BerkeleyDB and config files
by Michael A. Peters
I'd like to suggest a change to the way Fedora Core does some things.
I'm not talking about "standard" linux config files like /etc/passwd or /
etc/group or /etc/fstab (though the latter would be nice) - I'm talking
about Fedora (and RH I suppose since Fedora has become a RH test bed)
specific configuration files, such as /etc/security/console.perms and /
etc/sysconfig/network-scripts/ifcfg-[iface] type stuff.
Rather than store them in flat files, which have different
configuration parameters depending upon the file etc., use (maybe
optionally) BerkeleyDB.
bdb is already needed for the rpm database, so a Fedora system will
have bdb installed.
Both Python and Perl (as well as many other languages) already have
good bdb interfaces, and both Python and Perl also have gtk+ bindings
too.
If an embedded database was available for storing the configuration
information, it would be simpler for packagers to script additions to
those files, if necessary - IE the rpm for TiLP (software for talking
to a Ti graphing calculator) could easily add (if needed) a <ticalc>
class to console.perms with the necessary device node entries, and then
add the needed permission definitions for the device class.
It also would allow gui administration for some configuration nodes for
which there is not a gui. Kind of like how I can change some things in
gnome using gconf-editor for which there isn't an existing control
panel ready made for the task.
It also would make it easier for network administrators who set up box
after box with needs to customize some of the default configurations to
do so - after the install, a single documented script makes the needed
entries/changes to the database and it's done.
19 years, 7 months
CUPS and gnome
by Gérard Milmeister
I recently bought an Epson 1290 Photo printer, which has a lot of
options in CUPS. However, apart from GIMP (which is GTK2, not gnome),
gnome programs don't offer any of these options in their print dialogs.
So they have to be set externally. There are several possibilities:
1. Using system-config-printer or localhost:631, both of which require
root privileges.
2. Using lpoptions.
3. Using kprinter.
Clearly option 3 is the most user-friendly. However kprinter is a KDE
program, so it takes time to start in a gnome environment.
Are there plans to implement a gnome analog of kprinter and integrate
complete CUPS support for gnome programs?
--
Gérard Milmeister
Langackerstrasse 49
8057 Zürich
gemi(a)bluewin.ch
19 years, 7 months
rawhide report: 20041016 changes
by Build System
Updated Packages:
HelixPlayer-1.0.1.gold-6
------------------------
* Fri Oct 15 2004 Colin Walters <walters(a)redhat.com> 1:1.0.1.gold-6
- Require libbluecurve.so to fix 135709, since it apparently
cannot be done in comps at this point.
* Fri Oct 15 2004 Colin Walters <walters(a)redhat.com> 1:1.0.1.gold-5
- Ok, we can not Requires: redhat-artwork.i386. Reassigning
135709 to comps.
* Thu Oct 14 2004 Colin Walters <walters(a)redhat.com> 1:1.0.1.gold-4
- Pull in 32-bit redhat-artwork on x86_64 (135709)
- Rename hxplay.desktop to realplay.desktop, in order to set
default mime handlers for realplay.desktop, until we get
multiple application support in defaults.list (135705)
NetworkManager-0.3-1
--------------------
* Fri Oct 15 2004 Dan Williams <dcbw(a)redhat.com> 0.3-1
- Update from CVS, version 0.3
arts-1.3.0-4
------------
* Wed Sep 15 2004 Than Ngo <than(a)redhat.com> 1.3.0-4
- fix multilib problem #132576
* Fri Sep 03 2004 Colin Walters <walters(a)redhat.com> 1.3.0-3
- Fix trivial conflict in glib patch and reapply; this
removes dependency on glib2.0 and therefore prevents
symbol clashes with applications loading arts still
using glib 1.2, like XMMS.
