<     May 2017     >
Su Mo Tu We Th Fr Sa  
    1  2  3  4  5  6  
 7  8  9 10 11 12 13  
14 15 16 17 18 19 20  
21 22 23 24 25 26 _2_7  
28 29 30 31
00:11 tmh1999 joined
00:26 dave0x6d joined
00:57 ifbizo joined
00:59 <Shiz> https://twitter.com/swiftonsecurity/status/861364611773337600
01:10 ncl left
01:34 LouisA joined
01:52 s33se_ joined
02:06 grayhemp joined
02:07 grayhemp_ joined
02:15 mdillon joined
02:18 mdillon joined
03:05 <darkfader> ncopa: i forgot about the pv-grub thing, hopefully i can look at it next week when i have a few hours off
03:08 dirac4 joined
03:31 Emperor_Earth joined
04:00 emacsoma` joined
04:01 dirac1 joined
04:22 grayhemp joined
04:31 blackwind_123 joined
05:05 fabled joined
05:07 D630 joined
05:32 arnotixe joined
05:43 arnotixe joined
05:57 arnotixe joined
06:02 argulp joined
06:19 mmlb_ joined
06:19 Guest82584 joined
06:19 minimalism joined
06:19 fabled joined
06:20 robtec joined
06:21 yo61 joined
06:38 gogoprog joined
06:40 rnalrd joined
07:56 blahdodo joined
07:58 __0xAX joined
07:59 cyteen joined
08:05 skjoldr joined
08:09 kvda joined
08:26 mercurialchimera joined
08:32 blueness joined
08:42 thunfisch joined
08:44 thunfisch joined
09:04 ams__ joined
09:16 xsteadfastx joined
09:19 Spydemon joined
09:21 kvda joined
09:23 sparklyballs joined
09:50 lesion joined
10:23 <iron_houzi> Shiz: Sorry, been AFK over the weekend.. Sorry for the confusion, Ansible does not rely on `adduser`, but rather `useradd`. Is it possible to install useradd somehow?
10:27 <parazyd> iron_houzi: install shadow?
10:41 thxffo joined
10:53 sparklyballs joined
10:57 rollniak joined
10:58 argulp joined
11:01 LouisA joined
11:16 <iron_houzi> parazyd: That's it! Thanks!
11:16 mercurialchimera joined
11:21 sparklyballs joined
11:22 grayhemp joined
11:28 sparklyballs joined
11:32 kvda joined
12:00 argulp joined
12:23 farosas joined
12:23 thxffo joined
12:25 D630 joined
12:33 cyborg-one joined
12:34 TomJepp joined
12:50 rnalrd joined
12:51 argulp joined
13:04 Kruppt joined
13:22 ogres joined
13:25 cyteen joined
13:29 <Lucias> hey Shiz :)
14:00 argulp joined
14:01 blueness joined
14:05 <Shiz> hello
14:06 <Shiz> iron_houzi: shadow-utils?
14:06 <Shiz> sorry, it's just called shadow
14:06 <Shiz> the package
14:11 thxffo joined
14:13 fabled joined
14:14 cyborg-one joined
14:22 grayhemp joined
14:26 j0ni joined
14:29 Skele joined
14:34 peterrus joined
14:37 fekepp joined
15:09 mdillon joined
15:29 Emperor_Earth joined
15:37 dlaube joined
15:38 firstlast joined
15:40 arnotixe joined
15:46 psychi[m] joined
15:46 sergey_ joined
15:50 Kruppt joined
16:05 tobias1 joined
16:11 rsal joined
16:11 orbiter joined
16:12 kaniini joined
16:12 hl joined
16:12 ganlub joined
16:12 dhole[m] joined
16:12 budric[m] joined
16:12 xs[m] joined
16:12 Kooda joined
16:12 itsfemme[m] joined
16:12 copumpkin joined
16:12 fish_ joined
16:12 kaiyou joined
16:14 blueness_ joined
16:25 cyteen joined
16:28 ams__ joined
16:39 gromero joined
16:51 syx joined
16:51 <syx> Hi
16:51 <syx> Does this look okay? https://0x0.st/YyB.txt
16:53 <syx> networking | boot in rc-update
16:54 yo61 joined
17:02 robtec joined
17:05 gopar joined
17:08 fekepp joined
17:12 Somasis joined
17:13 Somasis joined
17:18 thxffo joined
17:29 blueness joined
17:32 cyteen joined
17:38 krainboltgreene joined
17:40 blueness joined
17:42 gopar_ joined
17:45 ingo__ joined
17:48 blueness joined
17:49 rsal joined
17:56 <_ikke_> syx: netmask 128 for ipv6?
