this is #debianan IRC-Channel at freenode (freenode IRC service closed 2021-06-01)
0[00:00:09] <sney> as long as there aren't missing build-deps it's usually pretty smooth, and devscripts will give you informative output when it doesn't
1[00:00:43] <sney> it can't be as automatic as installing binaries with apt, of course. but as long as you read what it says on the screen, success is likely.
24[00:14:48] <lifostack> hi all i cant get my laptops nvme ssd card to recognize in debian live. im trying to recover some data off a torched windows install
25[00:15:01] <lifostack> i've tried setting the SATA bios setting to AHCI no dice there
26[00:15:13] <lifostack> can anyone offer a suggestion what else to try? i'm getting a little frustrated and tired out
27[00:15:15] *** Quits: koniu (~koniu@replaced-ip) (Remote host closed the connection)
40[00:19:20] <lifostack> i've tried multiple different live environments to try and get this to work
41[00:19:45] <sney> lifostack: there are debian testing live images, that may or may not work. grml might be more capable (often is) for rescue stuff as well, and they are currently shipping a 5.6 kernel
42[00:20:04] <lifostack> thanks im not familiar with this tool
46[00:20:36] <ratrace> what I hate most is when people actively LIE they're using debian, so they could get help with kali..... or some other garbage derivative
47[00:20:48] <lifostack> lol thats dumb
48[00:20:54] <lifostack> my laptop at home does run debian buster
49[00:21:00] <lifostack> i also run a raspbian running pihole
50[00:21:09] <ratrace> something also runs windows at home, so why not ask in #windows
51[00:21:28] <sney> or the kids who think they have a new justification that they should be exempt from the debian-only rule, that nobody has ever thought of before
52[00:21:50] <lifostack> thanks for the grml recommendation
53[00:22:11] <brigand> sney: lol. but it could be a kali specific problem
54[00:22:27] <sney> indeed
55[00:23:28] <sney> (the people asking about kali in here are usually trying to use it as a general purpose OS, and asking for help with fundamental "how 2 wifi" stuff, so they also don't get much traction in the kali channel and generally have a bad time)
70[00:26:49] <lifostack> well i dont know if mint was out when i first started getting my feet wet with linux and i arrived late to mint anyways, but in all honesty most people using mint should prob just install debian from the get go
71[00:26:58] <lifostack> especially if you're planning on learning linux as a sys admin
72[00:27:27] <lifostack> all the training is under debian or fedora.... dpkg or w/e the redhat ones called
73[00:27:57] <brigand> rpm?
74[00:28:05] <lifostack> yeah i think thats the one
75[00:28:16] <brigand> gotta go to bed. good night
133[01:05:53] <Vatum> should i install first linux-headers, and then linux-image? i tried installing linux-image first and got some errors, from what i understand talking about missing headers
134[01:06:01] <sney> oxek: I assume the reason it's not on the man page is because there's just *so much* to document about aptitude. but if someone filed a bug that included a patch for the manpage to include this info specifically, it might be accepted
155[01:10:27] <Vatum> i mean everything is loaded in ram. worst case scenario, power outage and unbootable system, which i can fix with a usb live system and chroot
217[02:01:19] <sney> scroll down to the bottom, there is a download link under 'all'. put the resulting .deb on a usb drive, install locally on your debian system, and reboot.
218[02:01:58] <jennie> okay one minute, almost done
220[02:03:04] <runciter> i have an unstable installation with an incompatible set of packages installed that's preventing me from installing a compatible set. i want to uninstall everything and reinstall stuff from a saved selections file but keep most of /etc - is there a way to do this that doesn't involve debootstrapping from a rescue image?
221[02:03:40] <sney> runciter: as you can see from the topic, unstable support is in #debian-next on OFTC (note, right now you are on freenode)
222[02:03:45] <runciter> oops
223[02:03:47] <runciter> sney, thanks!
224[02:03:58] <sney> np
225[02:03:58] <runciter> having said that, this isn't really an unstable question
226[02:04:29] <runciter> i'm mostly curious if there's a better way to blow away all the installed packages and readd them by hand than booting into a rescue image and bootstrapping by hand
227[02:04:40] <sney> please respect the debian community's decision to split support by debian version in use, even if *you* think it doesn't matter what channel you are in.
228[02:05:02] <jennie> mount command for usb is giving error - " unknown filesystem type 'exfat; "
229[02:05:08] <runciter> sney, pretend i said stable
230[02:05:58] <runciter> it is an X Y problem - the X was something to do with unstable, the Y is "how do i hand bootstrap a debian system" - and i think i have my answer
231[02:06:14] <ratrace> so the question IS unstable after all. :)
232[02:06:30] <runciter> no, x y problem is the users asks x but actually wants y
233[02:06:41] <ratrace> and that makes it unstable. flips from x to y
234[02:06:51] <jennie> i cant mount this USB stick :( replaced-url
244[02:16:26] <TTT> do I need amdgpu-pro to get OpenCL running on Radeon 5700XT ?
245[02:16:34] <sney> !crosspost
246[02:16:34] <dpkg> Posting the same question in several places at the same time (IRC channels, news groups, mailing lists, forums) is impolite; your time is NOT more valuable than everyone else's. Your question might be answered elsewhere, meanwhile we are wasting our time doing research for a problem you've already solved. Cross-posting can also make you look like a spammer and get you k:lined. See also <multiple ask> <hurry>.
255[02:24:30] <jennie> sney installation done, whats next step please
256[02:24:32] *** Quits: dbristow (~dbristow@replaced-ip) (Remote host closed the connection)
257[02:25:09] *** Quits: rander2 (~rand@replaced-ip) (Remote host closed the connection)
258[02:25:40] <sney> jennie: you installed the package? now you reboot, and check 'lsmod |grep iwlwifi'. if the driver shows, then you can proceed with configuring your network
259[02:26:02] <jennie> yes, installed the package, done.
276[02:39:11] <mrjpaxton[m]> Hmm... considering that Brave already packages a .deb repo, I would hope to see it in the next stable release, too. You can still use the unofficial repo though.
277[02:39:14] <karlpinc> !wnpp
278[02:39:15] <dpkg> For information on packages which are not in Debian but you think should be, check the Work-Needing and Prospective Packages list at replaced-url
281[02:42:44] <lru> that's quite a list of software lol, thanks
282[02:42:52] <sney> there is this from a few years ago, but it looks like nobody is interested in packaging it, instead referring to the upstream repo replaced-url
283[02:43:23] <sney> if you want brave-browser in debian, and you are familiar with the software and have time, why not package it yourself?
