98[01:39:17] <doubletwist> Is there a way in dpkg or apt to
install a local file but only if it is newer? ie. an upgrade-only
option but where it will install if it's not installed at all?
147[02:39:07] <ld> I've configured dnsmasq and hostapd to
my interfaces specified in /etc/network/interfaces
148[02:39:24] <somiaj> I would look into nftables
149[02:40:07] <ld> but something prevents it from working... I
notice that running "systemctl stop networking" causes it
to begin working, partially
150[02:41:12] <ld> systemd is pretty vexing
151[02:42:05] <somiaj> Yea, I think it is more likely needing to
configure things properly, my understanding is networking isn't
this service that brings things up and down, it is more just a
service to start networking at boot.
152[02:42:35] <sussudio> ld: did you put net.ipv4.ip_forward=1
in /etc/sysctl.conf
153[02:42:38] <ld> that is how it describes itself too
156[02:44:08] <ld> most frustratingly although what I am doing
is not complicated at all and my installation is virtually all
defaults and no packages, it is impossible to succinctly and
precisely describe & inspect the system's state
157[02:44:53] <sussudio> i'd check the logs for any
messages
159[02:47:03] <somiaj> since stoping networking kinda works,
could it be that you need to not use the interfaces file to control
your interfaces and let dnsmasq/hostpad do this instead (I
don't know either, so unsure here)
160[02:47:07] <sussudio> also, if the interfaces are assigned to
the right ethernet ports
161[02:47:16] <ld> I am suspicious that it is at least one of
two things, (1) start order of "networking"
"dnsmasq" and "hostapd" services needs to be
controlled or changed, (2) "networking" somehow interferes
with operation of the other two
180[03:12:52] <ld> my answer is for "raspbian" which i
have no idea about
181[03:12:54] <miu5> Hi, when using gpg --keyserver
keyring.debian.org --send-keys 0x673A03E4C1DB921F, i get gpg: no
valid OpenPGP data found.. anyone help please?
182[03:13:09] <Casper26> ajoplin
183[03:13:18] <miu5> i follow instructions here
replaced-url
184[03:16:08] *** Quits: gabriel1__ (~darkangel@replaced-ip) (Remote host closed the connection)
197[03:24:32] <dpkg> If you have a question, just ask! For
example: "I have a problem with ___; I'm running Debian
version ___. When I try to do ___ I get the following output ___. I
expected it to do ___." Don't ask if you can ask, if
anyone uses it, or pick one person to ask. We're all
volunteers; make it easy for us to help you. If you don't get
an answer try a few hours later or on
replaced-url
199[03:24:50] <miu5> Hi, when using gpg --keyserver
keyring.debian.org --send-keys 0x673A03E4C1DB921F, i get gpg: no
valid OpenPGP data found.. anyone help please?
223[03:38:35] <mutante> miu5: you can check who else already
signed the key with --list-sigs . if that is a bunch of Debian
people then it's more trustworthy
343[06:38:38] <brizz> i have an issue with my terminal, in
python, when I press direction keys, [[A is output, and I can't
scroll up, or go right or left with the cursor. it seems to happen
after I paste with ctrl+shift+v. I have another server where this
does not happen. What could be causing this?
347[06:41:27] <oxek> and make sure your terminal correctly
identifies itself in $TERM
348[06:42:12] <brizz> even closing out terminal and re-opening
it stays broken
349[06:42:49] <brizz> it is installed
350[06:43:46] <brizz> whats weird is i can sudo (when it goes
broken), and it wont carry over to root. but im sure if i
ctrl+shift+v it will probably do it for root too
351[06:44:19] <oxek> terminal emulator are you using? What is
the value of $TERM? Are you using screen/tmux as well?
352[06:44:27] <oxek> are you on debian stable?
353[06:44:54] <brizz> yes on debian stable, using default
terminal, tho i tried 'FISH" tooo (i use fish shell)
354[06:45:08] <oxek> I don't know what "default
terminal" is
355[06:45:13] <brizz> xterm-256color
356[06:45:16] <brizz> gnome
357[06:45:20] <brizz> whatever is installed
358[06:46:28] <brizz> seems to only be doing it with python 3.9
term
359[06:46:34] <brizz> or interpreter
360[06:46:56] <brizz> 3.7 works too
361[06:47:29] <oxek> in that case, possibly an issue with the
python interpreter. Try asking whoever gave you python3.9, since it
is not provided by debian.