* Mon Aug 23 2004 Than Ngo <than(a)redhat.com> 1.3.0-2
- add missing epoch tag
authconfig-4.6.5-2
------------------
* Fri Oct 15 2004 Tomas Mraz <tmraz(a)redhat.com>
- force restart of autofs on firstboot call when using NIS (#133035, #124498)
* Thu Oct 07 2004 Tomas Mraz <tmraz(a)redhat.com>
- require python to install (#134654)
authd-1.4.2-8
-------------
* Fri Oct 15 2004 Adrian Havill <havill(a)redhat.com> - 1.4.2-8
- tweak setting of uid/gid for key file so systems with no prior
ident user/group don't generate a warning (#135837)
autofs-4.1.3-28
---------------
* Fri Oct 15 2004 Chris Feist <cfeist(a)redhat.com> - 1:4.1.3-28
- Fixed a bug which caused directories to never be unmounted.
* Thu Oct 14 2004 Chris Feist <cfeist(a)redhat.com> - 1:4.1.3-27
- Fixed an error in the init script which caused duplicate entries to be
displayed when asking for autofs status.
desktop-printing-0.17-2
-----------------------
* Fri Oct 15 2004 Colin Walters <walters(a)redhat.com> 0.17-2
- Backport patches from CVS:
- eggcups-int-gsize.patch (fixes type conflict on x86_64)
- eggcups-no-help.patch (rh bug 135919)
eel2-2.8.1-2
------------
* Fri Oct 15 2004 Alexander Larsson <alexl(a)redhat.com> - 2.8.1-2
- Fix crash on extensionless files (#135858)
firstboot-1.3.31-1
------------------
* Fri Oct 15 2004 Adrian Likins <alikins(a)redhat.com> - 1.3.30-1
- merge some updates from rhel3 branch
- try enabling tui stuff again
gnome-panel-2.8.1-2
-------------------
* Fri Oct 15 2004 Matthias Clasen <mclasen(a)redhat.com>
- Make dropping non-ASCII uris work. (#135874)
gnome-pilot-2.0.12-4
--------------------
* Fri Oct 15 2004 David Malcolm <dmalcolm(a)redhat.com> - 2.0.12-4
- use a higher-res icon for the panel applet (bz #135897)
gpdf-2.8.0-4
------------
* Fri Oct 15 2004 Marco Pesenti Gritti <mpg(a)redhat.com> 2.8.0-4
- rebuilt
hal-0.4.0-3
-----------
* Fri Oct 15 2004 David Zeuthen <davidz(a)redhat.com> 0.4.0-3
- Fix bad use of O_NONBLOCK as the 2.6.8-1.624 kernel exposes this (#135886)
- Never use the UUID as mount point candidate in the default policy
as it is unfriendly (#135907)
- Fix a trivial bug in fstab-sync so the syslog messages actually expose
the device name instead of just the word foo
httpd-2.0.52-3
--------------
* Tue Sep 28 2004 Joe Orton <jorton(a)redhat.com> 2.0.52-3
- add dummy connection address fixes from HEAD
- mod_ssl: add security fix for CAN-2004-0885
im-sdk-12.0.1-17.svn1994
------------------------
* Fri Oct 15 2004 Akira TAGOH <tagoh(a)redhat.com> - 1:12.0.1-17.svn1994
- iiimsf-fix-missing-sockdir.patch: fixed not working at all
if /var/run/iiim/.iiimp-unix or whatever is missing. (#135750)
- iiimsf-improve-segv-log.patch: improved the log messages. (Warren Togami)
kdelibs-3.3.0-5
---------------
* Wed Sep 29 2004 Than Ngo <than(a)redhat.com> 6:3.3.0-5
- add missing requires on libidn-devel
* Sun Sep 26 2004 Than Ngo <than(a)redhat.com> 6:3.3.0-4
- cleanup menu
* Mon Sep 20 2004 Than Ngo <than(a)redhat.com> 3.3.0-3
- fix a bug in ksslopen #114835
mailman-2.1.5-25
----------------
* Thu Oct 14 2004 John Dennis <jdennis(a)redhat.com> 3:2.1.5-24
- more FHS changes, matches with new SELinux security policy
* Wed Sep 29 2004 John Dennis <jdennis(a)redhat.com> 3:2.1.5-21
- move list data dir to /var/lib/mailman to conform to FHS
move lock dir to /var/lock/mailman to conform to FHS
move config dir (VAR_PREFIX/data) to /etc/mailman to conform to FHS
Thanks to Matt Domsch for pointing this out.
