[00:03] <pmatulis> Jeeves_Moss: ok
[00:04] <pmatulis> metalfan___: repair with a rescue cd maybe
[00:12] <metalfan___> thx
[00:12] <metalfan___> i did go down the the server a few minutes ago....then it worked
[00:12] <metalfan___> i aborted the prompt a few times
[01:00] <ziggles> hey guys, does anyone know if i reconfig mysql-server.... will i lose my databases?
[02:43] <Nhawdge1> Hey all i'm looking for some answers about the /etc/motd file
[02:43] <Nhawdge1> in particular I want to append an actual update to the standard dynamic information without loosing the dynamic stuff
[02:53] <qman__> Nhawdge1, add it to /etc/motd.tail
[02:53] <qman__> at least, I think that still works
[02:53] <qman__> that's how it used to work
[02:54] <Nhawdge1> actually
[02:54] <Nhawdge1> I found a bit better of a method
[02:54] <Nhawdge1> aparrently they added an /etc/update-motd.d folder
[02:54] <Nhawdge1> it runs all the scripts in there to create the motd
[02:54] <Nhawdge1> which is a much better solution
[02:57] <pschulz01> Greetings.. does the JeOS install still exist? I can't see  the install option on 10.04 server or alternate.
[03:22] <cs1> guys
[03:22] <cs1> where can i find my LDAP password??
[03:32] <cs1> guys
[03:32] <cs1> please help me
[03:32] <cs1> i need help badly
[03:41] <Jordan_U> !details | cs1
[03:42] <cs1> ooohh
[03:42] <cs1> ok
[03:42] <cs1> im having problem with setting up the LDAP server
[03:43] <cs1> i follow the documentations on the ubuntu server guide
[03:43] <cs1> then when i try to add the ldap script for the frontend part...
[03:43] <cs1> it ask me for my LDAP password
[03:43] <cs1> im running ubuntu server 10.04
[03:46] <pickles> anyone know of a nice guide for a proxy server + dns on xen or kvm?
[06:26] <henriquev> I need to set up a controlled/paid wi-fi hot spot, any idea? I'm looking for something like a access point (or software to be run in a computer) with an API that allows me to control the access to the Internet (like a firewall or something)... I intend to develop the connection management system.
[07:14] <larsemil> anyone here? i have some questions. first off - i have an iscsi mounted on the kvm host. for the virtual guests, should i a) have one partition with lots of images, b) do an lvm and create a logical volume for each guest or c) some other setup?
[07:36] <jussi> morning all
[07:38] <jussi> Ive got dovecot running (have had for a while), with squirrelmail. we need auto responders and filters etc, so I installed avelsieve. problem is, I cant get it to run.
[07:39] <jussi> I tried adding managesieve to dovecot.conf, as per this page: http://workaround.org/ispmail/lenny/server-side-sieve-filtering - but that gives me: @emailserver:~$ sudo /etc/init.d/dovecot restart
[07:39] <jussi>  * Restarting IMAP/POP3 mail server dovecot                                     Fatal: Unknown protocol managesieve
[07:39] <jussi> any ideas on whatws going wrong for me?
[07:40] <joschi> jussi: which version of ubuntu and dovecot are you using?
[07:41] <jussi> its 8.04 and let me just check - whatever is in the repos I expect
[07:41] <jussi> jussi.schultink@emailserver:~$ apt-cache policy dovecot-imapd
[07:41] <jussi> dovecot-imapd:
[07:41] <jussi>   Installed: 1:1.0.10-1ubuntu5.2
[07:48] <jussi> joschi: ^^
[07:51] <joschi> jussi: I think you're out of luck. the managesieve plugin/protocol is shipped since dovecot 1.1.
[07:51] <joschi> you are using dovecot 1.0
[07:51] <jussi> oh :(
[07:52] <joschi> but you can use the cmusieve plugin and use another managesieve implementation
[07:52] <jussi> hrm
[07:52] <jussi> can you point me to some docs or give me some ideas on what and how?
[07:53] <joschi> jussi: http://wiki.dovecot.org/ManageSieve
[07:55]  * jussi reads and scratches head
[07:56] <jussi> might pop over to #dovecot, see if theyve some recommendations
[08:00] <joschi> jussi: ehm, just read the wiki page. it clear mentions pysieved...
[08:00] <jussi> joschi: which does not appear to be in the repos?
[08:00] <joschi> jussi: so?