362[06:47:47] <brizz> i installed from source
363[06:48:03] <oxek> tell them that you're using debian
stable, gnome-terminal, $TERM is xterm-256color and you have
ncurses-term package installed.
379[07:32:18] <Arahael> brizz: I also use fish, and it's
unrelated to this issue - fish is not a terminal. It's a shell.
(And it's "fish", not "FISH" - unix systems
are case sensitive)
387[07:47:00] <somiaj> oxek: could also be a complication issue
in which their self complie didn't build with all the correct
libaries.
388[07:47:19] <somiaj> brizz: ^^
389[07:48:01] *** debhelper sets mode: +l 1015
390[07:48:33] <oxek> somiaj: yeah, that's what I meant but
expressed myself poorly
391[07:50:18] <nkuttler> zumba_addict: you link to the modules
from /etc/nginx/modules-enabled and restart nginx
392[07:50:31] <zumba_addict> it's good now
393[07:50:38] <zumba_addict> I had to install nginx-extras
394[07:51:00] <somiaj> zumba_addict: looks like if you have the
nginx-extras package installed you probably just need to add the
link to the module in /etc/nginx/mods-enabled (the module will be in
/usr/share/nginx/modules-available).
395[07:52:07] *** Quits: szorfein (~daggoth@replaced-ip) (Remote host closed the connection)
433[08:52:52] <jadelclemens> Hey all, I've searched all
over but haven't found a resolution to this. I'm trying to
mount an NFS share (Arch host) onto a Debian machine. I can mount
it, but for some reason my user cannot access the share ("Stale
file handle") even though the directory mode is sound (757).
434[08:53:13] <jadelclemens> Root can ls and touch files
reliably, only my user has this issue.
435[08:53:28] *** Quits: Nokaji (~Nokaji@replaced-ip) (Remote host closed the connection)
436[08:53:38] <jadelclemens> Standard advice is to unmount
share, restart the NFS service, and remount. I've had no such
luck.
437[08:58:15] *** Joins: mezzo (~mezzo@replaced-ip)
540[11:39:51] <TheBigK> for inotify to work properly, does the
file need be owned by the specific user? im using minidlna on my
raspi as the minidlna user. I added the user to my group for media
and when i switch to that user and try to read those files it works.
But when something is added. The deamon logs "Could not
access" and then the path of the new added directory
(permission denied)
541[11:40:23] <TheBigK> Do i actually need to change the user of
minidlna? That would be quite annoying since i need to change the
permission of everything... /run /var/log etc
542[11:41:08] <TheBigK> just inotify doesnt work properly yet
545[11:51:49] *** Quits: ChubaDuba (~ChubaDuba@replaced-ip) (Remote host closed the connection)
546[11:51:51] <jelly> TheBigK, uh, if you want the dlna server
to serve files, of course it has to be able to read them (read
perms) and access the directory (read and executable perms)
576[12:11:26] <TheBigK> jelly: i might give that a shot... but
read permissions the minidlna has.. but somehow the inotify doesnt
work ... when i restart minidlna to rebuild the db .... the new
added directories gets added and i can also access it
709[15:33:10] *** Quits: JohnML (~john1@replaced-ip) (Remote host closed the connection)
710[15:34:21] <petn-randall> ska: "stable" there means
the same as "stable" in Debian: The latest stable release,
which currently is buster, but will point to bullseye in a couple of
months.
730[16:02:18] <dostoyevsky> I am trying to build a debian-buster
image in a dockerfile but the build hangs at "apt install -y
--reinstall coreutils" ... since like 2000s; htop isn't
showing any cpu activity...
731[16:02:43] <oxek> dostoyevsky: I have the same issue
732[16:02:48] <oxek> it's a networking issue
733[16:02:51] <oxek> don't know how to solve it
734[16:03:01] <dostoyevsky> oxek: debian server problem?
