libera/#devuan/ Friday, 2019-05-10

XenguyThat seems reasonable DocScrutinizer0500:00
Xenguyonefang: OT, or w/e00:00
slvryeah, the clearly on-topic subject flag makes it work.00:00
Xenguy"Let the schism begin"00:00
slvrOther rooms I'm in are like that, informally. Chatter stops when questions from users show up.00:00
XenguyMakes sense to me00:01
DocScrutinizer05on the other hand it's quite silly to complain aboit some off-topic chat of hours ago, because "you flooded my IRC client with stuff that *I* don't like to read". If you don't like to read, just don't do00:01
DocScrutinizer05>><slvr> fair enough. I've found that rooms with some constant low level chatter are more friendly for newbies, but that's just me.<< no, that's a valid and correct observation00:07
sixwheeledbeastI'd sooner have a little off-topic than logs of people asking if anyone is there and leaving 10 minutes later00:18
DocScrutinizer05^^^00:19
DocScrutinizer05IRC still is a realtime medium with possibly rather long response time than any sort of streamed forum00:20
DocScrutinizer05point to consider: even when some of the 24/7 inhabitants keep logs, hardly any new user will00:22
DocScrutinizer05IOW it's frequently pointless to bother too much about backscroll more than maybe 2 or 3h ago00:23
DocScrutinizer05actually I receive some quite sarcastic comments quite often when I reply to a topic that been more than a 2 hours ago. In some channels you might get kicked for that00:32
MinceRthose channels don't sound fun00:33
gnarfacewell, earlier today there was only mtnman.  i responded to him within 1 minute, and 7 minutes later he left anyway without saying another word00:35
gnarfaceso, you can't save em all00:36
fsmithredHe's been here before00:37
fsmithrednot recently00:37
* golinux grimbles00:43
golinuxgrumbles really00:44
DocScrutinizer05why you grumble, dear? We're discussing this very channel's policies, couldn't be more on topic01:05
DocScrutinizer05and a 15 posts per hour would be perfectly allowable even if it were completely off-topic01:08
slvrI've got an on topic question01:17
DocScrutinizer05shoooot! :-D01:17
slvrgolinux: In the "help devuan" link in your forum sig, there is a report bugs with the bugreport utility thing. Where do those bugs go?01:17
slvrI'd like to see if any are squashable.01:18
gnarfacei am not golinux but i assume they go to https://bugs.devuan.org/01:20
gnarfaceor at least that's probably where they're supposed to go01:20
slvrcool, I'll look through them.01:22
DocScrutinizer05>>...and 7 minutes later he left anyway without saying another word<< I wonder if those are noobs on ORC who, after posting a question, clicked on #otherchan which takes them to there and they got no clue how to return to here01:25
DocScrutinizer05IRC even01:25
golinuxslvr: I think that report bug wants to send them to Debian.  I suggest sending reports by email to devuan directly.  There are instructions on the bugs.devuan.org site01:43
golinuxJust make sure it's a devuan not a debian bug which should be reported upstream01:44
Xenguysalmons swim upstream02:54
XenguyIs that haiku?02:55
xrogaansalmons swim upstream; leafs from green to red; bears feast02:56
xrogaan?02:57
Xenguyvery nice03:06
XenguyI think03:07
XenguyIs it 5-5-2 ?03:07
Xenguy5-7-5 apparently03:10
Xenguyfuck all of you and have a very good night, right there, officially logged03:11
fsmithredI want dinner first03:11
Xenguyoh no you don't03:12
Xenguy;P03:12
furrywolfno dinner?  and you're not even my type, unless you're female, mature, sane, practical, and good with a strapon...  which I'm fairly sure you're not.  :P03:12
XenguyJesus Murphy03:13
XenguyShort answer, no03:13
fsmithredlmao03:13
XenguyEither way, I had the first segment of the haiku, uh, pegged03:15
onefangThat escalated quickly.03:16
golinuxGet it together folks and move over to #debianfork03:17
XenguyAll it takes is a couple of drinks, and the sky comes crashing down03:17
XenguyWhoops03:17
agrishello, I'm having some trouble with PostgreSQL and redis-server in Devuan ASCII04:37
agriswhen I run rc-status, it says redis-server is stopped, but when I run rc-service redis-server status it say it is running04:38
agrishtop confirms redis is actually running04:38
agrisalso,04:38
gnarfaceis rc-status and rc-service part of redis?04:38
gnarfaceor are you using an alternate init system?04:39
agrisrc-status says postgresql is stopped, and rc-service postgresql status reports nothing at all, htop confirms postgres is not running04:39
agrisgnarface, I'm using sysvinit with OpenRC supervisory04:39
gnarfacei see, so openrc provided rc-status?04:39
gnarfacei'm guessing there's just something wrong with that script04:39
gnarfaceprobably a path or permission issue, but that's just a guess04:40