[08:01] <joschi> jussi: and BTW: http://packages.ubuntu.com/lucid/pysieved http://packages.ubuntu.com/karmic/pysieved
[08:01] <joschi> you can always do a backport
[08:01] <jussi> joschi: I would much prefer to use something from the repos. (Im sorry if this came across the wrong way, I did read the wiki page)
[08:03] <rahman> Hi, when I install openldap why there is no configuration questions asked like this : http://imagebin.org/105899 I'm on 10.04 btw
[08:06] <joschi> rahman: https://help.ubuntu.com/10.04/serverguide/C/openldap-server.html
[08:10] <rahman> joschi: I already read it the problem is when I do "sudo dpkg-reconfigure slapd" it doesnt ask for domain name, organisation name, admin password, backend db etc. It only asks if I want to keep db when I purge slapd and if I want v2 compatiblity
[08:10] <spajderix> hi
[08:11] <spajderix> is there a way to install php5.3 and php5.2 on one machine without having to compile it from source?
[08:11] <joschi> rahman: and where in the guide I posted does it say, that something like this should happen? or that there's a debconf frontend for openldap at all?
[08:12] <joschi> spajderix: if you find a .deb package providing a non-conflicting PHP installation: sure, then it's possible
[08:12] <joschi> spajderix: out-of-the-box with only official ubuntu packages: no
[08:14] <rahman> joschi: it says in the installation section : "The installation process will prompt you for the LDAP directory admin password and confirmation. .......If you require a different suffix, the directory can be reconfigured using dpkg-reconfigure."
[08:14] <joschi> rahman: the installation process of libnss-ldap... not slapd
[08:16] <joschi> rahman: and to be quite frank, I can't find the quote you posted on https://help.ubuntu.com/10.04/serverguide/C/openldap-server.html
[08:16] <spajderix> joschi: then compilation awaits me ... thanks for your help:)
[08:17] <joschi> spajderix: at least you only have to compile only one of the two versions yourself ;)
[08:18] <rahman> joschi: thanks for the hint it seems I am reading 9.10 docs.
[08:19] <joschi> rahman: oh, that's bad. the ubuntu 9.10 docs on openldap were/are really broken. there's already a "popular" bugreport on launchpad on this
[08:20] <spajderix> joschi: that's what i thought of, i guess 5.3 from repo and 5.2 from source will be easier since I want to use 10.04 :)
[08:20] <rahman> joschi: bad luck :) I did a google search and didn't notice the 9.10 part of the result url
[08:21] <spajderix> joschi: another question then: is there a way to get ./compile params for php5.2 used in ubuntu deb package?
[08:21] <joschi> spajderix: just get the source package and take a look at the specs in ./debian
[08:22] <spajderix> joschi: ok, thank you for your time :)
[08:41] <rahman> my ldap setup seems completely messed up: I get ldap_*: Internal (implementation specific) error (80) when I try to add ldif. opensldap FAQ says its a permission problem and when I do "chmod -R openldap:openldap /var/lib/ldap" I get chmod: invalid mode: `openldap:openldap' I did a purge and reinstall btw. Still no go
[08:42] <larsemil> rahman: chown, not chmod
[08:51] <joschi> hi, I have a problem with the a dynamic file template in rsyslog. it seems rsyslogd wants to write to the log files in /var/log/%HOSTNAME% before the directory was created.
[08:51] <joschi> at least in normal mode. in debug mode (`rsyslogd -c4 -d`), it creates the directory just fine.
[08:51] <joschi> I've posted details about my configuration, the error message and part of the debug output under http://pastie.org/1048389
[08:51] <joschi> it could be related to Bug #484336 and changing the owner of /var/log to the user 'syslog' works, but I'd rather avoid that
[09:06] <jussi> joschi: thanks for your pointers, Ive found a package and installed, just figuring out the configs now :)
[09:08] <rahman> larsemil: someone needs to fix openldap FAQ then :) thanks
[09:09] <joschi> rahman: be a hero and send an email to the author of that FAQ entry with your correction ;)
[09:26] <jussi> looking at: http://isp-control.net/documentation/howto:mail:install_avelsieve_for_webmail - how do I determine the correct UID and GID I need for section 4?
[09:29] <jcastro> zul: can you linke me to the server daily build spec for maverick?
