IRC Logs for #crux Monday, 2015-03-23

*** sh[4]rm4 has joined #crux00:14
*** Claus_ has quit IRC00:26
*** Claus_ has joined #crux00:26
*** lmarie has joined #crux00:30
lmarieEvening all.  Just reading over some of the recent logs and having a good laugh at onodera's response to the Firefox update.00:38
lmarieAll working well here using kernel 3.19.2, using nouveau instead of nvidia.  I fought the blob and the blob won, something wrong on my end, as I was able to get it to install correctly a month ago, but not this time round.00:44
lmarieNo big deal. I don't need nvidia driver as nouveau is fine for my use.00:45
jaegerlmarie: they're not great about keeping up with the latest latest kernels so you might try 3.14.x as it's a longterm version. Or stick with nouveau if it does the job :)00:52
lmarieHi jaeger.  Thanks.  Yeah, I wasn't sure which kernel to choose.  I probably should have chosen the long term one.  This was my first time to actually select a kernel.  I first tried 3,18,2 but that threw errors, and 3,19,2 did not.00:54
jaegerIt's always a bit of a balancing act. The new kernels may have awesome new features but nvidia doesn't update too quickly00:54
lmarieNouveau does the job on the asus a8ne board.  But it bothered me as I was able to configure the nvidia blob about a month ago, that is, gl-select use nvidia did what it said, without errors saying it cannot find nvidia.  I'll post the two logs from the two different Pkgfile outputs.  In case anyone is curious.00:57
lmarieI see.00:57
lmarieThis is while using kernel 3.12.24 http://pastebin.com/9V4tGyQ501:00
jaegerlmarie: regarding that first one I think we discussed that here a few days back01:01
lmarieYes.01:02
jaegerhttp://irclogs.shortcircuit.net.au/%23crux/2015-03-12.log.html#t20:55:13 <-- there, specifically01:02
*** tilman has quit IRC01:04
lmarieI tried the other suggestion that does not try to create the nvidiagtk.so's and prt-get was successful: http://pastebin.com/LDyDFat0  same success if I use 340.65 or 340.7601:04
*** tilman has joined #crux01:05
jaegerthe separate gtk libs didn't exist back in 340.65 or .76 so they're expected to be missing01:05
lmarieOkay.  Obviuosly I'm not getting it, which is why I gave up on it.  The problem came when trying to use gl-select use nvidia.  It would say no nvidia driver found.  That was the snag.01:06
jaegerlmarie: probably a naming issue. It looks like you called your port "nvidia-340"01:07
jaegergl-select understands: nvidia nvidia-sl nvidia-legacy-96xx nvidia-legacy-71xx01:07
lmarieHey, the light bulb turns on.01:07
jaegerthe last two don't even exist anymore, I think01:07
lmarieGot it.01:07
jaegerif you edit gl-select to look for nvidia-340 or rename to "nvidia" it'll probably work :)01:08
lmarieThat gives me an idea.01:08
lmarieYeah!  Hey thanks!01:08
jaegerno problem01:08
*** Claus_ has quit IRC01:09
lmarieYou take care,  and time to relax and laugh while I go and finish the movie I started last night, Alan Partridge.01:09
jaegerthanks, you too :)01:09
lmarie"Alan Partridge". Goodnight!01:09
*** lmarie has quit IRC01:10
Romster3.18.9 works on nvidia01:36
Romsterstill wonder how good nouveau is at 3D?01:37
*** sh4rm4 has joined #crux01:50
*** sh[4]rm4 has quit IRC01:50
*** sh4rm4 has quit IRC01:55
*** sh4rm4 has joined #crux01:58
*** JSchilli1 has joined #crux02:13
*** sh4rm4 has quit IRC02:37
*** JSchilli1 has quit IRC02:38
*** ryuo has quit IRC03:04
*** ryuo has joined #crux03:04
*** jue_ has joined #crux03:24
*** sh4rm4 has joined #crux03:25
*** sh4rm4 has quit IRC03:27
*** sh4rm4 has joined #crux03:28
*** sh4rm4 has quit IRC03:28
*** mavrick61 has quit IRC03:30
*** mavrick61 has joined #crux03:31
*** dsplayer14 has joined #crux05:00
*** ryuo has quit IRC05:01
*** dsplayer14 has quit IRC05:03
*** ryuo has joined #crux05:03
*** cezar has joined #crux06:23
*** cezar has quit IRC06:24
*** nilp has joined #crux06:39
*** nilp has quit IRC08:34
*** nilp has joined #crux08:34
*** mhe has joined #crux09:13
cruxbot[xfce.git/3.1]: xfce4-power-manager: updated to 1.4.409:34
