  Dflag = qflag = Nflag = Hflag = Rflag = hflag = siflag = FALSE;
   noindent = force_color = nocolor = xdev = noreport = nolinks = FALSE;
   inodeflag = devflag = FALSE;
   flimit = 0;
@@ -390,6 +390,12 @@ int main(int argc, char **argv)
 	  cmpfunc = dirsfirstsort;
 	  break;
 	}
+	if (!strcmp(--si,argv[i])) {
+	  j = strlen(argv[i])-1;
+	  siflag = TRUE;
+	  sflag = TRUE; /* Assume they also want -s */
+	  break;
+	}
 	if (!strncmp(--filelimit,argv[i],11)) {
 	  j = 11;
 	  if (*(argv[i]+11) == '=') {
@@ -1293,14 +1299,18 @@ void printit(char *s)
 void psize(char *buf, off_t size)
 {
   char *unit=BKMGTPEZY;
+  char *si_unit= kMGTPEZY;
   int idx;
 
-  if (!hflag) sprintf(buf, sizeof(off_t) == sizeof(long long)?  %11lld :  %9ld, size);
-  else {
+  if (hflag) {
 for (idx=size1024?0:1; size = (1024*1024); idx++,size=10);
 if (!idx) sprintf(buf,  %4d, (int)size);
 else sprintf(buf, ((size10) = 10)?  %3.0f%c :  %3.1f%c, (float)size/(float)1024,unit[idx]);
-  }
+  } else if (siflag) {
+for (idx=size1000?0:1; size = (100); idx++,size/=1000);
+if (!idx) sprintf(buf,  %4d, (int)size);
+else sprintf(buf, (size = 1)?  %3.0f%c :  %3.1f%c, (float)size/1000.0f,si_unit[idx]);
+  } else sprintf(buf, sizeof(off_t) == sizeof(long long)?  %11lld :  %9ld, size);
 }
 
 void html_encode(FILE *fd, char *s)
Only in tree-1.5.3: tree.c~
Only in tree-1.5.3: tree.o

Bug#562971: Wrong unit when using -h [patch]

[Ulrich Eckhardt]

Package: tree
Version: 1.5.2.2-1

Actually there are several aspects to this bug, which is also related 
to #556186[1]:

1. The prefix for thousand is k, not K, all others are correctly 
capitalized.

2. The prefix k means one thousand. Abusing it for 1024 is just that - 
abuse[2]. Note that the manpage 'documented' this bug using the 
offensive wording reported in #556186. This IMHO doubly asinine phrase 
was removed in one of the newer releases, but without replacing it 
with some honest explanation of the prefix abuse.

The attached patch fixes both issues, the latter by using the correct 
meaning of the SI prefixes and not by documenting the bug in the 
manpage. It was made against the pristine upstream 1.5.3 sources.

Cheers!

Uli

[1] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=556186
[2] I'm aware that some people disagree, this is kind of a religious 
matter.
--- tree.c.orig	2009-12-29 17:13:47.0 +0100
+++ tree.c	2009-12-29 18:09:36.0 +0100
@@ -1292,14 +1292,14 @@ void printit(char *s)
 
 void psize(char *buf, off_t size)
 {
-  char *unit=BKMGTPEZY;
+  char *unit= kMGTPEZY;
   int idx;
 
   if (!hflag) sprintf(buf, sizeof(off_t) == sizeof(long long)?  %11lld :  %9ld, size);
   else {
-for (idx=size1024?0:1; size = (1024*1024); idx++,size=10);
+for (idx=size1000?0:1; size = 100; idx++,size/=1000);
 if (!idx) sprintf(buf,  %4d, (int)size);
-else sprintf(buf, ((size10) = 10)?  %3.0f%c :  %3.1f%c, (float)size/(float)1024,unit[idx]);
+else sprintf(buf, (size=1) ?  %3.0f%c :  %3.1f%c, size/1000.0f,unit[idx]);
   }
 }