mc-4.6.1-0.7
------------
* Fri Oct 15 2004 Jindrich Novy <jnovy(a)redhat.com> 4.6.4-0.7
- update from CVS
- sync strippwd patch with upstream
- merged hp48.in patch to extfs patch (from Leonard den Ottolander)
- rebuilt
* Fri Oct 08 2004 Jindrich Novy <jnovy(a)redhat.com> 4.6.1-0.6
- update from CVS
- drop upstreamed vcsa and xtermaliases patches
- sync the rest of the patches with upstream
- update strippwd patch to eliminate passwords from dir hotlist
and chdir error messages
- update perl scripts (Leonard den Ottolander, #127973, CAN-2004-0494)
mkinitrd-4.1.17-1
-----------------
* Fri Oct 15 2004 Jeremy Katz <katzj(a)redhat.com> - 4.1.17-1
- run udevstart again instead of sleeping, this will ensure all
devices are created in all cases after modules have been
loaded (thanks to notting)
* Fri Oct 15 2004 Jeremy Katz <katzj(a)redhat.com> - 4.1.16-1
- mkinitrd: Sleep briefly to let udev finish creating devices
nautilus-2.8.1-3
----------------
* Fri Oct 15 2004 Alexander Larsson <alexl(a)redhat.com> - 2.8.1-3
- Slightly less bad error dialog when there is no handler for a file.
Not ideal, but this change doesn't change any strings.
* Tue Oct 12 2004 Alexander Larsson <alexl(a)redhat.com> - 2.8.1-2
- Fix open with menu on mime mismatch
- Create desktop links ending with .desktop (#125104)
- Remove old cruft from specfile
* Mon Oct 11 2004 Alexander Larsson <alexl(a)redhat.com> - 2.8.1-1
- update to 2.8.1
net-snmp-5.1.2-11
-----------------
* Fri Oct 15 2004 Radek Vokal <rvokal(a)redhat.com> 5.1.2-11
- Logrotate support added (#125004)
* Thu Oct 14 2004 Phil Knirsch <pknirsch(a)redhat.com> 5.1.2-10
- Extended the libwrap and bsdcompat patches
* Mon Oct 11 2004 Phil Knirsch <pknirsch(a)redhat.com> 5.1.2-9
- Droped obsolete lm-sensors patch and enabled lmSensors module
- Marked several patches to be removed for 5.1.3
newt-0.51.6-5
-------------
* Fri Oct 15 2004 Adrian Havill <havill(a)redhat.com> 0.51.6-5
- only do gpmclose if gpmopen succeeed (#118530)
oprofile-0.8.1-11
-----------------
* Fri Oct 15 2004 Will Cohen <wcohen(a)redhat.com>
- Additional ppc64 support for ppc64/970.
