IRC Logs for #crux Saturday, 2015-01-17

diversekori: actually #W is what I wanted00:00
koriheh00:00
diverseI need the title to show "WeeChat" in order for my notification thing to work00:01
diversealso fixed my unicode problem, so now I can see this: 水00:06
diverse俺はうれしいだ00:09
koridiverse: wrong one00:14
korithat's the kanji for water00:14
kori氷 is what you're looking for00:14
diverseもちろん、コーリさん00:14
diversei did it to see if you were paying attention, mr ice ;)00:15
koribut of course00:15
diversewell to be honest, when I first looked at your group logo, it looked like "water" kanji, but I did some googling :)00:20
koriI'm unhappy still00:23
korirepo's not growing fast enough00:23
korimight do work this sunday00:23
koriporting for the sake of porting00:23
diversewell, look at this list of applications and find stuff to port: https://wiki.archlinux.org/index.php/List_of_applications00:26
koridiverse: heh00:29
diverseif you are trying to hoard as much ports as possible, that would be a good place to start00:30
diverseplus you can quickly adapt their pkgbuild files00:40
diversejaeger: even though you were able to connect with openssl, did you try curl?01:01
diversejust a thought01:01
*** tilman has joined #crux01:06
*** xeirrr has joined #crux01:06
xeirrrmorning. I think it's bug: "switch user" and "Suspend" is not clickable.01:08
xeirrrI am inxfce401:08
diversewell for suspend, you need swap, but I assume you already have some form of swap enabled?01:08
*** xeirrr has quit IRC01:11
jaegerdiverse: yes and it failed. hence my comment guessing it's a bug in curl01:14
diversegotcha01:14
diversewhat a pain...01:14
jaegerespecially odd since curl upstream distributes that pem file01:15
diversethe said thing is, I can't find any curl alternatives, so I'm stuck01:19
diverses/said/sad/01:19
*** xeirrr has joined #crux01:21
diversexeirrr: did you get my message?01:21
xeirrrdiverse: Sorry I missed01:22
diversexeirrr: "well for suspend, you need swap, but I assume you already have some form of swap enabled?"01:23
*** xeirrr has quit IRC01:23
*** xeirrr has joined #crux01:23
xeirrrdiverse: Yeah01:23
xeirrrdiverse: 2g swap01:23
diverselet's see, I remember there being a "power" group, if you aren't already a part of that as a user.01:24
diversemight be related to the suspend feature01:24
xeirrrOk, I'll add and reboot to see01:25
diverseno reboot, just log out01:25
xeirrrok01:25
*** xeirrr has quit IRC01:25
Romsterlog out and back in is enough for new groups01:26
diversewell for anything except a new kernel upgrade01:26
diversebut unless it's too much work, yeah reboot01:26
Romsterwhat did cfe do again for llvm?01:27
diversecfe?01:28
RomsterteK_, why didn't you bump the release... i guess it wasn't that important for stray file removal.01:28
Romsterhttps://crux.nu/gitweb/?p=ports/opt.git;a=commitdiff;h=2ac4efbe63d5b3907e146f58e4cc2b8ffca16e8101:29
jaegerif the build doesn't change, no reason to make people rebuild01:29
Romsterprobably a clang thing than llvm01:29
Romsterjaeger, yeah i see the reason behind it.01:29
diverseit should probably go into the clang port01:30
Romsterah yeah cfe is needed for clang and it was with llvm originally.01:30
koriSHIT01:30
Romsteranotherwords soemone messed up when splitting that clang out of llvm port01:30
koriI just realized I never bump releases :301:30
korioh well01:30
Romsterkori, only on breakable changes.01:31
jaegerit was overlooked, that's all01:31
Romsteri know it happens01:31
Romstereven i didn't notice it.01:31
*** mavrick61 has joined #crux03:43
rmullIt wouldn't be too difficult for someone to write a script that syncs Arch's PKGBUILDs from git and converts them into crux Pkgfiles04:02
rmullRight?04:02
rmullpull, convert, prtcheck, publish...04:03
*** krue has joined #crux04:09
korirmull: it would probably involve a lot of sed04:17
koriand pain04:17
diverseprobably not a lot of sed, but maybe implement a buffered reader to parse the pkgbuild, collect data, and write out a Pkgfile using said data04:31
Romsterhttp://crux.ster.zone/ added logs list and version sort of distfiles. i got bored.04:35
Romsterrmull, there was some tool that half did the job of converting to Pkgfile but then i see soemone packaged packman if you search the portdb04:36
