IRC Logs for #crux Thursday, 2012-10-04

niklaswemorning!00:47
*** ulioidle has joined #crux02:03
ulioidleHowdy anyone around interested in helping abit?02:18
*** spaceninja has joined #crux02:21
spaceninjais it possible to compile the scsi_wait_scan in the kernel and not as a exernal module?02:22
ulioidleNone that I see02:27
frinnstmorning?02:28
frinnstmorning is never a good thing! :)02:28
frinnstulioidle: dont ask to ask, just ask :)02:30
ulioidleHa02:30
ulioidleWell then02:31
ulioidleI had abut of a syntax error this morning typed wrong and all02:32
ulioidleEnded up playing a very large number of mp3s at once02:32
ulioidleWell attempting to. It did not work out so well02:33
ulioidleLet me go paste my dmesg output for you02:33
ulioidlehttp://paste.lisp.org/+2U5202:35
ulioidleI am guessing I screwed my file system02:35
ulioidleVarious things no longer work02:36
frinnstyeah you seem to have to have some issues with your harddrive02:37
ulioidlevim did an outstanding job though, the only thing that was open that still works, even managed to pull my code from swap02:38
ulioidleHaha02:38
ulioidleThat much I gathered02:38
frinnst:)02:41
ulioidleThe few instances I found of this error online pointed to a near dead hard drive, but I find it hard to believe that I could kill my drive this way02:43
frinnstI dont really follow what you did to be honest02:45
frinnstyou ran a script that failed and then had these errors?02:45
ulioidleAhh02:45
ulioidleNo, it did not fail, I did02:46
ulioidleScript worked exactly as I typoed it02:46
ulioidleInstead of playing the list of music sequentially it tried to run them all at once02:47
ulioidleSo a few thousand instances of mplayer all trying to start at once and each play a different song02:48
frinnstoh ok. but that is probably totally unrelated to your disk issues. maybe the stress caused the drive to fail02:49
ulioidleThen a killall mplayer trying to battle it out with the lagging still spawning mplayers02:49
ulioidleThat's what I am trying to sort out02:50
frinnstyou use the old ata-layer? is it an old drive (i guess its an old ide drive)02:50
ulioidleYup and yup02:50
ulioidleOld but not much used drive02:51
frinnstif you have anything important on it and you still can access it, maybe start copying the data off the drive :)02:52
ulioidleI am just trying to find out if this could just be a filesystem error or not, as in worth the time to reformat and reinstall everything02:53
frinnstit complains about a specific sector. I would probably avoid using that drive again02:53
ulioidleI can access everything but some libs and programs so far02:53
ulioidleSounds good02:54
ulioidleI know little about hard drives02:54
ulioidleAnd I just got it all setup nice02:55
ulioidleOh well02:55
ulioidleTime to start anew, thanks for the help02:59
frinnstbackups, backups, backups :)02:59
ulioidleNah, I do not concern myself with ones and zeros much03:01
ulioidlethey are fun to play around with but I can not see myself going through any effort just to keep a few particular ones and zeros03:02
frinnstyou said it: <ulioidle> And I just got it all setup nice03:16
ulioidleSure I did, and a learned alot, and I will learn even more if I do it again03:18
frinnst:)03:19
ulioidleThere are a few mistakes that I would like to skip over as well03:20
ulioidleNot to mention my mysterious wandering glib.h03:21
Romstergood evening03:52
Romsterulioidle, i have expeareance with hdd's can you do a smartctl -a /dev/sdx x being disk you question is faulty.03:56
ulioidleSure, just give me a moment03:58
Romsteruse pastebin too not in channel :D03:58
ulioidleOhhhh03:59
ulioidleYup04:00
ulioidleIt's the drive04:00
ulioidleI heard that click click click! while the errors were printing during boot04:01
ulioidleAnd I can hear it go click now when it should be silent04:03
ulioidleThanks for the offer Romster but this drive will be dead soon04:04
cruxbot[opt.git/2.7]: wine: 1.5.13 -> 1.5.1404:06
Romsterhope you backed up04:06
