IRC Logs for #crux-devel Wednesday, 2012-06-20

*** mike_k has joined #crux-devel01:28
*** sepen has joined #crux-devel02:07
*** mike_k_ has joined #crux-devel03:22
*** mike_k has quit IRC03:26
jueQ: now that zlib stays in /lib we can add gzip support to kmod. I'd say it's more or less useless for us, because we only build the modules that are really needed. Anyway, is that something I should add?04:41
jueto give a number about what we are talking here: on my netbook I'm using a mostly modular kernel with ~350 modules, space needed for them is around 6MB04:45
teK_omg ponies!!11 ;)04:49
teK_what's to be done to make kmod support gzip?04:49
sepenjue: also xz support, but imho is useless04:49
juethe drawback is that kmod depends on zlib, not that bad IMO. We have a ticket with a request for adding xz compression, I wouldn't do that because xz ist installed in /usr/lib04:50
sepen*err, xz needs /usr/lib04:51
jueteK_: add --with-zlib04:51
sepenhttp://www.linuxfromscratch.org/lfs/view/development/chapter06/kmod.html04:51
teK_it won't hurt either04:52
teK_although it's a micro-optimization, obviously04:52
sepenjue: http://crux.nu/Wiki/Mdev04:53
juewell, adding deps to such low level tools like kmod is always a bit questionable04:53
juethat's the reason why I ask here ;)04:54
jueI'm against xz support, because moving xz to /lib for only that resaon seems a bit silly to me04:55
sepen~350 modules < 6MB is enough to take a decission imho ;D04:56
juethe whole module compression thing is important for distro kernels, where you have a lot of modules04:56
sepenwe're not releasing kernels04:57
sepenmaybe it would be more useful for crux-arm for example, where we released some distro kernels04:57
juesepen: normaly I have much fewer modules, that's only because I was lazy and copied the kernel config from http://code.google.com/p/kernel-netbook/04:58
sepenlol04:58
jue:)04:58
sepen$ du -hs /lib/modules/$(uname -r)/kernel05:01
juesepen: yeah, I saw your mdev page already, great05:01
sepen4.5M    /lib/modules/2.6.35.6/kernel05:01
teK_maybe we could add it to the isos kmod (yet to add?) and support all NICs and [SCSCI|(S|P)ATA|SAS] controllers :}05:01
sepenbit a bit, I've more work at home but I'm still at office05:01
juesepen: to save some space I'd suggest to use e.g.: 'ldd /bin/lsblk | grep libudev'05:03
jueof course not important at all ;)05:05
sepenyep05:07
sepenjust I pasted the notes I had in plain text format ;D05:08
jueteK_: not sure, but the ISO is using modprobe from busybox probably05:12
sepenjue: sure, initramfs contains busybox's modprobe, and loads squashfs module before switch_root to it05:14
sepens/to it/to crux.squashfs/05:16
juenice, busybox can use gz compressed modules with modprobe05:23
teK_so, support for all devices \o/05:24
jueso, we can lower the memory consumtion during install a bit by using compressed kernel modules ;)05:24
jue*comsumption05:24
sepenjue: well, we can lower the memory consumtion by returning back to .tar.bz2 instead of .tar.xz too05:25
sepenfor packages I meant05:26
sepenI don't see any great performance for that nowadays05:26
jueyeah, but for the cost of a bigger ISO05:26
frinnsti hope the next iso will fit on my tiny 256mb usb-stick :)05:30
sepenit would be easy if you remove crux.squashfs, I talked about that with jaeger but the problem is that you'll lost some useful apps like sshd, cfdisk, etc on install time05:31
sepenanyways you can have telnetd and fdisk from busybox ;D05:32
jueFWIW, 16MB kernel modules on our ISO05:32
sepenjue: so there is no need to zlib05:32
frinnstyeah thats not something you'll want to lose05:32
teK_telnetd?! /me cleans his eyes/ears05:33
sepenfrinnst: but I'm not 100% sure about that, you can try to pkgadd the packages you want to a temporary place05:33