[09:29] <zul> jcastro: sure
[09:29] <zul> jcastro: https://blueprints.edge.launchpad.net/ubuntu/+spec/server-maverick-daily-vcs
[09:31] <joschi> jussi: the user needs access to your sieve files (see http://wiki.dovecot.org/LDA/Sieve/CMU)
[09:38] <jussi> joschi: is there any big issue if I get it wrong, then go change it? is it going to eat things?
[09:38] <rahman> its realy getting on my nerves. I can't restart slapd, it complains No configuration directory was found for slapd at /etc/ldap/slapd.d/ even it is there and the permissions are openldap:openldap
[09:39] <joschi> jussi: if the uid and gid are wrong, pysieved simply can't read and/or write the sieve files - or cmusieve can't. but that won't blow up your server
[09:39] <jussi> joschi: ok, great. thank you
[09:40] <joschi> rahman: does it contain the cn=config.ldif file and cn=config subdirectory or is /etc/ldap/slapd.d/ empty?
[09:41] <joschi> rahman: I personally found the cn=config style configuration a major PITA and reverted to good old slapd.conf
[09:41] <rahman> joschi: both files are there
[09:41] <joschi> rahman: especially since the meta backend doesn't support dynamic configuaration yet
[09:41] <joschi> rahman: check /etc/default/slapd for which one is actually used
[09:42] <joschi> rahman: you can't use both at the same time
[09:44] <rahman> joschi: but the openldap wiki encoures to use dynamic conf as they say they can drop old style anytime. And in /etc/defaults/slapd SLAPD_CONF= is empty so it should use  /etc/ldap/slapd.d as described in that file. But I set it to SLAPD_CONF=/etc/ldap/slapd.d still same
[09:45] <joschi> rahman: the openldap wiki can say a lot of things. as long as the meta-backend can't be configured that way, I couldn't switch configuration style even if I wanted to
[09:55] <jussi> hrm, so problem now is that pyseived wont start. Ive tried starting it from the init.d script, but it doesnt work, when I run it straight, then: http://paste.ubuntu.com/465815/
[09:58] <joschi> jussi: "IOError: [Errno 13] Permission denied: '/var/run/pysieved.pid'" is pretty obvious, isn't it?
[09:58] <jussi> joschi: yes, but what do I do about it :D do I just chown the pid or?
[09:59] <joschi> jussi: either create a directory /var/run/pysieved, `chown` it by pysieved's user and create the PID file in this directory, or create the PID file in /tmp (or any other world-writable directory)
[10:00] <jussi> joschi: ahh, thanks
[10:34] <jussi> getting theremow, but Ive now run into: Fatal error: Allowed memory size of 16777216 bytes exhausted (tried to allocate 1048576 bytes) in /usr/share/squirrelmail/plugins/avelsieve/include/managesieve.lib.php on line 457
[10:37] <qman__> jussi, the solution to that depends on whether or not the script you're running is supposed to use more than 16 megs of RAM
[10:37] <qman__> if it is, increase the limit in php.ini
[10:37] <qman__> if it isn't, you'll have to find out why it's using that much
[10:38] <jussi> but the amount of bytes exhasted is more than the amount it tried to allocate? is that a cumulative figure?
[10:39] <qman__> that's a hard limit per-script
[10:39] <qman__> so it includes all the functions and data used
[10:39] <jussi> hrm
[10:40]  * jussi doesnt know
[10:40] <qman__> boosting it to 32 or 64 megs shouldn't hurt as long as you've got plenty of free RAM
[10:40] <qman__> if it still hits the raised limit it's probably a problem with the script
[10:41] <jussi> ok, which should I be working with? /etc/php5/apache2/php.ini
[10:41] <jussi> /etc/php5/cli/php.ini
[10:41] <qman__> is the script run from a web browser or the terminal
[10:41] <jussi> browser (so I know the answer ;) :D)
[10:42] <jussi> right, do I need to restart apache to get that to take?
[10:43] <jussi> ok, so theres an issue with the script. :(
[10:44] <jussi> Fatal error: Allowed memory size of 67108864 bytes exhausted (tried to allocate 1025 bytes) in /usr/share/squirrelmail/plugins/avelsieve/include/managesieve.lib.php on line 459
[10:44] <qman__> yeah, if it's using 64 megs there's definitely something wrong
[10:52] <RoyK_wrk> anyone that knows any open webinar software?
