Re: Reg: Error in apt-get install in libstdc++ package
Hi On Tue, 03 Sep 2013 22:08:51 +0200 Sven Joachim wrote: > Apparently, since there is no libstdc++ package, apt-get takes this as a > regular expression. The manpage does not really state that, although it > mentions regular expressions: Just a small nit - libstdc++ is a virtual package provided by libstdc++. One can see this by invoking: apt-cache policy libstdc++ apt-get tries to do what it is told - install all those (see apt-cache policy output) at once, and fails. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130904231822.e705ed452ed062ce50732...@gmail.com
Re: VPN failure
On Wed, 04 Sep 2013 13:03:22 -0400 Verde Denim wrote: Hi. > Any input is, as always, greatly appreciated. Would this be of any help? https://bugs.launchpad.net/ubuntu/+source/network-manager-vpnc/+bug/1107975 And you can also skip this unneeded layer of complexity called NetworkManager, and use vpnc directly. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130904232833.8c64bab1541e477e85812...@gmail.com
Re: Ownership of pluggable devices.
Hi. On Thu, 5 Sep 2013 11:19:25 -0700 peasth...@shaw.ca wrote: > /dev/KingstonUSB /home/peter/MY ext2 defaults,noauto,user 0 0 This line is the reason. ext2 filesystem stores information about file/directory permissions inside itself, and root of this filesystem (/home/peter/MY) is owned by root (uid=0 user). user=peter mount option allow this ordinary user to unmount filesystem, but has nothing in common with filesystem permissions. No amount of tinkering with mount options or device permissions will change filesystem permissions, you'll need to use chown(1) and chmod(1) for this. vfat filesystem (you use it for /home/peter/mylo) does not support file/directory permissions, so mount option user=peter also allows said user to do anything with filesystem contents. In short, invoke (after mounting the filesystem): chown -R peter /home/peter/MY it'll solve you problem. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130906003405.974f678a1e3a289a4e397...@gmail.com
Re: Ownership of pluggable devices.
On Thu, 5 Sep 2013 20:07:44 -0700 Gregory Nowak wrote: > On Fri, Sep 06, 2013 at 12:34:05AM +0400, recovery...@gmail.com wrote: > > In short, invoke (after mounting the filesystem): > > > > chown -R peter /home/peter/MY > > I do stand to be corrected, but I don't remember a situation where I > had to do chown like this after every mount. All I've ever had to do > in these situations is to mkdir /home/user/MY, mount /home/user/MY, > and chown user.user /home/user/MY. After each subsequent mount, user > could read/write /home/user/MY with no issues, and without using chown > every time after mount. > Hi. Take two different hosts, make user with the same name, but different uids on those hosts. Take USB stick, format it as ext2 (any filesystem keeping file permissions will do, actually). Now, move some files with this USB stick between hosts. This is exactly the situation that requires chown-after-mount unless you have root on both hosts. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130906192846.eb3f8209e39d4b01d6800...@gmail.com
Re: Re (2): Ownership of pluggable devices.
On Fri, 6 Sep 2013 08:28:23 -0700
peasth...@shaw.ca wrote:
> I've studied this note, installed udisks-glue and modified udisks-glue.conf
> as described.
> http://goshawknest.wordpress.com/2013/02/21/how-to-make-usb-disks-readable-by-all-users-on-raspbmc/
>
> Also noticed this.
> root@dalton:/etc/avahi/services# ps aux | grep "udisks-daemon: p"
> root 4177 0.0 0.0 6352 412 ?SSep05 0:07
> udisks-daemon:
> polling /dev/sr0 /dev/sdb
> /dev/sdb is the KingstonUSB; that is a good sign.
I'm not familiar with all these new fancy FreeDesktop gizmos. Still,
this part:
match disks {
automount = true
automount_options = { sync, noatime, "dmask=0", "fmask=0"}
Looks like mounting options (and, as I wrote before, tinkering with them
is useless), but this:
post_insertion_command = "udisks --set-spindown %
device_file --spindown-timeout 1800 --mount %device_file
--mount-options sync,noatime,dmask=0,fmask=0" }
Probably allows insertion of arbitrary command (maybe several), which
will be invoked with root privileges.
IMO, insert chown here, and you're set.
> Nevertheless, if root starts udisks-glue and then the
> KingstonUSB is plugged, access is again restricted to root as described
> previously.
> So I have at least one snag somewhere.
Hmm. In that blog they run udisks-glue with root privileges too.
> Once that is solved, I might invent an /etc/init.d/udisks-glue-start-script.
>
> Also,
> * From: Gregory Nowak
> * Date: Thu, 5 Sep 2013 20:07:44 -0700
> > ... don't remember a situation where I had to do chown like this after
> > every mount.
>
> Might this involve the replacement of hal with u*; still in progress.
YMMV, but I'd never use chown for this purpose. Besides, I haven't
meant you should run chown on each mount, running chown once should be
more than enough.
Reco
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/20130906204627.20634b96677ef922c935c...@gmail.com
Re: browser-plugin-libreoffice in wheezy
Hi On Sat, 07 Sep 2013 10:46:09 +0200 François Patte wrote: > Bonjour, > > How to install browser-plugin-libreoffice in wheezy: I get this message: > The following packages have unmet dependencies: > browser-plugin-libreoffice : Depends: uno-libs3 (>= 4.1.0~alpha) but > 3.5.4+dfsg2-0+deb7u2 is to be installed > E: Unable to correct problems, you have held broken packages. Apparently you took browser-plugin-libreoffice from backports repository, as said package is not included in main repository. This backported package depends on backported version of libreoffice and its libraries, so to install it, you'll need to execute something like this: apt-get install -t wheezy-backports browser-plugin-libreoffice ure > I was told that wheezy was a "stable" version... > They didn't lie to you - it is stable :) Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130907225104.45c74849ada249dc06940...@gmail.com
Re: Ownership of pluggable devices.
Hi.
On Sat, 7 Sep 2013 12:43:09 +1000
Zenaan Harkness wrote:
> On 9/7/13, Gregory Nowak wrote:
>
> Which filesystem to recommend for external USB portable drives, which
> move between 'random' hosts?
vfat, udf.
If you can stomach it, ntfs or exfat.
If moving said files is one-time activity, just tar needed files and
use USB stick as a raw device. Considering default block device
ownership for removable devices in Debian, no root is needed for this.