diverseoh wow, so distfiles is the collection of non-local stuff04:36
Romsterpacman*04:37
diverse*are04:37
Romsteri've had it forawhile and others are making there own distfiles collections.04:37
Romsteri just parse source= and wget04:37
Romsterfor now04:37
Romsterhttp://www.versionsort.com/list/crux-ster-zone-distfiles.txt that is the interesting one. i can see how many verisons i have in my directory.04:38
Romsteri just need to implement a --tail option to filter tool and then i can make it rm old versions. without relying on prtwash or other Pkgfile parsing tools04:39
Romsterhas the side effect of seeing what gets bumped the most.04:41
RomsterDistinct package names: 143704:42
*** xeirrr has joined #crux04:42
Romsterto top it off you can search though http://crux.ster.zone/distfiles/MD5SUMS04:45
xeirrrdiverse: Sorry for being late,(quite busy, get a break now). I indeed add $USER to power group, but "suspend" and "switch user" is still not clickable04:45
xeirrrRomster: do you know who maintain package "shadow"?04:46
Romsterprt-get info shadow04:46
Romsterit'll be the core ports team04:46
diversexeirrr: you are using xfce4-power-manager, right?04:47
xeirrrdiverse: yeah04:47
Romsterfrinnst, teK_ jue afaik not sure who else is in core. see core.git commits if you must.04:47
xeirrrRomster: Why our shadow doesn't provide command "gpasswd"?04:47
Romsterhttp://crux.ster.zone/gitstats/core/authors.html i should update this again04:47
Romsterwe have passwd why do we need that?04:48
xeirrrRomster: I think it's easy to maintain user and groups for some ways rather than edit /etc/group by ourselves04:51
Romstermaybe your trying to add a user to a group? use useradd -g for that.04:51
diversexeirrr: did you use "exec startxfce4 --with-ck-launch" in your .xinitrc file?04:51
xeirrrhttp://man7.org/linux/man-pages/man1/gpasswd.1.html04:51
Romsteryou don't need to edit /etc/group manually.04:51
xeirrrdiverse: you sure we need consolekit?04:52
diverseI guess... http://forums.debian.net/viewtopic.php?f=5&t=114412#p54070804:52
diversexfce does hard code some stuff for functionality04:52
diversexeirrr: I thought you use fluxbox though?04:54
xeirrrRomster: It's an idea though. I hope we do add it.04:54
Romsteri haven't ever had a need for it.04:55
xeirrrdiverse: nope, this is another machine, xfce404:55
diverseah gotcha04:55
Romsterdespite the ton of stuff i do.04:55
Romsterhard coding stuff is brain dead anyways04:55
Romsteralways leads to issues.04:55
xeirrrdiverse: will try it.04:56
diversexeirrr: all I can say is avoid the *-kit stuff and figure out the command tools for them and bind to them key presses04:56
xeirrrdiverse: Thanks man04:56
diversewell I guess that's all you can do for suspend, but for switching users, you are going to have to log out for that04:56
diverseand use "su - $USER"04:57
xeirrris Alan here?04:57
diversenope04:57
xeirrrbefore I am back to work, I hope someone can tell him and bump smplayer to 14.9.004:57
xeirrr3 minutes left, need to go back to work.04:58
Romsteror file a bug report for it.04:58
xeirrrRomster: Thanks04:58
xeirrrhave a good day04:58
Romstersince he doesn't idle in here.04:58
*** xeirrr has left #crux ("Leaving")04:58
Romsterhaving a lazy day04:58
diversehaving the distfiles is pretty handy though, in case if the mainstream server goes down05:06
*** SiFuh has joined #crux05:06
Romsteryeah and a few of us are starting todo it now too... or more official05:06
*** vlnx has joined #crux05:32
diverseI wonder if html5 youtube videos can display 720p or higher videos again?05:46
diverseI remember last time it got start at 360p05:46
koridiverse: on chromium you can05:50
korion firefox, just 360p05:50
diversedamn google05:50
diverseI guess I need to figure out a way to fake out youtube thinking I'm chromium, using user-agents like someone mentioned05:54
koridiverse: oh no05:59
koriit's related to technologies built in05:59
koriI'm pretty sure it's not about "boo hoo you're using firefox"05:59
diverseI thought it was more of google saying "hey, you 'other' browsers users, fuck you!"06:00
diversekori: so what technologies are missing?06:01
koriI am not 100% sure but youtube may be using blink-only things06:02
