IRC Logs for #crux Saturday, 2007-09-08

* predatorfreak returns from idle.00:04
predatorfreakn0yd: pkgmk doesn't support MAKEOPTS.00:04
predatorfreakn0yd: The closest you'll get is MAKEFLAGS.00:05
predatorfreakn0yd: Which some apps will support and others won't.00:05
n0ydmakeflags arent make opts00:05
predatorfreakYeah, I know.00:05
n0ydthey are nowhere near the same00:05
predatorfreakn0yd: MAKEFLAGS is obeyed by most configure scripts.00:06
predatorfreakand generally serves to start make with said options.00:06
n0ydThats pretty lame that I can't tell make to build with -jX00:06
n0ydmake will always accept that00:06
n0ydregardless of what its building00:06
predatorfreakYou mean MAKEFLAGS?00:06
n0ydSo then trying to get the compiler to take advantage of a smp machine is impossible00:07
n0ydpredatorfreak: -jX is considered a makeopt00:07
n0ydie, -j2, -j300:07
predatorfreakn0yd: It's a bloody option passed to the program.00:07
predatorfreakJust like any other.00:07
n0ydAlso, the vsftpd daemon script seems broke, the daemon will start but the script hangs00:08
predatorfreakn0yd: If make obeys MAKEFLAGS or the configure script obeys MAKEFLAGS, it doesn't really matter, as you start make to spawn children.00:08
predatorfreakSo long as the option is passed to the program it's good enough.00:09
predatorfreakn0yd: I couldn't help you with vsftpd.00:10
predatorfreakI've never used it.00:10
n0ydI'm looking at the script to see whats messed up00:10
predatorfreakI just work with ssh myself.00:10
predatorfreakn0yd: set -x might be more useful :P00:10
n0ydit starts fine and I can connect, the actual program itself is fine, but the script hangs00:11
predatorfreakThat's why set -x is useful.00:12
predatorfreakIf you stick that in the script.00:12
predatorfreakIt'll tell you what its hanging at.00:12
predatorfreakn0yd: Also, keep in mind CRUX isn't Gentoo, it's based around simplicity and most of the port maintainers don't have time to build packages 200000 times with differing -jX opts to see if make dies.00:14
n0ydWhere in the script should I put it.00:14
predatorfreakThe top.00:15
n0ydpredatorfreak: it would never die00:15
predatorfreakRight under #!/bin/blah00:15
n0yd-jX isn't a cflags00:15
predatorfreakn0yd: It does.00:15
predatorfreakn0yd: You've never encountered a badly written Makefile have you?00:15
predatorfreakIf the makefile is written wrong, things can be built when a depend file hasn't been built.00:16
predatorfreakLeaving you with nasty build errors.00:16
predatorfreakMakefile's get damn complex and if they're made wrong, the apps won't build right (or at all), leaving you with make dying in the build.00:17
n0ydAll -x gives me, is command not found errors.00:18
n0ydpredatorfreak: I understand that, but I still find it pretty damn silly.00:18
n0yd-jX is basic option for make00:19
*** jaeger has quit IRC00:19
predatorfreakn0yd: Pastebin the errors.00:19
predatorfreakn0yd: Read through a makefile for an app like say, pidgin.00:20
predatorfreakThey're huge.00:20
predatorfreakand developers obviously don't have the time to build stuff a billion times.00:20
n0ydpredatorfreak: I've seen plenty of makefiles, but not once have I ever had make fail because of -jX00:20
predatorfreakn0yd: Build libpth with make -j2 or whatever, then try to install it with make -j2.00:20
predatorfreakIt breaks.00:21
n0ydpredatorfreak: What errors? lmao. /etc/rc.d/vsftpd: line 2: -x: command not found00:21
predatorfreakIt's one command.00:21
predatorfreakset -x00:21
n0ydYou don't make install with -jX, thats pointless00:21
predatorfreakNo linebreaks in between.00:21
predatorfreakn0yd: MAKEOPTS, as used in Gentoo or whatever.00:21
predatorfreakBasically = alias make="make $MAKEOPTS"00:22
n0ydI have little experience with gentoo, so I have no idea.00:22
predatorfreakn0yd: Well that's where MAKEOPTS stems from.00:22
predatorfreakBesides, grep through the portage tree someday for "-j1".00:23
predatorfreakThey force it quite a bit.00:23
predatorfreakAnyway, the script should be00:24
predatorfreakwell, the top of it anyway.00:24
predatorfreakset -x00:24
predatorfreakThen when it's run it oughtta tell you everything it's doing.00:24
*** brian_ has quit IRC00:25
n0ydI fixed the script 5 minutes, ago.  As far as -jX goes, calm down.  Don't get your panties in a wad defending ol' crux. ;)00:26
predatorfreakn0yd: You didn't seem to indicate that you fixed the script.00:27
predatorfreakand I'm just telling you why and alternatives.00:27
n0ydI didn't feel a need?00:27
predatorfreakn0yd: Well considering I was talking about it, it would have been helpful to inform me that you'd fix it a tad be earlier.00:28
predatorfreakThe last bit scares me.00:31
predatorfreak"Let's make it so prior art means jackshit in patents!"00:31
predatorfreakThat was the only REDEEMING factor of the U.S. patent system.00:31
predatorfreakIf you could prove someone else made it first, the patent is invalid.00:31
predatorfreakFor once I'm hoping Bush will veto a damn bill...00:32
