[00:53] nijaba: How is the hosting for the survey coming along? [00:57] kirkland: I didn't see any further replies to our lsb list submission. There was one initial response to which I responded: http://lists.debian.org/debian-lsb/2008/05/maillist.html#00000 [01:28] owh: hmm, that's disappointing [01:28] owh: remind me... do we have a debian bug filed? [01:29] kirkland: Well, there are the two bugs that were there in the beginning, but they seem to have languished. Lemmie look for the numbers. [01:32] kirkland: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=291148 and http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=208010 [01:32] Debian bug 291148 in debian-policy "status action for init.d scripts" [Wishlist,Open] [01:33] http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=208010 [01:33] Debian bug 208010 in debian-policy "Require init.d scripts comply with LSB" [Wishlist,Open] [01:33] owh: hmm, i think that bug should be against lsb-base [01:33] rather than debian policy [01:33] owh: I'm going to open a new bug, and reference that one [01:33] kirkland: No, the discussion went right off the rails - in my opinion - and became a discussion about policy. [01:34] kirkland: A bit like it did here :) [01:34] owh: what's your "no" in reference to? [01:34] owh: "No" it should not be against lsb-base [01:34] kirkland: Ah, I see your confusion :) [01:34] kirkland: Yes, there should be a bug against lsb-base, but it was already that initially, it was changed to a policy bug. [01:35] IIRC [01:35] owh: ugh [01:36] See what I mean? [01:36] Or did I misunderstand what happened? [01:38] owh: okay, well, here's what I'm going to do.... [01:38] owh: I'm going to open a new bug against lsb-base in debian [01:39] owh: attach our patch [01:39] reference the policy bug for history [01:39] Both of the bugs I showed you? [01:40] As in, reference both? [01:40] Would it not be simpler to reference our LP bug with its attachments? [01:40] Or is that not a done thing? [01:42] kirkland: FYI, ours is https://bugs.launchpad.net/ubuntu/+source/at/+bug/203169 [01:42] Launchpad bug 203169 in samba ""status" function for init scripts" [Wishlist,Confirmed] [01:42] * owh had it open in a window :) [01:43] kirkland: Hmm, wonder why it shows as a samba bug - perhaps because that's the only one Confirmed. [01:54] Policy is the right place to end in Debian, but in Debian the Policy describes the facts on the ground, it doesn't generally lead them. [01:55] Sorry, ScottK, I don't understand what you mean. [01:56] No one will write it in the official policy until after there is a significant body of packages doing it. [01:56] Policy reflects the established consensus. It describes the current situation, not some future ideal. [01:58] ScottK: So, are you saying that kirkland and I will need to approach every single package maintainer and ask them to implement this before we get any traction? [01:58] owh: I think we need to get our 10-line patch accepted into lsb-base FIRST [01:58] kirkland: I agree. [01:59] owh: once that's there, we send our 3-4 line patches to a few dozen core service's init scripts [01:59] owh: and perhaps other people will patch others [01:59] owh: considering there hasn't been any movement on our patch to the debian-lsb list, i think we should file a bug (with patch) against lsb-base [01:59] kirkland: So, the lsb-base package maintainer hasn't responded, neither has the lsb-list, what's next? [02:00] owh: a bug against lsb-base with patch [02:00] kirkland: The pessimistic side of me suspects that it will be marked as a dupe and shifted to policy, but I agree with your process. [02:00] owh: perhaps [02:00] owh: Not all, but I think kirkland's got it about right. [02:01] owh: but here we're dealing with real and practical code, and not a policy document [02:01] ScottK: cool, thanks for the validation, i was wondering what advice you might have [02:01] Yes, and that is different from the two bugs so far. [02:02] I'd also check and see if any of the lsb-base maintainers are active in Ubuntu and if so