diverseblink?06:02
koritheir engine06:02
korifork of webkit06:02
diverseah ha06:02
diversedefinitely an implicit FU then06:03
diversebut nothing I can do about it06:03
diversekori: I guess, is blink compatible with other webkit browsers?06:06
korinot sure, probably noty06:06
korinot'06:06
diverseI would doubt it as well, but just to be sure.06:07
diverseGoogle - "Don't be evil" ... more like "Evil, Evil Everywhere"06:08
diverseRomster: I just updated my rust compiler after 4 days and my code still compiles ;)06:24
koriHEH06:25
diverseand heck, even before that, after updating it previously after 3 days, it still compiled06:33
diversewow it's been a week06:35
*** darfo has joined #crux06:55
Romsterdiverse, can you just stcuk your ports in a httpup or rsync repo and give me the url to try it. and Pkgfile metapkg too.07:07
Romstereasier for me to fire up a docker container07:07
diverseI could put them up on a github repo07:32
Romsterthat would be great07:33
Romsteri did grab the text last time but i later rebooted and yeah07:34
cruxbot[contrib.git/3.1]: smplayer: Updated version 14.3.0 -> 14.9.007:35
diverseI need to do one last code optimization, after giving another look, I realize instead of needing to reread the pkgadd.conf file and create a new allocation each time, I can just read it once and reference it many times instead.07:46
Romsteryeah memory is plentiful and reading from disk is expensive07:47
Romsternot like pkgadd.conf will change during a operation07:48
diverseless allocations means faster, I'm not doing it to save memory space07:52
Romsteryeah07:52
*** mhe has quit IRC07:54
diverseperhaps in some cases, I'm doing things a little bit faster than pkgadd08:04
diversebecause originally I was just porting the source without thinking about the structure08:04
diversekori: ping08:16
diverseah, nevermind, sorry about the ping08:19
diverseRomster: https://github.com/guruverse/cruxports08:29
Romsterbtw when you use git you also need to depend on git08:31
diverseyep08:31
Romsterah rust has it and you depend on that so thats ok08:31
diverseah, I forgot to add git to tar-rs, but since rust depends on git, it's not a big deal atm08:35
Romsteryeah i was just looking at that but rust pulls it in08:35
Romsterand you cant build tar-rs without rust anyways08:36
diverseor cargo for that matter08:36
diverseRomster: so keep in mind, there is stuff missing as commented in the `/* */` which I'm aware of, but what I want see is how other things work08:38
diverseas in other stuff breaking08:39
Romsterso what works and what does not so far?08:44
diverseRomster: https://github.com/guruverse/metagear/blob/master/src/metagear/gear.rs#L30708:45
diversehttps://github.com/guruverse/metagear/blob/master/src/metagear/gear.rs#L39108:45
diversehttps://github.com/guruverse/metagear/blob/master/src/metagear/pkg.rs#L29408:46
diversehttps://github.com/guruverse/metagear/blob/master/src/metagear/pkg.rs#L32108:46
diversebut of course you wouldn't be using the build command just yet anyway08:47
Romsterwas going too but guess we are jsut checking add and remove for now08:52
diverseand `info`09:00
diverseyou don't have access to use `build` yet either way09:00
Romsterk09:04
Romsterwell i can see what works and what doesn't09:04
diversethat's what I want09:04
Romsterin a container so i don't hose anything.09:04
Romsterif anyone can find a bug it'll be me.09:05
diverseI don't expect any less09:05
diversealso if you want that tar path thing solved quicker, bitch at acrichto at #rust09:05
Romsteroh he is the owner of tar-rs09:06
diverseyes09:06
diverseI even made this issue a long time ago: https://github.com/alexcrichton/tar-rs/issues/709:06
Romsternot even assigned yo anyone09:08
Romsterto09:08
diversehe's the owner, so he should be the one to do it09:09
Romsteri still need to open a ticket for cargo but ill have a play with cargo before i do.09:09
diverseyou can pull tar-rs and use cargo to build it like make, but if you are trying to communicate with the server at all, you will not be happy09:10
Romsteris that rust-curl wrapper viable to use?09:10
diverseprobably, put I need to create like 5 different ports for it, just don't try to build it explicity, because that will pull dependences from the central repo09:11