If you really need something more modern, ext4 without journal (beware
of file permissions).
> Perhaps fuseext2 would simplify this? I don't have experience with it yet.
man fuseext2 says that:
FUSEEXT2 options:
-o rw+
-o rw,force
enable read-write mount (EXPERIMENTAL).
'EXPERIMENTAL' is outlined with bold.
Still, fuseext2 worked for me in a short test on ext2. Didn't tried it
on anything else.
> The point is, a lot of us who would read this list, would be those who
> are able and/ or do install Debian for others.
>
> There should be an "easy for grandma" installation setup for the use
> ("by grandma") of externally attached/ portable drives.
But it is already "easy for grandma", as said devices are
pre-formatted at factory to vfat.
> Anyone know what The Right Way(TM)(R)(C) is?
> Or should be with some scripting/ configuration?
> Or could be with some hacking?
Choose one of these:
Using filesystem which doesn't support file permissions.
Not using filesystem at all.
Chown'ing on every mount.
Reco.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/20130907232632.8d2a305d41530a62da1bb...@gmail.com
Re: building from source for different arch
Hi. On Mon, 09 Sep 2013 23:20:34 +0200 Lucio Crusca wrote: > I'm trying to build opus package from source for both amd64 and i386, after > having applied the patch by Pino Toscano I've found here: > > $ dpkg-buildpackage -ai386 >From my limited experience in this area, invoking dpkg-buildpackage -a MAY create the package, but such package will have messed dependencies. Unless you intend to do something very strange (i.e. amd64.deb which contains i386 binaries), you'll probably better build package for i386 in a pure i386 chroot (easily made by deboostrap). Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130911230003.c7e076483231bb021c743...@gmail.com
Re: building from source for different arch
On Wed, 11 Sep 2013 21:03:47 +0200 Slavko wrote: > Hi, > > Dňa Wed, 11 Sep 2013 23:00:03 +0400 recovery...@gmail.com napísal: > > > Unless you intend to do something very strange (i.e. amd64.deb which > > contains i386 binaries), you'll probably better build package for i386 > > in a pure i386 chroot (easily made by deboostrap). > > Or in chroot created by the pbuilder, which is intented for packaging. > > regards Which, in turn, uses deboostrap :) But you're correct, of course. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130911233446.de5b40db4fc6ca04569b4...@gmail.com
Re: Security?
Hi. On Thu, 12 Sep 2013 18:44:42 +0200 Ralf Mardorf wrote: > On Thu, 2013-09-12 at 12:08 -0300, Beco wrote: > > Nice this startpage. I'm recommending to my students. > > I'm satisfied with it too :). We now have to trust > https://startpage.com/ by ixquick, that their claims are true. But I > never heard anything bad about ixquick. IMO startpage is better than > ixquick and duck duck go. TOR project seems to think that duckduckgo > startpage. Not that this 'enabled by default family filter' bothers me, but it can lead to unexpected results, as shown here: https://trac.torproject.org/projects/tor/ticket/8839#comment:7 Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130912232056.1660726e3752249f39c35...@gmail.com
Re: Security?
Hi. On Fri, 13 Sep 2013 09:05:35 +0200 Ralf Mardorf wrote: > However, when > searching for Linux related content I'm satisfied. It's ok for Linux IMO (and Solaris, and AIX), but sometimes these filters bite you when you least expect it. For example, I've not been able to locate pornview upstream (using startpage.com), and pornview was my favorite picture viewer (it's usage has nothing in common with the name) back in the Squeeze days: http://packages.debian.org/squeeze/pornview Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130913200415.1fad1aea41060bbdd5923...@gmail.com
Re: find'ing files containing certain words (all of them) ...
Hi.
On Sat, 21 Sep 2013 21:12:36 +0300
Alexander Kapshuk wrote:
> I could be wrong, but doesn't egrep, which supports extended regular
> expressions, fit the bill?
>
> =; echo two words > grep-AND-test1
> =; echo two > grep-AND-test2
> =; echo words >> grep-AND-test2
>
> =; egrep 'two|words' grep-AND-test*
> grep-AND-test1:two words
> grep-AND-test2:two
> grep-AND-test2:words
You're wrong indeed, as '|' means 'or', not 'and'.
$ echo two > grep-AND-test3
$ egrep 'two|words' grep-AND-test3
two
Please note that 'two.*words' won't be the solution (as 'words' can be
placed before 'two'), and even '(two.*words|words.*two)' won't be the
solution either ('two' and 'words' can be in a different rows of file).
