[06:37] mornin superfly and other peeps [06:43] heya Kilos [06:43] morin Kilos and superfly and all. have a good day all :) [06:43] hi kbmonkey [06:44] heya kbmonkey [06:51] Maaz, coffee on [06:51] * Maaz washes some mugs [06:51] Maaz, move it [06:51] Don't rush me Kilos . Making decent coffee is an art [06:52] Maaz: coffee please [06:52] superfly: Okay [06:52] hows the family superfly [06:52] same ol' same ol' [06:52] lol [06:55] Coffee's ready for Kilos and superfly! [06:56] thanks, Maaz [06:56] superfly: Sure [06:57] Maaz, ty [06:57] You're Welcome I'm sure [07:25] morning superfly and everyone [07:25] Maaz: coffee please [07:25] sakhi: There isn't a pot on [07:25] sakhi: coffee denied :) [07:36] heya sakhi [08:42] good morning everyone [08:43] inetpro: hi... Mr. Kilos was worried about you... seems you have been missing for a while... [08:44] yes nlsthzn stretching data as much as possible [08:44] so offline most of the time [08:44] lo inetpro [08:51] nlsthzn: sup [08:51] hi Kilos [08:55] inetpro: not to much... thanks :) [09:00] nlsthzn: how's the LoCo? [09:01] superfly: slowly starting to take shape... will be having a first IRC meeting this coming Sunday, to discuss some logistics etc... several passionate people in the loco currently... looking good [09:04] help [09:05] stinking help [09:05] how do you boot a win 7 pc into dos mode or command line mode please [09:05] Kilos: have you tried F8 at boot-up? [09:06] boet says it does nothing [09:06] i told him try that [09:08] Kilos: I haven't tried it with win 7 ... but if memory serves that still worked in vista... [09:08] * superfly hasn't used anything beyond XP [09:08] i used it with all of them up to xp. dunno what win7 has doen [09:08] done [09:13] boet put his win7 on his daughters laptop and didnt go register within 30 days [09:16] Kilos: but surely he should still be able to do it...? [09:17] it wont let him in because it has expired [09:17] grrr [09:17] not even to activate?! [09:17] nope [09:19] Kilos: that is silly... expected but silly [09:19] yip [09:19] winsucks as usual [09:40] whats wrong with chatting here instead of onna forum [09:41] * Kilos refers to lists [09:41] Kilos: huh? [09:42] they talking about starting a new forum [09:42] let them start one, if it works out great [09:42] Kilos: who? [09:42] we have even chatted about hoenders here [09:43] irc is very different [09:43] and some people don't like the barrier of mailing lists [09:43] they don't even have to start a forum [09:43] forums exist already [09:43] well they did last time I checked [09:43] yeah, i know, but they want their own domain [09:44] dont ask why :/ [09:44] and nobody was using them [09:44] s/nobody/very few/ [09:44] wikus and christof and maia and more peeps [09:44] * inetpro will continue using irc [09:44] https://lists.ubuntu.com/archives/ubuntu-za/2011-April/007446.html [09:44] * marcog too [09:44] irc is kiff and you can always go private for other info [09:45] there's nothing wrong with forums though [09:45] from irc it's easy to link to anywhere [09:48] it also seems a reason for them wanting a forum is so they can ask off-topic questions [09:48] they feel a mailing list can't solve that [09:50] my only issue with IRC is it feels fleeting... always nice to go back to a forum and search for a specific topic... specific info... feels more lasting [09:50] http://logs.ubuntu-eu.org/freenode/ logs are public [09:54] nlsthzn: it would be nice if people had more time to summarise and structure information for later reference [09:54] I don't have the time for sure [09:55] inetpro: info on the IRC logs you mean? [09:55] nlsthzn: even that [09:56] inetpro: sure, I have found on some of the better forums that many members will condense info into very handy how-to's brining the info of many threads into one handy place for all... but that needs time and commitment to be sure [09:56] nlsthzn: doesn't matter what medium is used, lots of information is available in bits and pieces all over the show [09:56] stack overflow solves that problem really well [09:57] marcog: stack overflow? [09:57] and there's an askubuntu.com which uses the same engine [09:57] inetpro: look at askubuntu.com [09:57] inetpro: google helps a little with that... but only a little [09:57] it's a Q&A