parted-1.6.15-4
---------------
* Fri Oct 15 2004 Phil Knirsch <pknirsch(a)redhat.com> 1.6.15-4
- Fixed dasd patch (had some duplicate file patches in it)
- Fixed problem with parted segfaulting on SCSI discs on s390 (#133997)
postfix-2.1.5-2.1
-----------------
* Fri Oct 15 2004 Thomas Woerner <twoerner(a)redhat.com> 2:2.1.5-2.1
- removed aliases from postfix-files (#135840)
- fixed postalias call in init script
rhgb-0.14.1-1
-------------
* Wed Oct 13 2004 Daniel Veillard <veillard(a)redhat.com> 0.14.1
- try to fix the DRI problem in conjunction with x.org update
- update languages
- fix for #134742
rhn-applet-2.1.15-2
-------------------
* Tue Oct 12 2004 Adrian Likins <alikins(a)redhat.com> 2.1.14-1
- fix #135392
- fix #135092
* Fri Oct 08 2004 Adrian Likins <alikins(a)redhat.com> 2.1.13-1
- the last change broke normal remind cases #135092
* Tue Oct 05 2004 Adrian Likins <alikins(a)redhat.com> 2.1.12-1
- be a bit more accurate about folks we remind to
register, this time including registered, but
not currently entitled systems
- add server capabilities support to the applet
rp-pppoe-3.5-22
---------------
* Fri Oct 15 2004 Than Ngo <than(a)redhat.com> 3.5-22
- Fix ip conflict in dsl connect, #135012
rpm-4.3.2-13
------------
* Fri Oct 15 2004 Jeff Johnson <jbj(a)redhat.com> 4.3.2-13
- fix: don't set handler if SIG_IGN is already set (#134474).
rpmdb-fedora-2.92-0.20041016
----------------------------
shadow-utils-4.0.3-35
---------------------
* Fri Oct 15 2004 Adrian Havill <havill(a)redhat.com> 2:4.0.3-35
- make the limit for the group name the same as the username (determined
by the header files, rather than a constant) (#56850)
squirrelmail-1.4.3a-5
---------------------
* Thu Oct 14 2004 Warren Togami <wtogami(a)redhat.com> 1.4.3a-5
- default_folder_prefix dovecot compatible by default
/etc/squirrelmail/config_local.php if you must change it
* Wed Oct 13 2004 Warren Togami <wtogami(a)redhat.com> 1.4.3a-4
- HIGASHIYAMA Masato's patch to improve Japanese support
(coordinated by Scott A. Hughes).
- real 1.4.3a tarball
system-config-securitylevel-1.4.12-1
------------------------------------
* Fri Oct 15 2004 Dan Walsh <dwalsh(a)redhat.com> 1.4.12-1
- Fix crash on badly formed /etc/selinux/config files
tcl-8.4.7-2
-----------
* Fri Oct 15 2004 Jens Petersen <petersen(a)redhat.com> - 8.4.7-2
- improve tcl8.3.5-tclConfig-package-path-90160.patch to look in libdir in
addition to datadir for packages, so that tclsh can load binary packages
in lib64 (135310)
* Fri Jul 30 2004 Jens Petersen <petersen(a)redhat.com> - 8.4.7-1
- update to 8.4.7
- replace tcl-8.4.5-no_rpath.patch by tcl-8.4-no_rpath.patch
- replace tcl-8.4.5-autoconf.patch by tcl-8.4-autoconf.patch
- no longer obsolete itcl
* Tue Jun 15 2004 Elliot Lee <sopwith(a)redhat.com>
- rebuilt
udev-038-2
----------
* Fri Oct 15 2004 Harald Hoyer <harald(a)redhat.com> - 038-2
- par[0-9] is now a symlink to lp
- MAKEDEV the parport devices
- now conflicts with older initscripts
vixie-cron-4.1-19
-----------------
* Fri Oct 15 2004 Jason Vas Dias <jvdias(a)redhat.com> - 4.1-19
- crontab -e should only strip NHEADER_LINES comments
- (NHEADER_LINES==0), not at least one header comment line.
- (bug 135845)
* Sat Oct 09 2004 Florian La Roche <laroche(a)redhat.com> - 4.1-18
- no need to make user installed crontabs readable
* Thu Sep 30 2004 Jason Vas Dias <jvdias(a)redhat.com> - 4.1-17
- Users not allowed to use 'crontab mycrontab', while
- 'crontab < mycrontab' allowed; this is because misc.c's
- swap_uids_back() was not using save_euid / save_egid .