Reco
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/20130921230559.007126c4783b01f08f766...@gmail.com
Re: systemd plans
Hi. On Fri, 27 Sep 2013 20:46:02 +0400 "Main Backup" wrote: > I just want to ask two simple questions about debian future. > > 1. Will systemd default init system in future? Why do you ask this question in the debian-user maillist? debian-devel or debian-testing seem to be more appropriate places for such fundamental questions. > 2. Why I just can't purge systemd-*** in my system now? I don't want it. Could you please specify exact package names you can't purge please? See, in my current Debian stable there is exactly one package which can fit your description is installed: libsystemd-login0 And I'm pretty sure I'm using good old sysv init. > I don't need it. I hate it by my all heart. > If I try to remove it, dbus depends on it as well. Why? ldd /usr/bin/dbus-daemon should shed some light on this mystery. > The usual way of expressing worries about upstream software quality or packaging quality of such software is filling bug report on an appropriate package. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130927221305.a555f79167aba37b5d837...@gmail.com
Re: systemd plans
On Fri, 27 Sep 2013 14:27:19 -0400 Tom H wrote: > libsystemd-login0 is the replacement of consolekit (which has been > deprecated) and is used for tracking sessions. Using it doesn't mean > that you're using systemd as PID 1. Even Ubuntu's using it. Oh no. What have you done :) You've mentioned ConsoleKit, which was designed by Poeterring, Satan incarnate himself. Worse, you've implied that libsysted-login0 is TRACKING user's sessions. And finally, you've mentioned the Dreaded Spaceboy Distro. > I don't think that dbus depends on consolekit on Debian 6 and I have no idea why dbus depends on libsystemd-login0 on Debian 7. Why, because dbus is linked against it. Clearly it is not a coincidence, it's a conspirancy. Just joking, of course. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130927230220.050524557d7d96f843518...@gmail.com
Re: Debian installer and raid0
Hi. On Thu, 3 Oct 2013 09:31:42 +0200 Francesco Pietra wrote: > Hello: > Did you use a recent version of the installer? What I would like to know - > before reinstalling everything on my servers - is whether the option to set > grub on both disks of raid 0 has now been introduced. Installing any OS on a RAID 0 (striping) is very strange thing to do - considering that disk failure in a RAID 0 = everything on this RAID0 is lost. Maybe you've meant RAID1 (mirroring)? Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131003115103.842a0b50f22ea27e9...@gmail.com
Re: ukash hijacked iceweasel
Hi. > On Sun, 2013-10-13 at 09:58 -0400, John Lindsay wrote: > > IIUC this is the thingy that doesn't allow to close a tab. When this > happens for Firefox here, I kill Firefox and then start it again. > Firefox usually asks, if the last session should be restored or not, so > don't restore the killed session. If Firefox shouldn't ask but > automatically restore the killed session, then it usually does work, to > close the tabs, before the site is loaded. After this Firefox is ok > again and can be closed and opened without issues. Disable javascript, reload offending tab (optional), close offending tab, enable javascript. I mean, why bother with complex solutions if there are simple ones? Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131014000626.cc1af515276cfc71844cc...@gmail.com
Re: virtualbox fails to compile module on 3.10
Hi. On Sun, 13 Oct 2013 11:20:07 +0200 Nickolay Todoroff wrote: > I'm stuck with the same problem as Kent. I'm running on Wheezy with the > 3.10-0.bpo.3-amd64 kernel (backports). > > The problem is actually that the virtualbox versions (esp. the OSE one) > from the repositories do not work with new kernels (newer than 3.2 as > far as I checked). It is clear that we can install the Oracle version, > however it is not optimal. The source of your problem is - you've installed backported kernel, but didn't install backported virtualbox module source. Please consider installing these packages (and upgrade your virtualbox to 4.2.16 from backports): http://packages.debian.org/wheezy-backports/virtualbox-dkms http://packages.debian.org/wheezy-backports/virtualbox Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131014001725.80991628ecd632411fffe...@gmail.com
Re: ukash hijacked iceweasel
On Sun, 13 Oct 2013 23:54:41 +0200 Ralf Mardorf wrote: > > > > Disable javascript, reload offending tab (optional), close offending > > tab, enable javascript. > > I mean, why bother with complex solutions if there are simple ones? > > I never tried it that way, because I couldn't find where to disable it > and I had no time to do research when this issue happened. Just for fun > I tried to find where it can be disabled for Firefox 24.0 right now, but > I couldn't find it ;). > > https://support.mozilla.org/en-US/kb/javascript-settings-for-interactive-web-pages This is outdated page indeed. Apparently Mozilla Foundation decided that it will be more user-friendly to double-click on 'javascript.enabled' at about:config page :) OT: Curiously, they decided that certain checkbox is 'killing their product': https://bugzilla.mozilla.org/show_bug.cgi?id=851698 Clearly they've never got into accound an evil JavaScript that renders browser unusable with no simple way to disable it. > I've got Ghostery installed, I'm using Ghostery to detect trackers, but > I can't find how to use it to disable scripts. > > IOW disabling javascript is much more work, assumed you won't install > many add-ons. I'm only using 2 add-ons, Ghostery and Anonymox. Probably because none of this extensions are designed to disable Javascript the way it should be done (i.e. whitelist for enabling JS on specific domains). As a NoScript user I don't find enabling and disabling Javascript that hard. > My Firefox ERS 17.0.9 (Vidalia/TOR) provides to disable javascript by > Edit > Preferences > Content, but I usually don't use the TOR browser or > outdated versions of Firefox. Good for you, I guess, but John (thread starter) uses stock Debian Iceweasel (judging from his MUA version), so Mozilla's KnowledgeBase solution should work for him. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131014071158.cba03d9651ff13a41f213...@gmail.com
Re: Jessie Minimum Kernel Requirement
Hi. On Wed, 16 Oct 2013 11:46:40 +0200 Florian Lindner wrote: > What is the minimum kernel version for the upcoming Jessie? Can I rely > on that this does no change after the freeze? (IIRC at Nov 5th) Considering one should be always able to do apt-get dist-upgrade from wheezy to jessie (and that means using wheezy's kernel with jessie's userland, at least temporary), it should be safe to assume that minimum kernel requirement for jessie's userland is 2.6.32. Since wheezy's kernel version is fixed for its lifetime, it's highly unlikely that such requirement will change in the future. Of course, wheezy's 2.6.32 and, say, CentOS's 2.6.32 are different kernels, and in the second case you can expect all kinds of strange behaviour. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131016155406.fad2fa9ed37e2e21f6909...@gmail.com