sepensee that, I wrote a little rc+setup stuff which avoids crux.squashfs file http://crux-arm.nu/gitweb?p=initrd.git;a=blob;f=filesystem/rc05:34
sepensee line 88 to see how we used pkgutils from package dir05:35
sepenwell, pkgutils is a static binary so its more easy to do that05:36
sepenbut anyways is possible to do some tasks on runtime :D05:36
sepenjaeger give me more reasons but I can't remember now05:37
sepenhere http://crux.morpheus.net/irc/?channel=crux-devel&date=06Jun201205:42
*** frinnst has quit IRC05:59
*** frinnst has joined #crux-devel06:13
*** frinnst has joined #crux-devel06:13
jaegerI think we could remove the dependency on squashfs while retaining ssh and the like, probably. A ramdisk root would be built either way but the modules could be in a tar.xz instead or even just a tar if they're already compressed07:09
jaegerIt's on my list to test07:09
juejaeger: your opinion to kmod with gzip support?07:43
jaegerI guess I don't have a strong opinion. If it doesn't cause any issues I have no objections08:11
sepenjaeger: there is no hurry about that, I think that is not a priority ATM08:12
sepenI was just saying some ideas, you know08:13
jaegerI haven't spent time on it yet, just added it to my todo list08:14
Romster$ du -hs /lib/modules/$(uname -r)/kernel08:34
Romster204K/lib/modules/3.1.2/kernel08:34
Romstermy firewall pc really is trim08:34
juejaeger: thx, cause nobody has missed such a feature until now, I see no real reason to add it, will just do the update08:35
Romster5.5MB on my desktop in modules, the time it takes to decompress would hurt loading time on a netbook perhaps.08:35
jaegerjue: ok08:43
jaegerlame... seems the rule generator stuff was removed from systemd-udev09:15
jaegerthis morning's comedy port: http://jaeger.morpheus.net/linux/crux/tmp/udev/09:19
jaegerIt has received ZERO testing09:21
frinnstdownloading a tarball named systemd-185.tar.xz makes me die inside09:22
jue:)09:22
rmullHow short is the roadmap for mdev inclusion?09:23
frinnstyou can run it today if you like, with some headache09:24
juejaeger: IIRC udev no longer copies /lib/udev/devices, but a 'cp -a /lib/udev/devices/* /dev' in start_udev should work09:25
jaegerok09:26
jueor you can use systemd-tmp something for it ;)09:26
jaegerI haven't really kept up with udev changes, honestly, just made this port to mirror the live one as much as possible09:26
Romsterguys when you pick up ports don't forget to update http://crux.nu/Wiki/OrphanedPorts09:26
Romsterwhat other distros are resting mdev?09:27
Romstererr s/resting/testing09:28
jaegergood call, forgot that, will remove the alsa ports from it09:28
Romstermay aswell remove what tek picked up just now too.09:28
juejaeger: yeah, that's great, I'm curious to hear how it works09:28
jaegerI guess I'll test it on my workstation here, I'm curious as well09:29
Romsterif no one disagrees i'd like to drop dosbox out of opt we have a newer one in emulators09:29
jaegerIt may break things09:29
Romsters/may/will their i fixed it :D09:30
Romsteror is it there in this sentence, i'm bad on their the're and there.09:30
jaegerin this case, "there"09:32
sepenhttp://thereifixedit.failblog.org/09:32
jaegerthey're is a contraction for "they are", when in doubt expand it09:33
jaegertheir is possessive, none of the others are09:33
jaegerthere is location or "there, I fixed it"09:33
jaegerthose 3 rules might make a decent cheat sheet, heh09:33
Romstersepen, thanks for fixing that port the other day.09:34
sepenRomster: chromium worked fine without red borders09:34
Romsterlol vice grips on egg slide09:34
Romsteri didn't get red borders i think only tek was seeing red?09:34
