IRC Logs for #crux Tuesday, 2012-09-25

*** joacim has quit IRC00:00
*** himynameisphil has joined #crux00:10
*** himynameisphil has quit IRC00:12
*** himynameisphil has joined #crux00:16
*** Rotwang has joined #crux00:41
*** tj_ has quit IRC01:05
*** lasso|qt has joined #crux01:10
*** Rotwang has quit IRC01:17
*** mike_k has joined #crux01:17
frinnstwee, yubikey01:19
frinnst hid-generic 0003:1050:0010.0004: input: USB HID v1.11 Keyboard [Yubico Yubico Yubikey II] on usb-0000:00:1d.1-2/input001:29
frinnstI fear i need to use pam if i want to auth with it on the local box01:30
*** tj_ has joined #crux01:39
*** pitillo_ has quit IRC02:21
*** pitillo has joined #crux02:21
niklaswedoes someone know who I can get the system information when i login (like ubuntu) ?02:43
frinnsthm, dont think i follow02:44
*** lasso|qt has quit IRC03:45
*** lasso|qt has joined #crux03:46
*** joacim has joined #crux03:56
Romsterniklaswe, Conky04:07
niklasweRomster: I mean when you login via ssh..04:11
niklasweRomster: like this ...
frinnstcheck /etc/motd04:30
frinnstor maybe a cronjob that outputs to it04:30
frinnstor even a script that runs with your shell04:30
*** zetsu has joined #crux04:34
*** zetsu has quit IRC04:40
Romsteryeah any of the ones frinnst mentioned.04:40
Romsteror a relay advanced shell prompt :D04:40
frinnstif a script outputs to motd you could also redirect it to /etc/issue to display the stuff before you authenticate04:41
cruxbot[opt-x86_64.git/2.7]: nvidia: update to 304.51.04:43
*** krue has joined #crux04:57
LukcRomster, you mean, zsh? :D05:11
* Lukc hides.05:11
Romsterwhat... you're still here ^_^05:14
LukcOf course I am. :D05:15
LukcI’m here h24.05:16
Romsteroh that's good just not 7 days a week then only 6 days worth.05:16
Romsterwork and stuff?05:17
LukcUniversity and parents. :D05:17
Romsteri'd be on more but work and stuff keeps me away from irc05:17
Romsterah those take alot of time.05:17
Romsterspeaking of time i have not looked over your stuff much yet.05:17
Romstermight get time this weekend.05:18
LukcThough I could skip some classes, I wouldn’t miss much…05:18
Romstersince i so far don't feel brain dead like last week yet05:18
LukcGood. :D05:18
LukcI currently have no web browser.05:19
LukcNor anything to watch a video from YT.05:19
LukcBesides, Google does not exist.05:19
ente_google does not exist?05:19
ente_like putting fingers in your ears and shouting LALALALALALA05:19
Romsterapparently web browsers don't05:20
ente_yget + mplayer05:20
Romsteror youtube downloader05:21
ente_yget is 200 lines awk or so05:21
ente_youtube-dl seemed weird05:21
LukcYeah, that or “prepare myself for a stupid English class”.05:21
LukcOr “eat”.05:21
LukcI think I’m gonna eat.05:21
ente_this counts as preparation for english class!05:21
LukcI’m gonna eat anyway.05:22
ente_and I eat too05:22
ente_while typing05:22
LukcMe too, but I only have an apple right now.05:22
LukcAnd PAAASTAAA won’t be ready in time.05:23
Lukc(what’s the word already?)05:23
ente_for pasta?05:23
ente_maybe pasta? :)05:23
Romsterfor pasta05:24
LukcOh, that’s good. I’m even better than I believed.05:24
LukcWhat’s the point of an english class, then?05:24
Romsteryou using unicode or something it's really irritating on your '05:24
Romster’ = '05:24
LukcYes, I’m using unicode.05:25
LukcTo have a perfect typography.05:26
Romsterargh so i need to fix xchat to nut suck05:26
Lukc(at least I hope I’m using unicode :o)05:26
Lukc(but, I don’t know what those characters could be if not unicode…)05:26
Romster even breaks on the log05:26
Romstermaybe it's just the font i'm using....05:27
LukcSomehow I doubt that.05:28
Romsterhmm lets see now.05:29
ente_why isn't there a second codepoint for ,05:30
Romsterhmm ' ' '05:30
ente_Romster: it's valid unicode, I see it correctly, though not what you sent.05:30
Romsteri'm on utf-8 and seems ok05:30
Romster/charset utf-805:30
Romsteror should i be on 1605:30
ente_my typography is so broken in so many ways, I don't even have to bother with unicode ♥05:31
Romsteror is that for like languages like Japanese?05:31