515[09:09:02] <toogley> hey. is there something like fdisk on the debian installer image?
516[09:10:03] <toogley> like before installing debian, i want to destroy the partition table to prevent debian from being confused, it tries to recognize a broken debian installation
526[09:26:32] *** Joins: mezzo (~mezzo@replaced-ip)
527[09:26:44] <archer86> hello guys I have a Debian 10 buster up and the php version is showing up as 7.3.19 when php.net shows it should be 7.3.26 is this correct?
528[09:26:58] <archer86> I can post my /etc/apt/sources.list if you need me too
529[09:27:09] <brigand> ,v php
530[09:27:58] <brigand> ,v php7.3
531[09:28:00] *** Quits: koniu (~koniu@replaced-ip) (Remote host closed the connection)
535[09:29:06] <brigand> judd. come back we love you
536[09:29:31] <archer86> apt list | grep php shows me php/stable,stable 2:7.3+69 all
537[09:29:46] <sney> changelog for buster-security php7.3, replaced-url
538[09:30:22] <sney> the package versions chosen by debian for a stable release don't always match with upstream's decisions for that series. usually due to time.
539[09:30:35] <archer86> ok so 7.3.19 seems to be the latest that is used in debian I would guess they do what CentOS does and backport fixes
540[09:30:42] <sney> yep, as needed
541[09:31:13] *** Quits: buzel (~buzel@replaced-ip) (Remote host closed the connection)
542[09:31:28] <archer86> stable is always the most secure in terms of getting hacked but might not have the latest features?
543[09:31:43] <sney> that's the idea.
544[09:31:56] <archer86> just like CentOS I do like security over say the extra bells and whistles
545[09:32:08] <sney> or, not necessarily "in terms of getting hacked" but more like centos, where it's always a predictable platform
546[09:32:49] <archer86> I actually recently dumped CentOS 6 and moved over to Debian 10 so I guess I caved in and now I am a user of potterings SystemD
548[09:33:09] <archer86> I think redhat was getting rid of CentOS anyways
549[09:33:26] <sney> redhat did some shuffling recently. for small time users it's more of a branding change than anything.
550[09:33:58] <sney> but anyway, let us know if you have more debian questions
551[09:34:06] <archer86> my very first linux distro was Redhat 7 back in the 90s so that made me tear up a bit
552[09:34:33] <archer86> I will be fine now though Debian has already been fun to use as well
553[09:35:16] <archer86> Ubuntu focuses more on Gaming and getting drivers more quickly to the Gamer but as this is just for a webserver I hardly see the need for the "bloat"
744[11:16:12] <HelloShitty> Hello. I have a question about rsync. I'm trying to keep 2 folders, in the same physical hard drive, to be in sync, meaning to have the same content over time. Can I do this with rsync, by just copying the difference in the files?
745[11:16:42] <HelloShitty> Or is this only possible over a network? Isn't this called the delta diff or elta copy or something like that?
762[11:26:43] <Lope> ratrace, hey bud, have you got any special tricks for passing thru a block device to a systemd-nspawn container? like is it possible to do as a parameter inline?
763[11:26:48] <brigand> if it's not for backup purposes symlink is the way to go, like nkuttler suggested
764[11:27:01] <ratrace> Lope: ah, Lope! Was waiting to find you again here. I did tests! Lots of tests!
766[11:27:13] <ratrace> Bottom line, systemd-nspawn sucks too. just.... doesn't work as expected.
767[11:27:32] <Lope> ratrace, great!, but no, that's just sad.
768[11:27:42] <Lope> Did I post a link for you for passing block devices thru?
769[11:28:12] <Lope> ratrace, I found this replaced-url
770[11:28:17] <HelloShitty> nkuttler: I only want it done once. because 2 different prograams will access the folder and might write down different data on them, but initially, I need them to be exactly the same.
771[11:28:20] <ratrace> this works: cd /mnt; mkdir foo; zpool import -R /Mnt/foo somepool; now this: unshare --mount ; mount -t proc proc foo/proc ; mount -t sysfs none foo/sys ; mount -R /dev foo/dev ; chroot foo source /etc/profile ; export PS1="(chroot) $PS1" and voila!
772[11:28:25] <ratrace> Lope: ^^^^
773[11:28:42] <ratrace> Lope: how do you tear it down? exit (chroot) then exit again to exit the namespace from unshare. NOTHING TO UMOUNT :))) win.
775[11:29:07] <ratrace> Lope: the key is to mount proc, sysfs and bind-mount dev, AFTER the unshare.
776[11:29:23] <ratrace> Lope: infact, you can also mount /boot and other things and exiting the namespace shell will tear those too
777[11:29:47] <Lope> ratrace, amazing!!!
778[11:29:49] <Lope> THANK YOU
779[11:30:00] <Lope> wow, so stoked to have an answer of how to do it without destroying my host.
780[11:30:15] <ratrace> yup. Host survives, no umount issues, no blockages, no funny business inside the chroot. grub works, initramfs-update works, cryptsetup works (with all its other caveats)
781[11:30:29] <Lope> I'm gonna print a tshirt that says exit\nexit
782[11:30:33] <Lope> hahaha
783[11:30:38] <ratrace> I just debootstrapped one such installation yesterday as a test, rebooted into it, works.
787[11:31:20] <Lope> ratrace, I had high hopes for systemd nspawn because supposedly you can pass block devices through to it, as per the links I posted above?
789[11:31:53] <ratrace> Lope: ah you can but: a) rbind being "default" is a lie b) you have to --bind each.and.every.device.you.need c) it still borks install-grub
790[11:31:59] <Lope> and the nice thing with nspawn and passthru (theoretically, if it works) is your container only sees the block devices that are passed thru, so you could even have os-prober and it wouldn't find all your other devices.
791[11:32:24] <Lope> well, then it's pointless eh :]
792[11:32:44] <Lope> ratrace, can you put stuff like exit\n exit in a script?
793[11:32:52] <ratrace> yes but something messes that up. I _think_ it's the seccomp filtering and/or capabilities drop. so it _could_ be possible to fully utilize nspawn this way but ...... holy hell unshare and done in two lines. one line and an exit :)
794[11:33:07] <Lope> cos if I paste commands that includes stuff like that into the terminal, after certain commands, bash drops executing what I pasted.
795[11:33:25] <ratrace> you can't double exit in a script unfortunately because they're different shells
796[11:33:27] <Lope> like if I write a script in a text editor, and paste it into a terminal emulator
797[11:33:44] <Lope> ah, pity there's no way to do it.
798[11:34:00] <Lope> I suppose one way would be to send the text into a terminal emulator.