Re: Jessie Minimum Kernel Requirement
On Wed, 16 Oct 2013 08:21:35 -0400 Tom H wrote: > Furthermore, Wheezy's kernel is 3.2 not 2.6.32. OOPS, you've got me. I mistook squeeze for wheezy. Reco. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131017135930.dd5d7ae9032904e5076d1...@gmail.com
Re: Jessie Minimum Kernel Requirement
On Wed, 16 Oct 2013 13:00:19 +0100 Darac Marjal wrote: > This isn't really a safe assumption. There have been transitions in the > past (such as udev) where the dist-upgrade should be performed as: > * Update sources.list > * Install new kernel and new udev > * Reboot > * Proceed with dist-upgrade Interesting. Don't remember running such procedure. Was it for testing or sid? Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131017140052.b18cc5b5b3c2632f3f6f3...@gmail.com
Re: Jessie Minimum Kernel Requirement
On Thu, 17 Oct 2013 10:23:30 + Bonno Bloksma wrote: > Hi Reco, > > >> This isn't really a safe assumption. There have been transitions in > >> the past (such as udev) where the dist-upgrade should be performed as: > >> * Update sources.list > >> * Install new kernel and new udev > >> * Reboot > >> * Proceed with dist-upgrade > > > > Interesting. Don't remember running such procedure. Was it for testing or > > sid? > > If I remember correctly it was in the step from Lenny to Squeeze. I know for > sure I had to do it on several machines. > Yup, just found it in my old notes, it was Lenny to Squeeze. Thanks, that explains why I don't remember it - I was running testing those days. Back to the topic of discussion - I'm pretty sure that this wasn't needed for squeeze → wheezy transition. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131017145221.616334bdb0c803760a34c...@gmail.com
Re: XFCE without GNOME/KDE parts (Debian Wheezy - HP Pavilion dm1)
Hi. On Tue, 22 Oct 2013 14:50:57 +0200 Ralf Mardorf wrote: > And I also don't use NetworkManager, my Ubuntus/Debian aren't different > to my Arch Linux, however, a default Xfce4 usually is used with lot's > Gnome applications. Out of curiosity, what are names of these Gnome applications? Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131022193411.632068fd8235b4d20e444...@gmail.com
Re: XFCE without GNOME/KDE parts (Debian Wheezy - HP Pavilion dm1)
On Tue, 22 Oct 2013 22:18:39 +0200 Ralf Mardorf wrote: > On Tue, 2013-10-22 at 19:34 +0400, recovery...@gmail.com wrote: > > Out of curiosity, what are names of these Gnome applications? > > Xfce does need gtk. XFCE needs GTK+2. Current GNOME needs GTK+3. GNOME2 depended on libgconf, and no XFCE part depends on it. > Xfce doesn't provide some software, resp. only > rudimentary software for some usages. I don't see this as a bad thing. Why bother implementing, say, DE-tied display manager, then there is an excellent nodm? Likewise, nothing beats good old xscreensaver. Unless you're trying to to something very uncommon, TightVNC client is as good as that GNOME gizmo. > Likely it's used with > NetworManager, Which doesn't even depend on X, not to mention any GUI toolkit. And I always failed to understand what's so hot about that NetworkDestroyer anyway. > Gedit etc. and as mentioned before, Mousepad is a graphical text editor for Xfce. > for showing devices > in Thunar, even trash, gvfs is needed. For that, you need thunar-volman, which in turn depends on libgudev, which in turn listens for udev events via libudev. Actual mounting can be provided by anything (which includes, but surely not limited to dreaded gvfs, which chews on harddrives). > Sure, there's no hard dependency to Gnome software (or KDE software), > but at least I install a DE to get a DE and Xfce4 is missing features. > Would you like to use mousepad for programming? Of course not. Vim covers all my programming and text editing needs. They say that Emacs will be good for that too, once they put a text editor inside :) Are you implying that Gedit is suitable as an IDE? > What are you using as > your MUA? Etc. pp.. Sylpheed, mutt, maybe Icedove (Thunderbird in Debian-speek). Evolution is not the only MUA provided in Debian, and there is nothing special about it IMO. > "XFCE without GNOME/KDE parts" isn't a subject from me, but I agree that > it's quasi impossible to use Xfce without stuff from Gnome or if > installing Qt doesn't matter, KDE. Why, it's perfectly possible as it's exactly what I'm doing for several years. I was merely curious if there's some 'killer GNOME app' that I'm missing, and it looks like it's still not. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131023005320.48d4779978692f656b72d...@gmail.com
Re: XFCE without GNOME/KDE parts (Debian Wheezy - HP Pavilion dm1)
On Tue, 22 Oct 2013 23:05:10 +0200 Ralf Mardorf wrote: > On Wed, 2013-10-23 at 00:53 +0400, recovery...@gmail.com wrote: > > I was merely curious if there's some 'killer GNOME app' that I'm > > missing, and it looks like it's still not. > > That's true, with one exception, I disagree regarding to gvfs. You > perhaps can explain the OP how to use udev or anything else to get back > mountable devices and trash by Thunar/on the desktop. Sure. I've just installed thunar and thunar-volman with the needed dependencies (I've disabled auto-install of recommended stuff long time ago). I've ensured that no GConf, DConf, gvfs or GTK+3 libraries are present (no QT libs too). I've used USB stick to do a quick test. Thunar helpfully showed me that it's a '1.89 Gb Volume' (um, whatever). Next, I've added that line to /etc/fstab: /dev/sdb1 /mnt auto noauto,user 0 0 Clicked on '1.89 Gb Volume' and … it mounted. Talk about magic. Next, I've selected 'Eject device' and … it unmounted! Awesome. That lead me to the conclusion that gvfs is a complex solution to a non-existent problem. Even without that /etc/fstab line Thunar showed me 'Trash', whatever that thing is. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131023020404.6ba086041dec5fa54d945...@gmail.com
Re: XFCE without GNOME/KDE parts (Debian Wheezy - HP Pavilion dm1)
On Wed, 23 Oct 2013 02:18:10 +0200 Ralf Mardorf wrote: > That makes me wonder, perhaps this works because your install is that > clean. There may be other differences. Versions, build options, etc. And there's always ltrace and strace to make sure that the software behaves exactly the way you want it. > Impossible for my usage, fortunately I anyway mount from a > terminal. I don't know if the OP could live without G and Dconf and > gtk3. I didn't install that stuff simply for the sake of an clean experiment. I'm pretty sure that thunar doesn't use GConf, DConf or GTK+3, so these libraries were not needed. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131023102116.8ee1b2e35bfca57d9a728...@gmail.com
Re: XFCE without GNOME/KDE parts (Debian Wheezy - HP Pavilion dm1)