ente_no, utf-805:31
ente_utf-16 is for windows05:31
Romsterso no 8 16 bit thing like 2 byte wide05:31
ente_8 is multibyte, dynamic width (up to four bytes per glyph afaik), 16 is fixed-width05:32
ente_i.e. always 16 bit05:32
Romsterso 16384 glyph than 25505:32
ente_i.e. not compatible with ascii afaik05:32
Romsteroh these numbers make no sense05:32
Romsterok so uuft-8 should of been utf-8-3205:33
ente_no, wait, this can't be. if unicode takes up to four bytes, every glyph fits into one 32-bit unsigned int. (which is utf-32)05:34
ente_dunno what the rules for utf-16 are. options: it supports only BMP (base multilingual plane) or you can somehow tell it your character is 32-bits wide (I didn't bother to look)05:34
ente_by flipping the highest bit or something like that05:35
ente_utf-8 is mostly just that: 8 bits for anything that's already in ASCII05:35
Romsterthese standards are never simple are they05:36
ente_anything else can be stuffed into up to 32 bits05:36
ente_nope :-)05:36
ente_try implementing utf-805:36
ente_try looping backwards over a string05:36
Romsteris it even possible.05:36
Romsteri know why tilman called it wtf-805:36
ente_not sure, but a possible solution is converting to utf-3205:37
ente_(and using that internally)05:37
Romstermakes sense.05:37
ente_actually, utf-8 is pretty well-designed compared with the other encodings05:37
Romsterperhaps, never bothered to learn how it functioned.05:39
ente_well, at least it replaces n+1 encodings by one05:39
ente_and usually deciding in which encoding text is encoded isn't trivial05:40
ente_e.g. for latin1 you have all of ASCII and some specials like ä, ö, ß05:41
ente_now for instance an utf-8 text using only these characters will also be 1 byte per glyph05:42
ente_because in utf-8, ä, ö, ß have the just the highest bit set (ascii being 7-bit)05:43
ente_(similar to how latin1 extends ascii)05:43
ente_(and yes, encoding is a mess)05:43
ente_at least the situation in germany is fairly ok, with text being usually either UTF-8 or ISO-8859-105:44
ente_but e.g. in russia, there are two encodings for cyrillic characters05:44
ente_actually three, KOI8-R, CP1251 (CP meaning Windows Codepage), which is similar to ISO-8859-5 which according to wikipedia never really caught on05:46
ente_and all of these are standardised since the 70s, when nobody could even imagine how we use them today05:47
ente_The 8-bit encodings KOI8-R and KOI8-U, CP866, and also Windows-1251 are far more commonly used. Another possible way to represent Cyrillic is Unicode.05:47
ente_and maybe not exactly the 70s05:49
ente_more like the 90s, except KOI8-R, which was already in use for teletype devices before I was born05:49
frinnsthaha wtf06:11
frinnstnice link romster :>06:11
Romsteri couldn't stop laughing... work safe06:16
Romsteryeah that's why utf-8 was invented to standardize06:17
*** pitillo has quit IRC07:01
*** pitillo has joined #crux07:01
LukcThe guy told me I should work on my pronunciation.08:04
LukcLike I couldn’t hear myself.08:05
LukcAnd after that I almost died of boredom.08:05
Romsterwhat guy?08:05
LukcThe teacher.08:06
LukcMore or less teacher.08:06
Romsterenglish teacher?08:06
Romstereh there always picky08:06
Romsteror is it their08:06
Romsteror they are...08:07
Romsteri hate english08:07
LukcI think it's "they are".08:07
LukcOr, btw, I'm gonna try to solve my UTF8 problem.08:07
*** Lukc has quit IRC08:07
*** Lukc has joined #crux08:08
LukcBetter? ™ é ế ǒ …08:08
Romstertry it's again08:08
Romster\o/ sweet jesus that's better08:09
Romsterand all them others work fine08:09
Romster™ é ế ǒ08:09
LukcCool. :)08:09
LukcAnd “…”.08:09
Lukc*Never* forget to use “…” instead of “...”. :D08:10
Romsteryeah i forget that is there08:10
Romsterno easy shortcut that i bothered to learn08:10
LukcRomster, is there no Dvorak-like for your language?08:11
LukcI mean, I’m French and I use Bépo, and there’s an easy shortcut for “…”.08:12
LukcIn fact, there are easy shortcuts for any character that should be used to have a correct typography.08:12
LukcIncluding the ones you never know when to use.08:13