sepennow I'm following this http://code.google.com/p/chromium/issues/detail?id=6398109:35
jaegerI'd better build myself a backup udev package just in case my workstation shits itself09:36
jaegerand done09:36
Romstersepen, i jsut made webkit optionally depend on geoclue09:36
sepenjaeger: you can keep udev installed, but you'll need to force some ports to ignore libudev on buildtime09:37
jaegerI'm not talking about mdev, I'm using udev 18509:37
Romsteri'm not really sure i like how dbus works myself either... seems the paths for the sockets is also messy too. surely someone is making a dbus replacement that's nice and neat and event driven.09:37
jaegerbrb, hopefully09:37
sepenjaeger: ah, sorry for the noise09:37
sepenRomster: I saw that yesterday09:38
Romstertell them to fix nspr to use the pkg-config script in there build system too.09:38
jueRomster: for what do you need dbus?09:40
Romsteri don't have any notification system currently, but something lightweight would be nice in some sort of bubble like windoze uses.09:41
rmullbubble?09:41
Romsterlike for pidgin users sign in notification etc...09:41
jaegerwell, the system booted, that's a good start. Going to rebuild xorg-server and see what else needs updating09:42
jaegertheir build system :)09:43
Romsterhttp://www.mydigitallife.info/wp-content/uploads/2006/04/wga-notification3.jpg bubble09:44
Romsterjaeger, off my xorg fork or your working on your own?09:45
jaegerI don't understand the question09:45
RomsterGoing to rebuild xorg-server and see what else needs updating09:45
jaegerUsing the default crux one at the moment09:45
jaegerudev 185 has a new soname09:46
jaegerbreaks a few things09:46
Romsterah k has anyone got access to fix a few minor issues like moving lidbrm to xorg and adding libpciaccess to libdrm?09:46
jaegerchrome, libdevmapper, lilo, mesa3d, nfs-utils, thunar, util-linux-ng, vlc so far09:46
Romsterah just looking though revdep09:48
jaegeryeah09:48
sepenjaeger: don't forget gst-*09:48
jaegerI have none of those installed09:48
jaegerlibdevmapper doesn't build against udev 185, awesome09:49
Romsterhmm that's also on my todo list look if i can move to the newer branch of gstreamer or i need to wait longer for other ports to catch up09:50
Romsteroh... error pastebin i'll poke the lvm devs09:50
jaegeralright, will have to try this another time, don't have time to troubleshoot it right now09:50
Romsterand test it in my chroot too.09:50
jaegerthe quick impression, though, is that the udev soname change is not trivial currently09:52
Romsterk09:56
jueanother biggest concern about building udev from systemd sources is, that we have to fulfill systemds additional build time dependencies, which are at least intltool, gperf and dbus AFAIK09:57
jues/biggest/big/09:58
jaegeryeah, I'm not a big fan of that09:58
juewhich means that we have to move all of them and their deps to core09:59
Romstereek no that's too messy, if anything i'd throw them in the one Pkgfile to just build and install udev if it came to that. but that is still messy10:00
Romsteri do get the impression they are forcing this on to us. but why let it into the kernel linus why...10:03
Romstersystemd is killing diversity10:04
sepenteK_: you need to update the orphaned list10:04
Romsterhe's gone to uni now10:04
sepenknow it, but I hope he'll read the backlog10:06
Romsterhttp://hackaday.com/2012/03/15/a-simple-touch-interface-for-music-player-daemon-and-more/ rmull might enjoy this random find i saw.10:09
rmullHaha, nice enclosure :P10:10
Romsteryeah i was like why not a jiffy box.10:10
rmullI like the LED notifier10:11
rmullPRobably depends on dbus -_-10:11
Romsterhehe no doubt.10:12
Romsterhttp://www.cbus-enabled.com/cbus-open-a-low.htm hmm c-bus10:13