Romsterwell i expect to be able to git clone or download an archive and cargo to build. and not expect cargo to talk to the server at all09:12
Romsteryeah this is the side of cargo we would like fixed. i was expecting some #cargo action. instead i got shunned or felt like it09:13
diversethat's how I felt too09:13
Romsterits been days not just 12 hours no one replied...09:13
Romsteracrichto> jdm: right now cargo may silently update submodules and not print out anything (and it could take awhile)09:14
Romsteroh nice...09:14
diverseso even if we got curl to work, cargo still has problems, sigh09:15
Romsterenne> Hi! Does anyone know if there are plans for "make install" type functionality in cargo? Or should I just use autotools to install programs?09:17
Romsterhaha09:17
Romsternot just us09:17
diverseoh wow, I got to join back to what how this goes09:17
diverses/what/see/09:18
diverseRomster: was that an old message?09:19
Romsternot that long ago in my backlog09:21
Romsterenne has quit though09:21
*** toriso has joined #crux09:22
*** SiFuh has quit IRC09:33
*** SiFuh has joined #crux09:36
*** hhhhhhhh has joined #crux09:54
*** leo-unglaub has joined #crux10:07
*** mhe has joined #crux10:15
tilmanhttp://imgur.com/fqjnK10:26
*** Pingax has joined #crux10:27
*** Pingax has quit IRC10:27
*** Pingax has joined #crux10:27
leo-unglaubhahaha, awesome10:27
diverse?10:28
diverseoh10:28
diverseoh10:28
diversethat is pretty funny10:29
joacimI use Windows Me, so I guess I'm safe10:30
Romsteromg is that for real10:30
tilmandunno, got a win98 system to test it on? ;)10:31
diverseone of the comments say you can try it on XP via the "handcap" menu10:36
Romsternot handy no.... but there are vm images of win 9810:59
Romsterwonder if that is second edition or not10:59
*** SiFuh has quit IRC11:18
*** SiFuh has joined #crux11:19
*** kls has joined #crux11:31
*** xeirrr has joined #crux11:41
*** xeirrr has left #crux ("Leaving")11:46
cruxbot[xorg.git/3.1]: mesa3d: fixed configure options13:31
torisoWindows Me, was the worst for me, i got the blue screen death of it, and that's how i started it on linux...so maybe i have to be thankful of Me :)13:33
*** leo-unglaub has quit IRC13:43
Romsteri went to me and went no drivers... noting but broken... went back to xp then i went to linux when i heard how awful vista was.13:44
*** BitPuffin has joined #crux13:49
kyubikoI may have accidently uninstalled flex, and it doesn't want to build again without libfl.so, a file it itself provides13:50
kyubikocould somone build it and upload the tar.gz so I can install that?13:50
*** xenkey has joined #crux13:55
Romsterkyubiko, you cna mount the iso and pkgadd flex#.... out of core/13:57
Romstercan*13:57
Romsterbut i'll upload this one i just built13:58
kyubikoRomster: Oh yeah, I didn't think of that13:58
kyubikothanks13:58
Romsterhttp://crux.ster.zone/packages/3.1/flex%232.5.39-1.pkg.tar.xz14:00
kyubikoRomster: thanks, that saved me having to wait for the iso to finish downloading anyway14:03
Romsterah14:03
Romsterthought you'd have it there...14:03
Romsteri've loop mounted the iso file before when i've did a bobo.14:04
kyubikoI dunno, I might have it somewhere14:04
kyubikoI don't think I do, unless it actually gets put someone on the fs on install14:04
Romsternah14:05
Romsterbut you can keep a copy of built packages before you do silly stiff <<14:06
Romsterstuff14:06
Romsterthat was its easy to fix and having the iso on cd or usb stick is wise.14:06
Romsterincase it wont boot at all.14:06
kyubikoI didn't actually use a cd for installation14:07
Romsteroh14:10
*** hhhhhhhh has quit IRC14:12
*** xeirrr has joined #crux14:28
xeirrrHi, hmmm, I am tired of doing a version bump bug report.  Our packages "libfm" "pcmanfm" are a little bit old. sourceforge url returns a 404 error. So please fix url. Btw, is rotwang still maintaining libfm?14:35
cruxbot[xorg.git/3.1]: mesa3d: configure cleanup14:36
xeirrrIf not, someone needs to take over14:37
xeirrrhave a good day.14:37
*** xeirrr has quit IRC14:38
teK_speaking of 404s..14:38
teK_we worked to provide distfile mirrors (in Europe, the US and somewhere down under, I think)14:39
