[gentoo-user] Re: python 3.1 safe as default python?
Adam Carter writes: >> Just installing it doesn't break portage. But using it does! :) >> >> 19:14:44 poke:~ $ eselect python list >> Available python interpreters: >> [1] python2.6 >> [2] python3.1 >> 19:17:24 poke:~ $ sudo eselect python set 2 >> 19:17:31 poke:~ $ emerge --info >> File "/usr/bin/emerge", line 41 >> except PermissionDenied, e: >> ^ >> SyntaxError: invalid syntax > > Did you try it as root? Is your user in the portage group? Yes, and yes. 02:03:10 poke:~ $eselect python list Available python interpreters: [1] python2.6 [2] python3.1 02:22:30 poke:~ $sudo eselect python set 2 02:22:39 poke:~ $sudo emerge --info File "/usr/bin/emerge", line 41 except PermissionDenied, e: ^ SyntaxError: invalid syntax 02:22:50 poke:~ $id uid=1001(che) gid=100(users) groups=4(adm),10(wheel),16(cron),18(audio),19(cdrom),27(video),35(games),81(apache),100(users),250(portage) I think that the comma on line 41 in /usr/bin/emerge is simply not valid in python-3
[gentoo-user] Re: python 3.1 safe as default python?
Albert Hopkins writes: > On Sun, 2009-08-02 at 11:17 -0400, Doug Hunley wrote: > I would think (hope) that the python maintainers were smart enough to > know whether installing Python 3 would break portage. Then again, this > is Gentoo ;-) Just installing it doesn't break portage. But using it does! :) 19:14:44 poke:~ $ eselect python list Available python interpreters: [1] python2.6 [2] python3.1 19:17:24 poke:~ $ sudo eselect python set 2 19:17:31 poke:~ $ emerge --info File "/usr/bin/emerge", line 41 except PermissionDenied, e: ^ SyntaxError: invalid syntax -- Christer
[gentoo-user] Re: Sync'ed my ~x86 system yesterday and now resolver stopped working
Timur Aydin writes: > Hi, > > I have synced my ~x86 system yesterday and after it completed, the > resolver doesn't work for some programs anymore. For example, ping > says "unknown host name". It doesn't even contact the dns > server, which is running on the same host. But dig works > fine. Also, using the IP address directly, I can access the internet. > > I am suspecting that the new glibc 2.10 is causing this. Anybody else > having this issue? I experienced similar problem when I upgraded to glibc-2.10, but the only thing I had to do was change from 'server' to 'nameserver' in /etc/resolv.conf (as told by my 'man resolv.conf'). I don't know why I had only 'server' before, perhaps that was allowed with earlier glibc. -- Christer
[gentoo-user] Re: Time to move on?
Dale writes: > Albert Hopkins wrote: >> >> Uh.. you don't "disable" it. You simply don't use the alias. >> > Oh, OK.If it is set up to add that > option, how do you tell it not to use it? You can type backslash before a alias to not use the alias. alias ls='ls --color' now ls lists with colors and \ls just runs the first ls found in your $PATH -- Christer
[gentoo-user] Re: Removing PAM from my system, is it adviseable?
"James Homuth" writes: > Will my system blow up at me if I remove PAM? I have used Gentoo at home for many years now. And I have never used pam. I even have # Don't want these, ever >sys-libs/pam-0 In /etc/portage/package.mask so that I detect if anything wants to pull in pam. A few times some ebuild tried to pull in pam dispite of USE=-pam, but that has allways been corrected by our devs. I don't actually dislike pam, it's quite good, and useful many times, it's just that I don't need it (yet). The most important point is choise. I can decide myself if I want it or not. And thats what Gentoo is all about (for me). -- Christer
[gentoo-user] Re: Disabling job control in bash
Alan McKinnon <[EMAIL PROTECTED]> writes: > Hi all, > > I need to disable job control on a machine with 650+ users because way > too many of them believe that Ctrl-Z is how you kill processes. So, I > want Ctrl-Z to basically do nothing. I've found a great solution: > > set +m One other way could perhaps be to tell the terminal not to send suspend on Ctrl-Z keypress stty susp undef -- Christer
[gentoo-user] Re: two identical /etc/sudoers -- only one works
maxim wexler <[EMAIL PROTECTED]> writes:
>
> If by "successful" you mean the call returns >= 0,
> then they're successful.
Yes,
>
> There is this, however:
> <...>
> open("/etc/default/nss", 0_RDONLY) =-1 ENODENT (No
> such file or directory)
> <...>
> connect(4,{sa_family=AF_FILE,
> path="/var/run/nscd/socket"}, 110} =-1 ENODENT (No
> such file or directory)
> <...>
> connect(4,{sa_family=AF_FILE, path="/dev/log"}, 110
> =-1 EPROTOTYPE (Protocol wront type for socket)
> <..>
I don't think those are important. Just standard tries to open
system-files if the exists, and connect to syslog.
I'm totally out of ideas now. Sorry I couldn't help you. Try to see if
it is some security-related problem as proposed by Arturo.