Romstergnu ddrescue if you haven't to new disk asap04:07
*** mike_k has joined #crux04:20
Romsterlooking up on the net on what kernel modules i needed on pcidatabase.net or something it was a very long time ago.04:24
Romsterugh wrong channel disregard that04:25
frinnstwtf. searched for keepassx2 on bing. first hit was my port04:39
frinnstgod they are useless04:39
*** spaceninja has quit IRC04:40
Romsterlol05:28
Romsterhttp://duckduckgo.com/?q=keepassx205:29
frinnstyeah, thats why I dont use duckduckgo.. silly bing results05:38
frinnstit's a shame05:38
*** pitillo has quit IRC05:52
*** pitillo has joined #crux06:05
*** kInOzAwA has joined #crux06:22
cruxbot[opt.git/2.7]: distribute: initial import, this is the succesor for setuptools06:26
cruxbot[opt.git/2.7]: fontforge: keep egg-info files06:49
*** sepen has joined #crux07:00
*** c0x has quit IRC07:34
*** c0x has joined #crux07:36
*** ulioidle has quit IRC07:42
*** pitillo has quit IRC08:23
*** sepen has quit IRC08:25
*** pitillo has joined #crux08:27
*** sepen has joined #crux08:27
cruxbot[contrib.git/2.7]: cython: fix deps08:33
cruxbot[contrib.git/2.7]: zope.interface: 3.8.0 -> 4.0.1, removed dep setuptools, new dep distribute08:33
cruxbot[contrib.git/2.7]: mako: 0.4.2 -> 0.7.2, removed dep setuptools, new dep distribute08:33
cruxbot[contrib.git/2.7]: deluge: allow deluge to detect 32bit and 64bit builds on it's own08:33
cruxbot[contrib.git/2.7]: pyopengl: 3.0.1 -> 3.0.2, removed dep setuptools, new dep distribute08:33
*** mavrick61 has quit IRC08:35
*** _mavrick61 has joined #crux08:38
*** lasso|qt has quit IRC08:47
*** jdolan has joined #crux08:50
*** ChanServ sets mode: +o jdolan08:50
*** tj_ has quit IRC08:55
*** tj_ has joined #crux08:55
*** sh4rm4 has quit IRC09:09
*** joacim has quit IRC09:16
*** sammi` has quit IRC09:29
*** sammi` has joined #crux09:30
*** niklaswe has quit IRC09:30
*** niklaswe has joined #crux09:32
*** lasso|qt has joined #crux09:35
*** cruxbot has quit IRC09:39
*** tj_ has quit IRC09:49
*** tj_ has joined #crux09:49
*** horrorStruck has joined #crux10:14
horrorStrucknice thread http://marc.info/?l=linux-kernel&m=134918302507859&w=2 :)10:16
horrorStrucklinus to kay sievers: Kay, you are so full of sh*t that it's not funny10:19
horrorStruckand al viro wanting to take udev back by force and maintain it. i can see the light.10:20
teK__"Two-faced lying weasel"10:24
teK__hahaha10:24
horrorStruckhehe that's just the starter10:24
jaegeramen, stop the crazy10:25
teK__still reading but they didn't even begin to talk about the whole include-udev in systemd _issue_10:28
teK__I think it would be10:28
teK__absolutely insane for the kernel to work around the fact that udev is10:28
teK__buggy.10:28
teK__Lennart and Kai may demand this :^)10:29
horrorStruckal viro wants udev in the kernel tree10:30
enteI'd be interested how heise covers the incident10:31
ente(if they do)10:31
frinnst:D10:32
* frinnst cracks open a beer and starts reading10:32
horrorStruckWhat kind of insane udev maintainership do we have? <--- linus <310:32
teK__this whole thred is so full of WIN10:32
teK__+a10:32
Lukc<3 too10:32
teK__We are apparently better off trying to avoid udev like the plague.10:32
LukcYes.10:32
ente... at least for now10:33
LukcShould we prepare the “100% systemd-free” icons, then? :p10:33
frinnstI thought about that yesterday10:33
enteiff they take it in the kernel tree, udev upstream might become reasonable again10:33
teK__I always wondered how Linus would silently accept the whole systemd+udev thing10:33
Lukcfrinnst, precisely.10:33
frinnsthow does it work with distrowatch? do they pick up releases by their own, or do "we" contact them?10:34
LukcBut someone answered we had udev, which makes us only ≃90 % systemd-free, or something near that.10:34
frinnstshould probably be included in the release announcements :)10:34
enteis 2.8 ready10:34
ente?10:34
teK__now they reference mdev as a working example.. hahaha10:34
enteuargh10:34
entelies!10:34