approach them directly. [02:02] Now that is a top idea. [02:02] This might be a good release goal for Lenny +1. [02:02] owh: i thought you emailed the maintainer directly as a first step? [02:03] kirkland: I did and received no reply. [02:03] kirkland: Twice. [02:03] owh: right, then you/we emailed the debian-lsb mailing list [02:03] Is there a single maintainer or are there several (including uploaders - those are co-maintainers). [02:03] * ScottK runs off. [02:03] kirkland: Yes. [02:04] * owh wonders if ScottK is talking about real running or virtual running :) [02:04] kirkland: Let me have a squiz at the changelog. Perhaps we can get some ideas from there. [02:04] owh: okay [02:05] owh: ScottK: http://pastebin.ubuntu.com/15201/ [02:06] attaching the patch [02:07] kirkland: In the gutsy changelog I have there are lots of ubuntu people, Lamont Jones, Tollef Fog Heen, Michael Vogt, SJR, Matthias Klose. In Debian there are Christian Perrier, Chris Lawrence. I emailed Chris Lawrence and he features most in the Changelog. [02:09] kirkland: That bug report looks good to me. When you've submitted it, I think we should add it to LP. Perhaps it should also contain a link to our LP bug. [02:09] owh: definitely [02:10] owh: I've submitted it, it takes 15 minutes or so to hit the system [02:10] As in, it should link both ways. [02:10] owh: yes, absolutely [02:10] I mean, there's no point on doing this kind of stuff in isolation :) [02:18] owh: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=483285 [02:18] Debian bug 483285 in lsb-base "lsb-base: lsb status_of_proc() function" [Wishlist,Open] [02:19] I'll add it to ours [02:19] kirkland: Do you think I should also add the other Debian bug to our bug? [02:20] owh: i'm doing it now [02:20] heya people [02:20] kirkland: I'll stop typing then :) [02:20] owh: updated https://bugs.edge.launchpad.net/debian/+source/lsb/+bug/203169 [02:20] Launchpad bug 203169 in samba ""status" function for init scripts" [Wishlist,Confirmed] [02:20] emgent: Salutations. What is your challenge today? [02:21] good day [02:21] owh: ? [02:21] kirkland: Do you think we should also add the other Debian bug: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=208010 [02:21] hi ajmitch :) [02:21] Debian bug 208010 in debian-policy "Require init.d scripts comply with LSB" [Wishlist,Open] [02:22] owh: hmm, perhaps [02:22] owh: but the debian-policy bugs aren't nearly as pertinent [02:22] emgent: Well, when a new (new to me) person turns up and greets the room, I try to be helpful :) [02:22] kirkland: No, but it shows that we mean business :) [02:23] As in, we've done our homework as best we can. [02:23] owh: sure [02:23] owh: you can add it then, if you want [02:23] owh: gh [02:23] owh: great, so you can help me with any problems I may have? :) [02:23] kirkland: some news from upstream and debian peope to cron ? [02:24] ajmitch: No, I know you - you're on your own :) [02:24] :( [02:24] emgent: i haven't seen anything, sadly :-( [02:24] argh :-\ [02:24] ok we will wait :P [02:24] ajmitch: Ask and perhaps yea shall receive. [02:25] emgent: http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=482284 [02:25] Debian bug 482284 in cron "crontab -e should use /usr/bin/sensible-editor" [Normal,Open] [02:25] yep i saw [02:25] owh: well I don't have any ubuntu server problems today [02:25] it`s in lp tracker [02:25] ajmitch: Well then you'll just have to take your chances :) [02:26] * lamont still has mixed feelings about the sanity of init.d status options [02:26] kirkland: Hmm, LP doesn't like two bugs for a distro about the same thing with a different ID. [02:27] owh: okay, then leave as the one with the patch to lsb-base [02:27] lamont: I recall you making a comment about set -e in the code, indicating that echo $? returned nada. If I recall that report from you correctly, can you elaborate because I don't understand it. [02:27] lamont: what are your objections? [02:27] owh: that was slangasek [02:28] kirkland: Yeah, it won't take two debian-policy things. [02:28] kirkland: Hmm, my memory