site, rather than forums [09:58] and that works really well for bringing everything together [09:58] marcog: hmm... I've seen it before and still have not really used it so much [09:59] inetpro: haven't used askubuntu myself, but i use stackoverflow heavily [09:59] i'm in about the top 2% of users [10:00] http://stackoverflow.com/ [10:00] marcog: nice [10:01] good to know that [10:01] most questions are answered within 10 minutes on SO [10:01] wow [10:02] marcog: so the idea is just to shoot with a new question regardless whether someone has asked it before? [10:03] inetpro: better to search for duplicates, but users are quick to mark duplicates if they find one [10:03] and 5 votes and the question is closed [10:03] it can be quite brutal for n00bs, but the end result is high quality [10:10] * nlsthzn hasn't been a fan of askubuntu... but stackoverflow sounds interesting [10:10] * nlsthzn sees it is specific for programmers... :/ [10:11] there are loads: http://stackexchange.com/sites [10:15] marcog: wow... that is a big list... [10:19] hmm... a simple search for 'Huawei E1820 ubuntu' presents useless results on both sides [10:35] I don't search for answers on a specific site. I don't really get that. I do a Google search, and see what looks relevant. [10:36] For a start, if you have a problem with an app on ubuntu, is it an Ubuntu problem, a Linux problem or a problem with the app? If you don't do a search that's general enough, you may be excluding the best sources of information. [10:37] It's like walking up to random people and asking "hey, can *you* help me with my problem?" [10:37] heh, i hate that [10:37] If I ask a question, I usually ask in the app-related forum. [10:38] however, on a ubuntu forum... asking ubuntu questions is a great way to get pointed in the right direction at least [10:38] Ubuntu is big enough that if there's a problem with app + Ubuntu the app people are likely to know about it, but not all apps are big enough that random Ubuntu people will know anything about them. [10:38] I would rather stab myself in the head with a fork than ask a question on the ubuntu forums. [10:40] confluency: sure, technical questions will not get much love their... but is great for us mere mortals just wanting to play back an mp3 file and such :p [10:40] Because the first five answers are going to be "Hey, i have no idea, but [here's a page of unrelated general instructions / here are my thoughts about a totally different app / here's how you do it on Windows]". [10:40] Unhelpful is always unhelpful, and bad advice is always bad advice. [10:40] Bad advice is *worse* when it's given to non-tchnical people. [10:42] If a newbie user cheerfully follows the first solution they find, and it happens to be a page of horribly convoluted instructions for doing something in a really hacky way, that is bad. Because two years later they will have no idea what they did, and neither will anyone else trying to help them. [10:43] And there is basically zero quality control on the Ubuntu forums to stop that kind of thing from happening. [10:44] true... I see what you mean... [10:47] this is why i prefer Q&A sites, the bad advice is downvoted and the good advice floats to the top [10:48] *sigh* waiting for this stupid array to background init [10:48] and heh, initializing arrays takes.... ages [10:48] Ongoing Progresses: [10:48] Background Initialization: Completed 8%, Taken 733 min. <== like, wow [10:49] at that rate, 152 hours :P [10:51] No matter what system you use, you need to have enough knowledgeable people using the site. If you had a Q&A site populated entirely by newbies, they'd keep upvoting stuff that worked even if it was a terrible idea, and you'd have the same problem. [10:52] yes, and SO really excels in that regard [10:53] confluency: I suspect that is the price you pay being so widely adopted by the masses [10:54] nlsthzn: not if the site is moderated effectively [10:54] marcog: hmmm... true... wonder if it is feesable on a forum the size of uf though? [11:13] wb Squirm [11:13] hey inetpro [11:17] hmm... how can I search my tweet history? [11:18] http://search.twitter.com/search?q=from%3Ainetpro+matrix [11:19] it doesn't go very far back though [11:19] marcog: it doesn't go back far enough [11:20] i'm not aware