jaegerente: no but it's in testing10:35
entemdev is for embedded systems10:35
teK__we considered switching..10:35
entemdev can't e.g. assign network interface names, iirc10:36
LukcIndeed.10:36
enteit only manages stuff in /dev10:36
Lukcente, anyway, this shouldn’t be mdev’s role to do that.10:36
enteIMHO, it's a misconception that network interfaces don't have a device node10:37
LukcIs that not what it is made for?10:37
LukcYep.10:37
LukcWhich is why mdev shouldn’t manage them. :D10:37
entealright10:37
LukcAlright?10:37
LukcJust like that? :o10:37
teK__horrorStruck: thanks for posting that link.. made my day10:37
enteis mdev separable from busybox?10:37
Lukcente, hum… maybe.10:38
horrorStruckteK__: np, made mine too so i *had* to share10:38
Lukc:D10:38
horrorStruckente: i guess you can chose to build mdev only from busybox config10:39
entehorrorStruck: it isn't a separate binary, no?10:39
entestatic /dev is almost as good as mdev10:40
horrorStruckno but your binary is mdev only10:40
frinnstusing mdev on my desktop felt like stepping back in time.. maybe 10 years or so10:41
frinnstits *very* basic10:41
LukcYep.10:41
LukcIt is.10:41
Lukcente made a point on static dev.10:41
Lukcmdev is also very slow.10:42
entewell, it's intended for embedded systems10:42
enteyou're using it on a desktop10:42
LukcSo it can be slow? :|10:42
enteno10:42
entebut I wouldn't do that :P10:42
entethe requirements are different10:42
LukcIIRC, mdev reparses its whole configuration file each time a new device node is created.10:44
LukcOr something like that.10:44
jaegerheh, even Alan Cox piped in against Kay10:44
jaegerthis thread is great10:45
enteLukc: so?10:46
enteLukc: same happens on connect()10:46
teK__maybe someone should mention the bundling with systemd as long as this thread is still hot10:46
entewith /etc/hosts and /etc/resolv.conf10:46
LukcOkay. :|10:56
Lukcente, but, wouldn’t it be more efficient to store the content of the configuration in a place it would be fast to access again?10:56
LukcIn binary form and preferably on a ramfs/tmpfs, for example?10:56
entehow slow do you think the network is compared with your disk?10:57
LukcSlow.10:58
enteexactly10:58
Lukcente, but, what if I had a 3000+ lines /etc/hosts?10:58
entedoesn't matter to parse some small files a few times10:58
LukcHow do you know those files are small? :|10:58
entethen maybe it takes 0.2 seconds longer to establish a connection10:59
ente$ time cat /etc/hosts > /dev/null10:59
LukcUh, mdev takes much more than 0.2 seconds per additionnal line in mdev.conf. :o10:59
ente    0m0.00s real     0m0.00s user     0m0.00s system10:59
entewell, then find out what it does :-)10:59
entejust reading the file isn't that expensive10:59
Lukcente, my guess is it is slower to read mdev’s config file because it contains regular expressions, a more complex grammar and stuff like that.11:00
LukcBut, I didn’t checked. :D11:00
entehm11:00
entelet's see11:01
enteregular expressions or glob()?11:01
enteuse the source, Lukc! :P11:01
Lukcente, http://www.mail-archive.com/busybox@busybox.net/msg15735.html11:01
Lukc:D11:01
Lukcxregcomp()11:02
LukcI guess it’s a regular expression. :311:02
enteyou could stat it instead of parsing it every time11:02
enteand keep the previous state cached11:02
enteno idea how much better that would be11:03
entei.e. only parse depending on mtime11:03
enteI'm not an expert on this kind of thing11:03
entefor lightweight and optimal design of C programs, feel free to join #musl :P11:03
Lukcmusl XDXD11:04
LukcSorry.11:04
entewhat?11:04
enteno, really, speak up :P11:04
LukcI spoke about it last week.11:04
LukcOr the week before, maybe.11:04
LukcOr whatever. :D11:05
entein this channel, you find mostly people interested in running a light-weight operating system11:06
entepossibly on embedded hardware11:06
entetherefore, it is not surprising that at least one of these people investigates alternative libcs11:06
enteI met musl's author on #uClibC before musl was even released11:06