Romsteror is that something totally different.10:14
Romsterare only available to partners under the C-Bus Enabled Program.10:14
Romsterdamn10:14
jaegerI'm starting to think we should remove Packager: lines from Pkgfiles if they contain Maintainer: lines :P10:24
jaegerJust got another email about a package I don't maintain (ffmpeg in this case)10:24
jaegerThe bug tracker is looking pretty bad, we should probably focus on cleaning up bugs sometime soon10:27
Romsterum why wasn't that sent to me i'm the maintainer...10:29
Romsterwhat's the issue?10:29
jaegerbecause people don't understand the difference between packager and maintainer, hence my comment10:29
jaegerI replied that he should file a bug at the bug tracker but if you like I'll forward you his email10:29
Romsteryeah forward the mail please10:30
sepenwell, people don't understand the diff between CRUX and CRUX-Contrib projects in flyspray, and etc. etc.10:30
Romsteri file bugs on the right one. on that note they OS selection. windows and stuff needs removing. so only Linux is the only option.10:30
Romsterand there is no crux 2.7.1 in there.10:31
Romsterfood brb10:32
sepenRomster: then you could file a ticket in our ticket system, lol10:32
juesepen: btw, what's the reason for using pm-utils 1.3.0 instead of 1.4.1?10:32
sepenhttp://crux.nu/Wiki/Pm-utils10:34
sepenhal maybe?10:34
sepenjue: anyways I should review again pm-utils, thanks10:34
sepenor at least the wikipage (which I adopted)10:35
juenp10:35
jaegerhrmm... definitely a lot of cleanup needs doing in the bug tracker10:35
frinnstheh, yeah10:42
frinnstsquashed a few10:42
jaegerI'll fix 790 and take over x11-fonts-dejavu, I think10:42
sepenjaeger: I picked this port10:42
sepenjust need to push it10:43
jaegerok, fine10:43
sepenwait I picked some more ports10:43
sepenx11-fonts-dejavu shared-mime-info talloc and git10:43
sepennow I'm updating git to 1.7.11, let me some time10:44
jaegerI think we should standardize on font package names10:44
jaegerxorg-font-dejavu would match the rest10:44
frinnstwhy not move dejavu to xorg?10:44
sepenwell, belongs to freedesktop10:45
frinnstah10:45
jaegerI was going to suggest moving it to xorg after we fix up the xorg repo10:45
sepenfor me that is ok10:45
jaegerto me the 'xorg' repo is CRUX's classification, not X.Org's :)10:46
sepenso we should start to move libdrm, mesa3d and friends too10:47
sepen+110:47
jaegermesa3d is already there but I agree10:47
sepenyep I forgot that10:48
jaegerDoes anyone object to me adding myself to the xorg group so I can fix FS#790?10:48
frinnstfuck no, go nuts! :D10:49
jue:)10:49
jaegerWe should probably also remove retired maintainers from the system groups for git10:50
sepenjaeger: 09:52:55 <jue> ok, until now we are four: jaeger, jue, sepen and Romster (in alphabetical order)10:50
jaegerThough I don't expect any malicious behavior10:50
jaegerHopefully teK_ won't be annoyed, don't mean to get in his way but we've got to start somewhere10:53
Romsteri thogght tek was after git from his eK_> grep -E '(cryptsetup|ctags|esmpt|git|llvm|openbox|rpm2targz|strace|valgrind|wireshark)' ~/.ck4up/ck4up.conf10:56
Romsterany malicious behavior is outside this group afaik10:57
sepenRomster: http://crux.nu/Main/Timeline10:57
sepenI'll update git and do some cleanups (man pages, etc.)10:58
sepennow it's time to go back to home, see ya'10:58
*** sepen has quit IRC10:58
Romsteri'd like to at some point rename the description to a less generic one so it reflects the descriptions of the intended ports in xorg. i just haven't got around to doing that in my xorg fork as of yet11:01
jaegerjue: do we have a specific format for FS tickets in commit messages like the notify stuff or would "comment (FS#790)" be sufficient?11:04