is fading, I'm getting old :| [02:28] owh: review the comments to https://bugs.edge.launchpad.net/ubuntu/+source/lsb/+bug/203169 [02:28] Launchpad bug 203169 in lsb ""status" function for init scripts" [Wishlist,In progress] [02:28] kirkland: I was just reading and noted it - my screen is on a satellite link and it was still refreshing. [02:29] kirkland: lamont commented that it needs a full spec and perhaps should be in the start-stop daemon. [02:30] lamont: owh: the functionality that is needed is described in the lsb spec itself [02:30] kirkland: I agree. [02:31] kirkland: To be fair, we're just providing step 1 of that, not even the full implementation, but at least the structure to hang the rest off. [02:31] lamont: owh: and I'm not clear on how this would work in the start-stop-daemon... this is a 3rd party evaluation of whether or not a service is running [02:32] owh: what do you know of that we're missing? [02:32] owh: if the script is running with 'set -e' active, then any non-zero return results in the immediate exiting of the script... (see any of the various posix-like shell manpages...) [02:32] lamont: Even in a function call? [02:32] the status of -e is inherited, yes. [02:33] kirkland: Well we're only providing the infrastructure for the LSB spec. We don't provide the actual return codes for all the codes that are specified. [02:33] owh: right [02:34] rather, I expect that the function would immediately exit with the non-zero status [02:34] lamont: Hold on, that makes no sense to me, we're not talking about an exit with a return code, we're talking about a function return. [02:34] lamont http://launchpadlibrarian.net/12710633/lsb.status.debdiff [02:34] That's the lsb patch we're talking about. [02:35] -mix 790 : sh [02:35] -mix \! : foo() { false; echo bar; } [02:35] -mix \! : set -e [02:35] -mix \! : foo [02:35] -mix 791 : [02:35] hmm [02:36] lamont: So, you're saying that the pidofproc $daemon >/dev/null will terminate and never get to the next line? [02:36] so, if the calling script is running with set -e (as is recommended for init.d scripts), and the daemon is not running, then you'll get nothing but effectively "pidofproc $daemon >/dev/null; return 1" [02:36] exactly [02:36] Crap [02:36] and this is why god invented if statements [02:36] If that is the case then, the rest of the lsb code is also borked. [02:38] given my experience with the folks who tout lsb initscriptage, I find that completely unsurprising [02:38] * lamont had the privilege of fixing a whole bunch of lsb-ized initscripts in the hoary timeframe [02:38] and that's the most annoying part about this whole experience [02:39] we're not touting lsb [02:39] i don't have any particular love for it [02:39] kirkland: I never said you were [02:39] So, why are we doing this thing? [02:40] i simply want to be able to determine if and when most any daemon is running or not [02:40] I mean, I realise that it's needed, but its also borked. [02:40] the lsb code just most-reminds me of something written by first-year CS students and debugged into existance: it generally works [02:40] lamont: It seems that our code has the set -e issue, but the pidofproc() function has the same issue. [02:40] the place that makes the most sense to put this functionality is in /lib/lsb/init-functions, as that is inherited by almost any init script [02:41] Hmm, so perhaps we need to "fix" pidofproc as well :( [02:41] fixing lsb stuff would be a major step forward for lsb, I believe [02:41] kirkland: and yes, in ithe init-functions makes lots of sense [02:41] of course, since that's sourced by everyone and his brother, making it slow would also get you attention from the "I can shave .37 seconds off the boot process" people [02:42] So, lamont, is what kirkland and I are attempting just really the thin end of the wedge in your opinion, or are we just wasting our resources? [02:42] s/process/time/ [02:43] I recognize the value of being able to see if a daemon is running. and putting it into lsb/init-functions makes the most sense to me, given the