LukcAh, cool. :D11:07
entecontrary to uClibC, musl does have non-embedded potential, though11:07
LukcuClibc doesn’t?11:07
LukcWhy?11:07
enteit even compiles on NetBSD out of the box11:07
LukcOh, great. :o11:08
entethe syscall numbers don't match, but porting it is cake11:08
enteprobably somebody is already working on it11:08
LukcCake is bad for health, but okay.11:08
enteuClibC: too limited, feature set specific to embedded systems11:08
LukcOk.11:09
LukcWell, I thought uClibc was more complete, to be honest.11:10
enteread http://www.etalabs.net/compare_libcs.html, ask #musl11:10
enteI don't want to spread FUD11:10
Lukcente, I also gave that URL last week. :D11:11
ente(take it with a grain of salt, the musl dev wrote it)11:11
LukcOr the week before.11:11
enteI am not an expert.11:11
enteI know people who are11:12
LukcI can see that…11:14
entethanks :P11:15
LukcEheh. ^_^”11:19
*** sepen_ has joined #crux11:23
*** sepen has quit IRC11:23
*** pitillo has quit IRC11:23
*** pitillo has joined #crux11:28
*** lasso|qt has quit IRC11:59
*** lasso has joined #crux12:09
enteRomster: apparently, UTF-8/16 can indeed only store only a subset of UTF-32: http://aiju.de/misc/unicode13:13
ente-only13:13
*** SiFuh_ has joined #crux13:30
*** SiFuh has quit IRC13:33
*** SiFuh__ has joined #crux13:33
*** SiFuh_ has quit IRC13:35
*** SiFuh has joined #crux13:36
*** Amnesia_ has quit IRC13:36
*** Amnesia has joined #crux13:38
*** SiFuh__ has quit IRC13:38
*** nthwyatt has quit IRC13:39
*** Rotwang has joined #crux13:46
*** suda has joined #crux14:13
sudahello14:17
sudais there a way to block out a certain port so that it wouldn't even be installed as a dependency?14:17
sepen_prt-get lock <port> didn't work?14:21
sudawith lock it can still be installed manually right?14:24
sepen_never try that sorry ;D14:24
sudaI guess, man page says it block out from sysup operation14:27
enteprt-get install `prt-get quickdep port | grep -v dep`14:27
sepen_so do $ prt-get unlock <foo> && prt-get sysup14:27
sepen_its easy, block the one you want the time you need it14:28
sepen_sorry need to go now, bbl14:28
*** sepen_ has quit IRC14:28
sudawell, I *want* to block it, from sysup and manual install14:28
Rotwanglock is for prt-get only14:29
Rotwangprt-get is a tool that aggregates pkgutils14:30
Rotwangyou can't lock things from pkgadd14:30
Rotwangonly thing that comes to mind is a wrapper script around pkgadd14:30
Rotwangor a patch for pkgadd14:31
sudaI'm new to crux so I will need some time before I get in to that :)14:45
*** sammi` has quit IRC15:16
*** sammi` has joined #crux15:27
*** mike_k has quit IRC15:31
*** Rotwang has quit IRC15:41
*** vaddi has quit IRC15:51
*** sepen has joined #crux15:56
*** lasso has quit IRC16:03
*** joacim has joined #crux16:22
krueIs there any reason that nfs-utils is built with tcp_wrappers support, but rpcbind is not?16:22
*** sh4rm4 has joined #crux17:10
*** suda has quit IRC18:09
*** suda has joined #crux18:12
entekrue: probably not, file a bug18:16
krueente: done. thanks.18:47
*** suda has quit IRC18:53
*** tilman has quit IRC19:43
*** tilman has joined #crux19:44
*** ChanServ sets mode: +o tilman19:44
*** horrorStruck has quit IRC19:55
*** sepen has quit IRC19:57
*** acrux has quit IRC20:24
*** acrux has joined #crux20:32
*** acrux has quit IRC20:32
*** acrux has joined #crux20:32
*** tj_ has quit IRC20:49
*** sh4rm4 has quit IRC21:00
*** tj_ has joined #crux21:01
*** sh4rm4 has joined #crux21:01
*** kInOzAwA1 has joined #crux21:03
*** kInOzAwA has quit IRC21:05
*** Romster has quit IRC21:10
*** _mavrick61 has quit IRC21:27
*** mavrick61 has joined #crux21:28
*** Romster has joined #crux21:31
*** Romster has quit IRC21:31
*** Romster has joined #crux21:31
*** kInOzAwA1 has quit IRC21:40
*** nthwyatt has joined #crux22:50
*** vaddi has joined #crux22:55
*** SiFuh has quit IRC23:11
*** SiFuh has joined #crux23:12
*** tj_ has quit IRC23:35
*** tj_ has joined #crux23:35

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!