735[16:03:09] <oxek> no
736[16:03:13] <oxek> docker problem
737[16:03:19] <oxek> meaning the docker version in debian
738[16:03:54] <oxek> if you restart the dockerfile build several
times, it will work one of those times
739[16:03:59] <oxek> I suspect some sort of race condition
752[16:18:48] *** Quits: cosimone (~user@replaced-ip) (Remote host closed the connection)
753[16:19:13] <Walex2> dostoyevsky: the difficulty there I
suspect is that 'coreutils' is "essential", so
other packages like 'apt' may depend on it without
declaring it, and complications may happen.
768[16:26:39] <jelly> These tags are an experiment in providing
a slimmer base (removing some extra files that are normally not
necessary within containers, such as man pages and documentation),
and are definitely subject to change.
775[16:33:10] <dostoyevsky> well, I use debian:buster now as the
FROM and removed the --reinstall and enabling of the man pages and
now the container is building again without hanging so far
776[16:33:48] <dostoyevsky> but the container built fine in the
last months... without hanging
847[18:13:31] <petn-randall> dostoyevsky: If you haven't
debugged the issue, you can't really blame it on a distro.
Correlation doesn't imply causation.
848[18:13:52] <petn-randall> dostoyevsky: ... which is why I
specifically asked for debug output.
858[18:28:22] <ValeraRozuvan> Is there an official guide on how
to make contrib and non-free packages (especially drivers) available
during a Debian install?
861[18:29:48] <somiaj> ValeraRozuvan: sounds like you need
firmware, easiest way is to use the firmware instller
862[18:29:51] <somiaj> !firmware images
863[18:29:51] <dpkg> There are <live> system and
<netinst> and DVD images containing non-free Debian
<firmware> packages available from
replaced-url
864[18:30:14] <somiaj> Otherwise read the install guide, there
are instructions to make firmware available during the install on a
second drive.
865[18:30:38] <ValeraRozuvan> And in the name
"unofficial" simply means that it's built be official
Debian maintainers but not officially endorsed by Debian?
867[18:31:07] <ValeraRozuvan> In the URL ...
cdimage.debian.org/cdimage/unofficial/...
868[18:31:15] *** Quits: nicopok (~nicopok@replaced-ip) (Remote host closed the connection)
869[18:31:22] <oxek> it means it's not part of debian
870[18:32:04] <oxek> and since it is not officially part of
debian, it doesn't strictly matter who built it
871[18:32:58] <oxek> there are no guarantees that it
doesn't contain rootkits, keyloggers, and all sorts of other
malware
872[18:33:03] <somiaj> ValeraRozuvan: the factoid stated this,
it just means debian's offical releases must be 100% free,
these are provdied for those who choose to want non-free firmware
with the understanding that they are non-free
874[18:34:22] <ValeraRozuvan> OK ... it's a bad situation
that you need non-free stuff in order to use my current hardware ...
875[18:34:26] <somiaj> ValeraRozuvan: But if you are going to
choose to install contrib and non-free packages provided by debian,
this installer is provided in similar a vain.
876[18:34:51] <ValeraRozuvan> For example my WiFi. I guess I
have to research and buy a system which is 100% supported by pure
Debian.
877[18:34:57] <somiaj> ValeraRozuvan: It is unfourntally too
common with certain hardawre (wifi is a big one and gpus second)
that need non-free firmware.
878[18:35:55] <somiaj> ValeraRozuvan: Here it is up to personal
choice, but it is hard to find 100% free wifi (note even wifi that
doens't provide userland firmware often has non-free firmware
already flashed onto it, there are a very limited number of chipsets
that allow you to flash free firmware onto them)
879[18:36:17] <ValeraRozuvan> One solution - don't use WiFi
;)
880[18:36:19] <somiaj> ValeraRozuvan: so though userland
firmware may have some additional trust, lots of hardware already
comes with non-free firwmare flashed onto it
881[18:36:35] <petn-randall> oxek: Eh, no? non-free doesn't
mean Debian gives a free pass on malware. It just means it's
not DFSG software, but is still redistributable.
882[18:36:45] <somiaj> There are sites out there for fully libre
machines that you can flash 100% open firmware onto them, but these
are very limited set of hardware.
