KNOPPIX v3.2 2003-04-18

**LinuxSam** · 04-20-2003, 11:12 PM

Since only moderators can start new topics I write it here...

* V3.2-2003-04-18 (more updates)
- upgrade to current testing/unstable snapshot of Debian
- OpenOffice 1.0.3.1 (custom printer fixed)
- XFree4.3 updated
- abiword 1.1
- updated mysql to version 4
- Removed fonts which seemed to be duplicates:
xfonts-intl-european, xfonts-intl-chinese (fonts already
in xfree 4.3 base fonts?), xfonts-cyrillic (replaced by xfonts-bolkhov)
- set default fontsize in KDE-konsole to large font (commonly used)
- configure-xawtv update (Fabian Franz and Kai Lahmann)
- Added hpijs for better HP inkjet printer support (suggested by Till Kamppeter)
- Added: speex
- gaim update, removed psi in favor of gaim's new ICQ and Jabber plugins
- Removed xsmbrowser in favour of linneighborhood

/LinuxSam

**Viro** · 04-21-2003, 10:41 AM

What updates were applied to XFree86? Anyone know?

**Sigmatador** · 04-21-2003, 11:59 AM

XFree86 4.3.99.2 (10 April 2003)
88. Fix file descriptor leaks in xman (Ted Unangst, OpenBSD PR#3186).
87. Fix bug in mode validation that occurs when the XF86Config doesn't
specify any mode or virtual resolution information (Marc La France).
86. Resetting 'bound' flag for an agp entry after undbind succeeded in
drm_agpsupport.h (Egbert Eich).
85. Ignoring hw_lock for drm device if lock was set by a different instance
(ie Xserver) to prevent second server to spin in driver release function
(currently only relevant for i8xx drm drivers) (David Dawes).
84. Adding extra delay before doing a slowBcopy in ia64 to prevent a MC
when saving/restoring VGA fonts (Egbert Eich).
83. Turn off needToSync in LeaveVT() in i830 driver to prevent server
crashes because xaa tries to sync blitter while switched away
(Egbert Eich).
82. Added the O_SYNC flag when opening /dev/mem to work around a kernel bug
which causes conflits between WC settings in MTRR registers and cacheable
attribute in page table (Egbert Eich).
81. Reduced retry count for mouse configuration to 4. This should speed
up server start for people who use a mouse repeater like gpm
(Egbert Eich).
80. Fixed xf86cfg to duplicate font names before they get freed when
unloading the module (Egbert Eich).
79. Added missing functions to xf86cfg (Egbert Eich).
78. Added missing ROP_NEEDS_SOURCE to Savage driver;
now all 2D accel functions should use this flag (Egbert Eich).
77. Reduced flicker with video playback in Savage driver (Tim Roberts).
76. Load vgahw submodule in RENDITION driver before using it
(Egbert Eich).
75. Magically speed up video playback on Neomagic (Egbert Eich).
74. Added drmAgpUnbind()/drmAgpBind() functionality on Leave/Enter for i810
cards to allow a second Xserver to be started if the first one has
DRI enabled (Egbert Eich).
73. Set NeedRingBufferLow for i855/i865 to avoid lockups on vt switches
(Egbert Eich).
72. Check for vtSema before accessing registers in I830BIOSSaveScreen()
(Egbert Eich).
71. Some sanity fixes for atiprobe to work better with xf86cfg: Handle
situation gracefully if the PCI config pointer doesn't exist
(Egbert Eich).
70. Changed behavior of fontfile: don't drop the entire directory if some
fonts cannot be rendered (Egbert Eich).
69. Fixed B&H bdf fonts: AVERAGE_WIDTH is not a string but an int (Mike Fabian).
68. Fixed type in radeon driver which caused display to go into
powersaving mode when Xcursor is disabled (Bugz: 90, Michel Dänzer).
67. Fixed typo in XKB-Config.sgml (Bugz: 53, Hiroyuki Bessho).
66. Fix SiliconMotion driver for mode switching and SEGV problem when
initializing Xv functionality (Bugzilla #50, Alan Hourihane).
65. Add Radeon DRI suspend/resume support (Charl Botha, #A.1431).
64. Rage 128 Xv fix for high resolution displays (Rik Faith).
63. PPC updates for Radeon (Michel Dänzer, Hui Yu@ATI).
62. Radeon patch from ATI (Hui Yu@ATI).
- IGP320/330/340 support (2D only)
- RV280 (9200/M9+) support
- Fix for some M9 laptop panels
- Improved version of monitor detection code previously submitted
- Fixed bug for two or more radeon cards
- Man page updates
- Workaround for double scan modes problem at high resolutions
- Overlay scaling problem when RMX is used
61. Fix xprop to get all windows (not only the first) in a property holding
a window list printed (#A.1676, Kim Woelders).
60. Fix for spontaneous repeated keyboard events during sync grab (#A.1713,
Michal Maruska).
59. Fixed DPMS problem on C&T 69000 due to incorrect LCD flag
(Bugzilla #101, Egbert Eich).
58. Added xkb symbols layout for BTC 5090 internet keyboard
(Bugzilla #57, Jack Angel).
57. Fix the Xinstall.sh script so that it can handle being run from a
directory with white space in its path name (#A.1690, Eric Branlund).
56. Allow AGPGART support to be enabled for OpenBSD (#A.1684, Brian Feldman).
55. Make the mouse driver check the SendDragEvents option (based on #A.1682,
Juuso Åberg).
54. Fix a double allocation in _XTextPropertyToTextList() (#A.1681,
Lubos Lunak).
53. Fix a typo that causes the 'yu' keymap to emit a lower case 'L' in
both shift states (#A.1675, Nikola Smolenski).
52. Allow the remote shell command for xon to be specified from the
command line. This allows ssh to be used (for example) (#A.1657,
Mike Urban).
51. Fix segfaults that can happen when using composition of RENDER and
non-RENDER fonts (#A.1656, Wu Jian Feng).
50. Add convenient ways to enter "double quote" and "cedilla" with the
en_US.UTF-8 compose rules (Alexandre Oliva).
49. Improve support for Wacom Cintiq boards (#A.1547, Sebastian Rittau).
48. Add support for Kensington's "ValuMouse" to the mouse driver (#A.1494,
Roy Wood).
47. Workaround for a crash that happens if a scalable font has a bitmap
entry in fonts.dir (#5687, Juliusz Chroboczek).
46. Single/dual rasterizer quiescence patch for the glint/gamma DRI driver
(#5685, Sven Luther).
45. Fix a FreeBSD/alpha build problem (#5679, Fred Clift).
44. Fix a bug in the GLU/project.3gl man page (#5676, Thomas Klausner).
43. Luit fixes:
- Fix a buffer overflow.
- Allow for holes in the pty space, and use the union of the pty
names in 4.4BSD and recent FreeBSD.
- Allow luit to be set-uid on all BSD releases, and introduce an
additional security fix on POSIX saved-id systems.
(#5672, 5674, ITO Tsuyoshi, Juliusz Chroboczek).
42. Fix memory leaks in ProcXF86VidModeModModeLine and
ProcXF86VidModeValidateModeLine, and unintialized fields of the mode
structure in ProcXF86VidModeAddModeLine, VidModeCreateMode, and
VidModeAddModeline (#5671, Miguel Freitas).
41. Fix SEGV in fbdev's mode validation (Alexandr Andreev).
40. Add two more variations of Thai XKB maps (#5658,
Theppitak Karoonboonyanan, Visanu Euarchukiati,
Supphachoke Santiwichaya).
39. Fix a twm crash when the window menu is used and no windows are open
(#5667, Matthias Scheler).
38. Restore the sign of UNDERLINE_POSITION in the FreeType backend (#5665,
Roman Kagan).
37. Check the return value of fclose() in xauth to prevent losing the
existing contents of an Xauthority file when there is insufficient
space to write the new one (#5663, Owen Taylor).
36. Fix case in xmag which would cause a BadMatch during a X_GetImage for
single child of root class InputOnly. Also do some null pointer
protection (#5657, Kevin Brosius).
35. Allow an arbitary message to be sent to the DDX by pressing a key
which has the appropriate action definition (#5519, Joe Moss).
34. Add a new request to the XF86Misc extension that allows a client
to send an arbitrary message to the DDX, which in turn can send the
message to the driver. The driver may also send a string back.
An example of this is implemented for the radeon driver (#5518,
Joe Moss).
33. Remove duplicate XShmGetEventBase() declaration (#5656,
Branden Robinson).
32. Fix drag lock when using a mouse button greater than 4
(Bugzilla #51, Paul Elliott)
31. Fix lockup on server shutdown/restart with the radeon driver
(Bugzilla #94, Michel Dänzer).
30. Add missing symbols to reference list that show up when DRI is not
loaded in r128 driver (Bugzilla #83, Leif Delgass).
29. Add missing symbols to reference list that show up when DRI is not
loaded in mga driver (Bugzilla #84, Leif Delgass).
28. Various build and warning fixes (#A.1703, Peter Breitenlohner).
27. Avoid overlapping strcpy() in imake.c (Dan Nelson).
26. Updates for building on OS/2 (#5650, Frank Giessler).