Hi. On Wed, 23 Oct 2013 11:15:34 +0200 berenger.mo...@neutralite.org wrote: > Don't you have gstreamer installer? Only that thing as a recommended dependency to webkit-gtk: $ dpkg -l gstream* | grep ii ii gstreamer0.10-plugins-base:amd64 Not that I need gstreamer. Once upon a time some kind soul told me about mplayer, and since then it's the only thing I ever use for video. sox and mpd play all audio I ever need to play. > I am interested in how to remove gconf2 and all it's family from my > system, since I do not understand why I need a configuration system for > stuff which is only used to play things in my web-browsers... which > anyway needs flash-player to really play stuff. (it's not the only > dependency I would like to remove, but it would be a good start. There > also tons of dependency for libsdl which do not seems to make lot of > sense) The way I see it, these guys (Maintainers of GStreamer packages) thought it is a good idea that gstreamer should depend on gconf (there was some bug about it, but I forgot the number). Next, they split gconf dependency to gstreamer0.10-gconf … only to make said package a dependency to gstreamer0.10-plugins-base. As for removing that stuff, there's a neat trick I currently use: $ cat /etc/apt/preferences Package: libdconf0 Pin: release n=wheezy Pin-priority: -1 Package: libgconf2-4 Pin: release n=wheezy Pin-priority: -1 By itself, it doesn't magically remove all that depends on that libraries, but prevents installing them. And gconf is not a 'configuration system'. To quote these ppl themselves (see /usr/share/doc/libgconf2-4/README): GConf is a configuration database system, functionally similar to the Windows registry but lots better. That alone IMO should be the reason do not install that thing ever. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131023213359.f1b483af53f82c6fdc5d3...@gmail.com
Re: XFCE without GNOME/KDE parts (Debian Wheezy - HP Pavilion dm1)
On Wed, 23 Oct 2013 09:45:07 +0200 "Ralf Mardorf" wrote: > What will you strace, if you're missing the Trash can icon? ;) For that task I'd use the source of thunar, gdb and ltrace. Nothing reasonable can be gained for tracing syscalls in this case. > Thunar doesn't need them, but other software does. Probably. As we discussed earlier, this 'GConf and/or DConf required' software is nothing special, and can be easily replaced with sane behavior equivalents. > Don't ask me what > software e.g. gtk3 needs, since I'm booted to > a very old Linux right no. > > > cat /etc/issue > Welcome to openSUSE 11.2 "Emerald" RC 1 - Kernel \r (\l). > > > uname -rm > 2.6.31.6-rt19 x86_64 What do you mean it is old Linux? I use RHEL5 (2.6.18) at office half of the time, and it's still considered 'new, full of untested and experimental features' OS by these people. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131023214523.7560ba3574ec6a766b22f...@gmail.com
Re: XFCE without GNOME/KDE parts (Debian Wheezy - HP Pavilion dm1)
On Wed, 23 Oct 2013 19:50:40 +0200 berenger.mo...@neutralite.org wrote: > I simply agree with everything you said here, but unfortunately, opera > depends on gstreamer0.10-plugins-good, which in turn... If you really don't need these libraries, there's a way. Check opera binary with ldd. If you see that it is linked with gstreamer0.10-plugins-good libraries, you're out of luck. Now, if they don't (i.e they use dlopen(3) instead of compile-time linking) - just use equivs to make a fake gstreamer0.10-plugins-good package, install it, and you're set. Repeat the process once Jessie becomes stable. > What is the difference between a "configuration system" and a > "configuration database system", excepted that the second one uses a > database (which are tools made to handle lot of data and not > configurations, as the name shows: data base, not configuration base > btw) ? You're both right and wrong IMO. You're right, as there's no noticeable difference between a 'configuration system' and a 'configuration database' both to a user (they're given tools) and for a developer (they're given API). You're wrong by defining a database as something that handles a lot of data. For example, /etc/passwd, /etc/shadow, /etc/group are their own database of users and groups (see getent(1)) with a real primary key - foreign key dependencies along each other. This database even has some utilities to work with it (useradd, usermod, passwd to name a few). And you've picked the wrong part of the quote. It's the 'functionally similar to the Windows registry' part which makes me feel uneasy :) Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131023224329.69c17175f24ee023ae655...@gmail.com
Re: maintained web-browser (
On Wed, 23 Oct 2013 10:02:27 -0500 Carson Chittom wrote: > Linux-Fan writes: > > You might take a look at Xombrero, although it's not in Debian. It's > based on Webkit and has the capability built-in to use a Javascript > whitelist. See https://opensource.conformal.com/wiki/xombrero Their implementation of 'Javascript whitelist' is questionable at best. I mean, you whitelist some site, go to it, and … you've also enabled all third-party sites to run Javascript in your browser as long as said site wants it too. A real implementation of 'Javascript whitelist' (see Noscript) will block any third-party JS from running by default in such scenario. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131024005013.94e69c2ed13e9fff20ee1...@gmail.com
Re: sudo and UNIXes (was: audacity export wma format[1 more question])
Hi. On Fri, 25 Oct 2013 12:31:55 -0600 Bob Proulx wrote: > Sudo has been on > HP-UX, SunOS, Solaris, IBM AIX and others for many years. It isn't > anything new. It is a good worthy tool. This is not entirely correct. Sudo is considered third-party software in HP-UX (HP merely builds it and doesn't install by default), AIX (not provided by IBM and therefore not supported) and Solaris (third-party software without any support in versions =< 10). About the only exception is Solaris 11 which provides sudo in default install (and it is configured the same way as in Ubuntu by default). Considering that primary usage of sudo is to provide controlled privilege escalation to uid=0, using unsupported (therefore - not updated unless local sysadmins care about security) sudo on these OSes is basically equivalent to giving everyone uid=0. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20131025234110.478c8065ddd992139a38b...@gmail.com
Re: kernels 3.0, 3.2, 3.4 won't boot in KVM
Hello. > Can anyone verify this, or is it just me? Is it possible I missed a kernel > config parameter required to allow the PAE-enabled kernel to boot in a KVM > session? Could it be a Debian build problem? An upstream kernel problem? Or a > qemu-kvm problem? I was able to boot the ISO you've provided using qemu-kvm from backports. Didn't test anything else, though. Host system is Debian Squeeze, qemu-kvm version is 1.0+dfsg-8~bpo60+1. I've attached full boot log taken from serial console just in case. Apparently, the trick is to tell qemu-kvm not to strip an 'lm' processor bit, i.e present a guest CPU as x86_64, not i686. boot.log Description: Binary data
Re: USB keyboard glitch