teK_you are all invited to test them by adding this to your pkgmk.conf: PKGMK_SOURCE_MIRRORS=(crux.nu/distfiles)14:39
teK_this works for core, opt and contrib ports14:40
teK_for now.14:40
tilmannot sure if that's the joke, but that url (http?) is a 40414:41
teK_:D14:41
RomsterteK_, did you include mine?14:45
teK_I did14:45
Romstercool14:45
teK_tilman: there, I fixed te listing just for you. Still worked for pkgmk before that :)14:45
*** leo-unglaub has joined #crux14:46
Romsterso do we have a wiki page on the syntax of mirror selection?14:46
teK_no14:46
teK_let's test it first before going official :)14:46
Romsterk14:46
tilmanhttp://crux.ster.zone/distfiles/0.9.4.tar.gz o_O14:47
tilmanwhat is that, a surprise tarball?14:47
teK_say thanks to github14:47
teK_you really did not follow developments in crux :-)14:47
Romstergithub14:47
teK_it's from someone braindead.. there are several ports with that issue14:47
Romstersurprise tarball dies14:48
Romstergithub does not supply names it's stupid.14:48
RomsterPKGMK_SOURCE_MIRRORS=(http://10.0.0.1/distfiles/ http://crux.nu/distfiles/)14:48
leo-unglaubhey :)14:48
Romsterthere i'll test that.14:48
teK_hey14:49
Romsterlo14:49
Romsterhttp://distfiles.serverop.de/ same deal tilman14:49
tilmanteK_: i'd refuse to package such crap14:49
teK_well14:49
teK_the alternative option is  to dl it, rename it and self host it14:49
teK_which sucks too14:49
Romsteronly way around it is either rename it in pkgmk source... or make it away of funky paths like http://romster.me/distfiles/hash/0.9.4.tar.gz/md5/daeee7dcccea33d6258a0a9d783470c4/0.9.4.tar.gz14:50
*** leo-unglaub has quit IRC14:50
Romsters/away/aware14:51
Romstersorry leo-unglaub hi wasm14:51
Romsterwasn't ignoring..14:51
*** jdolan has joined #crux14:51
Romsteror grab the git tree and tarball it ourselfs.14:52
Romstermaybe that's an option14:52
Romsterand stick it in on a server and let the mirrors sync.14:52
*** leo-unglaub has joined #crux14:54
leo-unglaubdamn14:54
leo-unglaubthis is an observation14:54
RomsterteK_, this isn't fool proof if distfile mirror hasnt' got the file it'll 404 and pkgmk will give up. until it is ran again and http://crux.nu/distfiles/ goes to another host (and thats a if)14:54
leo-unglaubwith every new kernel release, my computer gets more instable14:54
Romsterwb leo-unglaub14:54
leo-unglaubcan someone confirm that?14:54
teK_wtf Romster, we spoke about that already14:54
teK_it will try the original URL either way14:54
leo-unglaubmy chat tool just crahed and even killing it as root did not help14:54
teK_didyou kill -9 it?14:55
Romsterleo-unglaub, eh i'm happy with 3.17.7 i think 3.18 has problems still.14:55
leo-unglaubi tryed every kill level possibe14:55
leo-unglaubRomster: i am on 3.18 ...14:55
tilmanif its stuck in the kernel it just may not be killable14:55
tilmansuch is life :-)14:55
RomsterteK_, yeah i know... how it'll work.14:55
Romsteri've had that.. damn zombie processes14:56
leo-unglaubit's just ... a few years back i never had a program that i could not kill14:56
leo-unglaubnot i have this once a week14:56
tilmanthat's not a zombie14:56
leo-unglaub..14:56
Romsteri have had zombie state process that i couldn't kill. not saying leo-unglaub got that just now.14:56
Romsterbut that was a few kernel versions back14:57
tilmanzombies are already dead, no need to signal them14:57
Romsteryeah they sit in the process tree and wont go away... harmless though14:57
leo-unglaubyes, but you are missing the point14:58
leo-unglaubin my opinion linux has become a lot more unstable then it was years ago14:58
tilmani know what the point is, and you are probably right14:58
teK_I just wrote14:58
teK_to github wrt the release file names..14:58
tilmankinda difficult to deliver a stable product when things move at break-neck speed14:58
tilman:]14:59
leo-unglaubtilman: indeed14:59
leo-unglaubbut it still anoys me if someone in a meeting with windows 7 is having a better comuter experience than me ...14:59
tilmanjust laugh at them. they don't have a proper terminal :)15:00
leo-unglaubhaha, indeed *g*15:01
leo-unglaubpowershell my ass *g*15:01
Romstermaybe you can blame systemd?15:01