799[11:34:05] <ratrace> but you can prepare the "inner" shell, then do something like chroot /mnt/foo /path/to/the/inner/shell ; exit _before_ chroot and that semi-automates it all
800[11:34:11] <Lope> But how would you know when it's ready, etc.
801[11:34:22] <Lope> yeah, for sure.
802[11:34:39] <Lope> I've been thinking of semi-automating this stuff, cos it's a giant hassle.
803[11:34:53] <ratrace> sorry, I meant inner SCRIPT
804[11:35:23] <ratrace> I use ansible for debootstrapping servers and I do just that. "outside" script that prepares everything that has to be _before_ chroot, and there's "inside chroot" script that's invoked directly by `chroot`, so when it exits, the chroot exits automatically.
834[11:40:45] <ratrace> RedHat does one thing very well. they take FOSS designed by and for nerds and put a lot of lipstick on that pig, UI thingies, "enterprise value add-ons", and then they sell THAT as a service.
835[11:40:54] *** Quits: iterati (~vangelis@replaced-ip) (Remote host closed the connection)
836[11:41:03] <Lope> What about gpt partitioning a SSD, do you automate that too? like working out the size and then telling fdisk or whatever to make the partitions etc?
837[11:41:14] <Lope> cos that part I've always done interactively, and it's a pain.
838[11:41:55] <Lope> It's fine, I'm a nerd, so I'm happy to use ansible straight.
849[11:43:12] <ratrace> it's a test script I'm using for testing out provisioning for the actual one, which I can't show you because it's so specific to our deployment and I'm bound by NDA
863[11:45:29] <ratrace> but nobody offers raw ubuntu, among the hosting companies :)
864[11:45:38] <ratrace> (as rescue env)
865[11:46:15] <ratrace> OVH is a pita as they have custom kernels and building ZFS there was a lot of trial and error and tons of failed reboots ... becuse you're doing all that over ssh only
866[11:46:16] <Lope> yeah. I've been planning to make my own rescue thing for VM's, but haven't gotten around to it.
867[11:46:37] <Lope> yeah, for sure. But at lease SSH is a lot better than console.
871[11:47:23] <ratrace> note there are some hardcoded decisions in the setup script, because that's just for testing. so it isn't most perfect one, but it's a good example.
873[11:47:53] <ratrace> you'll see I have three different setups there. zfs, btrfs and raw ext4 atop of mdadm atop of dm-integrity, but that's broken atm I think... I did some changes and didn't finish it
876[11:48:35] <ratrace> point is.... such script is used both for provisioning the chroot and then it calls itself with a param that tells it it's run inside chroot so that function is run instead
877[11:48:36] <Lope> interesting, but you wouldn't put zfs ontop of dm-integrity?
878[11:48:57] <Lope> amazing, yeah I've done similar scripts before.
883[11:49:22] <ratrace> Lope: no, what for? we use dm-integrity only under mdadm as bitrot guard. actually testing that whole thing as it's pretty experimental even in the kernel
884[11:49:25] <Lope> But it's a pretty terrible language.
885[11:49:33] <ratrace> Lope: it's /bin/sh, not bash specifically
886[11:49:52] <Lope> geez, sh only, setting the bar REAL low there?
888[11:50:26] <ratrace> Lope: yes, the idea is to have MINIMUM requirements and dependencies . ansible prepares the script from config vars, shoves it onto the server and executes . python and /bin/sh are the only two requirements. could even do without python if you prepare the script locally and then manually push it out to the server
889[11:50:38] *** Quits: earthundead (~earthunde@replaced-ip) (Remote host closed the connection)
890[11:51:11] <ratrace> Lope: for compatibility reasons. at some point we were a freebsd shop, we migrated to debian, and at some point had two, three different environments to handle (ubuntu was in the equation too for a while)
891[11:51:39] <ratrace> and there's no bash on freebsd until you've installed it, set up ports, and built from ports.
894[11:52:38] <Lope> I like to be pretty much debian all the things. But I'm keen to try gentoo sometime.
895[11:52:45] <Lope> I tried clearlinux, but it's not for me.
896[11:53:13] <ratrace> Lope: also note one thing, this is just a testing provisioning script, lots of hardcoded decisions AND uses the old ZFS paradigm with legacy mounts for var. Today you don't need that, and we do more datasets (/var/{log,tmp,cache,spool}) by default
897[11:53:22] <Lope> Ubuntu for slapping on hardware when you just want to test that the hardware works, without needing to install anything, is nice.
900[11:54:01] <ratrace> yeah I use ubuntu live USB as a rescue thingy here at home and office
901[11:54:14] <ratrace> has ZFS, is fully functional and can debootstrap from easily
902[11:54:15] <Lope> okay cool, yeah the main point of interest in your script for me is the partitioning, and also will be interested to see how you call it differently from the chroot.
903[11:54:38] <Lope> But other than that I've got my own way of rolling my mdadm-luks-zfs shenanigans.
904[11:55:01] <ratrace> Lope: this way:
905[11:55:23] <Lope> Thanks for sharing it bud
906[11:55:24] <ratrace> out "Copy self to chroot and exec" ; cp /root/setup.sh /mnt/root/setup.sh ; chroot /mnt/ /root/setup.sh chroot
908[11:55:40] <Lope> you can't give the chroot environment variables, can you?
909[11:55:42] <ratrace> basically it copies itself into the chroot and executes itself recursively
910[11:55:49] <Lope> so if you need to give it info, you write it to disk?
911[11:55:55] <Lope> like /tmp/varname?
912[11:56:17] <ratrace> I don't know, maybe you can. but what for? the script executes itself recursively, inside chroot, so it already contains all the info
919[11:57:35] <Lope> what is the /usr/bin/env part of that command?
920[11:57:36] <ratrace> note that
921[11:57:44] <ratrace> export DEBIAN_FRONTEND=noninteractive is called once outside and once inside the chroot
922[11:57:53] <Lope> usually I've only ever done `chroot /tmp/foo` or `chroot /tmp/foo /bin/bash`
923[11:58:53] <ratrace> Lope: when you do FOO=BAR somecomand it depends ont he shell to interpret that as setting an env var. when you use /usr/bin/env, you explicitly set env vars for the command you give it as param to invoke.
924[11:59:04] <Lope> this is a nice one, do you do this? echo zfs-dkms zfs-dkms/note-incompatible-licenses note true | debconf-set-selections
941[12:03:02] <Lope> so export DEBIAN_FRONTEND=noninteractive removes all the hassles :) nice
942[12:03:15] <ratrace> it assumes defaults. nice thing, yes.
