libera/#devuan/ Wednesday, 2024-11-13

systemdletecant find a shim, but I see that nut wants avahi on freebsd...   hmmm.00:27
gnarfacelike a "recommend" or actually a requirement?00:28
systemdletewell, the startup complains that avahi isn't running00:28
gnarfacemaybe it wants it on linux too but that's just one of those accidentally omitted build dependencies?00:28
systemdleteso maybe not quite a requirement, since the nut install didn't break00:28
gnarfacehmm, i could only speculate00:28
systemdleteoh no, gnarface, there is actually an avahi meta package et al00:29
systemdletefreebsd => linux00:29
systemdletethere's even a thing called "launchd" that sounds an awful lot like the same sort of "project" that will eventually consume the entire bsd ecosystem00:30
systemdleteit's still talk though, from what I can gather00:30
gnarfaceafaik launchd is actually far older than systemd, i remember it from macosX decades ago00:30
* systemdlete would really prefer to just die...00:30
gnarfacei do remember it also being a pain in the ass though00:31
systemdletenot sure if it's the same project, or a new implementation00:31
gnarfacei assume it's what macosx inherited, but i haven't been following closely enough to be sure00:32
systemdleteIf architected well, and not a runaway train, it might be ok.  But I still believe in init freedom00:32
systemdletewell, nutconf ran and generated what looks like a config file--just like the ones I created on my linux system00:33
systemdleteactually, nut-scanner.  nutconf just returns to the prompt quietly00:34
systemdleteand there's no /etc/rc.d for it that I can see00:34
systemdletewell, at any rate, back to the build on linux.00:34
systemdleteI will try doing the 2.8.1 upgrade package and then the 2.8.200:35
systemdletebut I'd really rather just die00:35
systemdleteI'm gonna watch tv for a while.  This hurts my brain too much.00:35
gnarfaceit gets easier00:36
systemdlete(my problem is sitting for long times; need to take frequent breaks)00:37
systemdletebbl00:38
onefangJust got back from being out all morning.  https://mishka.snork.ca/apt-panopticon/results/Report-web.html went away last I checked.  http://veritas.devuan.org/apt-panopticon/results/Report-web.html is still good.  https://sledjhamr.org/apt-panopticon/results/Report-web.html is also good.  Both look OK to me.04:12
onefangI gotta rest up before I look deeper.04:12
onefangOr perhaps it was just a case of "package mirror herder is away all morning, we can play up now".  lol04:13
onefanggnarface: Was it just all the URL sanity errors you saw that made you think half of them are down?  That's normal.04:30
gnarfaceonefang: no, the http column04:30
onefangWhich reflects the URL sanity column.  Note that most of the time it has the same number of errors.  I have TODOs to sort these things out better.04:31
onefangLook at the green ticks next to the IPs in the DNS round robin column, that's a better indicator.04:32
onefangURL sanity is likely no longer needed, I haven't delved deep into the apt source code yet to check.04:32
onefangAnd apt-panopticon needs lots of work.  lol04:33
onefangToo many red crosses in that column though.  B-(04:34
onefangHas Debian updated their keyring again?04:37
onefangCoz this is odd.  A bunch are faillng only on their IPv4 incarnations IPv6 working fine, and the failure is a 404 on Debian's keyring, but most are fine.04:39
gnarfacei might not have proper ipv6 routing...04:40
gnarfacethat could have something to do with it04:40
onefangWhy would that have something to do with the mirror testing?04:40
gnarfacewell if they're failing on ipv4 and i only have ipv404:42
gnarfacethat was hours ago though, i haven't checked it recently04:42
onefangFailing for the mirror testers, which you are not.04:42
gnarfaceoh04:42
onefangLooks like Debian did update their keyring.04:42
gnarfacei misunderstood04:42
onefangAh apt-transport-https has gone away, one of the standard packages I test with.  HTTPS support is now included in apt itself, not a separate package.04:46
* onefang switches to testing apt-transport-tor instead.04:48
* onefang uncrosses my eyes and realises it's not IPv4 that's failing. lol05:06
m_mI'm not sure if you would support this, but I installed a devuan VM using debootstrap yesterday, and I have a problem with openRC. It tells me that I have ssh in the default runlevel, but it does not start it. When I start it manually (rc-service ssh start) everything works well. I do not have logs of ssh complaining at boot, and all the other services are starting well...11:21
gnarfacem_m: sounds like it might be a problem with the VM and openRC not playing nice together, something about pam and permissions or something. which VM?11:25
m_mit's a Qemu-kvm with libvirt11:26
gnarfacehmm11:26
gnarfaceyou starting the guest as root or as a regular user?11:27
m_mroot11:27
gnarfaceeh, stick around, someone might know the answer11:28
gnarfacewhich release on the host and guest?11:28
m_mI installed these packages for openrc to work :  elogind libpam-elogind openrc orphan-sysvinit-scripts11:29
m_mdaedalus on the both11:29
gnarfacedo you think it's possible the network is just starting after?11:29
gnarfacesshd would probably bail if it didn't find any network devices11:29
gnarfacethere should be a way to get logs... do you have rsyslogd and logrotate installed?11:30