*** predatorfreak has quit IRC01:13
*** n0yd has left #crux01:28
*** raw has left #crux01:45
*** zippyy has joined #crux02:01
*** zippyy has quit IRC02:01
*** Romster has quit IRC02:05
*** zippyy has joined #crux02:28
zippyyanyone used/using the core L distro ?03:00
*** pell has quit IRC03:01
*** pell has joined #crux03:01
*** zippyy has quit IRC03:14
*** zippyy has joined #crux03:17
*** mike_k has joined #crux03:21
* zippyy mmm some new ddr3 Ram....03:28
nipuLspeaking of ddr3, building a $8k pc next week03:42
*** treach has joined #crux03:52
*** Romster has joined #crux03:52
mike_knipuL: it is as easy as $2k pc plus budget plasma/projector =)04:03
mike_kwhat's the most pricy part?04:04
*** zippyy has quit IRC04:08
*** zippyy has joined #crux04:09
*** morlenxus has quit IRC04:10
*** morlenxus has joined #crux04:10
treachzippyy zapppped. :>04:10
*** morlenxus has quit IRC04:19
*** morlenxus has joined #crux04:21
*** RedShift has joined #crux04:36
*** patroclo7 has joined #crux04:46
*** Viper_ has joined #crux04:53
nipuLthis is without a monitor05:54
*** RedShift has quit IRC06:13
*** Viper_ has quit IRC06:34
rxinipuL: i did a server a few months with that amount in ram06:52
*** RedShift has joined #crux06:52
*** RedShift has quit IRC07:13
*** RedShift has joined #crux07:15
*** Deton8R has joined #crux07:21
*** Deton8R has quit IRC07:41
*** Viper_ has joined #crux08:03
*** lasso has joined #crux08:11
*** jjpk has joined #crux08:23
*** mxq has joined #crux08:48
*** zippyy has joined #crux08:50
*** zippyy has quit IRC09:02
*** zippyy has joined #crux09:06
*** zippyy has quit IRC09:52
*** zippyy has joined #crux09:52
*** jaeger has joined #crux10:01
*** ChanServ sets mode: +o jaeger10:01
*** zippyy has quit IRC10:03
*** zippyy has joined #crux10:03
*** zippyy has quit IRC10:23
jjpkzip zap.10:24
*** mxq has left #crux10:45
*** aon has joined #crux10:51
*** ChanServ sets mode: +o aon10:51
*** Viper_ has quit IRC11:02
*** RedShift- has joined #crux12:36
*** RedShift- has quit IRC12:38
*** RedShift- has joined #crux12:38
*** RedShift has quit IRC12:51
*** RedShift- is now known as RedShift12:51
*** n0yd has joined #crux12:57
n0ydIs there a trick to getting bash-completion working in crux? I installed the pkg, added the usual bash-completion stuff to ~/.bashrc, but it's a no go.12:57
tilmanit worked for me like that13:04
tilmanyou could try to manually run13:04
tilmansource /usr/share/bash-completion/init_bc13:04
RedShiftanyone play urban terror here?13:05
tilman5 years ago or so, yeah13:05
jaegerI play Crackdown, which consists of a lot of urban terror :)13:06
n0ydtilman: are you using this in your ~/.bashrc?
n0ydOr is that not the correct method for crux?13:07
tilmani'm using:13:07
tilman. /usr/share/bash-completion/init_bc13:07
tilmanjust like i pasted above :D13:07
tilmann0yd: see /usr/ports/opt/bash-completion/README13:07
tilmanit's even documented...13:07
n0ydtilman: Ahh, well what I pastebin'd is what it normally uses in every other distro I've ever used. :)13:08
n0ydLooks I found my first "cruxism" I guess.  I thought the "ism"'s only existed in Redhat. :-P13:12
n0ydHmm, I put that line in my .bashrc, still no go.  Lemme try sourcing it13:13
n0ydHmm, sourcing it worked. Weird13:14
n0ydI logged off and back on, could it be cause I still have another ssh session open with the same user? (This box is 1500 miles away)13:15
mike_ktilman: just run sysup and tried to rebuilt all xorg-xf86-input-* and xorg-xf86-video-* (just in case) after that. input-evdev failed:
tilmanmike_k: ugh, i need to update the port13:17
mike_kI guess, at least the old one should work for now13:17
n0ydargh, I hate mirrors that won't resolve/connect.  Having to edit the pkgfile for each one gets tediously annoying.13:21
tilmanmike_k: fixed, rsync should be updated in a few minutes13:22
n0ydtreach: yup13:22
mike_ktilman: thanks13:23
n0ydAny idea why I keep getting these empty "[033[0m]" files in every user home dir? I rm them and they reappear every time...13:23
n0ydtreach: thx13:23
tilmanfucked up bashrc/bash_login?13:24
treachsome colorcode, no doubt13:24
n0ydYea, I can't track the bastard down. :-P13:25
n0ydwoops, just found it in /etc/profile, I must've been really tired last night.  But the line is commented out.13:26
n0ydFixed that up, I had really messed up /etc/profile. I have no idea what I was doing last night. heh13:29
treachwhat you did is evident; what you were thinking of isn't.13:29
n0ydtis true. ;)13:31
n0ydick, is so slooow13:38
mike_ktilman: will you send that synaptics patch to the upstream? The latest git commit in their tree is 2007-07-06, so they update it from time to time without bumping tarball13:48
*** predatorfreak has joined #crux13:48
tilmanmike_k: i submitted the missing fix13:49
n0ydtilman: the line you are using in bashrc is _just_ ". /usr/share/bash-completion/init_bc"?13:52
n0ydThat's odd, I can't get it working, unless I source it manually.13:53