jueyep, that's ok, we don't have a specific format11:05
jaegerok11:05
jaegerRFC: should we remove ports on the OrphanedPorts page from the official repos after a certain timeframe has passed?11:09
Romsteri've been thinking about that.11:13
Romsterfew are really sitting for ages. i might pick up openldap wine can use that. need to review what's left after everyone has gone though timans stuff.11:13
jaegerIn my opinion, yes, with a short timeframe such as 30 days11:13
jaegerthis really only applies to opt11:14
Romsteri'd make it a requirment to be clean before crux 2.8 is released then 30 days after that for any others that appear?11:14
Romsterwhy not contrib too?11:14
jue+1, the ports are still in our git, so we lost nothing11:14
juefor the opt ports oc11:15
jaegerI don't mind doing it in contrib as well if it's feasible/desired, mainly just concerned about junk in opt11:16
jaegergoing to fix FS#739 as well11:16
jue.oO finally a stable version of nouveau is out, but no GLX with current mesa ;)11:17
* Romster curses11:18
Romsteri guess it works with mesa3d git?11:18
juebut I think mesa 8.1 is not so far11:18
jueyes11:18
Romsterhmm #7389 wasn't my bug report by any chance?11:19
Romster#73911:19
Romstercool so yeah mesa3d next release as you said shouldn't be too far away.11:19
jaegernope, was moritz wilhelmy11:19
Romsterah11:19
Romsteri did bug the hell out of tilman not intentionally. but nothing seems to get done.11:20
jaegerwell, now we know that tilman didn't have time11:20
jaegerI will also test FS#738 this afternoon when I get back11:21
Romsterit would of been nice if he did just say that than to get angry at me, oh well. it's the past now, i'm just glad things are finally being reviewed and fixed.11:22
Romsterjaeger, i was wondering after you got a email regarding ffmpeg, was there any ports i maintain your using and wouldn't mind picking up? or your time is limited or there isn't really anything that you did maintain before i picked them up? I feel a bit guilty i picked up some of your stuff. -_-11:28
*** joe9 has quit IRC11:46
teK_reading backlog (@sepen)11:46
teK_;011:46
teK_done11:49
Romsterrecalculating11:50
Romsteri should be in bed but i slept earlier if anyone is wondering why i'm still here.11:50
Romsteri'll probably nap again before work later.11:51
Romstersleeping cycle is skewed.11:51
*** joe9 has joined #crux-devel12:34
jaegerback now13:06
jaegerRomster: no reason to feel guilty, I did retire for quite some time13:07
Romsterk13:07
Romsterjaeger, what do you use to work with GPT partitons to create/resize? parted is a prick and wants to resize the FS with it when we only want to resize the GPT partition only.13:24
Romsteri got told awhile ago that gdisk wasn't really stable enough or is that ok now.13:24
Romsteror anyone else that knows GPT13:29
jaegerin a quick meeting, give me a few mins13:31
Romsterk13:32
*** sepen has joined #crux-devel13:32
sepenteK_: I got an error when trying to push 'git'13:33
sepennow I see that you adopted the port13:33
sepenI improved the Pkgfile a bit http://dpaste.com/761701/ (replace the Maintainer line by yours)13:37
sepenmakefile accepts mandir= and INSTALLSITEMAN3DIR=13:37
*** Romster has quit IRC13:38
*** Romster has joined #crux-devel13:38
Romsterwtf i only posted 2 lines13:39
jaegerRomster: I use parted/gparted, never needed to do it with a filesystem already in-place13:40
jaegergdisk is mostly usable but it does one thing I don't like: it automatically converts MBR to GPT13:40
jaegerIf I WANT to convert to GPT that's fine but I don't want a tool doing it automatically13:41
Romsterah is that all13:41
RomsterWARNING: you are attempting to use parted to operate on (resize) a file system.13:41