**Viro** · 04-21-2003, 12:42 PM

31. Fix lockup on server shutdown/restart with the radeon driver
(Bugzilla #94, Michel Dänzer).

YES!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Time to start using my Radeon card again

**sepht** · 04-21-2003, 06:21 PM

Would be great if my new audigy2 soundcard was supported.

**humdinger70** · 04-21-2003, 09:34 PM

How about fixing the hdd-install so that if your partitioning scheme creates several mount points (hda1 has /boot, hda2 has swap, hda3 has /, etc), the installer recoginizes this and uses them?

This, instead of lumping everything under one partition.

**eriixblaike** · 04-22-2003, 03:16 AM

Psi should be brought back. I have not used the Jabber plugin for Gaim, and was quite impressed. Until now, there hasn't even been a way to log in to alternative servers. This, however, does not mean they got it right. For all the things that can be done, there are some important ones that can't.

1) Server browsing is not implimented.
2) Even if you know the specifics of the agent you wish to register with, you can't.
3) Agents cannot be altered in any way (including removing an agent and the contacts associated with it).
4) The status of agents cannot be toggled (you cannot be logged in to Jabber, but have your agents offline, or decide that you want to get on ICQ at the moment, but want AIM off).
5) new account creation does not appear to exist.

There are more, but the point is it's fine to log in using Gaim for a little bit, but if a user wants to use any of the advanced features of Jabber, they're just out of luck. You can't even do searching for contacts with the JUD -- let alone searching through agents. Agent status isn't even displayed, so if something is wrong with an agent, you may not even know -- and you certianly can't do anything about it.

In short, it is a bad idea to remove the only true Jabber client from the distrib. (OK, many of these complaints are rendered moot for the live disk, since you're not expected to be doing anything with your account, but if it's _just_ a live disk, what is knx-hdinstall? Longer-term ideas are implimented, so the idea that one may wish to have some degree of control over their acct does not seem unreasonable (to me)).