IRC Logs for #crux Wednesday, 2019-07-17

*** john_cephalopoda has quit IRC00:01
*** john_cephalopoda has joined #crux00:15
*** tilman_ has joined #crux01:22
*** frydegg has joined #crux01:48
*** jolupa has quit IRC01:54
*** jakelyz has quit IRC01:55
*** jakelyz has joined #crux01:56
*** frydegg has quit IRC01:57
*** frydegg has joined #crux01:58
*** _________mavric6 has quit IRC02:46
*** _________mavric6 has joined #crux02:47
*** circ-user-qtxuz has joined #crux04:06
*** zimmer_Dl has quit IRC04:08
*** JanC has quit IRC04:49
*** circ-user-qtxuz_ has joined #crux04:53
*** circ-user-qtxuz has quit IRC04:56
*** zimmer_Dl has joined #crux06:29
*** zimmer_Dl has quit IRC06:45
*** slek has joined #crux07:15
*** pedja has quit IRC07:18
*** pedja has joined #crux07:19
*** guido_rokepo has joined #crux07:41
*** guido_rokepo has quit IRC09:36
*** guido_rokepo has joined #crux09:37
*** kalamaki has quit IRC10:19
*** kalamaki has joined #crux10:22
*** t58 has joined #crux11:09
*** jolupa has joined #crux11:17
jolupaHi everyone! I have a, don't know if it's a problem.11:18
jolupaI made a package for mailspring for my use, is from debian package11:18
jolupaand now everytime i update the system and make a revdep allways is showing the mailspring app11:19
jolupaany hint?11:19
*** xor29ah has quit IRC11:37
*** xor29ah has joined #crux11:37
pedjajolupa, what's the error (run revdep -vv)11:40
jolupaIs in some libraryes that comes with the pack files, is an atom application11:42
pedjayou'll need to add that lib directory to etc/revdep.d/$name11:44
jolupaAh! ok... Will do it11:44
pedjacheck out firefox-bin port for reference11:45
*** t58 has quit IRC11:45
jolupaAlso, is there any way to link a file in a package, I try it but the link is linking to /home/src/pkgmk/...11:47
pedjawhat do you mean by 'link a file'? symlink it from 'usr/lib/foo to /usr/bin/foo' in a package?11:53
pedjanvidia port does that a lot with the libs :)11:54
jolupayeah! symlink... I will check that too... Thanks!11:55
*** aref_ has joined #crux12:06
*** jolupa has quit IRC12:12
*** seiflek has joined #crux12:51
*** slek has quit IRC12:53
*** JanC has joined #crux13:05
*** JanC_ has joined #crux13:05
*** JanC_ has quit IRC13:06
*** JanC_ has joined #crux13:07
*** JanC_ has quit IRC13:07
*** JanC_ has joined #crux13:07
*** aref_ has quit IRC13:07
stenurirc via ssl, there it is13:08
*** JanC has quit IRC13:11
*** JanC_ has joined #crux13:16
*** nthwyatt has joined #crux14:43
*** darfo has quit IRC14:44
*** heroux has joined #crux15:07
*** ryuo has quit IRC15:10
*** nthwyatt has quit IRC15:17
*** darfo has joined #crux15:19
cruxbot[opt.git/3.5]: samhain: update to 4.3.315:19
cruxbot[opt.git/3.5]: mariadb: update to 10.4.615:19
cruxbot[opt.git/3.5]: poppler-qt5: update to 0.79.015:19
cruxbot[opt.git/3.5]: poppler-glib: update to 0.79.015:19
cruxbot[opt.git/3.5]: poppler: update to 0.79.015:19
*** ryuo has joined #crux15:22
*** tracer has joined #crux15:24
*** tracer has quit IRC15:25
*** onodera has joined #crux16:26
*** jolupa has joined #crux16:30
jolupaHi everyone. I found a extrange error at boot, the system is working perfect but about 3 lines, after loading the keyboard, show "scandir: no such file or directory" Any idea?16:31
*** guido_rokepo has quit IRC16:31
TimB_oh god, a new poppler version /me cries in LaTeX16:41
dlcusaThis is weird: -- it looks like four depends got concatenated into two (all packages are in the jolupalabs repo).  Has anybody ever seen anything like this by prt-get (yes, prt-get doe sthe same thing)?16:57
jolupa@dlcusa Woah!there's a problem with the repo or is prt-get? :O17:02