cruxbot[xfce.git/3.1]: xfdesktop: updated to 4.12.109:34
*** mhe has quit IRC09:44
*** doomicide has joined #crux10:40
*** doomicide has quit IRC10:42
*** doomicide has joined #crux10:43
*** mhe has joined #crux11:04
doomicide@Romster: Did you test building mpv with your external waf? I get the following error: http://sprunge.us/AXFa11:07
doomicideI think that's why the AUR PKGBUILD declares a seperate $waflib https://aur.archlinux.org/packages/wa/waf/PKGBUILD11:09
Romsteryes i tested it in a clean container and it worked.11:11
doomicideRomster, mhm ok, I'll try to figure out the problem then.11:12
Romsteri'll retest but it built as that.11:13
doomicideok, thank you11:13
Romsteri wouldn't commit it if it never worked.11:22
doomicidesure, just checking :) But I don't think that's the problem, I tried AUR's approach and I get the same error, looks like the problem is on my end11:23
Romsterto me that is extra stuff we don't need11:24
doomicideSeems like it. I aggree.11:24
Romsterhonestly why didn't they just use scons11:24
Romsterwaf jsut looks ugly.11:24
doomicideor just plain old make, like they used to11:24
Romsteryeah11:24
doomicideOh ok, i just need to run it as root. Jesus I can be stupid sometimes. But I used to test building as a normal user :(11:26
Romsteroh11:27
Romsterwell you can with fakeroot11:27
doomicideyeah I'll try that11:27
Romsterhttps://registry.hub.docker.com/u/romster/crux-base/ that is how i test now.11:28
rmullRomster's external waf worked fine for me too, fwiw.11:31
Romsterit should :)11:31
doomicideyeah I figured it out, works now :)11:31
Romsterworked on my system and in the container.11:31
Romsterjust how many build tools does there need to be.11:32
cruxbot[contrib.git/3.1]: mpv: 0.8.2 -> 0.8.311:34
Romsterif you like you can pick up waf port too.11:34
doomicidewhy not, you propably have enough projects already11:36
Romstertons11:36
doomicideI know ;)11:37
doomicideJust put me in the Maintainer field and I'll take care of it.11:39
Romsterk11:39
doomicidedoes waf have an rss feed?11:40
Romsterthey have a google page11:41
Romsterbut googlecode is closing down so i dunno where they will move too.11:41
Romsterand they have 2 sites main and a mirror and the main one has the source unreachable.11:42
Romsterthat whole project is in shambles.11:42
doomicidewhat a mess11:42
Romstertell me about it.11:42
doomicideI just checked rss hasn't updated since January 201411:42
cruxbot[contrib.git/3.1]: waf: hands over port to doomicide11:42
doomicidethanks11:43
Romsteri waf you... no not really but waf come on.11:45
Romsteri just hope mpv decides to got a more sane tool or waf dev(s) pick there game up.11:46
doomicideyep would be nice, mpv is a great project, I don't know why they'd use such a messy one11:47
Romsteri'm not that fond of cmake but it's solid i'd be happy if they even used that11:48
Romsterthey want Python based scons would be better. maybe someone needs to suggest that.11:48
vlnxI don't let pkgmk write to /usr/bin, so I just ran into that same waf error.11:51
Romsterit's not to write into... you mean to tell me it's installing outside of $PKG?11:51
vlnxI think it tries to. 'Cannot unpack waf lib into /usr/bin/.waf'...11:52
Romsterwtf build ssytem is this11:52
doomicideYeah apparently waf unpacks there when executed11:52
doomicideI know11:52
doomicideit's bullshit11:52
Romsterthis is why i need to get fakeroot into my root docker containers.11:53
Romsteri built it i never issued the install command.11:53
Romsterbut then i don't have a /usr/bin/.waf directory...11:54
Romsteron my system after installing waf11:54
Romsteruh wrong terminal11:54
Romster.../usr/bin/.waf-1.8.7-2d8c75bfac1b380bb14b9950e502f759/waflib/ mother of gad sigh11:55