treachget a better shell, like zsh. ;)13:54
tilmanyour bashrc is probably not sourced13:54
tilmaneasy to confirm wth echo or so13:54
* mike_k hopes a new intel driver will be more stable and reboots13:55
*** mike_k has quit IRC13:55
n0ydtilman: yup, thats the problem.  ~/.bashrc isn't sourced by default? What should I add to /etc/profile to get it source it?13:58
tilmani just ln -s .bashrc .bash_login13:58
n0ydoh ok, easy solution13:59
predatorfreakWhat the hell.14:02
predatorfreakpkgmk ALWAYS strips this package.14:02
predatorfreakEven when I tell it NOT to.14:02
tilmanmaybe it's not pkgmk, but the local Makefile14:02
predatorfreakDang evil developers.14:03
tilmanmaybe "STRIP= make ..." helps14:03
predatorfreakDon't break your code in the first place, but please don't freaking strip stuff for me.14:03
predatorfreakThe dang developer just wrote it with LDFLAGS = -s14:04
predatorfreakand install -s <_<14:04
jjpkLess bitching, more fixing.14:05
predatorfreakI forgot, how do you pass args to a program in gdb?14:06
tilmanr arg1 arg2 arg314:06
predatorfreaktilman: Thanks14:06
predatorfreak#2  0x0804a0dc in main (argc=1396586044, argv=0x20207065) at klogd.c:117114:07
predatorfreakYay! I know where it's dying now.14:07
predatorfreakSo I take it that's going fucky.14:08
tilmandoes it even *enter* InitMsyms()?14:09
tilmanargc is bogus btw14:09
tilmani'd rebuild with CFLAGS being -O0 -g14:09
tilmanand maybe try to valgrind it even14:09
predatorfreakThen it gets to doing fclose(ksyms)14:10
predatorfreakand dies at that.14:10
tilmanyeah, valgrind14:10
predatorfreakBlarg, here goes doom.14:11
predatorfreakI haven't used valgrind in so damn long.14:11
n0ydwth, I just installed xorg.  It said everything installed sucessfully.  But I have no /etc/X11/?14:15
predatorfreaktilman: valgrind needs a patch for glibc 2.6, btw :)14:16
predatorfreakOtherwise it refuses to build.14:16
predatorfreakI see EXACTLY why it's dying.14:18
predatorfreakBuffer overrun.14:18
tilmanheh, where?14:18
treachn0yd: no big deal, create it yourself, or use /usr/etc/X1114:18
predatorfreakand I've got nx on.14:18
predatorfreakSo it's killing it.14:18
* predatorfreak pastebins14:18
tilmann0yd: there's xorgconfig14:18
tilmanor Xorg -config or whatever14:19
n0ydI mean... I have no /etc/X11 dir14:19
n0ydat all14:19
treachsee ^14:19
n0ydAh, another cruxism ;)14:19
predatorfreakoh no wait, I'm close but wrong, sorry, I haven't read valgrind output in so long.14:19
predatorfreakIt's accessing memory that doesn't exist.14:19
tilmann0yd: the xorg-server tarball doesn't instlal /etc/X1114:19
tilmannewer x doesn't even *need* xorg.conf14:20
n0ydi know14:20
tilmanthat's all?14:21
n0ydxorg guys need to start following the LSB then.  not installing to /etc/X11/ is quite different, to say the least.14:21
predatorfreakIt seems to be reading memory which doesn't technically exist yet.14:21
predatorfreaktilman: I can pastebin more.14:21
tilmanpastebin everything14:21
tilmanplease :P14:21
tilmanoh the fclose error is actually the first14:22
tilmani'm currently copying large amounts of warez from an encrypted partition to an encrypted partition and i'm building xorg-server14:22
tilmanmy system is totally *not* responsive14:22
predatorfreaktilman: You dirty pirate.14:23
tilmanpredatorfreak: but i'll try it a few minutes14:23
tilmanload average: 4.8414:23
predatorfreakIf you have a CPU capable of it.14:23
treachtilman: I thought you had a pretty recent system?14:23
predatorfreakTry it with a kernel built with 64GB highmem.14:23
tilmanusing a software cursor in x doesn't help the responsiveness14:23
predatorfreakWhich enables PAE and NX-bit.14:23
tilmantreach: writing to a luks fs totally kills performance14:23
predatorfreakI THINK that might be in part why it's dying, because I've got the NX bit enabled.14:23
tilmanthe destination is aes with 128 bit key length i think14:24
tilmansource is aes with > 1024 bits (not my decision)14:24
tilmansource is also an usb disk ;)14:24
treachwatch out for the bundestrojaner. ;p14:24
tilman"Niemand hat die Absicht, einen Ueberwachungsstaat einzurichten."14:24
treachah, of course.14:25
treachabsolutely not. :|14:25
predatorfreaktilman: Figuring that NX support is causing it, that'd narrow it down as to why it's happening on my system only so far ;)14:25
tilmanpredatorfreak: it's trying to fclose(NULL)14:25
tilmanin line 19214:25
tilmannot surprising it's dying14:25
predatorfreakin klogd.c?14:25
tilmanif (ksyms == NULL)14:25
tilman  // blabla14:25
tilman  fclose (ksyms);14:26
predatorfreakerr not klogd.c14:26
predatorfreakksyms something14:26
tilmanoops, sorry14:26
tilmando you even have /proc/kallsyms?14:26
n0ydIs something wrong with the nvidia driver in the ports repository? I'm getting this error,
predatorfreakDisabled it in my kernel to save space :)14:27
predatorfreakThis would ALSO explain why it only hits my system.14:27
tilmanyou suck14:27
n0ydAlthough Xorg -configure is still acting like it works.14:27