17:56 <syx> yup
17:57 <syx> i'm just gonna use dhcp and configure ipv6 manually after boot
18:12 fekepp joined
18:14 <ingo__> where are set the timezone for timestamps in /var/log/messages? I'm on Europe/Berlin. date shows the correct time. but in the /var/log/messages i'm 2 hours back.
18:14 <clandmeter> logs are always utc
18:19 <ingo__> but only on alpine... have a few centos-kvms here, where all timestamps equal, so this is confusing. Thanks for this info
18:20 <avih> fyi, if someone is interested, tcc (tinycc) can now built and pass all tests on alpine (vanilla only). there's still some issue with va_args though, but it's already capable of building a lot of things (including itself)
18:24 Lloyd joined
18:42 D630 joined
18:58 <Shiz> ingo__: you would have to set the system timezone for that to work
18:58 <Shiz> avih: cool, nice work
18:59 <avih> Shiz: i [merely] didn't do anything ;)
19:04 <gromero> hi. does apk support socks5 proxy?
19:07 <feuerteufel> hi @ all
19:09 <feuerteufel> Shiz: What das that mean "system timezone"? I do have the same prob. with time in logs
19:12 czart_ joined
19:13 lnds joined
19:14 <Shiz> gromero: it uses libfetch which i believe doesn't, but you may be able to use torsocks to force it through socks5
19:15 <Shiz> feuerteufel: the present TZ variable
19:15 <gromero> Shiz: thanks. torsocks is my second options. I'll go for it :-)
19:16 <Shiz> I believe you can set it globally in /etc/profile, or just for the syslogger in /etc/conf.d/syslog
19:18 <Shiz> feuerteufel: probably better: ln -sf /usr/share/zoneinfo/Your/Timezone /etc/localtime
19:18 <clandmeter> afaik this is musl related
19:18 <Shiz> e.g. /usr/share/zoneinfo/Europe/Berlin
19:18 <Shiz> clandmeter: not really, it's not different from other libcs :P
19:19 <clandmeter> setting timezone does not change syslog timezone
19:19 <Shiz> it will if you set it at the right time
19:19 <Shiz> so setting it that the syslogd process actually has it set correctly
19:19 <Shiz> which means not in your local profile for instance :P
19:21 <clandmeter> i remember it was logging in utc, maybe the changed it overtime.
19:21 <Shiz> clandmeter: http://git.musl-libc.org/cgit/musl/tree/src/time/__tz.c#n127
19:21 <Shiz> :)
19:21 <clandmeter> http://www.openwall.com/lists/musl/2014/01/29/1
19:21 <Shiz> it first reads TZ, then /etc/localtime
19:21 <Shiz> ah
19:22 <Shiz> right, it's the syslog() function
19:22 <Shiz> didn't take that into account -- you're probably right
19:22 <Shiz> not used to syslog() being part of libc
19:22 <ingo__> Timezone was set on installation. so /etc/localtime is correct. Also set /etc/TZ. in /etc/conf.d/hwclock clock is set to local. on boot is see '[ 0.556926] rtc_cmos 00:00: setting system clock to 2017-05-08 20:09:44 UTC (1494274184)' which is not UTC, but localtime
19:22 <feuerteufel> The localetime is corekt, but the time in for exemple "/var/log/messages" is 2h late so I think there is UTC not locale time
19:23 <Shiz> ingo__: you very much want to set hwclock to utc, not local
19:23 <Shiz> /etc/TZ is not used by anything in the system i think
19:23 <Shiz> finally, what clandmeter linked above seems right: in musl, applications using syslog() will always log in UTC, not the local time
19:24 <Shiz> don't think this can be trivially changed
19:24 <clandmeter> seems some ppl patch it according the the thread.