doomicideI think arch contains this at /usr/lib or we go back to downloading waf everytime mpv is updated locally11:56
Romsterthe arch pkgbuild woulnd't prevent that either...11:56
doomicideoh ok11:56
doomicidei misunderstood it then, just skimmed it11:56
Romsteri'm not sure where in waf-light it uses /usr/bin/12:00
doomicideUsing bootstrap.py almost seems like the better choice imo.12:00
Romsterdoes that work on the new version of mpv?12:02
Romsterwaf is a POS12:02
Romsterlast time i looked mpv was at the most recent version12:02
*** jacob_ has joined #crux12:04
doomicideyes works for 0.8.3 although it uses waf 1.8.412:04
doomicidebut that could propably be patched12:05
doomicideor I'll open an issue on mpv's github12:05
Romster./waf-light configure --prefix=/usr build --make-waf12:05
Romsterthat fixed the writing to /usr/bin/12:06
Romsterdumb i know -_-12:06
Romsterbut eh new mpv works you could just nuke waf and keep the internal one in mpv12:07
Romsteryour call.12:07
Romsterfirst time i ever made a port that writes outside of $PKG sigh.12:08
doomicidemhm keeping waf seems more in line with crux' philosophy, on the other hand it really is a pain in the ass12:08
doomicideI think I'll keep it, this will enable others to add ports using waf more easily12:09
Romsteralso crux would choose not to keep a external dep that is bundled in for 1 port.12:09
doomicidebut downloading waf every time mpv is updated is so goddamn inefficient12:09
Romsteri'd say nuke waf, change waf back to python and let mpv deal with it12:09
Romsterhmm true12:09
Romsterwell just apply that fix and bump revision=212:10
Romsterno idea if waf devs will move off google code or what they have there own site.12:10
doomicidenah I'll follow your first suggestion for now  because waf seems to be such a pain in the ass to maintain atm, also you have a lot more experience maintaining ports12:12
doomicideso I trust your opinion12:13
cruxbot[contrib.git/3.1]: mpv: use bundled waf12:17
cruxbot[contrib.git/3.1]: remove waf port for now12:18
Romsterwell 1 port i see no reason to go with the pain of the extra dep so just drop waf12:18
cruxbot[contrib.git/3.1]: mpv: update dependencies12:19
Romsterwas about to say that waf -> python12:20
doomicideall done, thank your help12:20
doomicidethank you for*12:20
Romstersorry it turned out to be a pain.12:20
doomicidenp man, I actually tried using waf when mpv switched to it but that version back then wouldn't even install properly12:21
Romsteroh by the way12:21
Romsterfindredundantdeps -s mpv12:21
RomsterRedundant deps for mpv are: libass python12:21
RomsterRight is: # Depends on: docutils ffmpeg12:21
Romsterah12:21
doomicidepython is need for building with waf12:22
doomicidei'll see about the others, thank you :)12:22
Romsterxorg-xcb-proto depends on python and ffmpeg depends on that. so it is pulled in.12:22
doomicideohh ok, thanks again12:23
Romsterwell not exactly in that order but prt-get deptree ffmpeg12:23
Romsterwhat i usually do is finddeps filter out not needed ones omit glibc gcc and add them on depends on then findredundantdeps12:24
Romsterand review it all. note that some ports for build time wont by listed by findredundantdeps12:24
*** jdolan has quit IRC12:24
Romsterand then i do this in a clean environment so i don't pull in all the optional dependencies.12:25
doomicidemakes sense, I need to check out docker12:25
Romstervery handy12:26
cruxbot[contrib.git/3.1]: mpv: update dependencies...again12:26
doomicideahhh fuck12:27
doomicidewrong version12:27
Romsteri should of said this before.12:27
RomsterO_O12:27
Romsterum and you added a gz.paral file -_-12:28
Romstermpv/v0.8.4.tar.gz.partial12:28
Romstergit lot12:28
Romstergit log12:28
Romstergit revert <hash>12:28
cruxbot[contrib.git/3.1]: mpv: clean everything up12:29
Romsteryou haven't used git enough i think?12:30