tilmanpredatorfreak: i'll submit a patch for the fclose14:27
predatorfreaktilman: In the mean time.14:28
predatorfreakWhat would a fix be?14:28
tilmann0yd: nvidia is a 2nd class citizen in my evil little xorg world14:28
tilmanie i updated xorg-server to 1.4 (video driver abi 2.0) even though no fitting nvidia driver is released yet14:28
predatorfreaktilman: and I kept a 7.2 directory handy :)14:29
tilmani admit that this is hard to see for anyone who just installed the whole thing14:29
predatorfreakI know your evil ways!14:29
* treach cheers on tilman 14:29
n0ydxorg-xf86-video-nv should still work, right?14:29
predatorfreakn0yd: Yar.14:30
tilmanthe fix is to *not* try to fclose(NULL)14:30
tilmanif fopen returns null, there's nothing to close14:30
predatorfreaktilman: Well, then my quick-hack of commenting it out is probably proper :)14:30
treach..after all, you don't get to enjoy 3d over ssh anyway..14:30
tilmanpredatorfreak: it is14:30
* treach tries to imagine vim in 3d14:31
n0ydGood enough, this box is remote, and the person's who has my box sitting in their closet 1500 miles away is legally blind.  I doubt he cares for 3d accel.14:31
predatorfreaktreach: Bah! Quake over dial-up is awesome!14:31
treachpredatorfreak: kind of a moot example..14:31
predatorfreaktreach: But it's a fun example.14:31
predatorfreakYou use VNC.14:31
predatorfreakRun quake on a remote box.14:31
treach..aaaaand completely irrelevant.14:32
predatorfreakand you've got 3D over dial-up :)14:32
predatorfreaktreach: I say things that are relevant?14:32
jjpkCute but completely besides the point.14:32
treachI suppose shit happens, even to you14:32
jjpk3d slideshow vs lagmania.14:33
predatorfreaktilman: I suggest you mail the mailing list about it and put a quick patch in the 1.5 directory.14:33
treachjjpk: I don't think it would even qualify as a slideshow. :>14:33
predatorfreakand yeah, I can confirm that commenting it out works.14:33
predatorfreaktreach: You'd be measuring it in FPH (Frames Per Hour)14:33
predatorfreakYou'd be lucky if you got 0.5FPH :P14:34
predatorfreaktilman: One more mystery down :P14:34
* predatorfreak keeps on using rsyslog anyway.14:35
predatorfreakBlarg I always fuck up typing shit.14:35
predatorfreakI just did /m keeps on using rsyslog anyway XD14:36
tilmanpredatorfreak: lesson learned: valgrind ftw \o/14:36
predatorfreakI need to just alias all my comment typing mistakes to the proper things.14:36
predatorfreaktilman: Indeed.14:36
tilmani won't add the patch to core/sysklogd14:36
predatorfreaktilman: Why not?14:37
tilmanit only hits people with insane kernel configs14:37
tilmanbut maybe i'll reconsider ;)14:37
tilmanpredatorfreak: which kernel option/config switch is that btw?14:37
predatorfreaktilman: Well, I'm not using sysklogd anymore anyway, but I'd still recommend it.14:37
jjpkAlready catering to special interests? Not much lobbying was required! :D14:38
predatorfreakYou've gotta enable embedded options.14:38
tilmanmaybe they'll release 1.5.1 soon14:38
predatorfreakIt's uhh.14:38
tilmanhang on14:38
predatorfreakGeneral setup -> Configure standard kernel features (for small systems) -> Load all symbols for debugging/ksymoops14:38
tilmanfine, i'll add it14:39
predatorfreaktilman: It's not all THAT hard of an option to stumble upon :P14:39
treachI guess that's why I had the same problem.14:39
tilmanclemens will probably hit it, too ;)14:39
treachI don't load a bunch of debug symbols either.14:40
predatorfreaktilman: In addition.14:40
tilmantreach: note that this is the immediate-segfault-problem, not the it-takes-20-minutes-to-start-problem14:40
predatorfreakI still think there's some bug about it not realising the process died.14:40
predatorfreakSame thing.14:40
predatorfreakWhat happens with it-takes-20-minutes-to-start.14:40
predatorfreakIs that the thing forks.14:40
predatorfreakThe child dies.14:40
predatorfreakand the parent doesn't know it.14:40
tilmanthat should be SIGCHLD14:41
tilmani think14:41
tilmanpredatorfreak: wanna figure it out?14:41
predatorfreaktilman: Sure :)14:41
predatorfreakThe thing is.14:41
predatorfreakvalgrind keeps it from backgrounding.14:41
predatorfreakSo I can't figure out what's happening to the parent/child :\14:41
tilmanbtw, does uncommenting the fclose make it work *properly*?14:41
tilmanor do the missing kallsyms add other problems?14:42
predatorfreaktilman: You mean commenting it out?14:42
* tilman thinks14:42
tilmanpredatorfreak: err, yes14:42
predatorfreakYeah, it works fine after commenting it out.14:42
predatorfreakStandard klogd behaviour.14:42
tilmanklogd.c doesn't even check the return value of InitMsyms14:43
predatorfreakThat might also explain it.14:43
tilmanwell, no14:43
n0ydApparently the driver will still work with the new ABI14:43
tilmanfclose(NULL) *will* segfault (unless cosmic rays interfere or so)14:43
predatorfreaktilman: Give me a break, I'm not a C developer by profession :P14:44
n0ydif you use -ignoreABI and don't use the composite extension14:44