943[12:03:25] <Lope> one that comes up for me is it asks me to install grub from time to time... when grub is already installed.
944[12:03:31] *** Quits: Night-Shade (~TimF@replaced-ip) (Quit: My MacBook has gone to sleep. ZZZzzz…)
945[12:03:43] <Lope> So I have to say "no, don't install grub" in that interactive screen. THen say yes I'm sure, go away now.
946[12:04:07] <ratrace> yeah, grub wont install in noninteractive, because there's no default decision it can make. so you explicitly grub-install <somewhere>
954[12:06:23] <ratrace> Lope: oh btw... this setup script doesn't fully setup the installation inside chroot. the ansible script also uploads all the files that will end up in the installation, which is not visible here
956[12:06:53] <ratrace> this part: out "Rsync config files" rsync -va /root/rootfs/ /mnt/
957[12:07:32] <ratrace> (rsync in new line) that's the part that copies all the files ansible prepared under /root/rootfs/ into actual chroot. locale, network config, ssh keys, other things
958[12:07:44] <Lope> oh, nice!
959[12:08:01] *** debhelper sets mode: +l 1165
960[12:08:15] <Lope> it's always a funny thing. When pressed for time, I often have to decide if I should configure something manually, or write a script to do it.
980[12:11:59] <Lope> you would luks multiple partitions with the same key
981[12:12:09] <Lope> so you did something to unlock each one
982[12:12:23] <ratrace> ah yes, that's what the keyscript does. waits for the key to be delivered to specific location, and then uses it to unlock all luks containers found
983[12:12:39] <ratrace> the key can be delivered actively by ssh, or passively downloaded from the key server
984[12:12:44] <Lope> yes, what I'm saying is, before you taught me to do this, you didn't use keyscript
985[12:12:49] <ratrace> correct
986[12:13:27] <ratrace> now we do. it's same thing really in the end. the script that _unlocks_ is the same. what changed is that we no longer use local-top initramfs script to preapre it, but crypttab kescript
987[12:13:42] <Lope> oh, I see.
988[12:13:48] <Lope> It's amazing what you can do in initramfs.
989[12:14:41] <ratrace> yeah. once you discover the scripts and hooks .... there's no setup you can't do :)
990[12:14:55] <ratrace> and it's all very really simple. the whole initramfs concept and how debian's initramfs-tools does it
996[12:16:23] <Lope> ratrace, you can actually use initramfs as a rescue environment, but for some reason I've never been able to successfully run `update-initramfs -u` inside a chroot, from the initramfs environment.
997[12:16:29] *** Quits: Doo (~doo@replaced-ip) (Remote host closed the connection)
1000[12:16:54] <ratrace> and then you learn how simple and easy it is to download the tarball from kernel.org, untar, cd into it, make menuconfig or whatever way you want; make -j<NCPU> ; make bindeb-pkg ; and then apt install ./linux-image... ./linux-headers... and boom, you have your own custom kernel to boot into :)
1001[12:16:58] <Lope> I can't remember the details, but some /dev/fd/ stuff was missing IIRC. Something in /dev/ missing.
1005[12:17:38] <Lope> ratrace, how hard is it to compile a custom kernel on Debian? I used to do it regularly in ubuntu, but haven't gotten around to it yet in Debian.
1006[12:17:56] <ratrace> I just literally said how :)
1007[12:18:11] <Lope> hahaha
1008[12:18:25] <daysun> does anyone know what the 'master' of a network interface is all about?
1009[12:18:34] <ratrace> and you can start with debian's .config (copy from existing /boot/something) and tweak ... but note, no debian specific patches!
1010[12:18:38] <rudi_s> If you want a kernel with the debian patches, there's a source package available.
1032[12:21:32] <ratrace> Lope: you make deb-pkg IN the kernel sources from kernel.org
1033[12:21:42] <rudi_s> Lope: apt install linux-source; cd /tmp; tar xf /usr/src/....; cd /tmp/linux..; make deb-pkg (just like for the real kernel)
1034[12:21:57] <rudi_s> You can do the same in the Debian sources of the kernel
1035[12:22:08] <ratrace> Lope: the upstream is newer
1036[12:22:24] <ratrace> Lope: and debian specific patches are mostly, if not all, just backports of the fixes from the newer versions
1037[12:22:29] <Lope> ratrace, interesting, well, I'm on bullseye with access to unstable
1038[12:22:34] <Lope> I'm sure it's new enough? :P
1039[12:22:40] <ratrace> eg. I'm running on latest LTS now, Linux erebus 4.19.169 #1 SMP Fri Jan 22 11:07:25 CET 2021 x86_64 GNU/Linux
1040[12:22:49] <Lope> buster became too old for me.
1041[12:23:15] <ratrace> Lope: it's not new enough. debian doesn't backport ALL the fixes, until point releaes rebases to a newer version
1042[12:23:18] <rudi_s> All depends on the goal. If you need different options, the debian kernel is fine (it also contains a few patches which are not from upstream). If you need a more recent kernel either try backports or kernel.org
1043[12:23:28] <ratrace> so I find it better to just track the latest upstream
1044[12:24:08] <ratrace> (and yes, I use options specific to my machine; the default options are too generic and thus a tiny bit too crippled, to my taste)
1045[12:24:17] <Lope> Sounds good. Are the debian specific patches necessary to have?
1046[12:24:51] <Lope> So you gentoo-ify your debian kernel. Without the automagic compiling.
1047[12:24:53] <rudi_s> Depends, there are a few security hardening patches which are interesting.
1048[12:25:12] <ratrace> I'd have to look through ALL of them to ansewr tha, but from what I've glanced, they're mostly just backports that you get in the latest upstream LTS
1049[12:25:32] <ratrace> Lope: and then you can add patches from gentoo, they add some KSPP thingies
1057[12:27:58] <ratrace> unfortunately, the linux kernel devs would want us all to always run their latest kernels. always. many things don't get backported, many things don't get flagged as vulns. so imho there's this balance between !sns and getting all hte fixes, by tracking latest LTS
1060[12:28:38] <ratrace> and by latest LTS I mean, currently the 4.19 LTS. 5.10 is too young. I generally wait until 10-15 point releases before I try it. but in this case I can't because I don't want to upgrade to zfs 2.0.1 just yet
1069[12:41:00] <Lope> ratrace, I recently upgraded from my i5 4th gen slum to a monster 5000 series CPU so running on the bleeding edge here. 5.10 kernel and 2.0.1 zfs
1070[12:41:23] <Lope> I've still got my slumdog cpu though
1071[12:41:45] <Lope> Coincidentally it's SSD died the day after I copied all data to new build.