gnarface(for example, maybe it's looking in the wrong directory for PID files, or maybe your user just doesn't have access to that directory)04:40
agrisopenRC provides rc-status, rc-service, and rc-update04:40
agrisI am running this as root04:40
agriswell, root within an LXC container04:40
gnarfacehmmm. i wish i had more experience with openrc04:41
agrisso a permission related issue is not likely, unless root is not really root within this namespace04:41
gnarface"rc-service redis-server status" works as expected?04:42
agrisroot@nextcloud:~# rc-service redis-server status04:42
agris[ ok ] redis-server is running.04:42
agrisyes04:42
gnarfaceso rc-service works as expected in all cases, and rc-service ... unconfirmed in both cases?04:42
gnarfacesorry04:42
gnarfacei mean rc-status, unconfirmed in both cases?04:42
agrisno, rc-service is not working properly for postgresql04:42
gnarfaceempty output isn't a valid case? you're sure about that?04:43
agrisrc-service works fine for redis-server, but rc-status does not work properly for redis-server04:43
agrisgnarface, yes, it should say something04:43
agristhat's what OpenRC is for, daemon supervisory04:43
gnarfacefair enough.  i'm just not familiar with it, and a 0 output is a valid implied "success" message from a lot of legacy linux/unix command-line tools04:44
gnarfacei wonder if it could be the container04:45
gnarfacesomehow04:45
gnarfacethat would be one case where being root isn't necessarily sufficient privilege04:45
gnarfacedoes this look related? https://github.com/antirez/redis/issues/510504:45
agrisgnarface, maybe, but extremely unlikely. OpenRC doesn't do anything weird like that, it just calls sysvinit04:46
stiltrwas redis started from something other than the init.d script?04:46
gnarfacethat's probably what i'd be doing at this point, is hand-tracing the init.d script to see if it succeeds at everything it's trying (especially stuff like PID file location)04:47
agrisunlikely #5105 is related and I can ping redis and get a PONG back04:47
gnarfacehmm. ok, unrelated probably04:48
agris /etc/init.d/redis-server doesn't define a pidfile location, however it does use $PIDFILE, which may be inherited from the parent04:49
agrisoh wait04:50
agrisnvm i was wrong04:50
agrisRUNDIR=/var/run/redis04:50
agrisPIDFILE=$RUNDIR/redis-server.pid04:50
gnarfacethis looks related too: https://github.com/OpenRC/openrc/issues/21504:50
gnarface(at least contextually)04:50
agrisno, /var/run/redis/redis-server.pid contains a number, and that number does infact lead to the actual PID of the redis daemon04:51
gnarfaceok, so much for that theory04:52
agrisrc-status reports redis as stopped, not crashed04:52
agrisbut perhaps this could be related04:52
agrisHow old is OpenRC in Devuan?04:52
agris# rc-status -V04:53
agrisrc-status (OpenRC [LXC]) 0.2304:53
gnarfacewell it's the exact same version as debian04:53
stiltrhuh, I'm running 0.38.3 over in gentoo, fwiw.04:53
agrisI just updated a Gentoo Stable box two days ago04:54
agrishere's what it says04:54
agrisrc-status (OpenRC) 0.41.2 (Gentoo Linux)04:54
gnarfacedebian changelog says January 5, 201704:54
stiltrHuh, I updated mine a week or two ago... I thought... haha04:54
agrisIs there any reason we can't bump the minor version in the Devuan stable release?04:55
gnarfacejust parity with debian.  these packages are all inherited from them04:55
agrisWhen was the rc-config utility added to OpenRC?04:56
gnarfacei don't know that, sorry04:56
agrisDevuan only has the legacy rc-update04:56
gnarface0.40.3 in testing and unstable04:56
agrisWhat will it take to make it as stable?04:56
agrisI'm perfectly willing to test in as a backport04:57
agrison ASCII04:57
gnarfacealright, how much do you know about how testing becomes stable for debian?04:57
gnarfaceit's basically like that for devuan, only we have to wait on debian04:57
gnarfacethen fix extra shit04:58
agrisfuck04:58
gnarfaceand when i say "we" i really mean... not me.  i don't have anything to do with it04:58
gnarfacebut04:58
gnarfaceyour idea about the backport is good04:58
gnarfaceit's so good, that i'd actually be surprised if someone hadn't done it already04:58
agrisI ask because I've been running Gentoo servers for years and never had a problem with later versions of OpenRC04:58
gnarfacethere are a few people running openrc here, and someone may have actually already put the new build in some unofficial repo, if you want to try it.  you just have to figure out who of these sleepers know where it is04:59
gnarfacesomewhere around here there's some sort of unofficial proposed additions repo04:59
agrisBut at what point is there going to be enough 'testing' of OpenRC04:59
gnarfaceit might work now honestly, i don't know04:59
gnarfacebut devuan won't make testing stable until debian's testing goes stable05:00
gnarfaceso it looks like you have two options05:00