883[18:37:04] <ValeraRozuvan> By the way - are the new Raspberry
Pi completely open source? I.e. 100% pure Debian will work out of
the box?
884[18:37:39] <oxek> petn-randall: non-free firmware by
definition could contain anything, including malware
885[18:38:05] <oxek> it means debian would be supplying a black
box that nobody inspected, and hence it cannot be official
887[18:38:36] <somiaj> ValeraRozuvan: Again you might want to
clarify what that means, there will still be non-free firmware
flashed onto some of the chips it uses, so just because you
don't need userland non-free firwmare doesn't mean you
have 100% free hardware.
890[18:39:28] <ValeraRozuvan> I am looking at Google results -
can't find a concrete resource which will list parts for a PC
that will be covered 100% by pure Debian install.
891[18:39:33] <ValeraRozuvan> Any suggestions?
892[18:39:34] <petn-randall> ValeraRozuvan: The RPis are not
open source. But you can boot the RPi4 with stock Debian, AFAIR.
894[18:40:11] <petn-randall> oxek: What about non-free software
that also ships the source?
895[18:40:19] <somiaj> ValeraRozuvan: I personally prefer
userland firmware over having to flash new firmware on hardware if
there are bugs. But yes you do have to put trust in the manfacture
of the hardware/firmware.
896[18:40:43] <oxek> petn-randall: different meanings of
non-free. For firmware, it's not only not dfsg compliant, but
also closed source.
897[18:40:43] <jhutchins> ValeraRozuvan: Who would create such a
list?
903[18:42:18] <somiaj> ValeraRozuvan: Wifi is the biggest block
if you don't want to use non-free firmware. There are a very
limited set of wifi devices that don't have userland firmware
(though most will still have non-free firmware flashed on them). I
belive there are a few atheos chips you can flash purely free
firmware on, but these are fairly rare.
905[18:42:47] <ksynwa> Where can I find what kernel versions are
available in Debian? Was wondering if Testing/Unstable have newer
kernel versions. Thanks.
908[18:43:26] <somiaj> ValeraRozuvan: Again here it is personal
choice, but as for say intel's wifi, they ship additional
userland firmware vs having it flashed onto the device, this allows
easier updates. So if you trust intel, you can decide if you want to
trust their firmware or not. Though as you said, not using wifi is
an easy solution.
913[18:44:07] <somiaj> ksynwa: test/unstable/backports will
often have fairly recent kernels, though due to the freeze they
won't be that many new kernels until after the release of
bullseye.
914[18:44:37] <somiaj> ksynwa: and as oxek listed the current
status. Are you wanting a newer kenerl on stable?
915[18:46:08] <ksynwa> somiaj: Bullseye is the version after
Buster I assume?
916[18:46:20] <ksynwa> Thanks very much oxek somiaj
917[18:46:29] <somiaj> ksynwa: yes, and will be released in the
next month or two from what I hear
918[18:46:52] <somiaj> ksynwa: but you can use the backports
kernel on stable if you want to wait for the release but still need
a newer kernel
921[18:48:04] <ksynwa> somiaj: I am still looking into
stable/testing/unstable but I have no problem using testing or
unstable.
922[18:48:25] *** Quits: towo` (~towo@replaced-ip) (Remote host closed the connection)
923[18:48:34] <ksynwa> From what I am reading they are pretty
stable so I am open to them
924[18:48:42] <petn-randall> oxek: Sure, but we were talking
about non-free software in Debian. Which is not DFSG compliant. And
Debian wouldn't ship non-free software with known malware.
925[18:50:30] <somiaj> ksynwa: many do use them, though they do
come with additional issues. Things can and will break, testing gets
slow security support. I would personally suggest waiting for
bullseye's release, but if you can deal with slow security
support, its in a fairly frozen state right now, and you could
install it, and just keep it upgraded, when it releases you'll
be on the next stable release.
926[18:50:43] <somiaj> ksynwa: though if you just need a newer
kernel in stable, the backports kernel is the route I would go.
934[18:52:19] <somiaj> DJAnonimo: debsums is one such tool that
checks installed files from debian packages
935[18:53:08] <oxek> petn-randall: closed source firmware can
have unknown malware, and debian still unofficially distributes it
936[18:53:44] <jhutchins> DJAnonimo: Why do you suspect
corruption?