push both in ubuntu and debian towards lsb-ized init scripts [02:43] hrm... make that "I recognize that many people see value in being able to"... [02:43] ROTFL [02:44] Stop making jokes like that, my stomach hurts. [02:44] at the same time, I'm not sure of the overall value to the security of the internet in having my grandmother be able to think she's doing a good job of administering a public-facing server [02:44] pidofproc is a function that is provided by /lib/lsb/init-functions, and is used in various places [02:44] kirkland: But if what lamont showed us is true, then it is also borked, unless I can't read the code properly. [02:45] owh: test your patches [02:45] owh: apply the init-functions patch [02:46] kirkland: Specifically, this: /bin/pidof -o %PPID $1 status="$?" [02:46] owh: pidofproc needs a "if" in front of the /bin/pidof call, with the appropriate other changes, if it's to survive a -e script [02:46] owh: then patch apache [02:46] lamont: Are you talking about the same bit of code as I am? [02:46] exactly [02:46] lamont: That's a relief :) [02:46] just search for =.*\$\? [02:46] :=( [02:46] Yeah. [02:47] There's more :) [02:47] er, =.*\$? [02:47] I like the "... || status=$? [02:47] nice construct [02:47] see killproc() [02:48] that's a clean and fixed version of how to set status=$? [02:48] lamont: yes, that's concise [02:49] lamont: Which version of the code are you looking at, I'm looking at a gutsy one that is to hand. [02:49] is there any way to use a program as IDE on windows, and to work on files on a linux server? like u could do with the fish protocol? the only thing I need is a good IDE for windows which supports makefile projects (like eclips eg.) and I need a way to work on my files remotely in real time - does anyone know of a solution for that? maybe something like a webdrive but, i don't want to and can't [02:49] use webdav. [02:49] any idea? [02:49] hardy [02:49] c1|freaky: Samba [02:49] c1|freaky: samba [02:49] it's not in my LAN [02:50] i don't want to use a VPN [02:50] c1|freaky: Set up an SSH tunnel. [02:50] ssh tunnel for what? [02:50] c1|freaky: To tunnel ftp mount for example. [02:50] or sshfs [02:50] there sadly seems to be no support for fish [02:50] c1|freaky: Where? [02:50] can i mount ssh fs on a windows machine? [02:51] in windows [02:51] dunno, haven't used windows in 8 years [02:51] c1|freaky: Run a virtual machine or a cygwin. [02:51] no i dont want that [02:51] c1|freaky: http://people.ubuntu.com/~kirkland/search.html?cx=003883529982892832976%3Aly2fmeg302s&cof=FORID%3A9&q=sshfs+windows&sa=Search [02:51] c1|freaky: Or boot from a Live CD :) [02:51] ive got ubuntu installed (with gnome) on my desktop pc [02:52] but i want to be able to code in windows too [02:52] and in both operating systems at the same place and i dont want to format my other partition to fat32 [02:52] i could use SVN [02:52] but i dont have a simple method to manage svn repos [02:53] and besides, svn sucks [02:53] c1|freaky: Eclipse comes with one built in :) [02:53] lamont: There is that :) [02:53] owh: with what built in? [02:53] er, I mean, svn fails to work well for anything more than a single-developer [02:54] why should it fail its made for that [02:55] c1|freaky: it creates branches fast (so what). merges of any complex source project in svn are a multi-day (maybe just several hour) nightmare, since it doesn't actually track patchsets [02:55] the current generation of version control systems (bzr, git, hg) all track branchsets, and therefore give you fast merges (which is what we do care about) [02:55] lamont: I'm looking at the hardy version of init-functions, I cannot see what you showed us as the nice construct. [02:56] owh: line 129 [02:56] /sbin/start-stop-daemon --stop --signal "$sig" --pidfile "$pidfile" --quiet $name_param || status="$?" [02:56] (and friends) [02:56] and above that is "status=0", to complete the assignments [02:56] right, as lamont says [02:57] Hmm, while I believe you both, I must be going blind. [02:57] i gotta find a good solution for me and