IRC Logs for #crux Sunday, 2014-12-28

*** Romster has joined #crux00:13
mhethanks, I was building my LFS with pkgutils and needed python for X, but 2.7.9 failed because of some pyconfig.h and I could not find the right one00:16
mheso, it actually just skips that part that failed :)00:17
jaegeryeah00:20
mheis the pyconfig.h file in the Pkgfile and the directory needed? I think it is obsolete00:23
jaegerIt isn't needed in the i686 version, no, that was just a quick hack to get it working00:27
mhethanks, will get back to it tomorrow, got at the xorg-libs, with a package manager, that sucks00:29
mhegot all the xorg-proto stuff build in one package but the xorg-libs depend on one another00:30
mhewell, to bed it is00:30
diversehey jaeger thanks for helping me out yesterday00:40
jaegeryou're welcome00:43
diversejaeger: one interesting thing I wanted to mention (if you wanted to add this to the UEFI guide) that for asus boards, they look for the efi kernel in /efi/boot/bootx64.efi without having to use efibootmgr to specify where.00:47
jaegertechnically that's part of the EFI spec and how it's supposed to work for removable media00:47
jaegerIt's not in the docs, though, since you can be a lot more specific if you like, with efibootmgr00:47
jaegerI'll consider adding some notes about it, though00:48
diverseright, efibootmgr is a lot more flexible00:48
diversewith that tool, my uefi is my boot loader ;)00:50
diversejaeger: btw, do I still need to copy System.map to the /efi/boot along with the kernel?00:51
jaegerSystem.map is never *required* as far as I know00:52
diversein general or for efi only?00:52
jaegerin general00:52
diversedoes the handbook still need to say that we need to copy over System.map to /boot, or is that there for "just in case" in scenarios?00:57
diverses/in//00:57
jaegerIt's traditional to have it there but I don't think it's required. No reason NOT to do it, reallyu00:58
jaeger-u00:58
diversegotcha, just in case00:58
*** tilman has quit IRC01:04
*** tilman has joined #crux01:05
*** jdolan has quit IRC01:43
*** jdolan has joined #crux01:50
*** diverse has quit IRC02:27
*** jdolan has quit IRC02:45
*** diverse has joined #crux02:46
*** Feksclaus has quit IRC03:02
*** jdolan has joined #crux03:04
*** jdolan has quit IRC03:04
diverseanyone have experiences with Atheros' Killer eth on Linux?03:05
*** jdolan has joined #crux03:05
diverseah support for e2200 was added to the kernel since 3.10, nice03:09
*** BitPuffin has quit IRC03:24
*** Romster has quit IRC03:30
*** yang has quit IRC03:44
*** mavrick61 has quit IRC03:48
*** mavrick61 has joined #crux03:49
*** xeirrr has joined #crux05:09
*** Romster has joined #crux05:17
*** xeirrr has quit IRC05:34
*** yang has joined #crux05:50
diversewhere does prt-get store information about description, url, dependencies, etc?06:06
koridiverse: /var/lib/pkg/db06:08
koriwait06:08
koriI didn't read06:08
korinevermind.06:08
*** _root_ has quit IRC06:45
*** diverse has left #crux ("brb")07:24
*** diverse has joined #crux07:29
*** horrorStruck has joined #crux10:07
Romsterdiverse, parsed from Pkgfiles10:17
Romsterbut there is prt-cache that stores it in some file10:17
Romsterbut that only works if you use prt-cache ... instead of prt-get10:18
diverseI already got the parsing part, as you will see in my project. So I need to see how prt-cache works10:18
Romster--cache Use a cache file10:19
Romsterprt-get --cache ; prt-cache info deptree qt410:19
Romsterfar faster but everytime you do a ports -u ; prt-get --cache needs to be ran again. i see that as a design flaw.10:20
Romsterand anytime you manually edit Pkgfiles in certain circumstances like on dependencies. but can avoid that by using prt-get then and not prt-cache.10:20
Romsterso there is a few gotchas using prt-cache.10:21
Romsterparsing /usr/ports everytime is slow.10:21
Romsterbut genearlly the system buffers it after the first run so it's faster. but may never be faster than prt-cache.10:22
Romsterand you always need to parse first with prt-get --cache so the speed up isn't really worth the effort of prt-cache in my opinion.10:23
Romsterits like this 8 seconds to parse /usr/ports on first run, 2 seconds after that... prt-cache is 1 second or less.10:23
Romsterbut a royal pain if the cache is stale.10:24
diverseso prt-cache functions as a pre-cached prt-get?10:24
Romstersorry for the lengthy spiel on it diverse10:24
Romsteryep from a cache file.10:24
Romsterin /var/lib/pkg/...10:25
Romsteri think..10:25
diverseokay, but how does prt-get get the info for a package by itself?10:26