Romsterparted's file system manipulation code is not as robust as what you'll find in13:41
jaegerif parted refuses to do it without resizing the FS you can work around it13:42
jaeger1) change parted to sector display mode13:42
jaeger2) take a note of exact beginning and ending13:42
jaeger3) delete partition13:42
Romsterbs just need to fix a stupped program that wrote ms-dos on a gpt disk that has a existing lvm setup, note it's not mine some guy i'm helping.13:42
jaeger4) recreate it with exact same beginning and new ending sector13:42
jaeger5) resize2fs13:42
jaegerif it's lvm resize2fs isn't what you want but you get the idea13:43
Romsterwont need step5 probably need to pvresize for step5 in this case. ok13:43
jaegerAs with anything else, make sure there are bakcups13:43
jaegerbackups, too13:43
Romsteri was wanting to be sure parted wouldn't do something stupid13:43
jaegerI've done this with both MBR and GPT myself but ALWAYS make backups13:43
jaegerAn important thing to note is that if you're using GPT you probably want to leave a little space at the end of the disk13:44
Romsteryeah i always suggest to backup first.13:44
jaegerCan't be TOO careful13:45
Romsterthis was GPT before so there will be some space at the end now.13:45
teK_sepen: who shall take it? :)13:45
jaegersepen: what do you think of maybe renaming x11-fonts-dejavu to xorg-font-dejavu?13:46
sepennah, maintainer is not the problem13:47
sepenjaeger: when moved to xorg?13:47
jaegersepen: sure, just wondered if you liked or disliked the idea :) no rush at all13:47
sepen+113:47
sepenwe talked about that and I think is a good idea to group xorg ports13:48
jaegerWe did talk about it but I missed your response, sorry. My fault13:48
teK_I don't understand13:48
jaegerok :)13:48
Romstermind renaming x11-fonts-dejavu -> xorg-fonts-dejavu ?13:48
sepenyep13:48
Romsteroh jaeger already asked.13:49
jaegerthat's what we were just talking about, will happen when we redo the xorg repo13:49
sepenI think is a great idea13:49
Romsteri'm preoccupied helping, so i missed that line.13:49
jaegerok, great :)13:49
sepene.g: libdrm is required only by mesa3d13:49
Romsteri'd like to move glew to xorg at some point out of contrib.13:50
sepenwe could research a bit and also I think is a good idea to do the same for xfce whenever be possible13:50
jaegertime to test font packages per FS#73813:50
jaegerI've also already done FS#764 in my updated ISOs but was already planning it for 2.813:51
sepenI'm ok with xfce ports but help is always welcome ;D13:51
Romsterjaeger, most of that path issue stuff should be fixed in my xorg fork13:51
Romsterfor fonts.13:51
jaegeronly 1 port had the path wrong, I already fixed that13:51
teK_so I keep git?13:51
sepenyes teK_13:52
sepenyes I have a bit issue with git when trying to push but I'm ok that you adopted git13:52
sepenteK_: you're maintaing git repos :)13:52
sepenteK_: but please, see my changes to Pkgfile13:53
teK_as it seems=)13:53
Romsterwhile your active sepen libmpd needs glib added as a dependency13:55
Romsteri'm still testing other ports13:55
sepenjaeger: xorg-font-foo13:56
jaegersepen: that seems like a good standard to me, the ones in the xorg repo already conform to it13:57
sepenI mean to replace 'fonts' with 'font'13:57
Romsterxorg-font-larabie-coolvetica-ttf and stuff in my repo conform.13:58
Romster+1 sepen13:58
jaegersepen: the xorg-font-* packages are already singular, only x11-fonts-dejavu wasn't13:58
sepenthis could be added as a note in PackageGuidelines13:58
jaegergood idea, probably13:58
sepenhttp://crux.nu/Main/Handbook2-7#ntoc2614:00
sepenthat will help people to create standard ports14:00
Romsteragreed though we'll know what's going on.14:02