937[18:53:49] <oxek> we're going in circles anyway. The
point is that the 'unofficial' images that contain closed
source firmware are not official releases of debian.
938[18:54:56] <oxek> and official releases of debian have at
least some sort of guarantee to be malware free
953[19:04:05] <ValeraRozuvan> Should we really care about closed
source firmware - if Intel has the ME built in with closed source OS
running in ring -3?
965[19:12:07] <jhutchins> somiaj: Would you say closed firmware
is mostly wifi and video?
966[19:13:10] <somiaj> jhutchins: that is my experience, and
wifi is that way due to laws about wireless transmissions (compaines
are required to have more control over the hardware and not let
users go outside of power ranges/frequencies)
967[19:13:34] <somiaj> gpus are just because well high end gpus
are a cash cow and their creaters keep everything closed
968[19:14:19] *** Quits: fflori (~fflori@replaced-ip) (Remote host closed the connection)
969[19:14:55] <somiaj> even openwrt has issues with this,
basically they have a very strong stance and just will not help or
provide any information on how to adjust the frequencies/power of
the routers to avoid leagal issues with giving open firmware for
those.
971[19:17:01] <somiaj> intel use to be a decent gpu if you
didn't care too much about 3d gaming style stuff, but even now
their newer cards require firmware (though at least use open
drivers)
976[19:19:26] <mutante> Oh, i had never heard that reason for
wifi needing nonfree.. TIL
977[19:19:30] <somiaj> ValeraRozuvan: yea, and given the
standard practices, it is at least a decen compermise, open drivers
in the kernel and userland firmware to make it so updates are easier
than flashign the hardware.
978[19:19:36] <ValeraRozuvan> Like - if you look at MINIX - it
suffers due to lack of drivers & firmware - not gaining a large
user base because of this. Linux on the other hand - getting Love
from Big Tech.
979[19:20:32] <somiaj> mutante: that is my understanding why
there is so few wifi cards that dont' require non-free
firmware, there are lots of restrictions on wireless communications
(at least in the usa, though assume other countries do have some
laws about this)
980[19:21:06] <mutante> interesting, ack
981[19:21:12] <ValeraRozuvan> Anyways - Debian is awesome!
Started with it back in 2005-2006. Never looked back.
982[19:21:52] <somiaj> mutante: this even dates back to pre wifi
and hamm radio days, even those are quite restrictive.
984[19:25:20] <somiaj> ValeraRozuvan: and though there is this
non-free distinction, it is better than when the linux kernel us to
just ship the firmware directly, as opposed to having it as a
separate thing so users were clear that they are getting non-free
firmware and have to put their trust in the company that provides
it.
988[19:26:57] <somiaj> But that is one of the things I like
about debian, is it is at least clear on the distinction between
free and non-free, and gives users the choice and control of what
they choose.
990[19:28:31] <jhutchins> somiaj: I believe that's the
reason. I think the wifi chipsets that don't require firmware
have the code embedded in the hardware.
999[19:33:25] <ValeraRozuvan> Linux kernel has some
diferentiation between hardware drivers which don't require
firmware code to load, and hardware drivers which require firmware.
1000[19:35:00] <jhutchins> ValeraRozuvan: Like I said, the ones
that don't require external firmware have the code imbedded,
and like somaj said that makes updates difficult.
1002[19:35:22] <jhutchins> I've never re-flashed a wifi or
graphics card.
1003[19:40:52] <somiaj> jhutchins: I have heard of one version of
atheros chips you can flash open firmware on, but I to have the same
understanding, just because it doens't require userland
non-free firmware doesn't mean there isn't non-free
firmware in use.
1005[19:44:56] <Synaptic> Hello, i need help if possible:
i have a debian dedicated server and a debian vps,
the dedicated server is a storage server, i would like
to configure MFS in order to share the drive on the VPS, i
alredy installed it, and setup, but from VPS i cannot mount
it... i can ping the dedicate, i nmap it and i see the
port open, and i also
1006[19:44:57] <Synaptic> try to telnet and i connect to the MSF
port.