john_cephalopoda\me cries in \latex17:03
john_cephalopodaACTION cries in inkscape17:03
TimB_inkscape as well, true17:03
TimB_I'll have a look into that later17:03
TimB_latex that is, first I got work to do with latex17:03
TimB_john_cephalopoda: have you tried the system libs approach yet?17:04
*** heroux has quit IRC17:06
*** zimmer_Dl has joined #crux17:07
*** heroux has joined #crux17:07
pedjajolupa, I see the same. I have no idea what's that about.17:11
john_cephalopodaTimB_: Nope, not yet. I don't really get why poppler breaks things at _every_ release though.17:12
TimB_john_cephalopoda: same confusion over here17:12
jolupa@pedja So it may comes for las updates...17:12
dlcusajolupa, there are tabs in many of the # Depends on: lines--I'm not sure if such white space is supported.  However, it's hard to see how that could suddenly be confusing prt-get.17:15
jolupa@dlcusa Tabs? Weird, let me see in a few moments. Are the XFCE ports?17:16
dlcusajolupa, I don't understand the question.17:17
jolupa@dlcusar No I tend to put tabs in the first versions of XFCE, but I think that I removed all tabs, so I will check, just in case.17:17
*** heroux has quit IRC17:20
juebecause only the cpp, glib and qt APIs of poppler are considered stable17:22
juecups-filters has learned that since release 1.24 ;)17:23
cruxbot[xorg.git/3.5]: xorg-libpciaccess: updated to 0.1617:23
*** heroux has joined #crux17:25
pedjahm. 'scandir' seems to be from kbd?17:26
joacim30 minutes to download rust17:26
joacimwhat's in that thing?17:26
Anselmoiron is pretty heavy . . .17:27
jolupa@pedja kbd? There's different packages with kbd righ?17:27
joacimmore used to black metals myself17:27
joacimthose are pretty light17:27
dlcusaRust is produced by commiting iron through a water pipe, maybe?17:29
*** heroux has quit IRC17:34
cruxbot[core.git/3.5]: curl: update to 7.65.217:35
cruxbot[core.git/3.5]: bzip2: update to 1.0.817:35
jolupa@dlcusa I just updated some pkgfiles removing the tabs, if you wanna try.17:36
pedjajolupa, core/kbd17:37
*** heroux has joined #crux17:39
pedjadoes bzip2 requie rust yet :) ?17:39
pedjarust is slowly worming its way into core, I see17:43
joacimthinking about it, i rarely download bz2 files17:44
joacimmost are gz or xz17:45
dlcusajolupa, I'll tery, but I doubt it will make a difference, so I'll be looking into prt-get code shortly...17:45
joacimcould probably drop bzip2, and i wouldn't even notice =)17:45
pedjabetter compression then gz, faster then xz. I use it a lot :)17:48
pedjathat and zstd17:48
joacimhow much of the time saved compared to xz, do you spend compiling rust? =)17:49
pedjarust-bin ftw17:49
joacimyeah im still downloading that17:49
joacimthink half the internet must be broken on my end17:50
juejoacim: I doubt that, a grep through opt shows me 87 tar.bz2 files17:50
joacimmust be the swedes again17:50
pedjaI thought you are on a cruise, killing time, joacim :)17:50
pedjatheir internet kind of sucks and it's expensive af17:51
pedjajue, have you been seeing those 'scandir: no such file or directory' messages on boot?17:52
pedjaI suspect it's kbd update that's causing it.17:53
jueno, but I might missed that17:54
pedjayou people and your fast computers :)17:58
*** deep42thought has joined #crux18:12
deep42thoughtHi, after the upgrade to crux 3.5, gethostid() segfaults - I guess I need to recompile glibc, might this be the case?18:12
john_cephalopodaDidn't experience any issues with anything after the update.18:28
deep42thoughtI think, I accidentally had an old gcc (because I have my own port with ada enabled) during compilation of some packages18:33