doomicideok should be fine now...12:30
doomicidenah12:30
doomicideI'm a noob12:30
doomicideanyways, I'm gonna go eat something now12:32
Romsterwe all were once.12:32
Romstercool that looks ok.12:33
Romsterbuilds no redundant deps etc12:33
cruxbot[contrib.git/3.1]: sxhkd: 0.5.3 -> 0.5.512:48
*** hnode has quit IRC12:49
*** heroux has joined #crux13:04
*** mhe has quit IRC13:22
*** kori has quit IRC13:23
*** kori has joined #crux13:29
*** doomicide has quit IRC13:39
*** namra has joined #crux14:03
*** namra has quit IRC14:33
*** BitPuffin has joined #crux14:36
*** greduan has joined #crux14:40
greduanhey guys14:40
jawbomorning greduan14:44
greduanmorning jawbo14:46
jawbofinally updated my kernel to 3.19 - intel's "sna" accelmethod finally works!14:47
jawboi was having some trouble with the glx backend with compton, so it's very nice14:47
jawboi love the kernel naming schemes14:48
jawbo3.18 - shuffling zombie juror; 3.19 - hurr durr i'm a sheep14:48
*** pejman_ has joined #crux14:58
*** dougl has joined #crux15:03
*** sh4rm4 has joined #crux15:03
jawbono, 4.0 is hurr durr, 3.9 is diseased newt15:27
*** BitPuffin has quit IRC15:28
*** BitPuffin has joined #crux15:31
rmullThis is the second time my ALSA has crapped out after sitting inactive for a while... [ao/alsa] Playback open error: File descriptor in bad state15:53
rmullAnyone else see anything like this?15:53
frinnsthm, nope15:54
frinnstbut I had all sorts of trouble after re-reading a backup of / once15:54
jaegerrmull: not I15:54
frinnstso the backup was taken when alsa was running and re-read without alsa configured15:54
frinnstthat fucked shit uuuuup15:54
frinnstso purge all state files and try again - thats my advice15:55
frinnstobviously not a solution for the underlying problem though15:55
rmullAny idea where the state is kept? Not seeing anything in /var/run15:56
frinnstlet me check15:56
frinnstso /var/lib/alsa/15:56
rmullIt's empty for me15:56
rmullThe other interesting thing is that I can play stuff using pulse, but not alsa15:57
frinnstfredrik@farnsworth:/var/lib/alsa$ ls15:57
frinnstasound.state  asound.state.lock15:57
rmullhuh.15:57
rmullOne of the relevant strace lines: open("/dev/snd/pcmC0D0p", O_RDWR|O_APPEND|O_NONBLOCK|O_CLOEXEC) = -1 EBADFD (File descriptor in bad state)15:58
frinnsthuh15:58
frinnstbuggy kernel?15:58
rmullI suppose it's always possible, 3.19 here15:59
rmullI have never experienced a fd in a bad state before15:59
rmullThe first time this happened, I rebooted and it recovered16:00
rmullThis time, I was playing music this morning using mpv16:00
rmullI quit it (with 'q')16:01
rmullLike an hour or two passed, and I tried to play more music, and the error showed up16:01
frinnstI had problems with my old motherboard with sound just dying16:01
frinnstnever with the current board16:01
rmullOnboard sound? In this case I have a USB DAC16:01
frinnstyep16:02
rmullblargh16:02
frinnstintel hd in both cases. different codecs16:02
rmullIs there a way to "reload" those dev nodes without taking down the whole system?16:04
rmullI have never reported a kernel bug before16:05
rmullNor have I debugged a kernel16:05
rmullMaybe this would be a good a time as any16:05
rmullsound seems complicated to me though16:06
*** sh[4]rm4 has joined #crux16:09
*** sh4rm4 has quit IRC16:11
*** dougl has quit IRC16:34
*** itPuffinB has joined #crux16:49
*** Pingax_ has joined #crux16:52
*** BitPuffin has quit IRC16:52
*** lemmon has joined #crux17:07
lemmoncan anyone hit me up with some documentation about making a usb installer for crux? or should I just use dd?17:07
jaegerlemmon: unless you need to customize it, dd is sufficient. The ISO is hybrid17:08
lemmonthanks jaeger17:08
jaegernp17:09
*** itPuffinB has joined #crux17:13
rmullDoes anyone use the haskell package manager called 'cabal' ?17:20