Romsterit parses Pkgfiles directly10:26
Romsteras per prtdir lines in prt-get.conf10:27
diverseoh I see10:27
diverseI'm so stupid10:27
Romsterwhich is why a prt-get deptree is slow on the first run10:27
Romsteror anything that has to resolve dependencies and read a ton of Pkgfiles10:27
diversehmm, would there be any benefit to having that information stored in the database instead?10:28
Romsterports could update the cache but that really doesn't guarantee that a file has been edited after that point manually either.10:28
diversebut then again I got to stay compatible atm10:28
Romsterpersonally you could avoid implementing prt-cache for now.10:29
Romsternot a critical component10:29
diverseyeah, I'll just do Pkgfile lookup and parse it instead10:29
Romsterthat would be cool if you used a separate Pkgfile parser library too.10:30
*** hhhhhhhh has joined #crux10:30
Romsterthat way if someone else decided to use this framework they could add more parser stuff.10:30
Romsterfor there format.10:30
diversewell I could have that as a seperate module in libmeta10:30
Romsterwe don't but some distros use an API= thing in there build file to set how to parse it.10:31
diversean API var?10:31
Romsterfor breaking backwards compatibility. but then those that don't specify it use the old parser.10:32
Romsterlike gentoo ebuilds they have an EAPI=10:32
Romsterwe don't have a need for an API version but that can be done in there own parsers. we just need to look for Pkgfiles and know we are on a crux system. "crux" returns CRUX version ...10:33
Romsterso it's not gonna add any extra code to Pkgfile parsing.10:33
diverseoh I see what you mean. You want to also build packages for other distros?10:34
Romstersince we don't use it.10:34
Romstersome others may. and might wanna edit the Pkgfile layout. one thing comes to mind putting # Depends on: into a array variable or something.10:34
Romstercrux probably wont ever do that.10:35
Romsterthese # tag things that prt-get uses now.10:35
diversethe array varaible will have to be parsed by metapkg anyway, I like the way # tag style for parsing it and keeping the sh stuff only for bash10:36
diverseof course I would like to change "# Depends on:" to "# Depends:" or "# Requires:" depending on which reads better10:37
Romsteri'm lagging.10:40
Romsterseems some router on the net is being a pain.10:40
Romsterok again10:40
diversegetting your messages again10:40
Romsterjust looked at the log, no messages missing.10:41
Romstersee pvtmsg diverse10:41
Romsteraka query10:42
*** leo-unglaub has joined #crux11:00
*** Feksclaus has joined #crux11:15
*** Feksclaus has quit IRC11:15
*** xeirrr has joined #crux11:54
diversez3bra: how do you edit avatars in utox?13:24
*** fengshaun has quit IRC13:54
*** BitPuffin has joined #crux13:55
*** xeirrr has left #crux ("Leaving")14:07
*** EmoSpice has joined #crux14:15
EmoSpiceso glad I just built a Pkgfile for acpid2, only to look at the original Pkgfile for acpid and realize it's the same package - just one point release behind (locally - doing a sysup now)14:29
jaegeroops14:53
mheI noticed an error in libdevmapper's Pkgfile14:54
mhe113.tgz does not exist on that mirror, I used 114.tgz in the Pkgfile14:54
EmoSpicejaeger: It forced me to consider how long it'd been since I did a sysup, so that's good at least.14:54
mhecore/libdevmapper14:55
*** hhhhhhhh has quit IRC14:55
*** doomicide has joined #crux15:19
*** fengshaun has joined #crux15:20
*** jdolan has quit IRC15:50
*** hhhhhhhh has joined #crux15:57
juemhe: libdevmapper is using the lvm 2.02.114 sources, something wrong on your side?16:33
cruxbot[core.git/3.1]: vim: update to 7.4.56016:36
cruxbot[opt.git/3.1]: gvim: update to 7.4.56016:37
cruxbot[opt.git/3.1]: whois: update to 5.2.316:37
*** EmoSpice has quit IRC16:58
*** _root_ has joined #crux16:58
_root_hello folks;16:59
_root_I went for new install at last;16:59
z3bramorning16:59
z3bradiverse, no idea about utox, sorry17:00
_root_I have a question though; What should be the filesystem of my / and /boot partition. Could I use btrfs as my / ?17:00
_root_does it add to speed. I am installing crux for desktop use.17:01
_root_z3bra: Hello; top of the morning to you too sir;17:01
z3bra_root_, any filesystem would fit17:02
z3bracommon choices are ext4/xfs17:02
_root_z3bra: ythey should be some advantage over one another17:02
z3brabut yeah btrfs would do it too17:02
z3braobviously17:02
_root_so what would I gain or loss by going btrfs17:03
z3brasnapshots17:03