[10:54] <nimrod10> hi guys have you experienced the ubuntu-server 10.04 x64 installer that can't get a dhcp address during install ? It fails for me , but after I hit reconfigure again it gets a dhcp ip
[10:54] <Jeeves_> nimrod10: Yes.
[10:54] <Jeeves_> Do you have it connected with a switch with spanning tree enabled?
[10:55] <Jeeves_> The switch usually checks the port for 37 seconds if it won't break anything if it's really enabled
[10:55] <nimrod10> I have it connected to a cisco switch yes, at the moment I don't have a log in there
[10:55] <Jeeves_> during that time, you do have link, but no connectivity
[10:56] <nimrod10> Jeeves_, so what you're saying is to disable the spanning tree to fix this ?
[10:56] <Jeeves_> nimrod10: http://www.itsyourip.com/cisco/how-to-enable-spanning-tree-portfast-in-cisco-catalyst-switch-ios/
[10:56] <Jeeves_> Don't disable spanning tree, enabled portfast
[10:56] <Jeeves_> That will give you a warning
[10:57] <Jeeves_> because portfast introduces risks for spanningtree, but if you're not going to connect that port to another switch, that's ok.
[10:58] <nimrod10> yes I see , I was reading through the page you gave, me and I seem to recall something I read about blocking vs Forwarding on cisco
[10:58] <Jeeves_> That's spanning tree indeed
[10:58] <Jeeves_> If it's enabled, the port starts in Blocking-mode
[10:59] <nimrod10> thank you Jeeves_ I'll get the guy with the logins to change that and I'll try it now
[11:00] <Jeeves_> np!
[11:10] <kirkland> hallyn: see http://blog.dustinkirkland.com/2010/02/ksm-now-enabled-in-ubuntu-lucid.html for help on how to verify KSM
[11:11] <rahman> I see this insyslog when trying to start slapd:          slapd[8115]: daemon: bind(8) failed errno=13 (Permission denied)
[11:17] <nimrod10> rahman I haven't used slapd but that looks like a regular file/directory permission issue to me
[11:19] <rahman> nimrod10: which is annoying, I checked /etc/ldap/slapd.d and /var/lib/ldap all are set to openldap:openldap as supposed to be.
[11:20] <rahman> I will try to completely purge and reinstall
[11:20] <nimrod10> check the location of the pid file
[11:23] <Jeeves_> rahman: It gives a permission denied on bind()
[11:23] <Jeeves_> Are you root while trying to start it?
[11:26] <smoser> zul, sponsor https://bugs.launchpad.net/ubuntu/+source/xinetd/+bug/43574
[11:29] <rahman> Jeeves_: yes did it with sudo. But never mind. I purged it, completely removed all ldap directories and reinstalled it. It know works. So it seems it was a meesed up setup. Thanks anyways.
[11:38] <RoyK_wrk> anyone that knows any open webinar software?
[11:41] <alex88> hi guys, i'm doing ssh tunnel and after some time i can't navigate, wget on vps (ssh server) says
[11:41] <alex88> Connecting to www.google.it|209.85.135.104|:80... failed: No buffer space available.
[11:41] <alex88> any help?
[11:49] <RoyK_wrk> alex88, are you using ssh keepalives?
[11:56] <AnAnt> Hello, how can I make the value of gecos field always the same as cn in LDAP service ?
[12:27] <tola> Hi, are the AMI IDs on this page up to date? http://uec-images.ubuntu.com/releases/lucid/release/ I can't find those AMIs in the Community AMIs directory on EC2...
[12:28] <alex88> RoyK_wrk: dunno how to check?
[12:29] <alex88> btw, my own server has problems, lynx can't either navigate
[12:30] <RoyK_wrk> Alblasco1702, to enable keepalives, add these two lines to /etc/ssh/ssh_config on the client
[12:30] <RoyK_wrk> http://en.wikipedia.org/wiki/Lemon_curd
[12:30] <RoyK_wrk> erm - ops
[12:30] <RoyK_wrk> "ServerAliveCountMax 100" and "ServerAliveInterval 1"
[12:33] <alex88> RoyK_wrk: after reconnect i've the same issue..waiting for some time worked..
[12:36] <alex88> don't know why it happens, is possible because somebody was downloading from my server?
[12:36] <alex88> with download accellerators?
[12:37] <pmatulis> RoyK_wrk: making us desert?  yummy
[12:38] <pmatulis> *dessert
[12:44] <tola> Ah, is it possible to create an instance of an official Ubuntu AMI using Amazon's web-based console or do you have to use the command line tools instead because the AMIs are hosted elsewhere?