rmullWith all these language-centric package managers out there I wonder if it would be worthwhile to try to integrate an understanding of them into the distro's package manager17:22
rmullThat would take the burden of repackaging stuff distributed with pip, or cabal, or cpan away17:22
lemmonanyone know why I would have an error installing acl#2.2.52-1 on the setup script?17:30
rmullwhat error?17:34
lemmonits saying that all of the packages are already installed17:34
lemmonpretty strange17:34
lemmonim just gunna rewhipe the drive see if that changes anythin17:35
joacimnot if you make the same mistake again17:35
joacimshow us the mistake, and is it already installed?17:36
joacimshow us the error*17:36
lemmontoo late now, but it was on a drive with a really broken arch install from last year17:37
lemmoni dd /dev/zero it17:37
lemmonand presumed that would be ok17:37
lemmonI guess not17:37
joacimnot sure what you're saying. you installed crux without making a new filesystem on your target drive?17:38
lemmonI made a new filesystem17:39
lemmonor atleast I believed I did17:39
joacimthat's enough17:39
lemmoni prehapse didnt write the partitions to the disk17:39
joacimyou dont have to zero out the drive17:39
*** ryuo has quit IRC18:01
*** dougl has joined #crux18:20
*** lnds has joined #crux18:22
rmullpandoc is pretty cool, but in order to use it I had to install 705MBytes of dependencies.18:27
rmullBecause nothing else I do uses haskell18:28
*** mhe has joined #crux18:42
*** nullmark has joined #crux19:04
*** kori has quit IRC19:11
*** kori has joined #crux19:29
*** jdolan has joined #crux19:46
*** nilp has quit IRC20:03
*** Feksclaus has joined #crux20:13
lemmonI know this probably sounds really stupid of me, but how would one use prt-get to install something from the df collection?20:41
prologicsync the df collection to /usr/ports20:43
prologicusing ports -u20:43
prologicand an appropriate ports driver20:43
prologicand add the directory to /etc/prt-get.conf20:43
prologicprtdir /usr/ports/df20:43
Guest79894did I fuck up doing a sysup ._.?20:49
Guest79894my laptop has been updating for 7 hours20:50
Guest79894now20:50
jaegerDoes it appear from the output that it's hung on something or is it scrolling happily away?21:00
Guest79894it's running21:01
Guest79894looks like it's doing gcc atm21:01
jaegergcc is one of the ports that will take quite some time. Also firefox, llvm, etc.21:02
*** dougl has quit IRC21:03
jawboand libreoffice just came up to 4.4 too21:10
*** lnds has quit IRC21:16
*** jdolan has quit IRC21:17
lemmonexit21:38
*** lemmon has quit IRC21:38
*** lemmon has joined #crux21:39
rmullGuest79894: You might consider using ccache or distcc to reduce the amount of time spent waiting for compiles to finish21:39
lemmonprologic, I keep getting a Md5sum mismatch when installing your google-chrome port, is there anyway to just ignore the mismatch?21:40
frinnstpkgmk/prt-get -im (ignore md5)21:40
frinnstmanpages are a wonderful thing. they tell you all sorts of stuff :-)21:41
lemmonthanks a load dude, ahah21:41
frinnstbut make sure you understand why the md5sum has changed - as a general rule21:42
frinnsti think google-chrome silently updates the tarball so im sure its "safe"21:43
lemmonI just assumed there was an update that wasnt accounted for21:43
*** lemmon has quit IRC21:48
*** lemmon has joined #crux21:56
*** nilp has joined #crux21:57
Romsterhttp://crux.ster.zone/packages/3.1/ also i have some built packages.21:59
lemmonanyone know who to use lspci after tinstalling pciutils?22:07
*** dougl has joined #crux22:08
Workster/usr/sbin/lspic -k22:09
Workster/usr/sbin/lspci -k22:09
lemmonyeah, I you found it the before I realised I could just do whereis, thanks anyway22:10
*** lemmon has quit IRC22:12
*** nilp has quit IRC22:26
*** jdolan has joined #crux22:44
*** greduan has quit IRC23:08

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