1062[20:15:59] <ndroftheline> i'd like some help with root
on an mdadm device
1063[20:16:17] <ndroftheline> for some reason i just can't
get it to behave
1064[20:17:22] <ndroftheline> during the install, everything
looks great. but first boot, it always comes up from one of the
member drives instead of the mdadm volume
1065[20:18:07] <ndroftheline> once booted i can just mdadm -As
and the volume shows up
1066[20:18:22] <ndroftheline> but of course it's degraded
because i'm running off one of the members directly :(
1070[20:20:51] <ndroftheline> it's a supermicro machine
i'm trying to set up as a zfs backup system. it's got two
m.2 SSDs connected to its onboard controller. the mdadm volume is an
imsm (created by the onboard intel rste controller). i want to keep
it as an intel rste volume for consistency with other machines.
1071[20:21:22] <ndroftheline> i've tried to follow the steps
at
replaced-url
1078[20:26:59] <ndroftheline> seems to suggest the system has to
be already installed, in order to start - ie, "install package
mdadm" suggests it's already a running system. then once
the last thing is suggested, it says "boot into Rescue mode,
investigate" - as if the last step is boot to rescue which
i'm confused by.
1079[20:26:59] *** Quits: r3 (~r3@replaced-ip) (Quit: Je dois partir maintenant parce-que ma grand-mère
est flambée.)
1080[20:27:20] <ndroftheline> hints welcome :)
1081[20:29:03] *** Quits: ax56234 (~NickServ@replaced-ip) (Remote host closed the connection)
1082[20:30:06] <jmcnaught> ndroftheline: are you configuring RAID
in your EFI/BIOS or in the debian installer?
1096[20:33:50] <jmcnaught> ndroftheline: if the computer has a
hardware RAID controller then that would be the way to go, but if
not then I would set the SATA drives to AHCI and use Linux's
software RAID which can be configured in the installer.
1098[20:34:04] <jelly> ndroftheline, pastebin your current
/etc/mdadm/mdadm.conf, output of lsblk, and output of
"debconf-get-selections|grep mdadm/initrdstart" if you
have debconf-utils installed
1114[20:40:48] <jelly> roughly it sounds like initrd support for
md doesn't run or doesn't recognize the initialized array
1115[20:41:21] <jelly> no separate /boot?
1116[20:41:36] * jelly waits for other pastes
1117[20:41:49] <dostoyevsky> Ok, I made an image of the docker
container before it hung, and then ran `apt -y
-oDebug::pkgAcquire::Worker=1 update' to show where it hangs:
replaced-url
1118[20:42:16] <dostoyevsky> Last line is: 0% [Waiting for
headers] <-
gpgv:201%20URI%20Done%0aGPGVOutput:%20GOODSIG%2004EE7237B7D453EC%0a%20GOODSIG%20648ACFD622F3D138%
1135[20:45:39] <jelly> ndroftheline, okay, so first thing, mdadm
package is not configured to bring up any md array at boot
1136[20:45:51] <dostoyevsky> jelly: Should I open a ticket with
debian that `apt -y -oDebug::pkgAcquire::Worker=1 update' that
their debug output is unreadable?
1137[20:46:24] <jelly> dostoyevsky, show the output of "apt
update" without any other parameters, for starters
1139[20:47:58] <jelly> ndroftheline, it is unfortunate that this
recovery shell is ALSO booted mounting from a single member
apparently
1140[20:48:17] <ndroftheline> yes. i can get into a rescue shell
off install media if that's better
1141[20:48:18] <jelly> ndroftheline, can you boot a live image of
some sort instead?
1142[20:48:35] <jelly> not the installer, it's a crappy
rescue
1143[20:48:44] <jelly> grml.org or debian live
1144[20:48:48] <ndroftheline> well. hm. i actually haven't
successfully booted the "debian live" ever on this system.
it's like it doesn't support uefi?
1188[21:17:48] <ndroftheline> i want to nuke the drives incase
all the screwing around i've been doing has put confusing data
somewhere. recommended ways to reliably clear the member disks of
relevant data?