z3braand partition resizing iirc17:03
diverse_root_: if you use btrfs, prepare to backup a lot17:03
z3brahttps://en.wikipedia.org/wiki/Comparison_of_file_systems17:03
z3bra^ check it out17:04
_root_diverse: howdy; What do you mean?17:04
_root_UNstable?17:04
z3brayou can consider btrfs stable17:05
_root_So what was that comment about backuping alot ?17:05
diversewell it's stable, but if a bug were creep in, it could corrupt his data, so I rather not take the risk, unless you are like frinnst who backs up all the time.17:06
_root_:-[17:06
diverse_root_: talk to frinnst all about it17:06
diverse(or any other btrfs user which I don't know of)17:06
z3brasame goes for ext4/xfs and such17:07
_root_diverse: yep; you're right. I am most certainly not like frinnst17:07
_root_z3bra: xfs?17:08
z3braanother filesystem17:08
diverse_root_: you can still use it, just prepare for random bad things to happen because btrfs is still evolving.17:10
z3brabtrfs has been in the kernel for 7 years now17:10
z3braand is considered stable17:10
diversethat's not my point17:11
jaegeryou should keep backups of your important data regardless of your filesystem choice17:11
z3bra^17:11
_root_diverse: what would be the bad thing? system not booting because of corrupt filesytem on root partition? or waking up one day and see there is not a single file in my music collection on my music directory?17:14
diverse_root_: based on my research of btrfs a few months ago, it looked like a hassle to maintain, based on the recent bug reports I have found. My point is, to me, the frequency of problems didn't seem worth it, regardless of me backing up. Which is why I say do it more frequently, not that you shouldn't at all.17:17
_root_diverse: ok So ext4 then. and no hassle. yes?17:19
diversesigh, it's less hassle, still backup of course17:19
diversefrom a software developing point of view, btrfs still has a lot of missing features that it plans to implement, new additions depending on how they do them, can sometimes introduce problems more frequently17:22
rmullI used btrfs for my / partition and I started experiencing corruption that would cause the partition to automatically get remounted read-only17:23
diversebut maybe things have slowed down much since I last look at it, so it's probably safer.17:23
rmullI still have two other HDDs in BTRFS raid 1 which have not had an issue.17:24
rmullNot sure what was going on with the root partition but I converted it to ext4 and haven't looked back17:24
rmullI guess it's worth pointing out that I don't think I actually lost any data, or at least not any data that I needed17:27
rmullI backed everything up with the disk offline, and then restored it to the new ext4 partition.17:27
*** Feigrim has quit IRC17:27
diversermull: so how about for mounting it as /home?17:30
diversesounds fine that way?17:30
darfo diverse: klogd(8) uses /boot/System.map to label kernel oops if it available17:34
diversedarfo: ah okay, good to know17:35
darfoalso see /usr/src/linux-`uname -r`/Documentation/oops-tracing.txt for more in-depth coverage17:36
darfoIt used to be in the root but it was an attack vector17:36
darfodon't want the baddie knowing where kernel locations are17:36
darfoso it was moved to boot thinking it would not be mounted all the time17:37
darfo*/boot17:37
_root_funny thing. I formatted a 4GB USB-stick with G-parted. now I can mount it by /dev/sdd for example on sdd1. why?17:41
_root_what is wrong with my usb flash17:42
_root_o.017:42
diversedarfo: since I use efistub to create an efi kernel, can the System.map file be only in the ESP directory as my kernel, or does it have to be fixed at /boot/System.map always?17:42
diverse_root_: getting back to btrfs discussion, it seems mostly fine for none root partitions. Either way, do what makes you happy.17:45
_root_diverse: for using efi kernel; should I have GPT partition or MBR will do?17:46
diverseGPT17:46
darfodiverse: check klogd(8), it looks in several places. It's not really needed, just nice to have.17:47
diverseah nice17:47
darfoif you get kernel oops that need debugging you can always put it somewhere17:47
darfoklogd can find it17:47
darfousually I keep it with the kernel so it is current if it is needed17:48
_root_diverse: I do /join #gentoo17:49
_root_and ask how to change to GPT17:49
diversejust use gdisk and write it to disk17:50
_root_diverse: do I loose data?17:50
diversenope17:50
diverseor wait...17:50
_root_waiting ....17:50
diversejaeger: does using gdisk to change to GPT just change the partition table?17:51