*** t58 has joined #crux18:34
*** jolupa has quit IRC18:57
deep42thoughtnope, rebuilding glibc didn't fix it :-/19:01
*** onodera has quit IRC19:03
*** onodera has joined #crux19:07
*** onodera has quit IRC19:21
john_cephalopodadeep42thought: In which program does the crash happen?19:21
john_cephalopodaMaybe rebuild that one?19:21
deep42thoughtgethostbyid, for example19:22
deep42thoughtor a simple test c programm which uses gethostbyid()19:22
deep42thoughtmeant gethostid()19:22
deep42thoughtbash-5.0$ hostid19:22
deep42thoughtSegmentation fault19:22
AnselmoI get that on a recently upgraded 3.519:23
deep42thoughtbash-5.0$ gcc test.c && ./a.out19:23
deep42thoughtSegmentation fault19:23
john_cephalopodaHmm, 3.5, too. hostid doesn't segfault for me.19:24
deep42thoughtAnselmo: do you happen to have your own gcc for some reason?19:24
Anselmonew install or upgrade ?19:24
deep42thoughtupgrade, here19:24
AnselmoI appear to be using the gcc that came with the 3.5 iso on that system19:25
AnselmoI'm rebuilding coreutils19:25
deep42thoughtthis is next on my list19:25
deep42thoughtbut currently I rebuild gcc19:25
Anselmostill segfaults19:27
deep42thoughtAnselmo: do you use ccache?19:27
pedjaworks fine here, no segfault19:44
pedja3.4 -> 3.5 upgraded system19:46
pedjadeep42thought, install the glibc from the iso package?19:47
deep42thoughtOh, I should have thought of *that* ...19:47
deep42thoughtsegfaults, too19:48
pedjaanything in the dmesg?19:49
pedjaor kernel logs?19:49
deep42thought[140883.288300] a.out[5121]: segfault at 14 ip 00007fd3858ca19a sp 00007ffc526e11b0 error 4 in[7fd3857fb000+147000]19:49
deep42thought[140883.288313] Code: f1 4d 89 e8 4c 89 e6 48 89 df e8 31 a7 01 00 85 c0 75 b5 48 8b 44 24 18 ba 04 00 00 00 48 8d 74 24 10 c7 44 24 10 00 00 00 00 <83> 78 14 04 0f 4e 50 14 48 8b 40 18 48 63 d2 48 8b 38 48 85 d2 7419:49
AnselmoI just rebuild glibc and it still segfaults19:50
pedjaI wonder if gdb would shed some light on the issue19:51
TimB_test.c works fine over here19:51
deep42thoughtwhat should I do?19:51
deep42thoughtthe farthest I got with gdb was to blame gethostid instead of alpine :-)19:51
pedjaalpine, the mail reader?19:52
Anselmowell, thats a good start :P19:53
Anselmobuut blrp.19:53
Anselmostatement of confusion19:54
pedja'Inferior 1 (process 2465) exited normally]' test ran from gdb19:55
deep42thoughtProgram received signal SIGSEGV, Segmentation fault.19:56
deep42thought0x00007ffff7ee219a in gethostid () from /lib/
deep42thoughtwhat should I do now?19:56
deep42thought(with gdb)19:56
deep42thought#0  0x00007ffff7ee219a in gethostid () from /lib/
deep42thought#1  0x0000555555555152 in main ()19:56
deep42thoughtit only has that one call19:56
pedjathat's...interesting :)19:57
pedjaglibc from iso and rebuilt one both broken.19:58
pedjaI presume you ran revdep after the upgrade? a long shot, I know19:59
deep42thoughtI did19:59
deep42thoughtACTION runs it again19:59
deep42thoughtthese pop up: eagle, libreoffice, pidgin, rust-bin20:01
AnselmoI had also been having an issue with w3m and libssl, which I havent yet sorted out20:02
Anselmoapparently also qemu is broken, and qutebrowser (on my system)20:03
Anselmobut my machine isnt fast so I'll see when its recompiled things20:03
pedjanone of that explains glibc breakage, I think20:04
deep42thoughtwhich packages are involved in building glibc?20:05
deep42thoughtgcc, binutils, coreutils?20:05
deep42thoughtglibc itself?20:05
pedjatoolchain and glibc, afaik20:05
deep42thoughtthough, even the binary glibc from the iso is broken (or rather: looks broken)20:05