--
[EMAIL PROTECTED] mailing list
[gentoo-user] Re: two identical /etc/sudoers -- only one works
maxim wexler <[EMAIL PROTECTED]> writes:
>> root. Look for other occurences of geteuid in
>> strace output.
>
> Most similar instance is following the line:
> <...>
> readlink("/proc/self/fd/0, "/dev/tty2", 4095) = 9
> getuid32() = 0
You should aslo check for any of setuid seteuid setreuid or
setresuid. any call to any of those must be succesfull.
--
Christer
--
[EMAIL PROTECTED] mailing list
[gentoo-user] Re: two identical /etc/sudoers -- only one works
maxim wexler <[EMAIL PROTECTED]> writes:
> Most similar instance is following the line:
> <...>
> readlink("/proc/self/fd/0, "/dev/tty2", 4095) = 9
> getuid32() = 0
>^
> note: no 'e'
Yes that's also interesting, uid is zero, which is should be since you
are running strace as root. uid and euid is two different attribs of a
process, both should be zero.
>
>>
>> Check also that root actually has uid=0 on that
>> machine.
>>
>> $ getent passwd | grep ^root
>> root:x:0:0:root:/root:/bin/bash
>
> checks OK
Good, but unfortunate since i'm out if ideas now.
>
> FWIW, scrolling to the top of the output of strace --
> how do I write strace to a file?
strace -o file
> -- there is the line:
>
> close(393) =-1 EBADF (Bad file descriptor)
>
> followed by 630 similar lines ending in 'close(1023)
> ditto' then the "body" of the output.
That is perfectly normal. sudo just closes every fd's even the ones not
opened.
--
[EMAIL PROTECTED] mailing list
[gentoo-user] Re: two identical /etc/sudoers -- only one works
maxim wexler <[EMAIL PROTECTED]> writes:
> strace:
> <...>
> open("/etc/sudoers", O_RDONLY) = -1 EACCES
> (Permission denied)
> geteuid32() = 1
hmm, strange, geteuid should return euid which should be zero for
root. Look for other occurences of geteuid in strace output.
Check also that root actually has uid=0 on that machine.
$ getent passwd | grep ^root
root:x:0:0:root:/root:/bin/bash
--
Christer
--
[EMAIL PROTECTED] mailing list
[gentoo-user] Re: two identical /etc/sudoers -- only one works
I was wrong. Sorry.
I realize now that this cannot be your problem, sudo tell you that it
is not setuid if it's not.
$ sudo chmod -s sudo
$ sudo ls
sudo: must be setuid root
>
> Thanks Christer, never saw that command before, but
> like I told Walter, a listing for sudo is indeed:
> ---s--x--1 2 root root
> ^
> Is this supposed to be a
> one?
Did you type that line instead of cut'n paste? If not, I fail to.
understand the 1 in ---s--x--1
If you ask about the first number directly after the permission
string, it is the number of hard links to that file. If it is other
than one it means that the file has an other name also, you can find
that by using -i to ls to show the inode-number, and then find the
other with find -inum
Example:
$ pwd
/usr/bin
$ ls -li sudo
8803772 ---s--x--x 2 root root 107240 2007-05-21 11:11 sudo*
$ find . -inum 8803772
./sudo
./foo
$ ls -li foo
8803772 ---s--x--x 2 root root 107240 2007-05-21 11:11 foo*
Unfortunately I do not know what's wrong, try to strace sudo to see
what it does, remember that you have to bee root to strace a setuid
program. Look for
open("/etc/sudoers", O_RDONLY) = 4
The 4 is what filedescriptor open returned, and is -1 for a failed
open.
--
Christer
--
[EMAIL PROTECTED] mailing list
[gentoo-user] Re: two identical /etc/sudoers -- only one works
"Walter Dnes" <[EMAIL PROTECTED]> writes: > On Wed, May 23, 2007 at 06:14:53PM -0700, maxim wexler wrote >> Hi group, >> >> I connect to the web using >> >> $sudo /usr/sbin/pon >> >> on one machine(2.6.20-gentoo-r6). On another >> machine(2.6.19-gentoo-r5), I get >> >> :sudo: can't open /etc/sudoers: Permission denied. > The message suggest that the process of cudo actually don~t have permission to open the sudoers file. Check that the sudo-program (type sudo) is in fact owned by root, and has the setuid-bit set. $ls -l `which sudo` ---s--x--x 1 root root 107240 2007-05-21 11:11 /usr/bin/sudo* ^ ^ setuidroot -- Christer -- [EMAIL PROTECTED] mailing list
[gentoo-user] Re: Why are gentoo people so in love with colorized output?!?