[12:47] <alex88> RoyK_wrk: vps control panel says rcpsndbuffer is full, so i think there is no way to change that limit
[12:54] <nimrod10> alex88, who's eating all the memory needed by ssh ? who's filling that buffer ?
[13:04] <alex88> nimrod10: maybe a friend downloading with download accellerator..is there a way to check that buffer?
[13:05] <alex88> for example, what is using that, how much?
[13:06] <nimrod10> if you install iptraf you can see some statistics about your network cards , about the buffer I don't know how you'd see how much data is in it.
[13:07] <alex88> nimrod10: thank you, i'll check..because on the vps panel i can see the used buffer
[13:07] <alex88> so there is a way also on the system i think...maybe some /proc/ value
[13:08] <alex88> cause is useless that they give you 100mbit internet connection if everything stops with buffer full
[13:09] <alex88> i've tried that but it works on localhost and openvpn interface, not on venet0 that's internet... :/
[13:10] <nimrod10> I don't know if you should blame it on the net connection, your ssh buffer resides in memory, and that is were the problem lies , some people donwloading stuff can't fill the ssh buffer, they can only fill the memory outside of your buffer
[13:11] <alex88> it's not the ssh buffer, but the network buffer inside the kernel
[13:12] <alex88> i've just told that i can't navigate via ssh tunnel, when the buffer is full also lynx on the vps and wget can't work
[13:12] <alex88> then if the buffer is full other packets get dropped
[13:24] <Jeeves_> How can I connect a Debian bug with an Ubuntu bug?
[13:27] <jpds> Jeeves_: Via v6!
[13:28] <jpds> Jeeves_: I think that you can only do that in Launchpad.
[13:28] <jpds> Jeeves_: (+) Also affects distribution → Debian
[13:30] <Jeeves_> jpds: Yes, found it through #launchpad. Thanks anyway. :)
[13:30] <Jeeves_> And no v6 jokes!
[13:30] <Jeeves_> Ubuntu will be sorry if it can't be downloaded on it's ipv4 island :P
[13:31] <jpds> Jeeves_: What? releases.u.c has 3 4A records. :-)
[13:31] <Jeeves_> jpds: That's only two in a few minutes! :)
[13:32] <Jeeves_> jpds: By the way. All the nameservers are ipv4 only. :P
[14:38] <hallyn> kirkland: thx
[14:46] <remix_tj> does anyone knows where to put upstart scripts in lucid?
[14:48] <jpds> remix_tj: /etc/init/
[14:59] <remix_tj> jpds: thanks, i'll take a look
[15:13] <sommer> morning all
[15:14] <ttx> sommer: o/
[15:24] <smoser> cjwatson, "grub-pc grub-pc/install_devices multiselect WHAT_HERE" ?
[15:31] <cjwatson> smoser: just nothing after multiselect
[15:31] <cjwatson> smoser: and 'grub-pc grub-pc/install_devices_empty boolean true'
[15:31] <cjwatson> smoser: that's off the top of my head though
[15:32] <smoser> thanks
[17:32] <benedikt> how can i rename a kvm/libvirt virtual machine? (not the hostname, but the name libvirt/kvm konws it by, domain?)
[17:36] <resno> ive got a server that has no more hdd space, what can i do to easily remove unused apps/etc to get space back
[17:38] <ph8> hey all, I don't suppose anyone knows an auto responder I can use with postfix?
[18:04] <geneticx> when someone says "PCI experience required" do they mean peripheral component interface
[18:16] <ruben23> hi guys how do i check if my audio device is detected and properly install on my ubuntu
[19:36] <ND-work> hey guys
[19:36] <ND-work> quick question
[19:37] <VMTester> can anyone point me to a good guide to installing vmware server 2.02 on ubuntu server 10.04?
[19:37] <NightDragon> i am looking for a way to install windows on client computers with PXE , using a ubuntu box as my server
[19:38] <VMTester> I've run through http://hmontoliu.blogspot.com/2010/04/installing-vmware-server-202-in-ubuntu.html a few times, but have not been able to get it installed correctly yet
[19:40] <NightDragon> any ideas?
[19:40] <benedikt> how can i rename a kvm/libvirt virtual machine? (not the hostname, but the name libvirt/kvm konws it by, domain?)