On Wed, 20 Mar 2013 19:39:22 -0700 (PDT) Paul Zimmerman wrote: > A cheapie USB keyboard seems to cause a problem for the current kernel. It > DOES work, and I am typing this message with it. But every time the system > boots with this keyboard plugged into a USB port it takes an unusually long > time sitting at the "waiting for /dev to be fully populated" line and then > prints an error message: > > [ 12.661750] > /build/buildd-linux-2.6_2.6.32-48squeeze1-i386-F95osd/linux-2.6-2.6.32/debian/build/source_i386_none/drivers/hid/usbhid/hid-core.c: > usb_submit_urb(ctrl) failed > > Is there a way to figure out exactly what the chipset on the keyboard is, and > maybe get this fixed? > It's the "Windows Media pro USB keyboard" by Gearhead. Model# KB3600MPU. Hi. Try adding usbhid quirk to workaround annoying timeout: echo options usbhid quirks=0xb38:0x3:0x2000 >> /etc/modprobe.d/local.conf update-initramfs -k all -u reboot If it doesn't help - check vendor_id and device_id with lsusb -v, update usbhid quirk as appropriate. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130329004810.0a33bf8c.recovery...@gmail.com
Re: ssh corrupt packets with aes encryption
On Sat, 11 May 2013 09:41:49 -0600 Joe Pfeiffer wrote: > So... any idea where to go next? Hi. 1) Try to unload kernel module aes_x86_64 and it's dependants, restart sshd, try once more. 2) Boot from livecd, if possible, see if openssh works there. 3) If possible, swap problem host's disk to another, see if changing hardware help. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130511211524.a9e366b938b72f7039c8f...@gmail.com
Re: ssh corrupt packets with aes encryption
On Mon, 13 May 2013 08:48:44 -0600 Joe Pfeiffer wrote: > recovery...@gmail.com writes: > > > On Sat, 11 May 2013 09:41:49 -0600 > > Joe Pfeiffer wrote: > > > >> So... any idea where to go next? > > > > Hi. > > > > 1) Try to unload kernel module aes_x86_64 and it's dependants, restart > > sshd, try once more. Ok. My idea was that in-kernel implementation of Intel's in-CPU AES instructions corrupts sshd connection. Can you provide a result of tcpdump -nn -s0 -i lo tcp port 22 ? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130513214920.b4cb40bd6f4e63f13fd10...@gmail.com
Re: ssh corrupt packets with aes encryption
On Mon, 13 May 2013 13:05:41 -0600 Joe Pfeiffer wrote: > recovery...@gmail.com writes: > > > > Can you provide a result of > > tcpdump -nn -s0 -i lo tcp port 22 > > ? > > I don't have any experience with tcpdump, so I'm just blindly following > instructions here -- let me know if you need more! > Ok, good. I need three more things: 1) File /tmp/1.cap from: tcpdump -nn -s0 -i lo tcp port 22 -w /tmp/1.cap Interrupt tcpdump with Ctrl+C once done with `ssh localhost` (I assume you've enabled aes back). 2) /var/log/kern.log, which should include info from last boot. Any other kern.log.X will do, of course. 3) Output of /sbin/ifconfig -a -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130514002734.8eca28a6511b229740d0f...@gmail.com
Re: Why wheezy/Jessie kernel does not kill the app. that takes too much RAM ?
On Sat, 18 May 2013 14:11:38 +0700 Sthu Deus wrote: > Good time of the day. > > > I have only 2 GB of RAM and a RAM-hungry app.s like chromium (just ~30 > tabs) and game "BosWars" that do not run together unless hang my machine > completely. > > The question is, Why the wheezy/Jessie kernel does not kill the app. > that takes too much RAM instead of letting the app.s to hang the OS > almost completely? -- For though after several minutes i can launch > kill command i yet can not run it: i get allocation error. Nothing > remains but to reboot w/ reset button. Hi. Please execute: man 5 limits.conf Read it very carefully, especially the parts mentioning 'as' and 'stack'. Next, observe contents of your local /etc/security/limits.conf. The answer to your question is 'wheezy/Jessie kernel does not kill the second app because it is you who didn't told it to'. That, and those Google guys and gals who write Chromium thought that it is a good idea to disable Out-Of-Memory-Killer kernel facility for their browser. You can clearly see the results of such design by yourself. So, 1) Modify /etc/security/limits.conf . Don't forget to relogin. 2) Add more swap. 3) Possibly switch to less memory-hungry browser. Sincerely yours, Reco. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130519000338.864946104d77ee94d9c6f...@gmail.com
Re: Need help setting up bluetooth file transfers
> paired my desktop computer and my tablet but file transfers failed. The > tablet said the failure was because the computer does not use obexftp so > I installed obexftp and obexfs. hciconfig found the bluetooth device to > be hci0 so I entered the command Hi. Packages you've installed are 'client' ones, they are needed to transfer files from your computer to somewhere. You need to install 'server' variant of obex utils, called 'obex-data-server'. To simplify things, you may also consider installing 'blueman' package, which is the most sane GUI to bluetooth in Debian IMO. Reco. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130524214340.4b65cb4d16194393e5ae3...@gmail.com
Re: slow vim
On Mon, 27 May 2013 11:53:47 + låzaro wrote: > I'm using mutt with vim as default editor but it delayed so much before > startup. Any idea? Hi. 1) You have installed one of the following packages: vim-athena, vim-gnome, vim-gtk. Yet you do not have X available. On startup, vim tries to locate an X server, fails, tries again etc. Solution: remove these packages, install one of 'vim' or 'vim-nox'. 2) You're using any of libvte9-based terminal emulator (gnome-terminal, xfce4-terminal, lxterminal to name a few). It seems that wheezy's libvte9 has some kind of performance regression, which visibly affects speed of any ncurses-based programs. Solution: use a real terminal emulator, such as 'xterm' or 'rxvt-unicode'. 3) Run: strace -ttfo /tmp/mutt mutt edit any mail, exit. Strace's output (/tmp/mutt) with timings should give you some clues about a cause of a problem. Reco. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130527212804.575859f58c0b2ab688e73...@gmail.com
Re: /etc/shadow password hash format (migration from SuSE 9.3 to Debian Wheezy)
On Mon, 27 May 2013 16:23:20 +0200 "Andreas Meile" wrote: > A "apt-cache search blowfish" shows me a lot of Blowfish related packages. > So is there one on it which extends the login authentication routine also to > process Blowfish hashes in /etc/shadow or is that a much more complicate > procedure (compiling a new kernel for example)? Hi. Install 'libpam-unix2' package. Configure PAM as outlined in /usr/share/doc/libpam-unix2/README.Debian. It is that simple. PS Did you ever considered using a real e-mail client? -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130527214000.fb529b7f01f0bc8ccbe26...@gmail.com
Re: slow vim
On Tue, 28 May 2013 09:02:10 +
låzaro wrote:
> Hi, I removed these packages and install vim-nox, but I would like to use
> gnome-terminal with mutt... there are any solution for that?
>
Hi.
I'm not aware of any solution regarding libvte9, but it seems that libvte9 is
not the case here.
That strace file you've send me turned out interesting.
Basically:
1) You start mutt, pid 7703.
2) Mutt forks, and executes vim.basic, pid 7727.
Now we can rule out an X hypothesis, as vim.basic doesn't know anything about X.