predatorfreaktilman: The thing is,14:44
tilmanfclose will try to dereference the FILE pointer it got14:44
predatorfreakI think it forks BEFORE that.14:44
tilmandereferencing NULL -> sigsegv on linux14:44
tilmann0yd: that's what pedja described on the ml iirc14:44
predatorfreakand then that goes on in the child and the child sigsegv's.14:44
tilmanbut it didn't work for jaeger for some reason14:44
predatorfreakn0yd: Semi-worked for me.14:45
tilmanpredatorfreak: find out whether the parent receives SIGCHLD14:45
tilmani think you can get away with calling puts in the signal handler14:45
predatorfreakBut I could still make the thing die by using glxgears :)14:45
tilmaneg puts("olol got signal");14:45
predatorfreaktilman: Where is this at now?14:45
tilmanin the signal handler for SIGCHLD14:45
tilmanwhich is reapchild() according to grep14:46
predatorfreakThat's for syslogd though :\14:46
tilmanwell, if the fclose thing fixes everything, you're good anyway14:46
predatorfreaktilman: Yeah, but there's still another bug there worth fixing ;)14:47
predatorfreakand I'd say klogd just needs to check the dang childs return code.14:47
predatorfreakwell, signals.14:47
n0ydBut the only reason of install X was so if my legally blind friend needed to admin the box for me he could.  And it kinda requires him to use Compiz, so he can use the input-zoom plugin, so he can actually see anything.14:47
predatorfreakBecause it doesn't look like it's doing that at all.14:47
tilmanpredatorfreak: okay, copy from syslogd.c14:48
predatorfreaktilman: Blarg, exactly how?14:49
tilmanthis would be easier if i did it instead of you14:49
predatorfreaktilman: Yes :P14:51
tilmanpredatorfreak: mind if i report to crux?14:51
tilmanas in the ml14:52
predatorfreaktilman: I'm just happy I'm helping to pinpoint bugs :)14:55
tilmanthis is a cool bug14:55
n0ydumm libpng is outdated, and not only that, the pkgfile is broke because they dont have the older version on the server anymore14:55
tilmann0yd: sec14:56
n0ydjust updating it to 1.2.20 works fine though14:56
*** mrks has joined #crux14:57
tilmann0yd: thanks, fixing now14:57
n0ydtilman: no prob14:57
n0ydtilman: oops, guess I need to fix the md5sum14:57
tilmanyou can use pkgmk's -im option14:58
n0ydim using prt-get to install a lump of pkgs14:58
n0ydtilman: want the correct md5?14:59
predatorfreakn0yd: Just use pkgmk -im14:59
predatorfreakor prt-get -im14:59
n0ydFixing it worked fine...14:59
predatorfreaktilman: You need more people to bump packages stupidly and hope nothing goes wrong :P15:00
n0ydI'd rather just fix the problem than bypass it.15:00
treachn0yd: if you "fix" problems that way, the "fix" dissapears the next time you update your port tree..15:01
n0ydtreach: But he is fixing it right now anyway15:01
tilmanpredatorfreak: i need to put 2.4-wip on the laptop tomorrow15:01
treachso in order to really "fix" it, you either have to notify the maintainer, or create your own port15:01
n0ydAnd i realize that, hence why I "fixed" it15:01
treachn0yd: true, I was talking about the general case.15:01
n0ydBecause when I do ports -u, its gonna come down being the same15:02
n0ydI understand how the ports system works, it's very much like ABS in arch.15:02
predatorfreakn0yd: Only, we're too lazy to make packages unless it's for a CD :P15:03
treachsupposedly that's the wrong way to put it, but ok. That wasn't quite the point though.15:03
predatorfreakand really, ABS is descended from the CRUX ports style of source-packaging.15:04
jjpkFlexibility and simplicity are at the heart of it all.15:04
n0ydpredatorfreak: exactly15:04
predatorfreakArch stole our idea :P15:04
n0ydpredatorfreak: Not that Judd Vinet doesn't admit to that. ;)15:04
n0ydHe said arch was inspired by crux15:04
treach"our"? I don't seem to remember you from the <1.0 days. :P15:04
predatorfreakn0yd: I have no problems with Judd.15:04
predatorfreaktreach: Whatever :P15:04
predatorfreakn0yd: But I do have problems with the OTHER Arch developers.15:05
n0ydpredatorfreak: So do I. I have problems with their whole developer ideology15:05
n0ydIt's bullshit15:05
tilmancan you give an example?15:06
tilmani'm not familiar with their idiology, but curious :D15:06
treachmmh, arch users defecting to crux, is that a trend..?15:06
predatorfreaktilman: One of the particular examples that I remember is when a friend of mine, who's been a longtime Arch user.15:06
n0ydThe whole nominating developers, but you'll never be nominated unless you suck the higher developers on the trees cocks, and be part of their clique.15:06
predatorfreakAsked them why their kernel wasn't built with Forced Module Unloading15:06
n0ydIt's crap15:07
n0ydAnd the way they handle AUR is stupid15:07
predatorfreakBecause sometimes he'd have modules get stuck.15:07
predatorfreakand he couldn't unload them normally.15:07
predatorfreakand they told him he was doing something wrong.15:07
n0ydAnd they won't let users fix packages which have been outdated for months, and the maintainers are MIA.15:07