jaegerthat's how it's supposed to work. 99% of the time it will work fine17:52
_root_diverse: It seems I does. No loose od data; the otherway around have loose of data17:52
_root_*/it does17:53
_root_jaeger: hello17:53
jaegerhello18:01
*** Feigrim has joined #crux18:17
_root_jaeger: diverse I asked this before; but right now most of my system files (i.e rc.local rc.fix rc rc.single and config files in ~ directory) are all EXECUTABLE. Why? is it a malware on my system. why is this happening?18:19
_root_or they should be executable by nature?18:19
diverse...18:21
jaegerusually rc scripts are executable. if you're not sure, check the footprint of the port that owns them18:21
_root_ok18:28
_root_changing from MBR to GPT will effect on speed. I'm just asking?18:42
_root_*/will have effect18:43
joacimI don't think it makes a difference18:45
_root_joacim: I does solve the > 1TB problem as far as i know18:46
diverseit solves the 2.2TB partition limit, yes18:46
diversehowever if you use LVM, it's not a big deal18:46
_root_diverse: Speed?18:47
_root_diverse: LVM for desktop box? seems over-kill. or I am wrong. ?18:47
joacimit is a big deal. i rather have a proper fix, than to work around the issue with lvm18:51
*** jdolan has joined #crux18:51
_root_joacim: I don't get your point18:51
joacim"however if you use LVM, it's not a big deal"18:52
*** BitPuffin has quit IRC18:53
joacimlvm is nice, but for my storage drives, I prefer to use gpt instead of making my system more complicated by using mbr+lvm18:53
*** horrorStruck has quit IRC18:54
*** jdolan has quit IRC18:55
*** jdolan has joined #crux18:58
_root_diverse: I installed gptfdisk (gdisk) and I want to go from MBR to GPT withou loss of data. how?18:58
_root_gdisk /dev/sda <ENTER> r <ENTER> then?18:59
*** kori has quit IRC18:59
joacimI think it is best to find a html resource for converting from mbr to gpt without data loss19:00
*** kori has joined #crux19:01
diverse_root_: just do what makes you happy19:06
diversedon't overthink it19:06
_root_diverse: I want to go from MBR to GPT but how?19:08
diverseask the fine people at #gentoo perhaps19:08
_root_I did that in 2013 once with gptfdisk but I cant to get it this time19:08
_root_diverse: that was a joke m8.19:09
*** lnds has joined #crux19:19
_root_diverse: just gdisk /dev/sda <ENTER> w <ENTER> <ENTER>19:21
_root_thanks19:21
_root_Wellcome myself to 21th century :)19:22
_root_diverse: GPT+Grub2. in this case I would need a BOIS partition for my grub Yes? how much should it be? 2GB or 100MB?19:23
*** doomicide has quit IRC19:45
*** _root_ has quit IRC19:46
*** _root_ has joined #crux19:49
*** kori has quit IRC20:26
*** kori has joined #crux20:27
*** kori has quit IRC20:27
*** kori has joined #crux20:27
*** Feigr has quit IRC20:54
*** Feigr has joined #crux20:55
_root_kori: hello21:01
_root_I ceated ESP partition.21:01
_root_now how should I issue grub2 command for installing grub?21:02
koriESP?21:02
korialso, I don't use grub21:02
koriI use lilo21:02
koriso I wouldn't know, sorry.21:02
_root_lilo is old stuf friend; and doesn't support efi/GPT21:13
teK_elilo does.21:14
_root_teK_:  As of 2014-10-30 this project is no loger under active development21:17
teK_works for me :}21:17
koriI use compatibility mode21:27
korimy mobo is EFI21:27
korino GPT though21:27
*** Romster has quit IRC21:56
*** Romster has joined #crux21:57
*** leo-unglaub has quit IRC21:59
*** lmarie has joined #crux22:02
*** lisa_ has joined #crux22:07
*** lisa_ has left #crux ()22:08
*** leo-unglaub has joined #crux22:13
lmarieNo efi mobo here, but did use crux handbook to install grub.22:13
lmariegrub-install /dev/sdX22:14
lmarieand then:22:14
lmariegrub-mkconfig > /boot/grub/grub.cfg22:14
lmarieOn a side note, finally got enlightenmnet installed and booted into it.22:16
lmarieOrder: efl, elementary, evas_generic_loaders, e18, terminology22:17
lmarieTime to punch down dough. Later folks!22:18
*** lmarie has quit IRC22:18
mheI have this weird issue22:51
mheon my lfs/pkgutils box22:52
mheI am using the rc scripts from crux but put /etc/rc.d in /etc/init.d22:52
mheI installed dbus but the /run/var/dbus folder does not exist and if I create it and the system_dbus and dbus.pid files it gives other errors22:53
mhealso after a reboot the folder is gone22:57
*** teK_ has quit IRC22:58
*** teK_ has joined #crux22:58
*** hhhhhhhh_ has joined #crux23:03
*** lnds has quit IRC23:37
*** Romster has quit IRC23:46
*** hhhhhhhh_ has quit IRC23:59

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