[This is a followup to the whole thread, not any particular posting] Something interesting just happened to me that I would like to share. I realized that I have a program called usetool that I must have installed some time. What is that i thought, so my first try to find out was to try 'man usetool'. ...No manpage. Ok, next try, 'usetool -h' ...Good it has documentation, unfortunately the coloring made it unreadable, but with a lot of effort and eye-strain I could at least see that it has the -nc option. Ok, next try, 'usetool -nc -h' ...Dissapointment, still unreadable! I just wanted to let you know since it has relevance in this quarrel. I realize that the correct way to report this is to the bugzilla, and I will do that. -- Christer -- gentoo-user@gentoo.org mailing list
[gentoo-user] Re: [OT] Question about duplicate lines in file
Teresa and Dale <[EMAIL PROTECTED]> writes:
>
>
> Thanks, read the man page, it was short so it didn't take long. I tried
> this:
>
> uniq -u /home/dale/Desktop/hosts /home/dale/Desktop/hostsort
>
> It doesn't look like it did anything but copy the same thing over.
> There are only 2 lines missing. Does spaces count? Some put in a lot
> of spaces between the localhost and the web address. Maybe that has a
> affect??
The problem with uniq is that it (according to the manpage),
"Discard all but one of successive identical lines"
You need to have a sorted file for uniq to do what you want, or sort
it with the -u option
sort -u hosts > hostsort
If you don't want to ruin your original order you have to do something
else. This is one way of doing it with perl.
perl -ne 'print unless exists $h{$_}; $h{$_} = 1' hosts > hostsort
--
Christer
--
gentoo-user@gentoo.org mailing list
[gentoo-user] Re: CVS for kernel config?
"Anthony E. Caudel" <[EMAIL PROTECTED]> writes: > I am in the process of moving to an amd64 system and I anticipate a > lot of experimentation/tuning with the kernel. > > I was wondering if it is possible to set up CVS (or preferably > Subversion) so that I would be able to back up to any previous > configuration. It seems that if I just kept the .config file in cvs > and checked it in and out as needed that this might work. I suggest good old RCS. Its small, simple, and easy to work with. just emerge app-text/rcs then: $ co -l .config $ vi .config $ ci -u .config The revisions is kept as delta in a single file .config,v optionally in a subdirecdory RCS if that exists, you can even have RCS be a symlink to som other location if you like. -- Christer -- gentoo-user@gentoo.org mailing list
[gentoo-user] Re: System unusable after updating lvm
Alexander Skwar <[EMAIL PROTECTED]> writes: > Hm, did yo note the following warning? > > WARN: postinst > An old version of an installed library was detected on your system. > In order to avoid breaking packages that link against it, this older version > is not being removed. In order to make full use of this newer version, > you will need to execute the following command: > revdep-rebuild --library libdevmapper.so.1.01 > > After doing that, you can safely remove /lib/libdevmapper.so.1.01 > Note: 'emerge gentoolkit' to get revdep-rebuild No, I did not. And I don't think I did get that warning. $grep revdep-rebuild /var/log/portage/*lvm* ... nothing. And I don't have libdevmapper.so.1.01 I have /lib/libdevmapper.so.1.02 $equery l sys-fs/device-mapper [ Searching for package 'device-mapper' in 'sys-fs' among: ] * installed packages [I--] [ ~] sys-fs/device-mapper-1.02.03 (0) -- Christer -- gentoo-user@gentoo.org mailing list
[gentoo-user] Re: System unusable after updating lvm
[EMAIL PROTECTED] writes: > After updating lvm2 from 2.02.02 to 2.02.03, the system mount / which > is on a primary partition but frefus to mount /usr,/var,/home and /tmp. > On the console I see: > Setting up the logical volume manager > Volume group for uuid not found: m6reV.(four times) > Any idea? Happened to me too. I had to boot from livecd to copy lvm and some libs to get it up again. (I used ldd lvm to see which libs i needed also). -- Christer -- gentoo-user@gentoo.org mailing list
[gentoo-user] Re: Does openoffice really require pam?
Ciaran McCreesh <[EMAIL PROTECTED]> writes: > On Thu, 11 Aug 2005 20:32:26 +0200 Alexander Skwar > <[EMAIL PROTECTED]> wrote: > | I don't really want to install pam just because of OOo. Does OOo > | *REALLY* require pam? > > No, but tcsh does, and openoffice's build system requires tcsh. Is that really true. The simple experiment I just did (below), I interpreet as it's openoffice that wants pam. 23:52:25 poke:~ $sudo emerge -Dpv openoffice These are the packages that I would merge, in order: Calculating dependencies - !!! All ebuilds that could satisfy "sys-libs/pam" have been masked. !!! One of the following masked packages is required to complete your request: - sys-libs/pam-0.78-r2 (masked by: package.mask) - sys-libs/pam-0.77-r8 (masked by: package.mask) - sys-libs/pam-0.77-r6 (masked by: package.mask) For more information, see MASKED PACKAGES section in the emerge man page or section 2.2 "Software Availability" in the Gentoo Handbook. !!!(dependency required by "app-office/openoffice-1.1.4-r1" [ebuild]) 23:52:33 poke:~ $sudo emerge -Dpv tcsh These are the packages that I would merge, in order: Calculating dependencies ...done! [ebuild R ] app-shells/tcsh-6.14 +perl 0 kB Total size of downloads: 0 kB -- Christer -- gentoo-user@gentoo.org mailing list