- Thanks to Mads Martin Joergensen <mmj(a)suse.de> for pointing this out.
vte-0.11.11-6
-------------
* Fri Oct 15 2004 Matthias Clasen <mclasen(a)redhat.com> 0.11.11-6
- Fix a crash with input methods. (#131226)
ypbind-1.17.2-3
---------------
* Fri Oct 15 2004 Steve Dickson <SteveD(a)RedHat.com>
- Sped up the ypbind initscript by using fgrep
instead of grep (bz# 81247)
19 years, 7 months
GNOME Desktop CD icon weirdness
by Ralf Ertzinger
Hi.
I'm asking here since I do not know against which component to file a bug.
I've disabled all automatic CD mounting/playing/ejecting/whatever features
in gnome. Nonetheless gnome puts a "Blank CD" icon on my desktop whenever I
insert a blank CD in a writer. OK, I can live with that.
I am using k3b to write the CDs. When k3b is finished (and closed), the CD
in the drive is no more a blank CD. Nonetheless the icon persists. I can
even mount the just written CD in the burner (without taking it out). The
mount will succeed, but the "blank CD" icon is still there. Normally mounting
a CD will cause a new icon to pop up on the desktop, but this does not happen
(gnome seems to know that there already is an icon for this drive). Clicking
on the "blank CD" icon reveals an empty window, offering me to add files to
the no longer existant blank disk.s
Removing the disk from the drive causes the icon to disappear.
Where do I file this?
--
/* Sun, you just can't beat me, you just can't. Stop trying,
* give up. I'm serious, I am going to kick the living shit
* out of you, game over, lights out.
*/ -- linux-2.4.21-pre1/arch/sparc/lib/checksum.S (2.4.21)
19 years, 7 months
udev strangeness with latest rawhide
by Russell Coker
Running the latest rawhide I get AVC messages indicating that /bin/udev
(not /sbin/udev) is running in kernel_t during the early stages of system
boot.
/bin/udev is the file name used in the initrd! So it seems that after the SE
Linux policy is loaded (IE after /sbin/init has been run from the main root
fs) there is still a copy of udev from the initrd being run. This seems to
be a bug in initrd that could lead to inconsistent behaviour. I'm not sure
how this comes about (and of course apart from SE Linux messages in the
kernel message log all the evidence is gone by the time the system is ready
to login).
Any suggestions on how to debug this?
--
http://www.coker.com.au/selinux/ My NSA Security Enhanced Linux packages
http://www.coker.com.au/bonnie++/ Bonnie++ hard drive benchmark
http://www.coker.com.au/postal/ Postal SMTP/POP benchmark
http://www.coker.com.au/~russell/ My home page
19 years, 7 months
Fedora Core 3 Bug Status - 2004-10-16
by Bill Nottingham
Based on bug #123268 ('FC3Target') and bug #130887 ('FC3Blocker')
2004-10-16
Severity Total Closed Need Testing
BLOCKER 122 96 ( 78.69 %) 12 ( 12.50 %)
TARGET 916 501 ( 54.69 %) 52 ( 10.38 %)
Overall 1038 597 ( 57.51 %) 64 ( 10.00 %)
2004-10-15
Severity Total Closed Need Testing
BLOCKER 120 93 ( 77.50 %) 11 ( 11.83 %)
TARGET 914 495 ( 54.16 %) 52 ( 10.51 %)
Overall 1034 588 ( 56.87 %) 63 ( 10.00 %)
2004-10-14
Severity Total Closed Need Testing
BLOCKER 117 90 ( 76.92 %) 16 ( 17.78 %)
TARGET 923 484 ( 52.44 %) 56 ( 11.57 %)
Overall 1040 574 ( 55.19 %) 72 ( 12.00 %)
2004-10-13
Severity Total Closed Need Testing
BLOCKER 115 85 ( 73.91 %) 13 ( 15.29 %)
TARGET 914 478 ( 52.30 %) 57 ( 11.92 %)
