[00:26] <MraAlbertina> hi. how could i force the use of VESA on one machine? ubuntu 12.10
[00:29] <MraAlbertina> i mean on boot, trough grub
[02:44] <Fieldy> i installed roundcube, but I don't see anythying under /var/www related to it... did i miss something? i read all the output after install
[02:46] <patdk-lap> cause that isn't how ubuntu does things
[02:47] <patdk-lap> it got spread out and installed all over the place
[02:47] <patdk-lap> normally why I don't bother using ubuntu packaged web apps
[02:47] <patdk-lap> normally it will have like apache include a config from /etc/roundcube to do it
[02:49] <Fieldy> okay yeah that lead me to find it's actually installed in /var/lib and wants me to set up an alias to a url pattern which will point it there -- that is really odd
[02:57] <Fieldy> yea. this is messy... will mess with it more tomorrow. i already see a manual install in my future.
[04:07] <qman__> Fieldy, it automatically sets up a web directory and everything, all you have to do is turn it on
[04:07] <qman__> it might already be on, try visiting http://your.server/roundcube/
[10:18] <Catbuntu> hi
[11:11] <pelle2> hi
[11:12] <pelle2> doesn anyone have any idea of what causes this: booting with a certain usb-disk (not from it, just attached) does not work IF i do not have a monitor connected to the server
[11:12] <pelle2> if i have a monitor attached, the system boots
[11:12] <pelle2> if i remove that specific usb-disk, i can boot fine without a monitor attached
[11:12] <pelle2> ?!
[11:13] <ioria> i'd like to know if ISPConfig is necessary to install a full Ubuntu server ... (i'm preatty new about this) ? tx
[13:52] <Halite> I'd like to run a server on Ubuntu. How
[14:00] <andol> Halite: In this context, how do you define a server?
[14:56] <sliddjur> how do i change rep in cli?
[14:56] <sliddjur> i have a virtual server at my uni, and they use their own rep. and i cant dist upgrade. stuck on 10.10
[14:57] <sliddjur> if there are a swedish/nordic repositorie i'd be happy :)
[15:16] <qman__> sliddjur, in ubuntu, dist-upgrade stays with the same release version, use "do-release-upgrade" to upgrade to a newer release
[15:17] <qman__> sliddjur, that said, to change the repositories, you need to edit them in /etc/apt/sources.list
[15:17] <qman__> also, I'm pretty sure 10.10 is end of life now
[15:17] <qman__> !eol
[15:19] <sliddjur> qman__: crap.. i dropped connection on release upgrade.
[15:19] <sliddjur> what now
[15:20] <qman__> did it start actually upgrading yet? it should have started a screen/SSH session on a high port for your to reconnect to
[15:20] <sliddjur> yes, it disced when I was supposed to answer if i wanted to keep bash settings
[15:21] <sliddjur> i can still connect on same port.. but can i get back to my other session somehow? i can see it on pts/0
[15:21] <sliddjur> *** bash.bashrc (Y/I/N/O/D/Z) [default=N] ? Write failed: Broken pipe
[15:22] <qman__> sudo screen -d -r
[15:22] <sliddjur> dont have screen
[15:23] <sliddjur> and apt-get is now locked
[15:23] <qman__> well, you can kill off the process
[15:23] <qman__> did it do a lot yet?
[15:24] <sliddjur> qman__: yeah kind of
[15:24] <sliddjur> qman__: can i pause a process?
[15:24] <sliddjur> 18098 pts/1    00:00:00 dpkg
[15:25] <qman__> dpkg locks the database until it's done, the way to get out of it is to kill the process and then do a dpkg --configure -a
[15:25] <qman__> but in this case you probably don't want that
[15:25] <qman__> I would kill the process and try to apt-get install screen
[15:25] <qman__> then try to do-release-upgrade again
[15:26] <qman__> rather, screen do-release-upgrade
[15:27] <sliddjur> qman__: cant "move" process to my new connection?
[15:27] <qman__> no
[15:27] <qman__> it has to have already been set up to do so, with screen
[15:27] <sliddjur> ok.
[15:30] <sliddjur> qman__: i had to run dpkg --configure -a.
[15:30] <sliddjur> qman__: and now it continues where i was cut off :o
[15:32] <sliddjur> qman__: seems like it actually just finished release upgrade! wohoo
[15:35] <qman__> great
[15:36] <patdk-lap> ya, I have killed do-release-upgrade in the middle
[15:36] <patdk-lap> doing dpkg/apt-get  manually, and restarting do-release-upgrade suprissingly restarted it just fine for me
[15:37] <qman__> yeah, it's a surprisingly resilient process
[15:37] <qman__> I had one break in the middle but I don't remember how I resolved it, it did work out
[17:27] <sliddjur> I need to open one port for a specific IP in iptables, whats the command?