gnarfaceyou might want to try one guest with sysvinit instead11:33
gnarfacejust to see if that's working11:33
gnarfacelast i checked was chimaera but i didn't have problems with it11:33
gnarfaceoh, hmm, i guess i didn't try sshd on that one though ....11:33
gnarfaceyea, i dunno. seen something like this but not with qemu. turned out to be a pam/limits/permissions thing that shouldn't affect qemu.11:34
gnarfacemight be your guest's network configuration or what openrc is doing with it.11:35
gnarfacewithout some logging enabled i could only speculate11:35
gnarfacebut stick around, there's people who mess with qemu a lot more than me11:36
m_mrsyslog is installed and logging well... gonna try to find how to resolve this don't worry ;)11:37
m_mI think I misconfigured OpenRC... so let's get back to start!12:01
gnarfacefor what it's worth, qemu should be perfectly compatible with the netinstall. there should be no reason you'd have to resort to debootstrap.13:12
blizzowI see a bunch of gimp-help packages for devuan but, no gimp?15:13
blizzow(excalibur)15:13
n4dirfor zynaddsubfx i have lost the directory /usr/share/zynaddsubfx, which contains the banks with the instruments.22:25
n4diri tried all kind of things, purge it, reinstall it, remove files manually, what not, but no reinstall will give me that directory back22:26
n4dirwhat was the way to reinstall as if it was never there again?22:26
gnarfaceyou did "apt-get --reinstall install zynaddsubfx" ?22:27
gnarfacethat should do it, as should just doing "apt-get --purge remove zynaddsubfx && apt-get install zynaddsubfx"22:27
n4diryup, i did all that.22:28
gnarfaceif it's not, check the pre/post-inst scripts and figure out what they're looking for that's changed22:28
gnarfacemight constitute a package bug, though this is the type of thing that often slips off the table22:28
n4dirthis installation is a bit of a mess, so i wouldn't consider a bug yet22:29
gnarfacejust keep in mind that even the purge action itself is subject to the maintainer writing a postremove script that's capable of actually finding and removing everything22:29
gnarfacestuff that auto-generates files after first run tends to get left behind22:30
n4diryeah, i used "locate" and all and removed everything which got in my way22:30
n4diraptitude purge ~c; what not22:30
n4dirwhere do i find the pre/post-inst scripts?22:30
n4diri think there was a argument or parameter for apt-get though, which kinda says "just reinstall the whole shebang, no matter what"22:31
fsmithredn4dir, see /var/lib/dpkg/info/22:33
n4dirthere is a bunch of zynadd, but no pre or post22:33
gnarfaceheh, well that explains it22:33
gnarfacewho needs a good postrm script anyway? you were just going to delete the guest and re-spin it from your Windows Azure host right?22:34
gnarface(brought to you by the same people who said filesystems don't need to be cleanly unmounted anymore because we have journaling!)22:35
fsmithredwhere can I find a list of current prescribed use cases so I can conform?22:35
gnarfacei don't see this package in main...22:35
n4dirin path-to/zynaddsubfx.list it lists everything which gets installed? If so, no /usr/share/zyn/banks there22:36
gnarfacehmm, weird, it's in pkginfo but my local searches don't show it...22:36
gnarfaceoh, there it is, must have typoed the first time22:36
n4diralways use program names easy to typo. rule number one22:37
gnarfacewell, i don't know what's actually going on so i'll stop being angry without evidence22:37
n4dirsorry for the noise. zynaddsubfx-data seems to have the banks.22:41
n4diri sure don't remember i ever had to do anything but install zynaddsubfx, but i am not too sure22:42
n4dirwell, it is a dependency of zynaddsubfx. Anwyway, case closed, thanks22:42
fsmithredyeah, I just found that22:43
fsmithredlol22:43
gnarfacejust had to purge more packages?22:43
n4diri mean i installed version 3, third party package, of zynaddsubfx, and that created the mess. Not a fault of devuan packaging22:44
n4dirgnarface: i did apt-cache search zyn, then found data, then reinstalled it. I guess it was not installed at all22:44
fsmithredbiab22:54
AlverstoneCan somebody please tell me why /etc/rcS.d/S07cryptdisks is placed right after /etc/rcS.d/S06cryptdisks-early? What is the point? They're identical sans some sysvinit specific configuration at the beginning, like X-Start-Before etc. What exactly is the point?23:38
gnarfaceAlverstone: i don't know, but check the LSB headers in both those scripts and it'll probably make more sensea23:43
gnarfaceor wait, did you do that and you're still asking? my guess would be that one is for drives that are needed for actually booting and accessing the others (like they contain vital partitions, or partition mountpoints that are in the second bank)23:44
gnarfacewhereas secondary storage can wait23:44
gnarface(the LSB headers can override the "S#" symlink start orders based on LSB headers in other scripts)23:45
AlverstoneAh, now that makes sense. To be more precise, it does not make any sense at all simply because runit does `for script in /etc/rcS.d/S* ; do`. Amazing.23:57
AlverstoneSo much for the least supported init system. Guess one day I'll end up reimplementing the wheel all over again.23:58

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