[19:52] <RoyK> mmmhhh.... new number chruncher - this one with 2x12core opterons and 64 gigs of ram :D
[20:09] <cloakable> RoyK: But will it blend? ;)
[20:12] <oru_work> if I have the ip, how can I find out its location ?
[20:12] <jord> Hey. fsck hangs when it runs at boot on 10.04. It's been sitting there for a while and I can't cancel it or access at all. Any ideas why? / is ext4
[20:14] <RoyK> cloakable: blend?
[20:17] <failover> lol
[20:36] <pmatulis> i'm looking for a mailing list manager that integrates well with an LDAP backend for storing memberships
[20:37] <webPragmatist> is there a sshd log of sorts?
[20:37] <pmatulis> webPragmatist: try auth.log
[20:37] <webPragmatist> pmatulis: that only shows pam info
[20:41] <pmatulis> webPragmatist: not according to what i see.  increase you log level
[20:44] <failover> pmatulis, i'm think sympa could work with ldap
[20:45] <failover> oh, too late
[21:51] <jord> I'm setting up mdadm RAID5. I unplugged one of the drives in the hope of simulating a failed drive. The machine won't boot, just hangs with a fsck message. Any ideas where to look for clues? I've looked in /var/log but can't see anything.
[21:52] <jord> If I plug the drive back in, it boots.
[23:14] <TannerF> hi, i get "bash: php: command not found" even though php is installed
[23:15] <ajmitch> you probably want php5-cli
[23:18] <webPragmatist> is it possible to force an sshd "sftponly" user to a specific gid?
[23:21] <cloakable> make that gid their primary group?
[23:22] <webPragmatist> oh hrm
[23:22] <webPragmatist> what if they aren't a part of that group heh
[23:24] <failover> sudo chown user.group_you_want ssh_user_home_folder && sudo chmod g=+s ssh_user_home_folder
[23:26] <failover> after do that every file you create on "ssh_user_home_folder" will be in the "group_you_want" group
[23:26] <TannerF> thanks, ajmitch
[23:33] <webPragmatist> failover: but the issue is when they create directories those directories aren't +s
[23:33] <webPragmatist> g+s
[23:46] <webPragmatist> failover: yea like i said… g+s will not be on directories written in the home directory
[23:46] <qman__> webPragmatist, I've got a daily cron script as a workaround
[23:46] <qman__> it's hacky and slow
[23:50] <webPragmatist> qman__: nothing in sshd_config ?
[23:50] <webPragmatist> i have it chrooted
[23:50] <qman__> I don't think sshd has a force group option
[23:50] <qman__> but I could be mistaken
[23:51] <webPragmatist> yea nothing useful
[23:51] <webPragmatist> isn't there a way to use bashrc to set a specific chmod
[23:51] <webPragmatist> for newly written files
[23:52] <qman__> ah
[23:52] <qman__> "it takes the fsgid of the current process, unless the directory has the setgid bit set, in which case it takes the gid from the parent directory, and also gets the setgid bit set if it is a directory itself."
[23:53] <qman__> so if you recursively apply the group-owner and setgid flag, your system should theoretically maintain that state unless someone manually changes it
[23:53] <qman__> and with sftp-only access that shouldn't happen
[23:53] <webPragmatist> woh
[23:55] <webPragmatist> where do you see this?
[23:55] <webPragmatist> i don't see it in sshd_config
[23:55] <qman__> it's in mount's manual
[23:55] <qman__> it's not SSH specific
[23:56] <webPragmatist> qman__: whats the name of the command?
[23:57] <webPragmatist> just setting gid= ?
[23:57] <qman__> ?
[23:57] <qman__> you set the group-owner with chown or chgrp
[23:57] <qman__> and then set the setgid bit with chmod
[23:57] <webPragmatist> right but +s works on files created
[23:57] <webPragmatist> not directories
[23:58] <failover> work's on directories too
[23:58] <qman__> +s means two different things for files and directories
[23:58] <qman__> on directories it's setgid
[23:58] <qman__> err, nevermind, it's setgid on both
[23:58] <qman__> it just behaves differently
[23:58] <webPragmatist> failover: i mean… if you create a directory in a directory which is +s it doesn't keep +s
[23:58] <qman__> webPragmatist, yes it does
[23:59] <webPragmatist> i just tried it :(
[23:59] <qman__> if it didn't, something else happened
[23:59] <qman__> because it most certainly does
[23:59] <webPragmatist> well then crap