of anything that goes any further back [11:20] * inetpro looking for old stuff [11:21] sad that I can't even search my own tweet history, should perhaps keep a local backup [11:25] ahh, snapbird is what I need http://snapbird.org/ [11:29] then again, it's useless [11:30] can not even find the word reminder anywhere in my tweets [11:32] fortunately friendfeed looks like a good friend to have [13:23] * drubin shudders [13:24] * sakhi wonders [13:27] forums vs mailing lists. [13:28] Yeah. [13:28] marcog: who is adrian on irc? [13:28] confluency: like all for forums, problem is we have them no one uses them [13:28] drubin: adrian frith? [13:28] marcog: ya [13:28] drubin: htnl [13:28] irc nick? [13:29] *htonl [13:29] It doesn't really matter what protocol you're using as long as you have critical mass. If you don't have critical mass, you can complain about the protocol as much as you want, and people will continue not to use the one you prefer. [13:41] hrm, is there a way to manually set the modification time of a file? [13:45] touch maybe? [13:45] *nod* found it, touch -m [13:45] :d [13:46] trying to fix an rsync that went wrong so I don't have to resync an entire archive [13:46] :p [13:53] ugh whats the regex with sed to search and replace only the first 3 instances of something on a line [13:53] rather than the entire line [13:58] ^.{3} or something like that [14:14] root@mirror:/diskspace2/opensuse/opensuse/repositories# sh /diskspace5/touchscript.sh [14:14] Touched 20 thousand files [14:14] Touched 40 thousand files [14:14] heh wheeeeeeeeeee [14:14] this is gonna take a while [14:14] it needs to go through 580 thousand files :p [14:15] 200k down so far [14:15] lol [14:18] mass timestamp violation [14:18] :/ [14:18] lol, fixing an rsync that went wrong, I ran the wrong rsync command and it didnt keep the modification times when transferring between disk arrays [14:18] which resulted in a major fuckup :) [14:19] ouch [14:19] because now when you rsync from the source it wants to redownload everything [14:19] so resetting all the time stamps to the source time stamps [14:19] heh, its down 400k so far [14:19] 180k to go [14:19] not doing too bad [14:19] lol, fast machine [14:20] I did an rsync -r blah.blah::blah/ outputted the results to a text file, ran some sed/awk shit on it to turn it into a touch script and just running the touch script now [14:21] 500k down, 80k to go :P [14:21] then I retry the rsync from the source and hope I havent screwed that entire mirror doing this [14:21] LOL [14:21] will find out in a second [14:22] certainly be faster than trying to resync a 1.2 terabyte mirror [14:23] * Symmetria starts the rsync again :P [14:23] and now I pray lol [14:26] :P well, encouraging so far that it aint tried to download EVERYTHING again so far :P lol, will se when it gets to the point where I cancelled the rsync last time [14:33] wooot [14:33] it seems to have worked [14:35] congrats :) [14:44] heh, so apnic officially goes into v4 soft landing policy because they outta space [14:44] ouch [14:44] * Symmetria waits for the chaos to start === Superhuman_ is now known as noctem === Superhuman_ is now known as noctem [15:15] woot, heh, fixed 3 more mirrors with bad timestamps using that touch method [15:21] Ongoing Progresses: [15:21] Background Initialization: Completed 11%, Taken 1008 min. [15:21] wow [15:21] I am probably going to kill the perfomance on mirror slightly and crnak up that speed [17:09] Evening all [17:09] I am having a little problem with my OO [18:20] mazal whats the prob [18:20] maybe we can ask maaz [18:21] Maaz, coffee on [18:21] * Maaz starts grinding coffee [18:25] Coffee's ready for Kilos! [18:25] Maaz, dankie [18:25] Groot plesier Kilos my vriend [18:27] aw he be gone === acherv_ is now known as acr === acr is now known as acherv_ [20:33] *cough* forum or mailinglist? [20:33] g'evening btw :P [20:46] bah! when will quassel have a CLI client? [20:48] nuvolari_: hiho [20:49] nuvolari_: you can always try something other than quassel [20:49] inetpro: like? [20:49] hello btw :P [20:50] I'm having a 2nd look at irssiproxy [20:50] but that is a lot of work compared to quassel [20:50] nuvolari_: I guess what you want is to connect and be able to see the history? [20:51] inetpro: correct :P [20:51] nuvolari_: just connect with irssi or weechat via a server that is always on [20:51] with gnu screen [20:52] ye, used that in the past [20:52] hmm, wonder how lag is these days [20:52] I connected via GPRS the last time I tried [20:55] nuvolari_: IRC does not use all that much of data so even GPRS can be workable [20:55] inetpro: not the data usage, it's the lag [20:55] o.O server update. Hope everything is still ok. [20:56] bbiab [20:56] nuvolari_: true, can be very frustrating [20:57] not so slow after all [20:57] :> [20:58] I'm impressed actually [20:58] it used to be terribly slow === nuvimob is now known as nuvolari_ [21:05] nuvolari: wb [21:05] inetpro: thanks [21:05] just struggling to get my outojoin working [21:05] *autojoin [21:05] nuvolari: weechat? [21:06] inetpro: ya [21:07] nuvolari: /set irc.server.oftc.autojoin "#channel1,#channel2" [21:08] inetpro: I just noticed that I'm running on a very old distro here... :O [21:08] 9.10 [21:08] so I only have 2.6 available [21:19] yo Squirm [21:19] hmm. upgrade or bed? [21:20] hi [21:46] hi all . i should possibly lurk/listen a bit longer, but for those of you who have tried 11.04 - what are your thoughts? [21:47] haven't played with it [21:48] all my stuff is server based, so its all on LTS versions [21:48] k [21:48] that and I try and avoid upgrading things that are working :) [21:48] sure :) [21:49] but generally speaking, the LTS versions do what they are meant to, I mean, mirror.ac.za runs the LTS version and we've had no issues [21:49] (well, not since we put in the new hardware, but I think the old hardware had serious issues) [21:50] yes, i've come across issues with older hardware too, i think [21:50] heh, the 32 bit version was fine, but the 64 bit version was *NOT* happy [21:51] but it seems to be fine on the new system [21:51] hmm. k . thats good to know [21:52] talking about mirror, I need to mirror more stuff [21:52] LOL upgraded the disk space because we were running low, but the disk space upgrade was so large I now need to find some way to use SOME of it :) [21:52] :P I have 33 terabytes of free disk space right now [21:52] wow [21:52] :) [21:53] Using 31 TB [21:53] oh, 32 [21:53] so how large is a ubuntu mirror ? [21:53] (machine has 65TB total) [21:53] ubuntu? *hrm* lemme check [21:54] 1009 gigs for the entire thing [21:54] k. ta [21:54] so heh, 15 gigs shy of a terabyte [21:54] its not bad, sourceforge is the biggest of the mirrors [21:55] think sourceforge is 13 or 14 terabytes [21:55] whats on the 31TB at mirror.ac.za [21:55] there's a local sourceforge mirror? [21:56] noobish. sorry [21:56] k [21:56] went and saw. [21:58] doesnt seem to be a sourceforge [21:58] Squirm errr yeah [21:58] mirror [21:58] there is [21:59] if you try and download from sourceforge [21:59] you will be redirected local [21:59] tenet.dl.sf.net is an alias for sourceforge.mirror.ac.za. [21:59] tenet.dl.sourceforge.net is an alias for sourceforge.mirror.ac.za. [21:59] ah, not on mirror.ac tho? [22:00] oh [22:00] panphried its an automatic redirect when you actually pull the files [22:00] it gets redirected to mirror.ac.za [22:00] k [22:00] heh it took an insane amount of work to get that organized :) [22:01] for such a hugely popular download source, they are... more than a little disorganized and slow when it comes to sorting things out :) [22:02] trying to see how big that mirror is now, but LOL, the du is taking one hell of a long time [22:02] is the website also hosted there? or just the files. cause it comes up with page not found when I enter sourceforge.mirror.ac.za [22:02] not related .. but arent they closed infrastructure as well? [22:03] squirm, just the files, sourceforge never hosts its website anywhere but centrally [22:03] they host the central website then do geographic redirection for file downloads [22:04] http://sourceforge.net/projects/vlc/files/1.1.9/win32/vlc-1.1.9-win32.exe/download <=== try browsing there, and you'll see it will redirect you to the local mirror [22:04] :) [22:04] *hrm* panphried what do you mean by closed infrastructure? [22:05] cool, just doesnt work too well if you're stuck on local bandwidth without being able to browse their site [22:05] basically, to be a sourceforge mirror you apply, stating bandwidth, size