LukcSuch as — − - –.08:13
Romsteroh dash and space and what the ever the rest are...08:15
Romsterso how do i type thsoe on the meta key anyways08:16
*** mike_k_ has joined #crux08:16
Romsteri have a standard us keyboard.08:18
LukcI never had a standard US keyboard, so I couldn’t tell.08:18
*** mike_k has quit IRC08:19
LukcBut you should try to use Dvorak, you know… :308:19
Romsteryeah problem is every keyboard at work is qwerty08:22
Romstereven every notebook08:22
LukcThe physical layout does not matter.08:22
Romsteri get confused enough as it is with being dyslexic than to try and learn another layout08:22
Romsteryou sugest i change every keyboard setting to devork on every machne i work on?08:23
Romsterall the customers machines, i'll pass.08:23
LukcRomster, usually you can just change the layout on the machine you’re working on when you begin to work on it, and then put it back to whatever layout it was using before.08:23
Romsterand if i forget too? -_- customer wtf can't type...08:24
Romstermaybe i could try devork again at home but then it'll screw me up at work.08:25
LukcA lot of people use dvorak or bépo at home and something else at work.08:25
LukcMost of them ain’t complaining. O:)08:26
LukcBtw Romster, where are you from? US?08:27
Romsteryou have no idea at the skill level of thses customers the majority are, lets say technically challenged. and cn't even touch type whats changing the layout going to achieve when they don't even know the home keys?08:27
LukcI guess english is your native language, then. :o08:28
LukcThe discussion of earlier about english teachers now just seems so strange. /o\08:29
Romsteryes english au08:29
*** horrorStruck has joined #crux08:49
frinnstergo not english:
acruxfrinnst: were you using gcc + LTO ?10:55
acruxdamn i see that with -flto my binaries are bigger of 7x10:56
acruxinstead graphite on multicore system has good benefits10:57
Lukc7×. Ouch.10:59
*** lasso|qt has quit IRC11:02
acruxdoes maybe lto works fine only with gold?11:04
horrorStruckRomster: we talked about this few days ago, according to the bufferbloat guys, the txqueuelen trick is not needed anymore if you use codel or fq_codel11:04
horrorStrucklto works fine with gold, i use it as default and only a few ports do break11:04
frinnstacrux: im using the 2.8 toolchain11:05
acruxwell, 2.8 it isn't built with LTO support. Is it right?11:07
acruxmmh, perhaps was Romster the gcc+LTO user11:09
horrorStruckroot@blackout:~# du -sh /boot/vmlinuz-lto11:11
horrorStruck3.6M    /boot/vmlinuz-lto11:11
horrorStruckroot@blackout:~# du -sh /boot/vmlinuz-git11:11
horrorStruck3.5M    /boot/vmlinuz-git11:11
horrorStrucknot yet 7x11:12
acruxwell, your one is good enough11:13
horrorStruckacrux: with which port did you see that much difference?11:16
horrorStruck(so i can try on my side)11:17
acruxit was just a statistic number11:17
acruxanyway i'd like to see with a library port11:17
acruxlike libmpc11:17
acruxcould you build it as test?11:18
horrorStruckbuilding already :)11:18
LukcThat hurts.11:20
acruxmaybe lto must be used carefully11:20
horrorStrucki still have free space :P11:21
horrorStruckFWIW 1st one is lto+graphite+gold, second is graphite+bfd11:26
horrorStruckusing linux binutils, not gnu11:26
*** Rotwang has joined #crux11:27
LukcLinux binutils, not GNU. Eh? :o There are non-GNU binutils?11:29
horrorStruckdont ask about the difference, it's just because it was a requirement to build a lto kernel and it stayed on my machine11:30
acruxwith bfd you say the old ld11:30
horrorStruckgnu ld11:31
acruxthese binutils are simply a devel/trunk version11:31
LukcBut why do they even exist?11:31
horrorStruckin fact i think they're some guy's branch11:31
acruxi guess they have in that place to test some regression11:32
LukcSo, they’re GNU binutils.11:33
LukcWith some modifications.11:33
horrorStruckthey're different from gnu anyhow as lto kernels won't build with gnu binutils and do build with these versions11:33
acruxah, ok11:33
horrorStruckthis ---> H.J. Lu maintains a version of binutils with bleeding-edge features for Linux.11:34