1072[12:41:59] <Lope> power failures destroyed it's zpool.
1082[12:46:41] <ratrace> currently waiting for a hardware-savvy friend to get back to me on which nvmes and most importantly, which motherboard, as I want both ECC and two PCIe slots so I can gpupassthru one to the windows gaming VM
1091[12:49:50] <Lope> RAM prices are probably going to go up.
1092[12:50:14] <ratrace> and nvidia too. I want the rtx 3070ti one so I'm done for the next 5 or so years with upgrades
1093[12:50:25] <Lope> can you *EVEN* buy a 5900 in germany?
1094[12:50:36] <ratrace> why not?
1095[12:50:49] <Lope> hahahaha, that's crazy man, buying a 3070ti looking for 5 years
1096[12:51:02] <Lope> Not gonna happen
1097[12:51:07] <ratrace> what do you mean?
1098[12:51:14] <Lope> Rather buy a 2080 super or something and look for 2 years.
1099[12:52:09] <Lope> Nvidia was sleeping, AMD wasn't competetive on the high end. Now there's competition, and even intel is entering the market with Xe and it's probably going to be better than people expected.
1100[12:52:18] <Lope> Then there's also a chinese player looking to make GPUs
1101[12:52:33] <ratrace> but my primary concern here is that 3070ti is gonna suffice for high-end gaming for the next 5 years
1102[12:52:38] <Lope> so I don't think you can assume innovation will lag as it did in this last decade.
1103[12:52:47] <Lope> I'm saying it probably won't.
1104[12:53:12] <ratrace> it definitely will more than 2080
1105[12:53:33] <Lope> well yes, but you'll spend a lot more, and be less able to justify upgrading in 2 years.
1106[12:54:11] <ratrace> i'll admit I'm not an expert here, but the hw savvy friend said 3060 is currently the BestBuy in price/performance ratio, and 3070 will give me slightly more boost for slightly MORE money, so not really best buy, but should definitely hold me for the next 5years
1107[12:54:16] <Lope> eh, just my opinion, nobody can predict the future, But it's looking like innovation is on.
1108[12:54:34] <ratrace> right now I've got gtx960, 6 years old. I can't play any of the modern games that require more than 2GB of VRAM that it has
1113[12:55:24] <Lope> I mean if you just want to be able "run" games, yes, 3070ti should be fine.
1114[12:55:29] <ratrace> just kidding, I don't minecraft.
1115[12:55:34] <Lope> I think get the 3060 with the 12G VRAM
1116[12:55:45] <Lope> that'll future proof you best IMO, the VRAM.
1117[12:55:51] <Lope> without over spending
1118[12:56:35] <Lope> I don't play games even. I used to play overwatch, before having kids haha.
1119[12:56:41] <ratrace> thing is... even if this becomes obsoleted _performance_ _wise_ in the next 2 years, i'm sold. I wanna play the games NOW, at max quality. I don't want to ...... AGAIN ....... compromise and get shitty performance within 2 years
1120[12:56:49] <ratrace> and inability to play at ALL new titles in 4-5
1123[12:57:07] <Lope> If I had your goals I would try get the max VRAM
1124[12:57:08] <ratrace> couldn't play Wolfenstein: Colossus, and that's a few years old game now, because it demands more than 2GB
1125[12:57:23] <wsky> get an rx580
1126[12:57:30] <wsky> if it' available
1127[12:57:33] <ratrace> now Serious Spam 4 is out and I want that puppy at MAX :) DooM Eternal, I want it at MAX
1128[12:57:44] <wsky> or perhaps a 1650\
1129[12:57:53] <ratrace> srsly, no
1130[12:57:58] <Lope> ratrace, check out Tech Deals on youtube for GPU recommendations.
1131[12:58:08] <Lope> He's pretty pragmatic.
1132[12:58:11] <ratrace> I did, and I'm sub'd to Linus Tech tips, I'v seen the benchmarks
1133[12:58:23] <Lope> Haha, linus tech tips is a bit of a joke channel though.
1134[12:58:24] <wsky> doom eternal work in full hd @ 60fps on rx580
1135[12:58:30] <ratrace> 3060/3070 is very good thing. vastly superior to 10xx anything.... somewhat superior to 20xx anything
1136[12:58:32] <wsky> works*
1137[12:58:36] <Lope> If you want GPU recommendations, best is hardware unboxed and Tech Deals.
1138[12:59:35] <Lope> I would unsub from linus tech tips if you want GPU recommendations haha
1139[12:59:46] <ratrace> I watch it mostly to see him drop thing s:)
1140[12:59:59] <Lope> Linus tech tips is basically a channel all about Linus flexing and acting low testosterone.
1141[13:00:18] <Lope> "Look how much money I have"
1142[13:01:06] <Lope> I bought 6 12K cameras bro, time to upgrade the SSD array for the 100th time this year bro, upload on the 10Tbit internet connection.
1143[13:01:19] <ratrace> anyhoo, if you scour the net for benchmarks you'll see 3060/3070 doing very good against 2080 and much much very good against 10xx. yes, it's a bit more money per performance... but I care about absolute numbers here, not relative
1160[13:04:23] <Lope> Cos currently, the reason Nvidia GPUs are so expensive is the die is MASSIVE and if there's any defects, they throw it away.
1161[13:04:35] <ratrace> also note they can't cram inifinte number of chiplets in there. they still have to remain within reasonable TDP, reasonable amps and form factors
1163[13:04:52] <Lope> So if AMD can make a monster GPU out of a ton of chiplets, there's no real limit to the perf they can get, and it'll be very cheap, relatively speaking.
1170[13:05:35] <ratrace> Lope: no limit except the above mentioned TDP, cooling profiles and amps. and most importantly: PCIe/mobo/chassis form factors
1172[13:05:40] <Lope> Basically they spend a certain amount of silicon, then how hard they push it, largely determines the TDP
1173[13:05:50] <Lope> So if they can make the silicon much cheaper (which they will)
1174[13:06:00] <Lope> they can push it more gently and get like 3x the perf per watt.
1175[13:06:18] <ratrace> if you ask me.... instead of gettint faster consumer GPUs, the future is in streamed games. google stadia, nvidia's whastitcalled, amazon is in the equation now, and amd might get too
1176[13:06:38] <Lope> I don't think streaming games will ever replace FPS
1177[13:06:49] <ratrace> it will. they just have to get a dc in your town :)
1178[13:06:50] <Lope> only tomb raider sort of stuff
1179[13:06:57] <Lope> Even then, too much lag
1180[13:07:08] <ratrace> don't bet on it
1181[13:07:10] <Lope> They'd have to basically put a DC in your apartment building.