of hardware, location, etc, etc, then you go through this hectically long process that requires a lot of paperwork that drags on for 6 months or so, and eventually if they approve the mirror, they let you start synching it [22:05] root@mirror:/diskspace5# du -sm /diskspace4/sourceforge/ [22:05] 11537152 /diskspace4/sourceforge/ [22:06] actually slightly smaller than I thought, only 11.5 TB [22:06] still fairly large [22:07] heh still the largest mirror on the box, but only JUST [22:07] 11058528 /diskspace4/hg/ [22:07] that one is coming up on its heels fast :) [22:07] (human genome project mirror) [22:08] I suspect that the astronomy dataset mirrors will overtake both of those though soon, to the point where I suspect we're going to have to triple or quadruple the disk space in the system [22:08] scientific data is insanely huge :) [22:08] shu. [22:08] dev/sda3 532G 35G 470G 7% / [22:08] dev/sdc 13T 9.1T 3.8T 71% /diskspace3 [22:08] dev/sdd 26T 23T 3.2T 88% /diskspace4 [22:09] ^^^ thats our current disk space [22:09] (preceeding / stripped so that irc doesnt try make those all commands) [22:09] whats cern's size? [22:10] diskspace3 is due for an upgrade to 39 odd terabytes at the end of the year, won't do that before the 3TB drives are available for those arrays though [22:10] heh all the LHC data? [22:10] multiple petabytes [22:10] heh, so not many mirrors of that then :) [22:11] heh, to mirror the complete LHC data, I'd need a minimum of 4 racks with nothing but disk space in them [22:11] (I can do approximately 312 terabytes per rack) [22:11] just checking how many zeros in a peta. .. [22:12] petabyte = 1024 terabytes [22:12] (so 1024*1024 gigabytes) [22:12] k. so not too many orders of magnitude greater .. [22:13] heh, you could put together a really decent petabyte system with proper performance and decent hardware for about 3.3 million rand :P [22:14] thats with the raid cards and san boxes etc, to allow you to connect a petabyte to a single server without using NFS or other such bullshit [22:14] brb :) smoke [22:14] do you use zfs for that [22:14] k [22:18] using xfs [22:19] oh yes, no zfs on linux .. yet [22:19] *shrug* xfs works pretty well for us [22:20] i remember, ... if it aint broke :) [22:20] had no catastrophic failures on it yet :) [22:21] Symmetria: how do you handle making backups of all that data? [22:21] heh [22:21] we don't [22:21] its a mirror [22:21] if it dies, we resync [22:22] *shrug* its not like we're short on the bandwidth to resync [22:22] what if something goes wrong at the source? [22:22] I mean, we synched the entire sourceforge in 3 days [22:22] inetpro use another source :) [22:23] backing up a mirror is pointless, you want your mirror in sync with the primary, if you restore from your own backup its not in sync anymore [22:23] sure, was just wondering [22:23] besides which, the disk arrays in mirror.ac.za is heavily redundant anyway (they are all running raid-5 + hotspare) [22:23] lots of mirrors.. is there versioning with mirrors. [22:24] i mean, is there any versioning at the source [22:24] depends on what you're mirroring, I mean, mirrors do get outta sync on occasion when something goes wrong with the sync process, but the bigger projects have mechanisms to poll the mirrors to let you know if something is broken [22:25] sourceforge runs constant polls on our system and if it disappears they pull it from rotation automatically [22:25] and ubuntu emails me if our mirror becomes unavailable to its poller pretty quick [22:25] if source is poisoned .. do all the mirrors sync with that .. such that data is lost ? [22:26] panphried pretty much yes :) then its up to the source to fix itself so the rest of the mirrors can sort themselves out [22:26] k [22:26] which is why primary sync mirrors for things like sourceforge are so heavily restricted, if you aren't an official mirror you arent getting anywhere near it [22:26] What's the noise? [22:26] Oh it's my pillow calling... [22:27] anyway, gnight [22:27] goodnight [22:27] im off too [22:27] gnight :) [22:27] oh. i hear that too. good night [22:27] night all [22:27] heh :) sleep well guys [22:27] Im off to read a book while I wait for maintenance ot start [22:28] Symmetria: have fun watching you script [22:28] your* [22:29] enjoy :)