1189[21:18:49] <ndroftheline> oh, i missed that - grml assembled
the array as a dmraid instead of using mdadm :o
1190[21:22:47] *** Quits: Mister00X (quassel@replaced-ip) (Quit: "I'll be back" — Arnold
Schwarzenegger)
1208[21:43:56] <urk> Is there anyone on here who figured out how
to get the wifi working with kernel 4.19, and a Dell XPS 15 7590?
Previously I was led to believe that wifi would only work with
kernel 5.10 and above.
1216[21:48:57] *** Quits: tangarora_ (~tangarora@replaced-ip) (Remote host closed the connection)
1217[21:49:37] <ndroftheline> so now i'm back into the
installer at the partition disks page. i've chosen to use an
entire disk. the first option on the list of disks to partition
reads "RAID1 device #126 - 950.2 GB Software RAID Device"
which is what it's shown in the past.
1218[21:50:13] <ndroftheline> dropping to busyboxy, mdadm
--detail /dev/md126 indicates the array is currently doing a
resync...wut
1220[21:51:50] <ndroftheline> anybody know if a sataraid is
supposed to start life by doing a resync? maybe this is the
initialisation step, which mdadm automatically starts when first
activating the array?
1221[21:53:33] <ndroftheline> i think i'll let it finish its
initialization before proceeding with the installation...
1268[22:30:53] *** Quits: Gerowen (~Gerowen@replaced-ip) (Remote host closed the connection)
1269[22:31:14] <somiaj> all init.d scripts should be converted to
a systemd unit file. Is this a script provided by you or a debian
package?
1270[22:31:30] <jmcnaught> unixbsd: a quick look at
hostapd's file list shows it ships a native systemd
hostapd.service, as well as an instanced hostapd@.service unit
1271[22:31:46] <greycat> There are still Debian packages using
init.d files. exim4 is one.
1272[22:32:25] <somiaj> I meant systemd will convert/run init.d
scripts, not that packages need to have done this.
1273[22:32:45] <jmcnaught> unixbsd: check
/usr/share/doc/hostapd/README.Debian it mentions needing to unmask
hostapd
1281[22:40:04] <somiaj> unixbsd: note, it is useful to check
/usr/share/doc/packagename to see info about stuff that might need
to be done after installing a package, there is often useful info
there.
1282[22:40:11] <somiaj> just for future reference.
1294[22:49:47] <unixbsd> all done, I have made a script to
install and transfrom the debian machine into a debian router
DEBIANROUTER, automatically, :
replaced-url
1301[22:54:48] <urk> jmcnaught: I didn't find 5.10 to be
that stable, and it crashed a couple of times after running the
chmod command.
1302[22:55:28] <jmcnaught> urk: chmod can nuke a system if you
are not careful.
1303[22:55:34] <urk> jmcnaught: I am going to wait until the new
installer comes out, and hopefully 5.10 will be in that. For now,
4.19 seems to be pretty solid.
1304[22:55:58] <urk> jmcnaught: I was careful and it nuked it
anyway.
1305[22:56:08] <somiaj> urk: The 5.10 kernel installer is already
out, and there isn't going to be too many changes between now
and release time.
1306[22:56:09] <jmcnaught> urk: 4.19 is solid, but it's
never going to support your wifi
1315[22:58:01] <unixbsd> but amd64 is not, pae = ok
1316[22:58:28] *** Quits: tangarora (~tangarora@replaced-ip) (Remote host closed the connection)
1317[22:58:28] <somiaj> urk: yes, and the rc1 installer is out
and ready to use, I used it just fine on my new system (there are
still some installer bugs, but I didn't run into them)
1318[22:59:06] *** Quits: DJAnonimo (~senior@replaced-ip) (Remote host closed the connection)
1319[22:59:21] <urk> somiaj: Yeah, I took note of the installer.
I am going to wait just a few more days before using it. Maybe I am
paranoid, but can't afford any more downtime at work. 4.19 is
rock solid.
1326[23:03:19] <urk> somiaj: I picked up an adapter from a local
computer store, and running wifi off of that and Cat 8 cable.
However, its a nuisance because the cable isn't long enough to
get to my kitchen table so I ended up making a make-shift table from
boxes.