acruxtoo bleeding edge also for me :D11:37
horrorStruckworks great, i use it on 3 machines, not issue at all so far11:38
acruxi still use bfd (but with graphite/cloog/ppl)11:39
acruxand i'm testing lto11:39
horrorStrucki use latest 2.22 snapshot ( as firefox would build with 2.2311:39
horrorStrucki use graphite but never benchmarked it, did you?11:40
acruxjust now on a 4core (arm) and i saw11:41
acruxa better cpu time usage11:41
horrorStruckok good to know. was the difference significant?11:43
acruxabout 25% but i guess it also depends from how the code is optimized in loops etc..11:44
acruxit's not so "magic"11:44
acruxthese are my cflags for a 4core cpu11:45
acrux-ftree-loop-linear -ftree-loop-distribution -fpeel-loops -floop-interchange -floop-strip-mine -floop-block -fgraphite-identity -floop-parallelize-all -ftree-parallelize-loops=411:45
horrorStruck25% is huge11:45
acruxwhen it optimizes is good otherwise nothing change11:52
acruxthus for me is good to have it11:52
*** joe9 has joined #crux11:52
horrorStruckyou have some flags i wasn't using, rebuilding hard now :P11:56
*** joe9 has quit IRC12:02
frinnst lol12:09
frinnstthe oss-security mailinglist IS FUCKING DEPRESSING12:10
frinnst~20-25 mails per day in my inbox12:11
frinnstjust people requesting CVE assignments12:11
frinnstjust the amount of exploits/issues is what's depressing me12:11
jseAnother fine example of first world problems.12:19
*** horrorSt1uck has joined #crux12:24
*** horrorStruck has quit IRC12:27
*** pitillo_ has joined #crux12:28
*** pitillo has quit IRC12:30
*** pitillo_ has quit IRC12:30
*** horrorSt1uck is now known as horrorStruck12:31
*** pitillo has joined #crux12:34
*** pitillo_ has joined #crux12:39
*** pitillo has quit IRC12:39
*** pitillo_ has quit IRC12:45
horrorStruckfrinnst: just in case, gtk 2.24.13 is out :)12:55
horrorStruckand gdk-pixbuf 2.26.412:58
*** pitillo has joined #crux13:06
frinnstah cheers13:06
frinnstforgot about those13:06
*** joacim has quit IRC13:07
cruxbot[opt.git/2.7]: gdk-pixbuf: updated to 2.26.413:08
cruxbot[opt.git/2.7]: gtk: updated to 2.24.1313:08
horrorStruckthat was fast, thanks13:09
*** pitillo_ has joined #crux13:11
*** pitillo has quit IRC13:12
*** Rotwang has quit IRC13:12
*** Rotwang has joined #crux13:13
*** joe9 has joined #crux14:40
rauzdamit it jsf is such a shit ...14:41
*** lasso has joined #crux15:01
Rotwangthe last page from learn python the hard way is depressing15:04
frinnstnot that bad. and it's true for every profession probably15:13
frinnstie dont make your hobby into a career15:13
frinnstpersonally i think its a bunch of bs15:13
*** joe9 has quit IRC15:24
*** mike_k_ has quit IRC15:34
*** lasso has quit IRC15:35
Rotwangby the way, in the end I've used cluster ssh15:48
Rotwangas my budy described it: cfengine for poor15:49
*** vaddi has quit IRC16:07
*** vaddi has joined #crux16:18
*** sammi`_ has quit IRC16:19
*** vaddi has quit IRC16:35
*** sammi` has joined #crux16:39
*** Rotwang has quit IRC16:48
*** joacim has joined #crux16:50
*** s44 has joined #crux17:14
*** spider44 has quit IRC17:17
*** pidsley has joined #crux17:19
*** pidsley has quit IRC17:24
*** poulecaca has quit IRC18:30
*** poulecaca has joined #crux18:37
*** pidsley has joined #crux18:51
*** tilman has quit IRC19:38
*** tilman has joined #crux19:39
*** ChanServ sets mode: +o tilman19:39
*** horrorStruck has quit IRC20:16
*** pidsley has quit IRC20:28
*** horrorStruck has joined #crux20:28
*** horrorStruck has quit IRC20:33
*** Lukc has quit IRC20:52
*** Lukc has joined #crux20:52
*** pidsley has joined #crux21:14
*** pidsley has quit IRC21:16
Romsteracrux, i was using LTO for awhile on gcc 4.5.x series few ports broke but the majority worked fine21:30
Romsternever tried LTO on kernel that's relatively recent.21:30
*** mavrick61 has quit IRC21:37
*** mavrick61 has joined #crux21:38
Romstertelnet 1721:44
RomsterQuote Of The Day21:44
*** sh4rm4 has quit IRC23:07
*** nthwyatt has quit IRC23:53
*** nthwyatt has joined #crux23:53

Generated by 2.11.0 by Marius Gedminas - find it at!