19:25 <Shiz> tbh i don't disagree with dalias, i think any timezone adjusts should be done in syslogd itself, not depend per application logging to it
19:31 cyborg-one joined
19:34 argulp joined
19:35 <feuerteufel> Shiz: I agree, time has to be set from one point!
19:35 <feuerteufel> But why are there 2 diffrent times on one system?
19:42 <darkfader> example: systems might have users from different continents
19:42 <darkfader> it's the same time, just different time zones
19:42 <darkfader> if you put the system itself in utc and they use TZ it's a lot easier
19:43 <syx> Hi
19:43 <syx> https://0x0.st/YtL.txt should this work to bind the addresses and stuff?
19:45 vectr0n joined
19:47 <feuerteufel> darkfader: you right!
19:48 <feuerteufel> But it looks like TZ dosn't work, correct?
19:51 <ingo__> but this works only if it's consistent. had activated the rc.log. Timestamps here are localtime. So syslog are UTC other logs are localtime.
19:54 <ingo__> if the time offest was also printet, this makes sense. so what time is printet in which log?
19:55 <feuerteufel> ingo__: I'll check ... yust a sec ...
19:56 <Shiz> as mentioned above, it will always be UTC regardless of your timezone
19:56 <Shiz> because of how musl's syslog() call works
19:57 <dalias> syslog is not permitted by POSIX to inspect the TZ variable much less behave as if it called tzset()
19:57 <dalias> the syslog() function that is
20:01 <feuerteufel> Ok apache take's the locale time like that: [08/May/2017:21:58:33 +0200]
20:01 <ingo__> dalias: that is what i have understood.
20:05 <ingo__> the most logfileoutputs are configurable. Problem is if here, if the offset is not in the timesstamp, and you load the logfiles in a system like graylog or splunk, you have don't know anymore, how to correlate the lines for error search.
20:05 <ingo__> think here, the solution is, to install rsyslog or so.
20:05 <feuerteufel> A nsca-log take's also the locale time, but no daut about that! That script is build by me ;)
20:06 <Shiz> the issue is that the syslogd format should just pass UTC timestamps
20:06 <Shiz> and that the syslogger then can adjust it to some timezone if it desires
20:06 <Shiz> :P
20:06 <* Shiz> just wrote a patch that makes busybox syslogd do that, maybe
20:08 <dalias> yes, the syslogd should convert utc timestamps from the client sending the log message into whatever format it wants them saved in
20:08 <ingo__> Is there some page on the wiki, where this is described? didn't found it.
20:09 <dalias> but really everything on computers should be done in utc...
20:09 <dalias> saving local times is a mess of ambiguity
20:09 <dalias> esp. at daylight time change boundaries
20:10 <ingo__> hm, i have here arround 300 server, which produce arroucht 100GB logs per day, all on localtime. it would an absolut mess to find something, if they where on UTC.
20:11 <Shiz> i'm confused what mktime() assumes about the struct tm given
20:11 <Shiz> :P
20:12 <Shiz> wrt timezines
20:12 <dalias> local
20:12 <dalias> :(
20:12 <dalias> that's why the timegm function (nonstd iirc) exists
20:13 <Shiz> bleh
20:13 <Shiz> i just want to convert a UTC struct tm to a local... thing
20:13 <Shiz> but timegm is bleh
20:13 <dalias> fwiw you only need time conversion functions in stdlib for local time
20:13 <ingo__> If there is a good documentation about it, that in the syslog output is UTC used, there should be no problem.
20:13 <dalias> for utc the formulas can just be used
20:13 <Shiz> yeah but i'm implementing a -Z flag for busybox that converts presumed-UTC incoming timestamps to local time
20:13 <Shiz> :)
20:13 <Shiz> busybox syslogd*
20:14 <dalias> well you can also use the globals that tell the tz offset
20:14 <dalias> extern long timezone;
20:14 <Shiz> oh?