jjpkIt's called democracy, haven't you noticed?15:07
predatorfreakand that they wouldn't just put forced module unloading in their kernel package.15:07
n0ydI can't explain it very well, but it's very broken, the way they work.15:08
treachwell, having developers that doesn't play nice with the other devs sounds like a bad idea.15:08
jjpkControl freaks with a strict sense of work-to-rule.15:09
jjpkThat will fly a long way.15:09
* treach shakes at the thought of having e.g han as a crux-dev..15:09
n0ydAnd then things that are obviously broken, or just plain messed up, they just turn a blind eye to, and say "things are fine the way they are", they have no desire to actually develop the distro to make things better, ie, enhance things, or even fix things that have been broke for ages.15:09
predatorfreaktreach: Welcome to #archlinux.15:09
*** mrks_ has quit IRC15:09
predatorfreakAlso known as "Our Developers are Han."15:10
tilmanthey are legion?15:10
treachpredatorfreak: heh, well, I've never tried it, and I have no desire to.15:10
predatorfreaktilman: Yes, but not like /b/ ;)15:10
predatorfreakn0yd: They also claim they're going to SOMEDAY clean their repos out.15:11
predatorfreakand get rid of the old, shitty packages in there.15:11
predatorfreakBut they never do.15:11
jjpkPublic, written plans and talk are very cheap. ;)15:11
n0ydThey fucking piss me off.  I remember one time, all the developers threw a shit fit because I made a channel called #archlinux, on the OFTC network, just to help spread the distro around a bit.  I even noted it wasn't the official channel in the topic.  But they still all cried to the OFTC  network staff, and the staff laughed at them.15:12
jjpkMight be astute illusion management.15:12
n0ydThey could use some reform, but it will never come.  Judd is quite lazy, to be honest.  His theory on everything is "Fuck documentation! STRIPALLDOCS!"15:12
treachcorporate mindset strikes again.15:12
jjpkzomg. copyright infringements!!!!15:12
treachn0yd: well, crux does the same in that regard, except for the manpages15:13
treachdocs are already on the web, so why have a local copy?15:13
n0ydtreach: But it's funny, he won't even write general docs for things that actually need them, that he created.15:13
treachsounds a bit inconvenient.15:14
jjpkOh right. "good source documents itself"15:14
n0ydjjpk: haha, exactly.15:14
predatorfreakn0yd: But the code is SO easy to understand when it's written by crazy people!15:14
treachisn't all code? ;)15:15
n0ydOh, and they still refuse to admit pacman gets retardly slow when the local package set is larger than 200 packages.15:15
n0ydAnd that they would be better off using a real database for its backend15:15
predatorfreaktreach: Of course :P15:15
predatorfreakn0yd: It depends on the FS.15:16
n0ydInstead of a bunch of 1kb files for each package, all sitting in one dir, and it gets fragmented.15:16
predatorfreakWith something like reiserfs I got it to be blisteringly fast.15:16
predatorfreakBut with say XFS it's slow as fuck.15:16
predatorfreakext3 with dir_index is slightly better.15:16
n0ydpredatorfreak: Yes and no.  But you'll notice it on every FS, the best way to take care of it is, make the partition for /var/lib/pacman about 50MB or so.15:17
predatorfreakHowever, yeah, an actual DB file would be nice.15:17
*** hp_tux has joined #crux15:17
treachn0yd: we actually had a little discussion on that topic a while a go, but there are problems with that approach as well.15:17
predatorfreakSo it doesn't fuck around thrashing through the filesystem.15:17
n0ydI recommended berkdb, they scoffed at that, and said "things are fine the way they are, why inovate? Inovation!?! Thats Crazy! Ban him! Ban him!"15:18
treachfor instance the entire syncing mechanism is based on files being copied around, and if you start relying on a db, the httpup concept probably goes out the window.15:18
treachwe were discussing around sqlite15:18
predatorfreakn0yd: Kind of like how they were pissed I knew cdparanoia didn't rip to FLAC ;)15:18
n0ydOk, aspelkl won't build...15:19
predatorfreaktreach: Nah, I think he means more like a DB as CRUX has now.15:19
n0ydNot sure why...15:19
predatorfreakJust something that's faster to parse than ripping through the FS.15:19
tilmann0yd: pastebin the full error message?15:19
treachmmh, ok.15:20
predatorfreakWith most modern filesystems it's easier to just load one big file.15:20
predatorfreakand parse through that in memory.15:20
predatorfreakThan thrash the disk looking for information on disc.15:20
tilmanre. the package db thing:15:20
tilmanour db format is *not* a bottle neck15:20
treachnot as such.15:20
predatorfreaktilman: I noticed :P15:21
treachbut having the stuff in a organized db would be more flexible.15:21
n0ydtilman: there really isn't a error, beside its failed to install. But here it is,
n0ydI have no idea why it's failing.15:21