1182[13:07:20] <ratrace> lag is very much overhyped
1184[13:07:50] <ratrace> 5-10msec doesn't do that much of a difference .... unless maybe in very pro competitive gaming but ..... we're not talking about that.
1185[13:08:15] <Lope> 10ms lag, are you crazy?
1186[13:08:16] *** Quits: mezzo (~mezzo@replaced-ip) (Quit: leaving)
1187[13:08:18] <Lope> that's insane
1188[13:08:21] <ratrace> I get 10msec from here to my server across the border, over residential ADSL....
1189[13:08:27] <Lope> consider 60fps is 16ms per frame
1190[13:08:40] <Lope> There's a HUGE difference going from 60fps to 200fps
1191[13:08:57] <ratrace> the lag doens't add up between frames, lol, it's just a systemic offset
1192[13:08:59] <Lope> I do vastly better with framerate unlimited
1193[13:09:19] <Lope> Yes I know
1194[13:09:28] <Lope> but I've run 60hz game on 60hz panel
1195[13:09:34] <Lope> and 200hz game on 60hz panel
1196[13:09:39] <ratrace> but anyway.... we can criticize it from the standpoint of _today's_ networking norms....
1197[13:09:40] <Lope> the difference is insane
1198[13:09:55] <Lope> and what changes, is the input lag.
1199[13:10:10] <Lope> input lag is critical in FPS
1200[13:10:29] <Lope> you don't realize how important it is until you slash it.
1201[13:10:41] <Lope> then you won't want to go back to higher latency.
1205[13:11:23] <ratrace> i've been gaming since..... well... lets see... when did Zork come out? :) and I'm a huge FPS fan. played a lot of it online, until the kids and their reflexes became too much for no-longer-a-kid me
1206[13:11:30] <Lope> Then if you play at 200hz on 60hz panel. You'll say "wow, I'm SO much better than I thought I was!!!"
1218[13:13:13] <ratrace> so back ontopic.... bullseye's timing is great for my plans to go zen3. 5.10 is all I need, as I'm not playing on linux any more, so latest mesa and stuffs don't mean much to me
1221[13:13:57] <ratrace> and phoronix has some benchmarks about schedulit perf regressions for amd on 5.10, fixed in 5.11, but we can bump the kernel easily, so that's not the issue iethre
1228[13:18:05] <ratrace> ideally I'd get a 2 NUMA zones mobo, so I can fully separate the hardware between the host and gaming VM. give it its own cores, its own memory chips. the only pain point is virtio, because it's vastly superior to handle the RAID on the linux side
1229[13:20:06] <Lope> interesting, so you're saying 5.11 has even better 5000 series perf than 5.10?
1230[13:20:14] <Lope> Here I was thinking that 5.10 was the pinnacle...
1231[13:20:15] <ratrace> according to phoronix
1232[13:20:22] <Lope> so when is 5.11 coming to unstable?
1233[13:20:35] <ratrace> I usually take his numbers with a big burlapsack of salt, and then run my own tests where possible to confirm
1234[13:20:46] <ratrace> Lope: I think not before the bullseye release
1235[13:21:28] <Lope> ratrace, how can I see how many slots I've used for luks?
1241[13:23:08] <ratrace> the linux kernel LTS "promise" is a lie really. a lot of NEW FEATURES too get backported, not just bug/security fixes
1242[13:23:39] <ratrace> LTS here really only means Long, nothing else. aka, we won't EOL this thing for a few years and you can source from us, won't have to patch yourself.
1258[13:44:03] <Lope> ratrace, yeah for sure. Often in software, you don't bother to fix the old version that has a bug, often you fix a bug and improve something simultaneously.
1260[13:45:04] <Lope> I think it's quite an ignorant mental model to think of software in terms of always patching an "old stable version" because that's not realistic. Often the fixes in the new version aren't compatible with the old version. So if you want the fix, you need the new version.
1261[13:45:36] <azeem> in particular with the kernel, you also need new/updted drivers for new hardware
1263[13:45:55] <ratrace> Lope: I prefer the context of "Stable" to mean API/ABI stable. as long as those don't change in a breaking manner, I'm fine with !sns
1269[13:47:55] <ratrace> they broke ZFS... mid release of 4.19 by backporting POLITICAL PATCHES that altered the ABI and ZFS could no longer build or work. so debian had to adjust by dropping SIMD in stable's zfs.
1276[13:49:31] <Lope> I just find it funny, with how popular ZFS is (I mean popular enough, that it should work, right?)
1277[13:50:29] <ratrace> the part I find funny is why don't kernels upgrade in the middle of Debian's release. the API/ABI stability promise is BROKEN anywya, the code gets backported anyway, so.... wth..... just gie us LTS kernels in the stable
1292[14:05:02] <tangarora> su - $SUDO_USER -c gsettings set org.gnome.shell favorite-apps "$(gsettings get org.gnome.shell favorite-apps | sed s/.$//), 'myapp.desktop']"
1293[14:05:11] <tangarora> any reason this would not work?
1294[14:05:50] <tangarora> It does not... its like su - -c is not taking the whole command...
1296[14:06:56] <ratrace> I'm guessing that second subshell is not run as the SUDO_USER
1297[14:08:03] <ratrace> $() is a shell construct, run by your shell before you invoke su, so I'm thinking you're getting gsettings as one user and applying it as another .... unless that's exactly what you want?
1317[14:15:35] <ratrace> and just to be clear, you want one unprivileged user (the one running su or sudo) to exec a command as another unprivileged, $SUDO_USER user, right?
1318[14:16:37] <tangarora> ratrace: actually it is a script that install software, and I want to also set configuration items on that users desktop.
1327[14:18:40] <tangarora> so if I ran the script; it would elevate privs, and install software, create shortcuts in /usr/share/app* and then add them to the faforites in the users task bar... or whatever it is called.
1328[14:18:52] <ratrace> so I'm guessing you want one more sudo inside that $()
1333[14:19:45] <ratrace> if I understand you correctly, sudo -iu $USER "gsettings .... $(sudo -iu $USER 'gsettings ...')"