3) Vim.basic, in turn, loads an impressive bunch of its' plugins, and finally:
7727 08:52:51.424329 stat64("/bin/sh: 1: ispell: not found\n", 0xbfbbbed0) =
-1 ENOENT (No such file or directory)
7727 08:52:51.424441 stat64("/usr/lib/ispell/", 0xbfbbbed0) = -1 ENOENT (No
such file or directory)
7727 08:52:51.441003 stat64("/usr/local/lib/ispell/", 0xbfbbbed0) = -1 ENOENT
(No such file or directory)
Tries to locate ispell.
4) Failing that, it executes find (pid 7732), checking out something in the
depths of /usr/local/lib:
7732 08:52:51.445337 execve("/bin/bash", ["/bin/bash", "-c", "(find
/usr/local/lib -name \"*.ha"...], [/* 43 vars */]
5) Since then, vim waits - waitpid(7732, ...) - for find from 08:52:51 until
08:53:16 (time format is hh24:mm:ss).
So, something (one of vim plugins, probably), launches find at vim start, and
that probably explains the whole mess you have.
I cannot point to the offending plugin, but you can easily disable them one by
one, and see which makes the difference.
PS Please don't CC me, I'm subscribed to this list.
PPS strace(1) is your best friend. Have it, love it, use it.
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/20130528200828.1afd7b8eb92afe67fd26e...@gmail.com
Re: strange authentication request (GNOME)
On Tue, 28 May 2013 16:16:36 +0100 Matej Kosik <5764c029b688c1c0d24a2e97cd7...@gmail.com> wrote: > Is this some new dpkg/apt frontend or something unrelated? Hi. Yep, new one. Called PackageKit, brought to you by the same people, who are responsible for PulseAudio, ConsoleKit, GNOME3 and systemd. Whenever is a good or a bad thing is in the eyes of beholder. The story is like this: long time ago Debian got itself dpkg(1). RedHat got rpm(1). Next, Debian got tasksel, and it was good. RedHat thought it is not enterprisey enough to install a bunch of packages at once, so they passed. Then, Debian got apt-get, and it was excellent. RedHat guys and gals wrote a horrible abomination called up2date. Short of being enterprisey, it lacked any positive qualities. Then, Debian got aptitude, and couple of years later RedHat was like 'up2date is bad, we need something else, let's borrow yum from one of our derivatives'. yum usually gets the job done, if you don't mind wait a little. Finally, Debian got GUI for dpkg as synaptic (for GTK people) and adept (for KDE people). RedHat got nothing worthy of speaking. Four or five years since then RedHat said 'we need GUI for installing rpms, as it is new enterprisey trend now' (add usual 'lowers TCO' and 'adds ROI' speeches). So they wrote that PackageKitty, which claims to be a universal package installer for any Linux distribution known. tl;dr; version: install synaptic, use it. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130528211606.76e76311f69a60934856d...@gmail.com
Re: strange authentication request (GNOME)
On Thu, 30 May 2013 01:32:27 +1200 Chris Bannister wrote: > I am fairly sure that aptitude came AFTER synaptic! While I don't argue with that (I started using Debian when etch was testing), on my Debian system changelogs for aptitude and synaptic clearly show that aptitude was first, and synaptic was second: aptitude (0.0.6a-1) unstable; urgency=low * Initial Release. -- Daniel Burrows Sun, 6 Feb 2000 12:58:13 -0500 synaptic (0.16-1) unstable; urgency=low * Initial Release (closes: #115167) -- Michael Vogt Tue, 13 Nov 2001 23:17:20 +0100 -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130529203900.0bf1af29a4e2dfa341d32...@gmail.com
Re: WDIDLE.EXE use on WD10EZRX disk
On Thu, 30 May 2013 12:34:58 -0700 Gary Roach wrote: > Hi all, > I have found multiple warnings about Linux and the short head parking > time of the WD Green drives. I have seen multiple instances of the use > of the wdidle.exe package to reset this to say 30 seconds. But Western > Digital emphatically states that that utility in only for about 3 > "not-WD10EZRX" drives. Is this utility safe to use on the drive in > question or not? Hi. It's really simple - do you have the source code of this wdidle.exe utility? What about running this wdidle.exe on Linux? If both are 'yes', well, good for you. If both are 'no' - for all you know, it WILL brick your harddrive, steal all food from your fridge, and will do unspeakable things to kittens. Never use this thing. Instead, go to http://idle3-tools.sourceforge.net/, and get yourself an excellent idle3-tools, which is the free software, and actually works. I've used it myself to disable this questionable head parking on WDC WD7500AARS-00Y5B1 with success. Reco. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130531001925.1657cd6df23b47cb76b54...@gmail.com
Re: User unable to umount
On Thu, 30 May 2013 11:50:06 +0200 Erwan David wrote: > I do not use gvs (nor any g*) because of dependdencies and I do not trust it. > > As a grpahical tool I use smb4k, but it seems unable to do kerberos > authentication nor automatically mount a mount point at start of > session Hi. Looks like you've been hit by Debian bug #660431: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=660431 Basically, umount.cifs and possibly other umount helpers are deliberately broken upstream to comply with some obscure systemd design oddity. A workaround seems to be: a) umount cifs filesystem b) remove symlink /etc/mtab c) create an empty file /etc/mtab d) mount cifs filesystem In my case I said 'screw this', and started using smbnetfs, which: a) Definitely can be used without root and /etc/fstab entries. b) Features automatic mounting and un-mounting cifs filesystems. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130531004425.c136ac82da106d8228445...@gmail.com
Re: udev & automounting
On Mon, 24 Jun 2013 21:37:43 -0700
peasth...@shaw.ca wrote:
> Is there a better way of preventing the automounting
> than by adding entries to fstab?
> /dev/sdb1 /nowhere auto noauto 0 0
> /dev/sdb2 /nowhere auto noauto 0 0
> /dev/sdc1 /nowhere auto noauto 0 0
Hi.
Why bother with symlinks, when you can just rename device
pseudo-files?