Romstereven though we dont' run it... the kernel has been tarnished with all the changes to make it compatible.15:01
leo-unglaubactually i do15:01
Romsterbecause ya know. the kernel was buggy when systemd broke the kernel.15:02
leo-unglaubdebian has become a lot more unstable15:02
Romster</joke>15:02
leo-unglaubso unstable that i tryed systemd15:02
leo-unglaubäh, crux15:02
Romsteroh gawd15:02
Romsterhow did that fair... not well your back here.15:02
leo-unglaubsystemd just introduced a lot new complexity to the entire thing15:03
leo-unglauband is partly responsible for the entire problem15:03
Romsteri'm sure it's not all the fault but a big part of it.15:04
Romsternot even well tested15:04
leo-unglaubsome ideas behind systemd are not bad15:04
*** jdolan has quit IRC15:05
Romsteri think it's more the implementation than the ideas that's really bad15:05
leo-unglaubyep, some ideas are good in there15:05
Romsterif it didn't try todo everything then there be far less resistance15:05
leo-unglaubfor example i still dont know why all our rc scripts are manually written15:05
leo-unglaubif we could abstract it15:05
*** jdolan has joined #crux15:06
Romstereh that depends on the init rc stuff used.15:06
leo-unglaubyes, but the entire find the pid file, get the pid, kill the process15:06
leo-unglaubis mostly the same every time15:06
Romstermy bigest concern is other init systems.... there getting the arse over systemd... diversity, isn't that linux's biggest asset?15:07
Romsterrunit had that pretty much covered with bearly any code required...15:07
Romsterwas basically run the program, if the process dies it'll restart it on the next 30 second scan of watch dog timer15:09
Romsterthere is no one way to write a service file.. but looks like there is now... systemd style.15:10
cruxbot[contrib.git/3.1]: dvb-firmware (NEW): Various DVB firmware15:13
leo-unglaubbrb kernel update15:32
*** leo-unglaub has quit IRC15:32
*** leo-unglaub has joined #crux15:33
leo-unglaubso, new try :)15:33
leo-unglaubgreat, now suspend is broken ...15:33
leo-unglauboh for fucking sake ..15:34
leo-unglaubi am this close to download an openbsd iso ...15:38
Romsteroh noes15:38
*** booly-yam-271 has joined #crux15:41
tilmanleo-unglaub: steam and opengl drivers have kept me in linux land the last couple of months. i'm quite tempted15:43
tilman:D15:43
leo-unglaubtilman: *g*15:43
leo-unglaubi fully understand why15:44
tilmanand dual booting is for suckers :)15:44
leo-unglaubmy computer here is the best of the best because i like to play computer games .. but i just extracted the kernel archive ... and my chat window lagged ... i mean ... seriously??15:44
tilmani'm a bit conflicted about the gpl dislike (hate?) at the other side though15:44
leo-unglaubi have 16 GB RAM, two SSD drives in raid 0 ... and it laggs ..15:45
jaegerI have a sparc LX in a closet with openbsd on it, I think... if you want to play with that :D15:45
leo-unglaubCrux for openbsd ... that would be nice *g*15:45
tilmanor pointless15:45
leo-unglaubtilman: yes and no15:46
tilmanconsidering crux took a lot of ideas from bsd15:46
tilman:p15:46
leo-unglaubi know ;)15:46
kyubikoleo-unglaub: how many cores does your cpu have?15:46
leo-unglaubkyubiko: 4 real ones with HT15:46
leo-unglaubso 8 cores for the os15:46
tilmanenjoy _one_ of them with openbsd15:46
tilmanmuhahaha15:46
leo-unglaubtilman: lol, that is a myth15:46
tilmanoops, didn't mean to spread FUD. i thought openbsd doesn't have SMP?15:47
kyubikoweird15:47
leo-unglaubwth ... it's still unpacking that archive ...15:48
*** booly-yam-271 has quit IRC15:49
leo-unglaubis that a bsdtar bug?15:49
joacimThe mouse cursor on my old C2D machine would lag when I played flash stuff15:49
leo-unglaubmaybe bsdtar does not like unpacking the linux kernel archive *g*15:50
joacimnever seen irc lagging tho =)15:50
leo-unglaubjoacim: i cannot move the window15:50
leo-unglauband typing is delayed by 10 secounds15:50
joacimthat's not bad. you're expecting too much of your i7/xeon!15:51
jaegerperhaps the underlying storage is failing15:51
tilmanleo-unglaub: how are you unpacking it? verbose mode?15:51
tilmanie stdout spam or not?15:51
leo-unglaubwell you are right .. i can play battlefield and crysis in high resolution, but unpacking a fucking archive is to much?15:51