Overall 1029 563 ( 54.71 %) 70 ( 12.00 %)
2004-10-11
Severity Total Closed Need Testing
BLOCKER 105 78 ( 74.29 %) 10 ( 12.82 %)
TARGET 892 438 ( 49.10 %) 49 ( 11.19 %)
Overall 997 516 ( 51.76 %) 59 ( 11.00 %)
2004-09-30
Severity Total Closed Need Testing
BLOCKER 90 67 ( 74.44 %) 11 ( 16.42 %)
TARGET 831 374 ( 45.01 %) 45 ( 12.03 %)
Overall 921 441 ( 47.88 %) 56 ( 12.00 %)
2004-09-23
Severity Total Closed Need Testing
BLOCKER 74 43 ( 58.11 %) 9 ( 20.93 %)
TARGET 703 286 ( 40.68 %) 38 ( 13.29 %)
Overall 777 329 ( 42.34 %) 47 ( 14.00 %)
2004-09-08
Severity Total Closed Need Testing
BLOCKER 35 14 ( 40.00 %) 3 ( 21.43 %)
TARGET 591 194 ( 32.83 %) 33 ( 17.01 %)
Overall 626 208 ( 33.23 %) 36 ( 17.00 %)
2004-08-18
Severity Total Closed Need Testing
TARGET 415 61 ( 14.70 %) 16 ( 26.23 %)
19 years, 7 months
FC3T3: iiimf-libs-devel gets installed but iiimf-libs doesn't
by Louis Garcia
Just noticed while updating that iiimf-libs-devel is installed but not
iiimf-libs. Their are a few hanging symlinks in /usr/lib/im/lib.
Shouldn't the devel package depend on iiimf-libs?
Should I file a bug against im-sdk or anaconda?
--Louis
19 years, 7 months
gnupg 1.9
by Lars G
hi
any chance of updating gnupg to 1.9 in fedora core?
problem: i'm using kmail, which needs gpg-agent,
but gnupg 1.2.6 only provides this via the
package newpg from fedora extras, which is obsoleted.
L
19 years, 7 months
prelink/ExecShield/PIE interactions?
by James Ralston
I'm confused as to how prelink and ExecShield interact, and
unfortunately, neither of these two docs really clarifies the issue:
http://www.redhat.com/f/pdf/rhel/WHP0006US_Execshield.pdf
http://people.redhat.com/drepper/nonselsec.pdf
The nonselsec.pdf document states that prelinking and load address
randomization exclude each other. Prelink does not appear to conflict
with either randomizing the stack location or making the stack
non-executable.
But if that's the case, then what exactly does the --no-exec-shield
option to prelink actually *do*? What functionality of ExecShield
(that it doesn't already conflict with) is left for it to disable?
The reason I'm curious is due to dealing with programs that load
Windows DLLs and want to load them at specific addresses. Here's what
mplayer recommends as a work-around:
http://www.mplayerhq.hu/DOCS/HTML/en/faq.html#id2895374
If I compare prelink.log files with and without the --no-exec-shield
option, it's obvious that --no-exec-shield does in fact free up
address 0x400000:
--- prelink.log.old 2004-10-15 04:03:35.000000000 -0400
+++ prelink.log 2004-10-15 17:38:15.549151696 -0400
@@ -1 +1 @@
-/usr/sbin/prelink -av -mR -q
+/usr/sbin/prelink -av -mR --no-exec-shield -q
@@ -12,2 +12,2 @@
-Laying out 1 libraries in virtual address space 00101000-50000000
-Random base 0x3d5ee000
+Laying out 663 libraries in virtual address space 41000000-50000000
+Random base 0x4d93c000
But I still don't understand why --no-exec-shield has that effect, or
what I actually loose by using it.
The nonselsec.pdf document also states that PIEs are excluded from
prelinking. That makes sense. But if code segments aren't writable
(and I don't believe they are, IIRC), what benefit does PIE actually
bring? Is it trying to make it harder for an attacker to leverage
pre-existing code in the application to launch an exploit?
TIA for any insight...
19 years, 7 months