So, this:
> KERNEL=="sd?", ATTR{size}=="1000944", SYMLINK+="SanDiskCF", \
> OWNER="peter", GROUP="users"
> KERNEL=="sd?1", ATTR{size}=="205569", SYMLINK+="SanDiskCF1", \
> OWNER="peter", GROUP="users"
> KERNEL=="sd?2", ATTR{size}=="795312", SYMLINK+="SanDiskCF2", \
> OWNER="peter", GROUP="users"
becomes:
> KERNEL=="sd?", ATTR{size}=="1000944", NAME="SanDiskCF", \
> OWNER="peter", GROUP="users"
> KERNEL=="sd?1", ATTR{size}=="205569", NAME="SanDiskCF1", \
> OWNER="peter", GROUP="users"
> KERNEL=="sd?2", ATTR{size}=="795312", NAME="SanDiskCF2", \
> OWNER="peter", GROUP="users"
Reco
--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive:
http://lists.debian.org/20130626234131.7959baaf37b54a1f306ad...@gmail.com
Re: Re (2): udev & automounting
On Wed, 26 Jun 2013 17:22:32 -0700 peasth...@shaw.ca wrote: Hi. > Originally I did just as you recommend and noticed a message > in /var/log/syslog instructing to use SYMLINK+=. > Of course it says so. See, no block device named sd? = no automounting by all those fancy freedesktop toys, and udev was hijacked by freedesktop long time ago. Of course, NAME= part of udev rule can break in a future, but as of wheezy's udev - it works. > The automounting isn't so bad after all. The user > can umount the device and ignore the spurious error > dialogue. And that is a bad thing, IMO. It is not a user who should get used to the software, it is the software who should do what man says. Personally, I've got rid of all those automounters, and use good old autofs. As a bonus, I've got an auto-unmounting, something that no other software can provide. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130627201854.a4bb4d3f89e690e8f442d...@gmail.com
Re: How to use the debian installation iso for installing packages using aptitude
Hi. On Tue, 13 Aug 2013 10:36:44 +0530 Anubhav Yadav wrote: > root@Innovator:/home/neo1691# apt-cdrom -d=/media/apt1 add > > Using CD-ROM mount point /media/cdrom/ ... > Any help? I want to say some bandwidth as it is not cheap here, and is very > slow > > Thanks! According to apt-cdrom(8), '-d' means: Mount point; specify the location to mount the CD-ROM. This mount point must be listed in /etc/fstab and properly configured. Configuration Item: Acquire::cdrom::mount. Note /etc/fstab part. Apparently you have entry for /media/cdrom in your fstab, but don't have one for /media/apt1. Reco -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130813135911.8593249fb1103c25bbba8...@gmail.com
Re: How to use the debian installation iso for installing packages using aptitude
On Tue, 13 Aug 2013 18:03:49 +0530 Anubhav Yadav wrote: > > I added the following line to my etc/fstab file > > /home/neo1691/dvd1-mountpoint /media/apt1 iso9660 rw,user,noauto 00 This is wrong, IMO. Should be something like (see fstab(5)): path_to_iso /media/apt1 iso9660 loop,ro,user,noauto 0 0 or /home/neo1691/dvd1-mountpoint /media/apt1 none bind,ro,user,noauto 0 0 Last one assumes you want to mount an iso manually to /home/neo1691/dvd1-mountpoint before using apt. Both examples assume that '/media/apt1' is the directory, not a symlink. > And my first line in /etc/apt/sources.list > > # deb cdrom:[Debian GNU/Linux 7.1.0 _Wheezy_ - Official amd64 DVD #Binary-1 > 20130615-23:06]/ wheezy contrib main > > deb cdrom:[Debian GNU/Linux 7.1.0 _Wheezy_ - Official amd64 DVD #Binary-1 > 20130615-23:06]/ wheezy contrib main Seems ok to me. > But now if I run say for example apt-get install xchat > > This time it never ask me to insert the cd rom it will search the mirrors. > > I guess I am close to achieving what I want. Did you run 'apt-get update' or 'aptitude update'? Also, if you really want apt not to use public Debian mirrors, you should comment out anything not related to 'deb cdrom' in sources.list. PS OOPS. Resending to the list. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130813175825.b793d1da9d701725c0f51...@gmail.com
Re: How to use the debian installation iso for installing packages using aptitude
On Tue, 13 Aug 2013 20:26:36 +0530 Anubhav Yadav wrote: > This is what I did. > Different roads lead to the same roads I guess! Well you must have > guessed by now why I want to use CDROM for installation of packages. > > > On 8/13/13, Anubhav Yadav wrote: > > dvd1-mountpoint is a folder where I mount the iso and /media/apt1 is a > > symbolic link to dvd1-mountpoint. > > > > So now what I will do is I will just mount the iso directly into > > /media/apt1 where apt1 is a directory not a link. Will this work? Ok, try it like this: 1) Add to sources.list: deb file:/media/apt1 debian main contrib non-free 2) Comment out anything else in sources.list. 3) Run apt-get update. -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20130813193725.d65531dfa61f718ba3950...@gmail.com
Re: KVM problem
On Sat, 10 Mar 2012 00:12:11 +0100 "tv.deb...@googlemail.com" wrote: > I'll leave you there with your judgment call, but Blender is considered > a bit more than "gruff" in the 3D modeling and video community, Yum is > Fedora package manager, the developers will be delighted to know they > are writing "stupid buggy python scripts". YMMV, but I would not bring yum as a fine example of python program. yum's memory footprint makes it nearly unusable on low-end VPS'es, and implementation language is not the last reason of that. Yum's speed at resolving package dependencies is abysmal comparing to apt, too (whenever python is to blame here, or not). IIRC modern Fedoras are unable even to install given less than 512M memory, and yes, you have to thank yum for that. Yum has some redeeming qualities, and the first of it that it is much better than up2date. Which is, in turn, implemented in C, and (surprise) originally developed by RedHat. yum, on the other hand, was adopted by RedHat from YellowDog Linux distribution. A good example of python-based package manager, IMO, is Gentoo's emerge. Surprisingly, it wasn't hit that hard during theirs python2 -> python3 transition. That thing pkg they ship with Solaris 11 is also written in python, and mostly does whatever one needs from package manager. Back on topic, personally I use virsh to manage qemu-kvm. I've used virt-manager last time about two years ago. Quit using it when virt-manager's interface was reworked once more, and started to look like they developed it for people with heavily impaired vision. Those days, the most meaningful way to do something with libvirt, was virsh dumpxml -> edit -> virsh define. Mine's today way of interacting with libvirt is basically the same. As an alternative to libvirt, one can always use this: http://current.workingdirectory.net/posts/2012/managing-kvm/ -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/20120310230203.7e9baa90.recovery...@gmail.com