1334[14:19:45] <tangarora> this works as me: gsettings set org.gnome.shell favorite-apps "$(gsettings get org.gnome.shell favorite-apps | sed s/.$//), 'myapp.desktop']"
1335[14:20:43] <tangarora> what I want to do is elevate privs to install, then for one command in the setup script de-elevate.. to me again
1336[14:21:08] *** Quits: tagomago (~tagomago@replaced-ip) (Remote host closed the connection)
1337[14:21:28] <tangarora> Right! Where the user in this case is me... the "sudo user"
1338[14:21:29] <ratrace> or sudo -iu $user /bin/bash "gsettings dance here | piped into | sed or whatever" if I'm not mistaken
1342[14:22:16] <tangarora> where $user is the one who called the script.
1343[14:22:22] <ratrace> right
1344[14:23:51] <ratrace> just don't forget -c for that /bin/bash as I did
1345[14:24:35] <ratrace> also... you'll want to resolve the $SUDO_USER before you invoke the sudo from within the scrip that's already called via sudo
1346[14:25:18] <ratrace> as running this will return root: sudo /bin/bash -c 'echo $SUDO_USER' (note ' and not ", so that's $SUDO_USER env as seen from within the shell run by root)
1347[14:25:37] *** Quits: tagomago (~tagomago@replaced-ip) (Remote host closed the connection)
1351[14:30:05] <tangarora> this worked as a function: sudo -iu $SUDO_USER gsettings set org.gnome.shell favorite-apps "$(gsettings get org.gnome.shell favorite-apps | sed s/.$//), '${NewApp}']"
1361[14:33:52] <ratrace> so either you exec that as a _string_ given to /bin/bash -c (and note, don't use " but '), or you first get gsettings into a var, then echo that var into second invocation of gsettings
1378[14:40:51] <tangarora> so this command: echo sudo -iu $SUDO_USER gsettings set org.gnome.shell favorite-apps "'$(gsettings get org.gnome.shell favorite-apps | sed s/.$//)', '$myapp']"
1419[15:00:08] <jennie> i have no clue, i just want to connect to wifi
1420[15:01:42] <ratrace> then you'll need 'wpasupplicant' at minimum .. but wait... what do you intend to do? download .deb elsewhere and then copy via usb stick or something?
1427[15:04:02] <ratrace> jennie: I don't know how to get a recursive list of dependencies, so you can download them all.... but... if you can, just download the DVD1 ISO and use it in your apt sources.list
1457[15:34:37] <ratrace> incidentally, that's how I got into Linux in 2006, via OpenSuSE. Debian couldn't install on the laptop I had then, OpenSuSE did and worked great. Debian is really running itself into the ground with all this purism.
1458[15:35:15] <jim> ratrace, hi, by way of explanation, yesterday jennie got a netinstall image for buster (the one with firmware), but it couldn't get the interface going
1459[15:36:07] <jim> so, we manually got some .debs for the firmware, and installed those, then we had the interface
1460[15:36:30] <ratrace> right and now they need wpasupplicant, with dependencies
1461[15:37:22] <jim> after that, tried to get wicd curses version, but we got bogged down into locating all the dependencies, which it turned out jennie didn't want to do...
1463[15:37:45] <jim> so, jennie started the conversation with you
1464[15:38:20] <ratrace> well, if they can get the DVD ISO (wrt bandwidth considerations etc...), it's better to get that and not guess and copy over file by file
1467[15:38:31] <jim> I think there are a couple of choices at this point... she could store the dvd image file on the new minimal debian, and mount it from there
1468[15:38:47] <jim> or rufus the dvd image, and reinstall everything
1469[15:39:05] <ratrace> no need for reinstallation, they can use the USB stick in apt sources.list
1470[15:39:06] <jim> the latter choice will be less "tinkering"
1471[15:39:38] <jim> right that would be the first choice, which depends on getting the dvd image mounted
1484[15:45:17] <ratrace> but irrelevant in this case. jennie alraedy has debian installed, they just want some packages to get wifi going. I recommended getting the entire DVD iso, if possible, so they don't fiddle with pinpointing dependencies
1485[15:45:33] <BCMM> oh sorry, it says "the one with firmware" above
1488[15:46:03] <tangarora> sudo -iu $SUDO_USER "gsettings set org.gnome.shell favorite-apps \"\$(gsettings get org.gnome.shell favorite-apps | sed s/.\$//), '${NewApp}']\""
1489[15:46:03] <tangarora> -bash: gsettings set org.gnome.shell favorite-apps "$(gsettings get org.gnome.shell favorite-apps | sed s/.$//), 'rtd.desktop']": No such file or directory
1490[15:46:14] <tangarora> I am pulling my hair out!
1491[15:46:25] <tangarora> bash think its a file...
1492[15:47:23] <tangarora> and if I remove the outer quotes: bash: syntax error near unexpected token `('
1493[15:47:43] <ratrace> tangarora: sudo -iu $SUDO_USER /bin/bash -c "gsettings set org.gnome.shell favorite-apps \"\$(gsettings get org.gnome.shell favorite-apps | sed s/.\$//), '\${Newapp}']\""
1494[15:48:01] *** debhelper sets mode: +l 1180
1495[15:48:53] <jim> BCMM, right, that's what we did yesterday, and it worked well as far as it went (able to boot into minimal debian)
1496[15:49:32] <BCMM> yeah sorry i misread the whole situation. this is, like, debian has been net-installed successfully but no wireless-related packages were selected at install time, right?
1497[15:49:33] <ratrace> tangarora: infact, I think that \$ in sed and \${NewApp} (whatever that is) would have to be double escaped
1498[15:49:53] <jim> ok, we got a minimal debian, and the dvd image... how can we mount the dvd image?
1499[15:50:17] <jennie> Dvd1.iso moved to debian system from USB stick, and contents of DVD1.iso extracted on to USB stick and USB stick is mounted in debian
1501[15:50:46] <jim> BCMM, yes, and, wireless interface didn't show up yet
1502[15:51:29] <tangarora> ${NewApp} is supposed to substitute for $1. ${NewApp}=myapp.desktop
1503[15:51:34] <jim> jennie installed some firmware, and then the interface showed up
1504[15:51:52] <jennie> jim iso is now in /root/wl
1505[15:51:57] <jim> that's where we are
1506[15:52:08] <jim> ok
1507[15:52:15] <ratrace> tangarora: or just don't use double quotes on the outside .... sudo -iu $SUDO_USER /bin/bash -c 'gsettings set org.gnome.shell favorite-apps "$(gsettings get org.gnome.shell favorite-apps | sed s/.\$//), \'\${Newapp}\']"'
1508[15:52:19] *** Quits: thiras (~thiras@replaced-ip) (Remote host closed the connection)
1509[15:52:19] <BCMM> oh sorry, it's an ISO not a physical CD
1510[15:52:23] <BCMM> mount with -o loop
1511[15:52:27] <jim> what is the exact name of the image file?