[17:27] <sliddjur> tcp
[17:30] <Armenia> hi
[17:40] <sliddjur> can someone port scan one ip for me?
[17:41] <sliddjur> nmap 84.210.170.149 -p5666
[18:05] <Tzunamii> sliddjur: http://nmap.online-domain-tools.com/
[18:06] <Tzunamii> Black belt in google-fu is a beautiful thing
[21:51] <rwb> Hi, I am having a problem with a server (ubuntu 12.04) getting past grub.  It is frozen on "Starting up ..."  It has just been through a long fsck..  How can I get it to boot?
[22:49] <Quest> hi
[22:49] <Quest> when a visiter visits a page. what information can be grabbed of the user? ie its ip,  operating system? what else? can a unique identification of its browser be grabbed or its mac address of his system and not the network router/swtich behind which it is?
[22:53] <shauno> Quest: you might find https://panopticlick.eff.org  interesting.  but no, you won't get a mac address.  you'll have the mac address of the last router on the path to you.  usually your own router.
[22:55] <stgraber> not completely true, you can get the MAC address of the visitor if you're using IPv6 without privacy extensions (as EUI64 addresses contain the MAC address in the last 64bits)
[22:57] <patdk-lap> shauno, that assumes his uplink uses ethernet
[22:57] <shauno> well, it's a safer assumption than ipv6 at this point :)
[23:04] <Quest> shadeslayer,  stgraber  ipv4 dont show mac?
[23:05] <stgraber> Quest: they don't, only the router closest to the source knows the MAC and the information isn't passed on
[23:05] <patdk-lap> mac is only used on the local *ethernet* cable
[23:06] <stgraber> not only ethernet, infiniband can also be used to carry IP packets and does have MAC addresses too, which is why I didn't limit my statement to ethernet
[23:07] <patdk-lap> infiniband doesn't use mac, except if using ethernet emulation on infiband
[23:07] <patdk-lap> but ya, it's not used for t1, ax25, ....
[23:07] <Quest> you said you can get the MAC address of the visitor if you're using IPv6 without privacy extensions (as EUI64 addresses contain the MAC address in the last 64bits)
[23:07] <Quest> stgraber, ^
[23:08] <stgraber> Quest: correct
[23:08] <Quest> so it will give mac of machine to the http server?
[23:08] <stgraber> for example, if you get a connection from 2607:f2c0:f00f:2700:2677:3ff:fe89:87bc, you can find that my MAC address is 24:77:03:89:87:bc
[23:09] <patdk-lap> you have to assume ff:fe means that, and not someone that just used that in their ip
[23:09] <patdk-lap> it's a pretty good guess though, currently
[23:10] <patdk-lap> I have no idea why people love mac addresses though
[23:10] <Quest> the ipv6 isnt the routers ipv6 incase the request(s) are coming from a network behind that router?
[23:10] <stgraber> usually not, until very recently IPv6 didn't have NAT and most implementations still don't use NAT (and as far as I'm concerned, really really shouldn't)
[23:10] <patdk-lap> the only thing guarrenteed (and not if using nat) is the ip address
[23:11] <Quest>  the ipv6 isnt the routers ipv6 incase the request(s) are coming from a network behind that router?  or the ip of the router will be sent for all requests?
[23:11] <patdk-lap> stgraber, using a lot of ipv6 nat for ipsec networks between our clients
[23:11] <stgraber> IPv6 has so many addresses that using NAT just causes pain. However most operating systems nowadays (including Ubuntu) use privacy extensions by default to avoid the issue of making everyone easy to track by just looking at the last 64bits of an IPv6
[23:13] <stgraber> patdk-lap: I never found any case where I really needed IPv6 NAT, I've had a few cases where I found it useful to use some of the non-routable (RFC1918-equivalent) addresses for point to point VPN links but the rest I always managed with properly designing the network so that everything is properly routed (even if some paths are routed over VPNs)
[23:13] <patdk-lap> yes, that assumes you can manage it
[23:13] <Quest> so theres no real way for tracking a pc uniquly that is using pv4 and is behind  a router?
[23:13] <patdk-lap> like I said, ipsec vpn between are clients networks, out of our control
[23:13] <stgraber> one case where NAT will likely be used is for tethering on cell phones as doing on-demand prefix delegation for those is rather tricky and most providers probably won't bother
[23:15] <stgraber> Quest: right. For http you can get away with some clever javascript+cookies combination like google and other ad agencies do, but it's not that easy and web browsers often evolve making it harder and harder (as far as I'm concerned, this is a very good thing, I don't like being tracked :))
[23:23] <Quest> hm
[23:23] <Quest> thanks
[23:24] <Quest> stgraber,  thats contrary in pv6 right?