gnarfacetwo options other than just trying to fix the version you have anyway05:00
gnarface1) backport it yourself 2) find someone in here who did already05:00
stiltrThe only dependency difference I saw was libselinux1. You can try just installing the deb, if you want to live dangerously. = )05:01
gnarfaceyea, i recommend against that, but people have been known to get away with such risks in a pinch05:01
agrisI don't exactly feel comfortable waiting on Debian to actually finish something regarding a non-systemd init05:01
agristhey seem hostile to it05:01
gnarfacei don't blame you05:01
gnarfacedon't panic though, don't get angry.  this is a solveable problem...05:02
gnarfacebut i would also hate for you to go through the trouble of backporting it if it turns out there's a simple known fix for openrc in lxc containers already floating around out there05:02
agrissure05:02
gnarfaceit doesn't seem possible that you'd be the first person to run into this05:03
agrisbut if your ever looking for a OpenRC maintainer, I am your guy05:03
gnarfacethat position may actually be open05:03
gnarfacedevuan is welcoming volunteers05:03
gnarfaceare you aware of the mailing list?05:04
agrisconsidering I use Debuan and OpenRC in prod so much05:04
agrisyes05:04
gnarfaceok, good05:04
agrisI follow the mailing lists.05:04
agrisI am also aware of all the spam regarding people getting butthurt over the april fools prank05:04
gnarfaceyou might want to mention this issue on there in lieu of just sticking around here in IRC until morning, if you have somewhere else to be05:04
agrissure thing05:05
gnarfacei think my best advice to you right now is try to see if you can backport the beowulf src package to ascii05:05
gnarfaceif it builds with no trouble against ascii dependencies, there's a good chance upstream debian will take it into the backports repo i would think05:05
gnarfaceand if not, there's always that unofficial devuan additions repo... wherever it is (i forget what they were calling it)05:06
gnarfacejust out of curiosity, have you tried this outside the LXC container?  does it fail the same way on bare metal?05:06
gnarfacei use linux-vserver containers a lot, and i know that weird stuff can fail just because "root" in guests still doesn't have permissions to do stuff like create it's own network devices05:09
gnarfaceso sometimes some additional pre-setup on the host is necessary05:09
agrisI use OpenRC on bare metal and QEMU vms a lot as well, but have not run redis-server out of a LXC container with OpenRC before05:10
gnarfaceso 3 possible culprits05:10
gnarfacebut leaning towards it not being redis-server itself, because rc-status also fails weird for postgresql, just ... differently05:10
agrisIn this case, It's a very lage CentOS 7 LXC hypervisor, running the official LXC Devuan scripts05:10
gnarfacethat might be a clue actually05:10
gnarfacethat it fails differently for postgrequl05:11
gnarfacepostgressql however you spell that05:11
agrisjust say postgres05:11
gnarfaceheh, ok05:11
gnarfacei really should have made a point of memorizing who else was using openrc so i could ping them when questions come up05:13
agrisnvm on postgresql05:20
gnarfaceworking now?05:20
agrisDebian's postgres seems to be setup a bit weird05:20
gnarfacehmm05:20
gnarfacesounds plausible05:21
gnarfacethey do some odd changes on certain things for security05:21
agrisinstead of status outputting OK from OpenRC, it runs pg_lsclusters05:21
agrisfor some reason when I installed postgres it didn't initalize a default cluster05:23
agrisSo I just made one manually now and it seems to work05:23
gnarfacealright, one mystery solved there at least05:25
gnarfacehmm. reminds me of a weird issue i had in vservers recently.  been having to reinstall cron on fresh guests or it doesn't run.  not sure wtf is up with that.05:25
gnarfacecertainly someone in here had already built the new OpenRC for ascii.  who was that?05:31
stiltrI just gave it a go. If anybody wants to give it a try: https://stiles-engineering.com/ascii-openrc/05:43
xkr47good day all and thanks for Devuan!11:13
unixmanesr16:19
unixmanWhoops, wrong window. Time for more caffeine. :D16:19
g4570n!ping16:41
infobot1 packet transmitted, 1 packet received, 0.0% packet loss16:41
fsmithred!ssh16:48
infobotfrom memory, ssh is http://www.webos-internals.org/wiki/Application:OpenSSH, or http://www.openssh.org16:48
fsmithredhrmph16:49
djphfsmithred: eh?16:53
fsmithredI was hoping for a better response16:54
fsmithred...like a shell16:54
arminfsmithred: err what18:18
_abc_Hi. I'm trying to format a manual page manually using groff -man Tascii $src.man >out.txt and the system says it can't find man.macros . What package do I have to install?20:55
_abc_Strangely it is present but part of tcllib?20:56