predatorfreak./.libs/ undefined reference to `acommon::HashTable<acommon::StringMap::Parms>::erase(char const* const&)'15:22
predatorfreakThat maybe? :P15:22
predatorfreakAlso rebuild WITHOUT -O3.15:22
n0ydI totally overlooked that whole line15:22
predatorfreakThen tell us it's not working :P15:22
n0ydOk, although I've built aspell plenty of times with -0315:23
predatorfreakWell, CRUX doesn't do any fancy filtering like other distros.15:23
predatorfreakor any fancy patching for stuff like that.15:23
tilmanit's unlikely that's it an optimization-induced problem, but it's worth a try15:23
n0ydpredatorfreak: Neither does LFS, or T2.15:23
predatorfreakor any anything about CFLAGS really.15:23
n0ydpredatorfreak: The only source distros i use are LFS and T2, both of which are from scratch, no filtering involved.15:24
predatorfreakMight just be G++ fucking up on you.15:24
n0ydI'm not a Gentoo type of guy.15:24
predatorfreakWouldn't be the first time :)15:24
treachtilman: again wrt the db thing; if we used a db, we could basically throw away all those hackish little scripts we have to keep the system tidy, and do everything via the "ATME" tool. :P15:24
n0ydGentoo is the opposite of simplicity imo.  It's more of a giant migraine.15:24
predatorfreakn0yd: Finally! Supporters!15:25
tilmantreach: are you talking about the local ports db now?15:25
treachlike for instance finding orphans should be a snap.15:25
n0ydI find it funny though, most Arch users are ex-Gentards15:25
tilmani'm not opposed to that15:25
tilmanbut seriously, you shoud start discussion about it on the list15:26
predatorfreakJust don't screw it up like Arch with their FS-DB thing, where they note how the packages where installed.15:26
predatorfreakThat pissed me the fuck off.15:26
n0ydOMG?! Developers having a open ear to inovation?!15:26
predatorfreakBecause ABS would explicitly install build deps.15:26
n0ydI've never heard such a thing15:26
treachtilman: I doubt that would be particulary useful, since I'm no hacker myself.15:26
* n0yd praises tilman 15:26
treachtilman: also, there are other, serious drawbacks with the approach as well.15:26
predatorfreakand then pacman wouldn't show me that they were orphans <_<15:26
treachtilman: so I'm not sure the net gain would be that big,15:27
n0ydpredatorfreak: I think they changed some of that with pacman3.15:27
predatorfreakn0yd: They did.15:27
predatorfreakThey made it worse.15:27
n0ydI will admit, pacman3 did become a little bit faster15:27
tilmantreach: seems you understand enough of it to start a discussion15:28
tilmanbut maybe i'll do it instead15:28
tilmanand you can chime in with thoughts15:28
treachI think that would be better.15:28
predatorfreaktilman: This is why I like CRUX! People can contribute ideas :)15:28
n0ydHmm, aspell compiled it appears.  It looks like it's on python.  Odd, I always use -03 on LFS, and never have had a problem with aspell, other things like my basic toolchain I've had problems with.15:28
predatorfreakn0yd: -O3 ricer :P15:29
n0ydpredatorfreak: It's not as bad as some guys.15:29
predatorfreakn0yd: Bah! I used to do all that fancy CFLAG ricery and now I just use a basic set of flags-that-won't-fuck-me-over.15:30
tilmann0yd: care to file a bug?
n0ydtilman: Is that really a bug? I'd consider that an upstream bug, if anything.15:30
tilmanwell, maybe it can be fixed by bumping opt/aspell's version15:31
tilmanif it's the latest version, it seems it needs a patch15:31
predatorfreaktilman: for -O3 to work properly? :P15:31
tilmani understood it was related to python?15:32
predatorfreak<n0yd> Hmm, aspell compiled it appears.  It looks like it's on python.  Odd, I always use -03 on LFS, and never have had a problem with aspell, other things like my basic toolchain I've had problems with.15:32
predatorfreakHe said it built fine.15:32
predatorfreakand that he's building python now.15:32
predatorfreakor was.15:32
predatorfreakMaybe he's on flyingpants now :P15:32
tilmang++ bug maybe :/15:32
predatorfreaktilman: Time to push 4.2.1 to 2.3 to fix the evil 4.1 -O3 bug? :P15:33
treachand get an amount of other evil bugs instead..?15:34
tilman4.2.1 has other evil bugs15:34
predatorfreaktilman: Bugs are fun though! :P15:38
tilmanyeah, i remember how much fun you had with sysklogd15:38
n0ydbbiab. I have 2 sinks full of dishes to clean.15:39
predatorfreaktilman: Ahh yes, the genocidal urges.15:39
treachbugs are only fun after they die.15:39
predatorfreakHehe, looks like I'm actually going to manage to get to a concert this year.15:44
predatorfreakand without spending too much on tickets :D15:45
tilmanby sleeping with security?15:45
predatorfreaktilman: I wish.15:45
n0ydI went to ozfest a few weeks ago15:45
predatorfreakI'm not Marilyn Manson dang it!15:46
predatorfreakI can't get away with that.15:46
predatorfreaktilman: Tickets for the show I'm going to are only 33$ per ticket with fees.15:46
tilmanonly ._o15:46
predatorfreaktilman: Hey.15:46
predatorfreakMy last concert was like 75$ a ticket.15:47