leo-unglaubtilman: no, no std spam15:52
leo-unglaubjaeger: no, the drives are fine, i already checked15:52
jaegerterminal output speed can matter, yeah. I can hang some of my machines unpacking archives on the efi framebuffer15:52
*** jdolan has quit IRC15:52
jaegeror an x terminal when acceleration isn't working15:52
jaegerif it's not a visual issue, what about your filesystem and its mount options? the block scheduler?15:53
jaegeror even the CONFIG_HZ_* setting in the kernel?15:54
*** jdolan has joined #crux15:54
jaeger(just guesses)15:54
leo-unglaubthe filesystem is btrfs and the drives are two ssd drives in a RAID015:55
leo-unglaubso they should have double write speed15:55
jaegerew. is TRIM even supported in that config?15:55
leo-unglaubnone of the drives is full and trim is disabled15:55
jaegerare they sandforce controllers?15:56
leo-unglaubi disable trim on all ssd drives to make them life longer15:56
jaegeroh god15:56
leo-unglaubwhat do you mean by CONFIG_HZ?15:56
jaegertell me you're trolling, please15:56
leo-unglaubno, i follow the official kernel mounting flag proposal15:56
joacimI've been running without trim on my mac for about a year now. it is fine.15:57
joacimwrites are a few MB/s slower now, but reads are what they've always been15:57
jaegerThe controller type makes a big difference. There are some that can behave well with TRIM disabled, some that really shit the bed15:58
jaegerThere's no good reason to disable TRIM on a system which supports it15:58
leo-unglaubhmmm15:59
leo-unglaubjaeger: do you mount with the realtime flag?16:00
joacimI just checked. writes aren't slower at all. this is an 840 pro.16:00
jaegerI use relatime,discard16:00
leo-unglaubrw,relatime,ssd,space_cache16:00
jaegeran 840 pro is one of the best in terms of read/write consistency. It's NOT a shitty drive16:00
joacimi looked at the benchmarks when i bought my 850 pro. it seemed like it was one of the few that didn't shit the bed when you stressed them.16:01
jaegerleo-unglaub: btrfs enables the "ssd" flag by default16:02
jaegernot familiar with the space_cache option16:02
leo-unglaubspace_cache uses the drive for swaping and caching16:03
leo-unglaubif the ram is full16:03
jaegerbtrfs-specific?16:03
leo-unglaubnot sure16:04
jaegera quick search suggests it is16:04
jaegerDepending on your SSDs' controllers, disabling TRIM could actually lead to enough write amplification that you shorten the life of the drive rather than stretch it16:05
Romsterext4   defaults,relatime,discard16:06
Romsteron my intel ssd16:06
jaegerwith all this said, "shorten the life of the drive" usually means jack shit, at least with modern SSD controllers16:07
jaegerbecause you'd replace the drive with something bigger or newer long before it actually failed16:07
joacim5.25" SSD!16:07
jaegeranother thing about which I wonder - are you using whole disks or partitions for the RAID0? If the latter, are they aligned properly?16:08
Romsterdon't SSDs align to 1MiB?16:13
jaegernot automatically. the partition software can, though16:14
jaegertechnically you want to align to a multiple of the erase block size but sometimes that info is hard to find16:14
jaegerusually 1MiB or 4MiB are safe16:14
Romsterdoes since lately not like the early stuff16:14
jaegerparted does by default, for example. no idea on fdisk recently16:15
Romsteri had to be careful with my lvm setup as it did default to 384Kb than to 1MiB16:15
Romsterfdisk since a few versions ago does too16:16
*** hhhhhhhh has joined #crux16:41
*** doomicide has joined #crux16:47
*** hhhhhhhh has quit IRC16:50
leo-unglaubokay, suspend now works again17:15
leo-unglaubbut hibernate is still broken ..17:15
*** c0x has joined #crux17:25
leo-unglaubdoes anyone of you know a better way to list all hard drives than fdisk -l?17:54
rmullWhat's wrong with fdisk -l in your opinion?18:10
rmullHow about ls /dev/sd* :P18:10
Romsterfor a in /dev/sd[a-z]*; do fdisk -l $a; done18:11
Romsteror something18:11
Romsteror maybe you want blkdev18:12
Romsterblkid18:13
*** Feksclaus has joined #crux18:16
*** kori has joined #crux18:19
korihey crux18:20
*** Roosevelt has joined #crux18:21
Roosevelthey all18:21