teK_I seem to use a different TAB-setup than tilman or sepen does -_-14:02
Romsterwhat's so hard with a single tab14:02
Romstertilman uses no tabs14:03
Romsteruse what you prefer i guess, no one seems to mind as long as it work.14:03
Romsterbut i did get told off for using tabs in the meta tags for prt-get use so i only use a single space on that.14:04
Romsteralso i really dislike commas on Depends on:14:04
Romsterspaces work equally well on there own.14:04
jaegerI'm the exact opposite, much rather have commas than spaces14:09
Romsteryeah but i only see commas with spaces on the majority of ports.14:26
Romstersilly doing both.14:27
jaegerIf nobody objects I'll push http://jaeger.morpheus.net/linux/crux/tmp/xorg-font-bdftopcf.diff.txt to fix FS#73814:27
jaegerIt adds xorg-bdftopcf to many of the xorg-font-* ports' dependency line14:28
teK_sepen: objections wrt git? :>14:28
Romsterlooks fine to me jaeger i've already done that in my xorg fork.14:28
jaegerI wonder if I should push that as one large fix or push each port separately14:29
jaegerone large push is easy but it'll look like shit in the logs14:30
teK_nobody using identd anymore? strange :)14:33
jaegerI think I'll do each port individually, it'll look cleaner that way14:34
jaegerfor P in $(git status | grep modified: | awk '{ print $3 }' | xargs dirname); do git add ${P}; git commit -m "${P}: added missing dependency xorg-bdftopcf"; done14:36
jaegerYour branch is ahead of 'origin/2.7' by 25 commits.14:39
jaegerheh14:39
Romsterhmm14:40
Romsterthat's one way to do it.14:40
jaegeramusingly git shortlog shows tilman having done 1337 commits14:40
rmullhaha :D14:41
Romsterwas not expecting that14:41
jaegerRomster: it was the easiest way I could think of to do all 25 at once14:41
jaegerwhile keeping them separate commits14:41
Romsteri have a c program to parse the git diff, but i lost where it's on my hdd.14:42
jaegerprobably overkill in this case14:42
Romsterdarn it's got tilman that maintains cmake i was gonna pick it up and roll back the version, it's got a regression14:46
jaegerok, pushing my font commits14:49
Romsterok i'm going ot pick up flac if no one objects since i'm dong audio stuff.14:50
Romsterdoing*14:50
jaegerfine by me14:50
Romsternot like it's had a update in a long time.14:52
sepenteK_: great work!15:01
jaegerteK_: FS#830 seems like a safe change to me, though I don't use btrfs... have you run into any issues there?15:18
jaegeror frinnst15:18
jaegersepen: what group did we decide is willing to work on the new xorg repo?15:22
*** mike_k_ has quit IRC15:26
juenice to see that the separate-udev-from-systemd-build thread is still alive ;)15:36
teK_ah yeah rc-guy15:49
teK_sorry I 'missed' that15:49
teK_but I think I already gave my +115:49
teK_there was this 'ugly' message that /sbin/btrfs emitted15:50
teK_i.e.15:54
teK_/sbin/btrfs dev scan  1>/dev/null15:54
teK_failed to read /dev/sr015:54
teK_:)15:54
jaegerheh15:54
teK_but I was told they planned to fix that15:54
teK_it's not thatmuch of a deal but I'd ask  myself if everything would beokk15:54
jaegerWould be nice if upstream fixed it, yeah15:55
teK_00:31 < frinnst> i'll look into it a bit more tomorrow, maybe ask the #btrfs people if they have a best practice for it15:58
teK_14:03 < frinnst> teK_: looking at the btrfs source, the error we see with /dev/sr0 is the only type of error it can throw15:58
teK_14:04 < frinnst> correction, 2 errors15:58
teK_14:05 < frinnst> still, it's only a cosmetic and temporary error (will be more intelligent in future versions)15:58
teK_14:11 < teK_> so we'll stick to the current proposal and wait what the future brings15:58
teK_so we should go with it15:58
teK_http://crux.nu/bugs/index.php?do=details&action=details.addvote&task_id=83016:00