predatorfreakWith fees.15:47
tilmanwhich band is that anyway?15:47
predatorfreaktilman: Alabama Thunderpussy and Obituary.15:47
predatorfreakI'm going more so for Alabama Thunderpussy than Obituary.15:47
n0ydAnd before that, I went to rock the bell music festival down in new york, with rage against the machine15:47
n0ydAnd bunch of other kickass bands15:47
predatorfreakn0yd: I'm getting to the point where I just tend to go to concerts by bands I really lurve.15:48
tilmanpredatorfreak: i've never heard of alabama thunderpussy, and i'm totally clueless on oldschool death metal. or florida dm in general ;)15:48
predatorfreakand Alabama Thunderpussy (henceforth shortened to ATP so I can be lazy :P)15:48
n0ydI really love RATM15:48
predatorfreaktilman: Stoner/sludge/something metal.15:48
n0ydAnd seeing cypress hill, wu tang, and the roots was really cool also. plus all the other bands15:49
tilmanpredatorfreak: sounds boring :D15:49
predatorfreakand not surprising, they're not even that big within an underground movement :P15:49
predatorfreaktilman: It's not, they execute it with a raw, Motorhead-esqe fury.15:49
predatorfreakIt's fortunate for me that I'm gonna get to see them right near me too, heck, they're like 1 mile away.15:50
predatorfreaktilman: Now, if only I had the money and Clutch were coming to this area, then I'd go to two concerts this season.15:55
predatorfreakAh well.15:55
*** Romster has quit IRC16:17
*** Romster has joined #crux16:20
*** patroclo7 has quit IRC16:38
*** lasso_ has joined #crux16:48
*** treach has quit IRC16:48
*** lasso__ has joined #crux16:50
n0ydtilman: Are you responsible for gnutls also?16:56
tilmannope, that's jaeger16:57
tilmangrep Maintainer /usr/ports/opt/gnutls/Pkgfile16:57
n0ydk, it's outdated and broke.  aparently they don't keep old versions on server either16:58
n0ydjaeger: ping16:58
*** lasso_ has quit IRC17:06
*** lasso has quit IRC17:12
*** lasso_ has joined #crux17:25
*** lasso__ has quit IRC17:41
*** hp_tux has quit IRC17:46
*** RedShift has quit IRC17:56
*** n0yd has quit IRC18:06
*** jjpk has quit IRC18:06
*** predatorfreak has quit IRC18:19
*** n0yd has joined #crux18:21
n0ydDamn, gtk is broke also.  And I'm not sure how to fix it.18:22
n0ydtilman: You around?18:22
thrice`that's not broke.  you just have cups installed18:25
n0ydOh, so I need to force it to overright?18:25
thrice`nope.  just pkgadd it18:25
n0ydI'm trying to install it with prt-get, and it says failed.18:26
thrice`prt-get -if ? :)18:26
n0ydI understand the -f, whats the i mean? if you don't me asking...18:27
thrice`-if is ignore footprint18:27
n0ydGuess I'll just use that for the independent package, or is it safe to ignore all the footprints for a bunch of packages?18:29
jaegern0yd: a footprint mismatch doesn't necessarily mean the package didn't work, it just means it built differently than the package maintainer's build18:29
jaegersince cups adds optional support to gtk, that's why you see it here18:29
n0ydjaeger: Ah, ok.  Makes sense.  Did you see that other package earlier which is outdated and the mirror in the Pkgfile doesn't work?18:33
n0ydgnutls iirc18:33
jaegerI was AFK at the time but I actually just updated that18:33
jaegerthanks for the notice18:33
n0ydNo prob. Does crux have a place to flag packages out of date to let the maintainers know?18:34
n0ydJust curious18:34
jaegeryou could file a bug report on the bug tracker but I prefer just to get an email about it, myself18:34
n0ydI was just curious, cause Arch has an option on there website when you look through the package list, to flag packages out of date, which I assume emails the maintainer.18:35
jaegerprobably so... might not be a bad feature to have18:35
n0ydWouldn't be to much hassle to code into the current ports db site actually.18:36
n0ydJust have a link next to he package, no user input or anything.  And it automatically sens an email.18:37
jaegerthere's been talk for a long time that the ports db would be replaced with a better one, though, so its features don't get messed with much18:37
jaeger@seen tillb18:38
clbjaeger: tillb was last seen in #crux 17 weeks, 6 days, 6 hours, 54 minutes, and 27 seconds ago: <tillb> hi everybody18:38
n0ydlong time no see? :-P18:38
jaegerso it seems18:38
*** lasso__ has joined #crux18:45
*** Jason5876 has joined #crux18:50
*** lasso_ has quit IRC19:01
*** _mavrick61 has quit IRC19:03
*** Dudde has quit IRC19:03
*** _mavrick61 has joined #crux19:04
*** Dudde has joined #crux19:05
*** n0yd has quit IRC19:10
*** schniggie has quit IRC19:20
*** rxi_ has joined #crux19:26
*** rxi has quit IRC19:46
*** roliveira has joined #crux19:51
*** predatorfreak has joined #crux19:52
*** predatorfreak has quit IRC20:35
*** lasso__ has quit IRC20:41
*** predatorfreak has joined #crux22:38
*** jaeger has quit IRC22:50
*** morlenxus has quit IRC23:05
*** morlenxus has joined #crux23:05
*** predatorfreak has quit IRC23:35

Generated by 2.11.0 by Marius Gedminas - find it at!