1328[23:04:33] <somiaj> here I was suggesting a wireless one, but
yea that will be the concession you'll have to make if you want
to stick with buster and 4.19 for a while.
1329[23:05:22] <longshot> What happened to rc.local? Is there an
easy way to do something like that these days?
1330[23:05:54] <greycat> !rc.local
1331[23:05:55] <dpkg> /etc/rc.local may be used to run simple
commands at boot time. It exists by default in jessie or older; in
stretch or newer you need to create it. Don't forget the
<shebang> and be sure to chmod 755 it. rc.local is considered
a hack, a stopgap, or a temporary band-aid; see <systemd>
1334[23:06:18] <somiaj> longshot: it is still there, though it
has always been a hack, it has always been better to write a proper
init file, so now a systemd unit file.
1335[23:06:27] <greycat> Support for it is still there, even if
the file isn't yet.
1336[23:06:49] <petn-randall> longshot: TBH I'd make a
systemd unit out of it. Much more robust than rc.local, as it
won't block booting when you do something wrong.
1337[23:07:11] <somiaj> longshot: If you let us know what you
need to run at startup, we maybe albe to better help point you in
the right direction.
1338[23:07:47] <longshot> somiaj: It's just some debugging
commands, stuff I'll throw away quickly.
1341[23:08:21] <ndroftheline> resync complete...now i'll try
to let the installer do its thing on the software raid volume...
1342[23:09:05] <ndroftheline> using defaults, with all data in
one partition.
1343[23:09:22] <longshot> Yeah, but not because anything has gone
wrong or anything. I'm working on a kind of appliance type
system and just need to see what the state is at certain points
1345[23:10:01] <longshot> I know rc.local runs later in boot, but
it's much earlier than me being able to log in
1346[23:10:03] <somiaj> longshot: systemd has some nice tools to
help you anaylize startup and its units.
1347[23:10:24] <longshot> That might be helpful as well
1348[23:10:28] <somiaj> longshot: might be worth looking into
some of those tools, but either way, rc.local is there, just create
it
1349[23:10:44] <dob1> sed 's/\s+|$/|/g' file this shoud
remove extra spaces before | or all the lines that ends with |
right?
1350[23:10:49] <dob1> *should
1351[23:11:39] <somiaj> longshot: systemd-analyze might be a good
starting point
1352[23:12:57] <greycat> dob1: no, because \s and + are not
supported by sed in its default mode.
1353[23:13:13] *** Quits: _till_ (~till@replaced-ip) (Remote host closed the connection)
1354[23:13:17] <greycat> well, maybe \s is, whothehellknows,
it's GNU crap
1355[23:13:28] <greycat> but + is an ERE thing, not a BRE thing
1356[23:13:54] <dob1> I have to use -E ?
1357[23:14:43] <dob1> no
1358[23:14:56] <greycat> you literally used the word
"spaces" as well, so you could simply use a space
character instead of \s which is GNUperl for "more than just
space"
1359[23:15:26] <greycat> sed 's/ *|$/|/' and you
don't need the g either because you can never match the end of
line more than once
1373[23:34:25] <urk> unixbsd: How come I don't get a
notification of upgrade from 4.19? I run update, and upgrade
frequently, including distribution upgrade.
1374[23:35:00] <somiaj> stable doesn't get upgrades that
often.
1375[23:35:05] <urk> unixbsd: The link you provided is from
Duvian, and not Debian.
1379[23:36:29] <jmcnaught> I would not go using random kernels
provided by strangers. That link doesn't even provide the
kernel in package form, just some loose files and no modules?
1396[23:51:17] <urk> jhutchins: For now I am inclined to just use
4.19 until a few more bugs are removed from the new installer. 4.19
works great and is super solid even though I have no wifi support.
1397[23:51:52] <jhutchins> urk: Well, we could work on that.
1398[23:52:29] <jhutchins> We've seen a fair number of
problems with the 5.10 backports.
1400[23:53:36] <urk> I like the Dell XPS 15 7590. Its super
bright, great case, great keyboard, runs Linux like a champ, but the
battery could be better, and doesn't seem to be too durable.
Also, the color gamut/quality could be better even though the laptop
is super bright.
1401[23:53:54] <urk> too much blue light in the screen.