20:14 <Shiz> nice
20:14 <Shiz> that's.... good enough
20:14 <dalias> it's posix not plai nC
20:15 <dalias> you have to add/sub it from the timestamp and set tm_isdst==0
20:15 <Shiz> i think/hope busybox assumes posix
20:15 <Shiz> if not, whatever
20:15 <Shiz> :P
20:15 <dalias> *nod* it does
20:15 <dalias> anyway just using the posix formula is equally portable and doesn't depend on global state :-p
20:17 <Shiz> https://txt.shiz.me/NzM4NjFjOW
20:17 <Shiz> something like this is what i have then
20:19 <Shiz> with an appropriate call to tzset() im init
20:19 <Shiz> in init
20:21 thxffo joined
20:30 <feuerteufel> Shiz: What I have to do with that patch?
20:30 <Shiz> what i linked isn't a patch
20:31 <Shiz> hmm, above code snippet doesn't seem to work either
20:31 <feuerteufel> That's why I don't know what to do with that ;)
20:46 <Shiz> dalias: what was that posix formula? ;p
20:59 <Shiz> hmm, somehow timezone isn't getting set properly...
21:06 <Shiz> nvm
21:06 <Shiz> timezone is seconds WEST of UTC....
21:08 consus joined
21:14 jenkins joined
21:37 <nmeum> I just use this patch to overwrite the time instead of converting it http://sprunge.us/DjhF
21:37 <Shiz> nmeum: that breaks for remote syslogd
21:37 <Shiz> :P
21:38 <nmeum> I don't use a remote syslogd so I don't care :p
21:38 <nmeum> but if you write a proper patch that would be great
21:39 <Shiz> nmeum: https://github.com/alpinelinux/aports/pull/1360
21:39 <Shiz> ;)
21:39 <nmeum> did you suggest that patch upstream?
21:39 <nmeum> as in: busybox upstream
21:39 <Shiz> not yet, considering i just wrote it 15 minutes ago
21:40 <nmeum> :)
21:44 blueness joined
21:54 <Shiz> nmeum: http://lists.busybox.net/pipermail/busybox/2017-May/085437.html
21:54 <Shiz> :P
22:06 nidan_ joined
22:06 <nidan_> After upgrading to 4.9.26-0-hardened, my ALSA sound is now crap, is this a known issue?
22:10 <nidan_> ASUS Z97 MB with Realtek ALC892 btw..
22:13 MH0815 joined
22:22 syx left
22:50 <nidan_> Nm, it's not actually ALSA, it's mplayer. Playing .wavs work nice.
22:55 <nidan_> mplayer's mp3 decoder seems broken, it can play .flac, .wav, aac..
22:56 <nidan_> Lot's of high-pitched artefacts when playing mp3s though.
22:56 <Shiz> may i recommend mpv
22:56 <hiro> disable alsa's resampler and mixer
22:56 <hiro> it might still be using these horrible linear interpolation
22:57 <nidan_> Shiz: I'll check it out.
22:57 <nidan_> hiro: Don't know about alsa's resampler.
22:57 <nidan_> hiro: alsamixer settings seem fine.
22:58 <hiro> alsamixer is not what i said
22:58 <hiro> good night
22:58 pickfire joined
22:59 <nidan_> hiro: Nah, I noticed. Care to explain?
23:03 pickfire joined
23:05 sigmoidbigfoot joined
23:08 pickfire joined
23:13 pickfire joined
23:17 Emperor_Earth_ joined
23:19 dirac1 joined
23:29 dirac2 joined
23:29 Emperor_Earth__ joined
23:30 <nidan_> Shiz: fyi: mpv plays the mp3s fine. Something must be broken in mplayer or perhaps ffmpeg atm.
23:30 <Shiz> i'd always recommend mpv over mplayer
23:30 <Shiz> it's an mplayer fork that's actually maintained for the better
23:30 <Shiz> :p
23:31 Emperor_Earth_ joined
23:33 <nidan_> Shiz: Yeah, I glanced through the options and the man-page, felt more structured than mplayer's
23:34 Emperor_Earth joined
23:36 emacsoma` joined
23:36 dirac1 joined