other people. the best thing would be direct remove access but webdav fails to support https on win vista [02:57] *remote not remove [02:58] apparently hg (aka mercurial) provides some pretty good windoze support [02:58] * lamont has never used hg though. [02:59] damn [02:59] i guess the only and one solution is using linux then? [03:00] hello people! [03:00] ._. again [03:00] lol [03:00] kirkland: Line 129 is: if [ -z "$sig" ]; then [03:00] well thank you for the help [03:00] * owh feels very dumb. [03:00] c1|freaky: hg has a windows client (possibly a server), and provides a UI on both platforms [03:01] owh: you're looking at /lib/lsb/init-functions ? [03:01] kirkland: Yup. [03:01] lamont: yes but trac doesn't support it. is there any software like trac for hg? [03:01] owh: grep "|| status" /lib/lsb/init-functions [03:01] kirkland: Did it change between the final beta and current? [03:01] owh: lsb-base 3.2-4ubuntu1 [03:02] c1|freaky: bugzilla comes to mind... [03:02] or what does trac do? [03:02] lamont: Ah ha. Different version :( [03:02] owh: welcome to hardy... :-) [03:02] lamont: example for trac: http://scm.code-1.de [03:02] lamont: Nah, the VM I launched is running hardy, final rc, no updates since. [03:03] owh: I hope that VM isn't doing anything with SSL then... .:( [03:03] * owh wasn't expecting any updates in lsb given the flack we got for even thinking about asking to add our function :) [03:03] c1|freaky: ah - wiki+bug tracking... [03:04] with an almost content free description, FTW! [03:04] lamont: No, it's a machine on my laptop, just to test bits, it's not connected to anything, provides no services, is not in production in any way. [03:04] lamont: yes, a scm and it only supports subversion and bzr with a plugin [03:04] But I'm not going to download 135Mb to update it as it's now asking me :( [03:05] c1|freaky: sounds like an interesting challenge then... windows tends to not be well supported in opensource stuff, simply because the bulk of opensource people don't use windows, so there's little motiviation to deal with it [03:05] lamont, kirkland, given the set -e comments, I'll download the latest lsb source and make some patches. [03:06] owh: sounds good [03:06] * owh has to go visit a client, so I'll be doing this tonight at the earliest. [03:06] lamont: yea ... i could use subversion ... but if it's true that subversion isn't good when working in teams that sucks [03:06] lamont: Thanks for taking the time to explain, appreciated. [03:06] owh: pub lunch with a client? ;) [03:06] ajmitch: I wish :( [03:06] owh: latest lsb source is definitely where you want to base your changes, and it doesn't help that there's a kernel update in hardy-updates... [03:06] ajmitch: Apart from that, it's 10 am around here :) [03:07] anyone in here familiar with accessing a non-encrypted shared network? [03:07] owh: glad to help [03:07] its 4 AM here [03:07] owh: ah, you're a bit further west, I take it [03:07] linuxmademecrazy: depends on what you mean by "shared network" [03:07] ajmitch: Further west than me gets wet fast. [03:07] I mean, we all plug our machines into something that exactly matches your description, but I doubt that's what you mean,.. [03:07] owh: okay, yeah, fixup pidofproc [03:07] what exactly is the question? [03:07] owh: i'd say you should patch that separately [03:08] owh: open a LP and a Debian bug, link them together [03:08] lamont, Umm, the network authentication is "shared". It's a shared network key. I have a lame motorola router that has been setup in a pretty unique way =/. [03:08] as lamont said, that should be fixed regardless of whether or not they take our status_of_proc() function [03:08] kirkland: Yeah, I'll give the whole thing the once over. Make our patch separate and make a bug/fix/patch for ubuntu and debian on the issues raised by lamont. [03:09] owh: and remember: don't call the muppets. :-) [03:09] lamont, the network authentication is "shared" and the data encryption is "disabled" and I have to provide it with a network key and an associated key index