teK_any comments?16:00
teK_sepen (frinnst obvious) .. anybody else?16:00
jaegeryour +1 is for waiting to see what future updates upstream provides?16:03
jaegerinteresting... flyspray tells me "you voted for this task"16:04
jaegerI don't know what that means or how I would have done that16:04
jaegerapparently I did it today16:05
teK_+1 for applying the diff16:05
teK_more specific now16:06
jaegerWe could mute the output about /dev/sr0 manually, I suppose16:08
jaegerIt might annoy some users in the boot messages16:10
teK_yeah we considered that too, buthow? 2>/dev/null? bad incaseof other errors. patching out? worthless because itwill be changed16:11
rmullDid we ever fix the sysfs mount crap during boot?16:11
teK_the only reasonable but ugly thing would be to grep this one error out16:11
jaegerbtrfs device scan 2>&1 | grep -v "failed to read /dev/sr0"16:11
teK_hehe ;)16:11
teK_rmull: ?16:11
jaegerit's rather specific but it would be a temporary fix until upstream sorts it out16:11
jaegerI don't like it, honestly, but no idea when they might fix it16:11
jaegerit's ugly as hell in /etc/rc16:12
rmullteK_: Complaint about /sys already being mounted16:12
rmullIt was present in 2.716:12
rmullI haven't rebooted in a while16:12
rmullnot big deal, was just noticing it on my crux-arm system.16:13
teK_jaeger: it is indeed16:14
jaegerI guess my preference would be to wait for an upstream fix16:14
teK_can anybody confirm the /sys message? I don't quiteremember16:14
jaegerI vaguely remember discussion about the /sys issue but I don't remember the cause and haven't noticed it in a long time16:15
teK_I will pay attention when I reboot16:16
rmullIt would be from the mounting of sysfs in /etc/rc16:17
rmullLike others, I don't remember if it's still happening, only that it existed at one point16:17
jaegerhttp://crux.morpheus.net/irc/?channel=crux&date=04Jul2011#05:44:2216:17
rmullYeah, that's the one16:18
jaegerYou've had this happen very recently?16:20
rmullNot that I can remember16:20
rmullYou can ignore it for now16:20
rmullI'm making unnecssary noise16:21
jaegerwell, if it pops up again save the message and try to duplicate it :)16:21
rmullwill do16:23
jaegeroh, another question - are you running the default crux rc or a different one?16:25
rmulldefault. You can really ignore the question - I saw this on an older machine recently that has'nt been updated16:26
jaegerheh, alright16:26
*** sepen has quit IRC20:04
*** mavrick61 has quit IRC21:35
*** mavrick61 has joined #crux-devel21:36
jaegerDoing a test build of xorg 7.7 now21:52
jaegerremoved 44 ports from the repo that don't exist in the release sources, will see if they're obsolete or what21:52
jaegerthere were a few drivers mentioned in the release notes that are unmaintained so that will be some of them21:52
jaegerwill post details when I'm done21:53
jaegerthere are also 3 ports in the repo that don't match the versions in the release tarballs but I doubt that's a problem21:57
jaegerinput-synaptics, input-vmmouse, video-ati21:57
jaegerthey're newer rather than older21:57
jaegerer, 4... I forgot mesa3d22:05
jaegerwhy the hell does xinit not exist in X11R7.7?22:21
jaegerrunning into problems due to missing libpthread-stubs.la now, will have to work more on this tomorrow22:56
jaegerI don't like using the older xinit but no idea what happened there. I didn't find anything in the release notes about it22:57
jaegerno more libpthread-stubs.so breaks a lot of things... going to bed, will look more into it tomorrow23:07
*** mike_k_ has joined #crux-devel23:29
c0x /lib64 points to ../lib, is this normal?23:56
c0xmaybe /lib64 -> /lib, or more flexible /lib64 -> lib?23:56

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