koridiverse: did you need anything earlier? I know you told me to nevermind the ping but now I'm curious18:22
*** kls has quit IRC18:23
rmullkori: ping18:23
korirmull: pong18:24
rmullnever mind18:24
Rooseveltanybody want to tryout the new awesome(3.5.6) find it here.18:24
kori:18:24
kori:<'18:24
rmullgotcha!18:24
tilmanrmull master troll18:24
rmullYes, yes, fear me18:24
leo-unglaubrmull: Romster: i am asking because fdisk -l is missing a harddrive ...18:34
joacimhow about parted?18:38
joacimyour motherboard have multiple sata controllers or something?18:39
leo-unglaubit's in the laptop18:39
leo-unglaubso i cannot simply swap the port18:39
joacimthe idea was that maybe you were just missing the driver for the second sata controller or something18:45
joacimunlikely when most controllers are ahci compatible these days18:45
leo-unglaubjoacim: no, it's ahci18:45
leo-unglaubworked for the last 6 month18:45
leo-unglaubmaybe the secound laptop ssd broke18:46
rmullDoes dmesg show anything related to it?18:52
leo-unglaubnope, nothing18:52
leo-unglaubi dont even find it in the bios...18:52
rmullWow, that is unusual18:52
rmullMaybe you're on the wrong computer and you only think you have a HDD and controller18:52
leo-unglaubhehe, you meight be laughting ... but i really checked for that18:53
leo-unglaubbecause it seams so strange18:53
leo-unglaubbut it's the right laptop, because in my init file is a cryptsetup decrypt command for /dev/sdb18:54
leo-unglaubbut there is only /dev/sda to be found18:54
*** crash_ has joined #crux18:54
rmullAny kernel updates? Maybe the drive letters changed?18:55
leo-unglaubrmull: no, if the drive letters changed, there still would be two drives shown18:56
leo-unglaubbut the former sdb is completely gone18:57
leo-unglaubaccording to all tools i find there is only sda18:57
rmullDo SMART utilities report anything?18:58
leo-unglaubno, they also dont find the drive18:58
leo-unglaubWTF18:58
leo-unglaubthats new for me ... the drive completely gone ..19:01
leo-unglaubi need to open my laptop19:01
leo-unglaubi want to check if someone took my second drive ...19:01
*** dougl has joined #crux19:04
*** darfo has left #crux ("Leaving")19:09
*** dougl has quit IRC19:12
*** leo-unglaub has quit IRC19:39
*** leo-unglaub has joined #crux20:07
*** doomicide has quit IRC20:17
*** hhhhhhhh has joined #crux20:29
*** crash_ has quit IRC20:51
*** xenkey has left #crux ("Leaving")21:56
diversekori: I wanted to know, if you had set the .gitignore file to ignore *.tar.* files in your ports repo, but after getting impatient, I went to your repo and saw you only the ignore the work dir, therefore I concluded you wouldn't know the answer to my question, thus "nevermind"22:31
diverseI wasn't sure if the asterisk wild card would work either, so I ended up having to do some googling22:32
diversealthough I did learn thing that could be of interest to you, if you use "**" pattern, it scales up all directories to the search point, so "**/foo" could ignore e.g. "/bar/foo" "/baz/foo" "/baz/bar/foo" etc22:35
diverse*one thing22:35
diverseor "/my/awesome/path/to/the/one/and/only/foo"22:39
diverseyou get the idea22:39
*** jdolan has quit IRC22:39
*** jdolan has joined #crux22:40
koriaye22:42
koriwell I don't use that work dir anymore22:44
*** toriso has quit IRC22:47
diverseI started uploading my ports, but I wanted to know how to ignore any tar files, because I don't want to throw them away22:47
diverseor move them22:47
joacimhttps://crux.nu/Wiki/SettingUpAnHttpupRepo22:48
*** crash_ has joined #crux22:48
joacimyou probably want to ignore a possible work directory too22:48
joacimif you use git, work with the .gitignore file22:49
*** toriso has joined #crux22:52
korimy goal is to eliminate the need for the git pull rodeo23:04
*** hhhhhhhh has quit IRC23:20
*** joacim- has joined #crux23:25
*** joacim- has quit IRC23:25
rmullRomster: ping23:46
rmullHave some questions about the wxgtk port23:46
rmullthen again, maybe I don't - sorry23:54
cruxbot[xorg.git/3.1]: xorg-iceauth: updated to 1.0.723:55
cruxbot[xorg.git/3.1]: xorg-mkfontscale: updated to 1.1.223:55
*** b1n0 has joined #crux23:55
b1n0:) hi !23:56
diversehi got a question?23:59

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