pedjayeah, that's the 'wth' part :)20:06
pedjatried running strace?20:06
pedjait's a guessing game at this point :)20:07
Anselmoyeah thats why I noted the other weird things I'd run into o.o20:08
pedjaapparently, hostid is just the massaged host ip address. hm.20:26
deep42thoughtthough my ip does not show up in the strace log20:29
deep42thoughtonly the ip of my router20:29
deep42thoughtI changed the ip of my router to in /etc/resolv.conf AND IT WORKS20:30
deep42thoughtACTION loses faith in his systems20:30
pedjathat was the weird part, I haven't seen it trying to parse resolve or hosts or nss in that strace log20:32
deep42thoughtputting an invalid nameserver into /etc/resolv.conf makes the test program hang20:33
deep42thought... and then segfault20:33
deep42thoughtso maybe it's something about my dns's response which makes it hickup?20:34
AnselmoI switched the nameserver in resolv.conf and it didnt fix it (changed to from a set of opennic servers)20:34
deep42thoughtit reads /etc/resolv.conf20:36
Anselmoand let my router tell it what to do over dhcpd, and still hostid segfaults. . .20:36
pedjaAnselmo, what's in your hosts file?20:37
Anselmothe only uncommented line on this machine is' localhost '20:38
pedjaAnselmo, and if you add another line, ' nameOfHost'20:42
pedjasegfaults here too if I comment it out20:43
deep42thoughtpedja: then it works over here, too, with my regular name server20:43
deep42thoughtand the outcome of hostid actually depends on the ip which I enter into /etc/hosts20:44
Anselmostill segfaults pedja20:44
Anselmogah. has this occured on systems that arent upgraded from 3.4 ?20:49
pedjaI don't know what to tell you, works fine here with that added line.20:54
AnselmoI'm just thinking my system seems broken in other ways, I'll prolly just do a reinstall20:55
AnselmoI havent always treated things carefully and its potentially some weird thing I did a long time ago comming back to bite me :P20:55
pedjarun rejmerge, maybe you missed something after the upgrade :)21:05
deep42thoughtthanks for your assistance, cu when my next problem arises ;-)21:07
*** deep42thought has quit IRC21:08
pedjathe Crux support contract only covers so much :)21:11
*** darfo has quit IRC21:12
*** kalamaki has quit IRC21:12
pedja'supports ECC RAM (in non-ECC mode)'. what21:14
*** kalamaki has joined #crux21:14
*** darfo has joined #crux21:16
pedjaI am missing the point of getting (and paying extra) for ECC, if it doesn't actually work.21:16
Anselmoha ha21:31
dlcusapedja, two bits flipping it cannot catch--that's a very low probability, though.21:38
ryuothere's two versions of ECC.21:39
ryuosingle bit and multi-bit.21:39
dlcusaMy understanding is the multi-bit implementations only allow notification of corruption, not recovery.21:48
pedja'zfs will kill your kittens without ecc' threads are fun :)21:50
*** seiflek has quit IRC21:51
stenurI cannot make my hostid crash.21:59
dlcusa sems to be useful.21:59
ryuoso is this:
john_cephalopodaI once read some article by somebody working in a datacenter, who manages a TB or so of ECC RAM and can monitor how often they do corrections.22:47
john_cephalopodaFor that TB, it is actually quite frequent.22:47
john_cephalopodaAll the cosmic radiation has quite some impact.22:48
ryuothat's also much more RAM than a regular server has.22:48
john_cephalopodaAs I said, it's a whole datacenter.22:50
dlcusa sems to be useful.22:50
*** Romster has joined #crux23:26
*** leah2 has quit IRC23:42
*** SiFuh has quit IRC23:49

Generated by 2.14.0 by Marius Gedminas - find it at!