number. [03:09] * owh is not familiar with that reference :) [03:09] * lamont waits for linuxmademecrazy to describe the underlying authentication mechanism [03:09] lamont, this is the windows properties panel http://img249.imageshack.us/img249/3467/47452358cn6.jpg [03:09] owh: don't worry. [03:10] lamont: I suspect it has some rubberneck connotations though :) [03:10] linuxmademecrazy: ah. 802.11-ish network [03:10] lamont, I'm kinda, not that smart when it comes to linux, my strong suite is php T_T [03:10] lamont, you're familiar with it?!?! =D [03:10] familiar with 802.11. nfc what windows means with what they've said on the control panel though [03:11] linuxmademecrazy: mentioning that it's a wireless network is relatively important :) [03:11] linuxmademecrazy: I expect that it's some incarnation of WPA or WEP [03:11] ajmitch, you might be right.... xD [03:12] Later all. [03:12] linuxmademecrazy: so the question you want to be asking is "how do I configure linux to talk to a wireless network in a manner that windows describes as 'shared authentication, data encryption disabled'?" [03:12] yeah, i'm calling it a night too [03:12] adios [03:12] and I don't know what the answer to that question is [03:12] g'night kirkland/owh [03:12] lamont, yeah, pretty much. [03:13] linuxmademecrazy: to make your life more challenging, people who know linux well tend to not know windows well, and vice versa [03:14] lamont, ah, makes sense, I've never ever used linux before today, but I figure I should start learning, because I want to know how professional server management is done, and windows sucks when it comes to serving up some data. [03:15] lamont, having no internet on my ubuntu installation makes it a lot more challenging. [03:15] heh. yeah [03:16] I've pretty much spent all day going through all the settings in the wireless network configuration GUI [03:17] The biggest problem I keep running into, is that the GUI doesn't want the associated index key that is used in association with the network key. [03:17] So, I figure my solution would have to deal with the terminal/console [03:18] which, I have no idea about the available commands. [03:23] linuxmademecrazy: google can sometimes be your friend... dunno [03:25] lamont, I love google, but linux documentation is very messy, and cluttered, compared to what I'm used to dealing with. I think I might have come up with one possible solution, but I don't want to switch back to my other harddrive until I have 2 or 3 possible solutions, becuase I hate wasting time for switching. This would be soooooo much easier with a second computer xD [03:26] yep. [03:26] linuxmademecrazy: or even vmware installed on the windows box with linux running inside that [03:26] once I get this all working and I start figuring out linux, I'm going to start documenting this crap <_< [03:26] * lamont tries to decide if he really just said that... [03:27] I can't just VMware my wireless card =/ [03:27] linux needs more people who are willing to write docs [03:27] Yeah [03:29] * lamont waits for bind9 to finish building so that he can figure out if his current challenge is trivial, or a royal pita [03:29] First thing I'll be doing is writing a TUTORIAL on how to start off, from scratch, setting up a fully operational apache2 web server with php5, and RoR, networking (both wireless and lan) installation will be included. [03:30] After that, I'm scratching down all the terminal syntax and comands in a similar documentation style that php.net has. [03:31] Well, time for the moment of truth, I'll be back =D [03:33] hello , dhcp3-server dont start after reboot, how to fix this corect way ? [03:45] m1r: depends on what the error is behind "dont start" ... [03:46] what does /var/log/daemon.log have to say on the subject? [03:46] New bug: #235378 in samba (main) "hardy samba critically slow" [Undecided,New] https://launchpad.net/bugs/235378 [03:49] larmont , one sec [03:51] lamont: it is working fine after i start it manually , dont see nothing strange on daemon.log [03:53] m1r: meh. /var/log/syslog [03:53] sorry about that [03:53] ok, sec [03:55] lamont: this one is strange line : May 28 04:20:40 m1r0-desktop dhcdbd: message_handler: message handler not found under /com/redhat/dhcp/wlan2 for sub-path wlan2.dbus.get.domain_name [03:55] yeah, but it's also not dhcpd [03:55] that's dbus having a small fit [03:56] sigh. so much for my challenge being trivial. [03:57] m1r: that'd be "not your issue" [03:57] lamont , yes i check further, long list in syslog :) [03:58] m1r: so anything from dhcpd? [03:58] dont see :/ [03:58] only for wlan2 address reciving [03:59] lamont: i think that it isnt enabled to start at all [04:00] lamont: where can i check for dhcp3-server being run on start ? [04:01] Well, that was a fail [04:02] m1r: oh. when you said "run it manually" you didn't mean "/etc/init.d/dhcp3-server start" ? [04:03] see /etc/default/dhcp3-server as well - that says what interfaces to listen on, for starters... [04:03] lamont: i can start it normaly with start command, but it dont start at boot , if i managed to explain good [04:03] lamont: it is set to eth0 in default/dhcp3-server [04:03] ok. what does ls /etc/*.d/*dhcp3-server say? [04:04] should be a K file in rc1.d, and S files in rc[2-5].d [04:04] if the links are there, then I suspect that it's trying to start before eth0 exists... if the links aren't there, then that's the issue [04:05] ok one min to check [04:06] it says /etc/init.d/dhcp3-server - INTERFACES="eth0" [04:07] rc1.d have K20dhcp3-server file [04:08] should i check S files also ? [04:08] from 2-5 ? [04:08] 2 also have @K file [04:09] all have dhcp3-server info [04:10] lamont: how u mean : then I suspect that it's trying to start before eth0 exists... [04:10] does rc2.d have S40dhcp3-server? [04:11] sec [04:11] lamont: seems that it dosent [04:12] lamont: only K@... [04:13] interesting.. I expect that "mv /etc/rc2.d/K40dhcp3-server /etc/rc2.d/S40dhcp3-server" will make the problem go away, at least until maybe the next time you upgrade dhcp3-server [04:13] as to how/why it got that way, no clue [04:14] lamont: i was instsalling EBOX , so maybe that is problem [04:14] lamont: i turned dhcp server OFF on ebox and maybe it deleted it [04:15] actually, it changed it from start to kill for runlevel 2 [04:15] which is to say, "don't start dhcp3-server for runlevel 2, kthx" [04:15] hence the behavior you're seeing [04:16] m1r: runlevel 2 is the default, you see... [04:16] ah [04:17] and K files get told 'stop' and S files get told 'start' upon intering runlevel N, based on /etc/rcN.d/ [04:17] sounds little messy but i think i got the point :) tnx for this tip lamont [04:18] m1r: man update-rc.d - the right way to fix it is burried in there _somewhere_, I expect [04:19] lamont: so , i am not sure should i edit it manually or try through ebox interface [04:20] well, if you do it manually, there's a chance that some future dist-upgrade will undo it... not sure [04:20] * lamont has never used ebox, nor does he plan to use it. [04:23] lamont: yes i see, i will try first over ebox, then if it dont work i do manual add to rc2.d just for temp fix. tnx for help m8, really appriacte it [04:24] np. have fun [04:24] yes you too , tnx m8 :) [04:31] i go reboot to test [04:35] lamont: it's fixed, tnx [04:35] the dirty way :P [04:37] heh [05:50] hi ? [05:51] hi [05:51] iam searching a ubuntu-server [05:51] apropiated [05:51] for my pc its a old cpu [05:51] intel pentium mmx at 233 mhz, 32 ram [05:51] Christian: that might be a bit low even for server [05:52] Christian: but it could run i think [05:52] Christian: I have a PII/233 with 192MB of RAM acting as a print server [05:53] yeah [05:53] lamont, that is powerfull machine :D loads of ram [05:53] amazingly, it hasn't actually used any swap [05:53] m1r: it's acting as a CAT5<->USB adapter. :) [05:54] :D [05:54] what little gadgets can do ;) [05:55] Christian: with 32MB, you may find that you have too little RAM [05:56] m [05:56] lowest i installed was 350mhz 128 ram if i recall [05:56] i have a hd [05:56] and have installed hardy heron [05:56]