_abc_Guys, groff on devuan ascii does not work as the manual says it should? source loading a macro package with so man.macros searches only in local dir, ignores -M and GROFF_TMAC_PATH= env. Both tried. I had to strace to see it. Is this a new feature of gnu groff? "feature" ?21:32
_abc_I am sort of fed up using my ninja skills (strace etc) to find holes in open source stuff which presumably passes regression tests? Is there a groff specific channel on freenode? Texttools in general?21:34
gnarface_abc_: i don't know much about it, but 2 minutes of googling infers to me that man.macros is a file full of a bunch of undocumented macros, and you're generally expected to just find an example, and feed groff a man.local you've written yourself based on that.  i don't know if that's true, that's just what a few minutes of googling suggests to me.21:40
gnarface_abc_: someone on some forum (maybe stackexchange) suggested looking at the old Solaris documentation for the macros to get an idea of what is available21:41
gnarfacemy own suggestion would be to try to find the man.local file used to make some popular GNU manpage you like the layout of and borrow examples from that21:42
_abc_gnarface: the way it works is, tcllib provides a man.macros which is used by all tcl manpages21:42
_abc_gnarface: the manpages invoke it using so man.macros in the source.21:42
gnarface_abc_: right, but that is probably not "the" man.macros file, it's probably just "a" man.macros file.21:43
_abc_gnarface: groff is supposed to look in the path passed with -M or GROFF_TMAC_PATH= for this21:43
_abc_gnarface: strace shows that a) it does not do that and b) there is no GROFF_TMAC_PATH env related string in the binary, contrary to what the manpage claims21:43
_abc_gnarface: yes it is "a" man.macros file, the one to be used in this case.21:43
gnarfacehmmm. my guess is the man page is out of date21:44
_abc_gnarface: so the question is, why does groff ignore the -M and env21:44
_abc_gnarface: it is out of date but it is also written in the same style as modern ones are21:44
gnarfacehmm.21:45
gnarfacelooks like this is not the only outstanding bug in groff21:45
gnarfacehttps://bugs.debian.org/cgi-bin/pkgreport.cgi?dist=stable;package=groff21:45
gnarfacei would think there's another way to make man pages21:46
_abc_Sure but this one is somewhat eyestabbing21:46
_abc_strace shows it does look in the "close" proximity of where the file is located under tcllib (see apt-file search man.macros output)21:47
gnarfaceis there an info page for it?21:48
_abc_nope21:48
_abc_I fixed it manually, copying the man.macros to where I run groff, then it works.21:49
_abc_This after reading strace21:49
gnarfaceyes, that seems to be the workflow everyone online suggested21:49
_abc_So it was not amusing for me but now I am curious why there is a desync between groff manpage and binary21:49
gnarfaceit looks like there might be some other undocumented environment variables too21:50
_abc_Specifically, GROFF_TMAC_PATH is not in strings $(which groff)21:50
gnarfacemaybe there's a webpage with more up-to-date documentation.  sometimes the man pages get left behind21:50
_abc_I ran strings on it and looked. The other manpage indicated env vars are in21:50
_abc_Perhaps GROFF_TMAC_PATH was deprecated21:50
gnarfacethere is a info page for it21:51
gnarfaceit's more like a whole info book actually21:52
_abc_Perhaps I am misinterpreting so macro scope21:52
_abc_yes the info "21:52
_abc_page" tends to be a book with gnu stuff21:52
_abc_I do not have the info pages installed for groff21:52
gnarfacehttps://www.gnu.org/software/groff/#documentation21:52
gnarfacesomeone also apparently wrote a pdf called "Groff and mom: an overview"21:53
_abc_apt-file search groff.info -> installed ; yet info does not find it here...21:53
gnarfacethey additionally encourage you to check in /usr/share/doc/groff*21:53
gnarfacehmmm, getting info pages working required an extra step but i did it so long ago i forget what it was21:53
gnarfacei'm using emacs as my info reader though so it might be just related to that21:54
gnarface"info groff" doesn't give you anything?21:54
gnarfacei wasn't aware apt-file even worked anymore21:54
gnarfaceit was broken for a long time21:54
_abc_So I had groff-base installed but groff not. Fixing it now.21:54
_abc_Also all the files you alluded to were not installed.21:55
_abc_gnarface: fwiw the info says the search path for .so can be set with -I cli option. Bummer did not try that. The manpage could really use some updating.22:17
helios21https://git.devuan.org/WIP-init/user-services/blob/master/README.md22:33
helios21see my post on dng-ml22:33
* helios21 is now off to play some Neverwinter Nights Enhanced Edition (with violence level set to zero)22:33

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!