1521[15:53:51] <BCMM> oh hang on, [14:50:17] <jennie> Dvd1.iso moved to debian system from USB stick, and contents of DVD1.iso extracted on to USB stick and USB stick is mounted in debian
1528[15:54:40] <ratrace> i'm guessing this is how. been long time since I did it but ...... the installer itself should've left the commented out hints in the sources.list ?
1529[15:54:40] <jennie> BCMM please check this [09:40:46] <jennie> replaced-url
1530[15:54:41] <tangarora> retrace: ${NewApp} should become the name of the shortcut passed to the function: $1
1531[15:55:20] <ratrace> tangarora: oh crap, then you do need double quotes on the outside, and then double-escape \\$ for sed
1560[16:02:07] <ratrace> "add is used to add a new disc to the source list. It will unmount the CD_ROM device, prompt for a disc to be inserted and then proceed to scan it and copy the index files."
1567[16:03:16] <BCMM> will just want to mount it again
1568[16:03:52] <jim> howbout we try that, with apt update
1569[16:03:56] <ratrace> indeed
1570[16:04:24] <jennie> ratrace sudo mount -o loop -t iso9660 /dev/sdc1 /media/cdrom this gave me error from reddit but this apt apt-cdrom add -d /media/usb-drive worked fine
1571[16:05:05] <ratrace> jennie: well I hope you didn't copy paste that verbatim, but you actually used appropriate devices, and an existing mountpoint directory :)
1572[16:05:06] <BCMM> yeah that first command would be for if sdc1 had a raw iso image on it
1573[16:05:23] <ratrace> BCMM: no, if sdc (no 1) had
1574[16:05:26] <BCMM> (which would be a weird thing to write to a partition, imho)
1575[16:05:31] <jennie> yes, i did use sda1 for my usb
1576[16:05:40] <ratrace> the ISO actually contains a number of partitions
1577[16:05:56] <jennie> now this apt-cdrom add worked, whats next step>?
1587[16:08:01] <jim> you were a noob once too :P it's harsh to judge
1588[16:08:39] <longshanks> I remember when I installed Slackware '96 I spent at least an hour trying to work out how to get teh CD visible on the filesystem :)
1589[16:08:55] <jennie> apt-update gives error ' updating from such a repository cant be done securely and disabled by default" see apt-secure(8) manpage
1590[16:09:26] <ratrace> jennie: try the step #4 in that reddit comment
1591[16:09:51] <ratrace> adding [allow-insecure=yes] to that deb cdrom:... line
1592[16:10:59] <jennie> meanwhile please check the picture replaced-url
1594[16:12:39] <ratrace> jennie: the network repos will error out as long as you don't have the network set up, so you might actually want to temporarily comment out other deb entries
1595[16:12:54] <jim> it looks like you tried to add the same dvd image again
1596[16:15:14] <jennie> how can i find my cdrom name?
1597[16:15:22] <jim> I'm not quite understanding how this is working
1726[17:23:10] <foxide> DDDD: Use the -p flag to change your passphrase on the key. That'll cause it to convert the private key into the necessary format.
1727[17:23:22] <foxide> -e only outputs the public key.
1728[17:23:23] *** Quits: JohnML (~john1@replaced-ip) (Remote host closed the connection)
1729[17:24:00] <jim> jennie, hi... are you still around?
1874[18:24:58] <Seb95passiongnul> Hello Debian family, I'm going to ask a bit silly question, but I'm having errors with Lintian, I'm having trouble seeing how: The source of the following file is missing. Lintian checked a few possible paths to find the source, and did not find it. Please repack your package to include the source or add it to "debian / missing-sources" directory. Please note, that very-long-line-length-in-source-file
1875[18:24:58] <Seb95passiongnul> tagged files are likely tagged source-is-missing. It is a feature not a bug. The package is on mentors: replaced-url
1876[18:25:35] <Seb95passiongnul> what should I do to stop having errors?
1948[18:50:45] <SerpentSpeech> or is it at home pc?
1949[18:51:12] <Akuw> home pc
1950[18:51:19] <Akuw> reboot necesary?
1951[18:51:25] <jim> jennie, right now, on your linux laptop, you already have partitions set up... you'd have to arrange to format them again, and use them, and mount them, all in the partitioning tool
1952[18:51:28] <SerpentSpeech> try logout then reboot if not
1973[18:57:50] <jim> jennie, if you install with the dvd image, it should put a source to itself in /etc/apt/sources.list, so then you should be able to install those packages
1974[18:58:35] <jim> when you get to the partitioning tool, let me know
1975[18:58:37] *** Quits: n4dir (~n4dir@replaced-ip) (Remote host closed the connection)
1976[18:58:50] <jennie> i dont understand about the partitioning tool
2053[19:16:45] <jim> relipse, ok, if that's done, you just change the /etc/apt/sources.list so it's on buster inetead of stretch, then do the same thing again (which will be MUCH more): apt update ; apt upgrade ; apt dist-upgrade
2054[19:18:10] <jim> basically an upgrade like this will replace your entire system
2064[19:25:16] <jennie> okay its asking for software installation 1 debian desktop environment 2 print server 3 standard system utilities and they are already checked
2081[19:32:32] <jim> also remember there are other dvd images, which probably have other DEs, and once we get the net up (which it's looking good for), then you can install what you want
2082[19:33:45] <jim> relipse, yeah, gimme a min...
2083[19:33:54] <sney> relipse: the new file doesn't add anything of note so you may as well keep the current file
2084[19:34:51] *** Quits: darunesh (~darunesh@replaced-ip) (Remote host closed the connection)
2109[19:45:54] <jhutchins> jim: wicd is obsolete and deprecated and fails with many cards.
2110[19:46:03] <jim> ok
2111[19:46:14] <jim> nm is not?
2112[19:46:32] <jhutchins> jim: Network Manager replaced it years ago. NM is still maintained and is the preferred solution.
2113[19:46:50] <jim> ok
2114[19:47:16] <jhutchins> wicd was geat when it first came out, it was the first real solution to portability.
2115[19:48:19] <jhutchins> jennie: So you'd want to make sure that Network Manager is there. There are things like nm-cli if you don't want to use a desktop applet to manage it..
2117[19:48:44] <jennie> jhutchins nmcli command not found
2118[19:48:57] <zutat> nmtui also
2119[19:49:12] <SerpentSpeech> click and add wireless network?
2120[19:49:13] <jim> jhutchins, here's the current situation... jennie wrote buster dvd 1 to a usb stick, then booted from that, and ran the installer to completion... she's now rebooted into that, and now where should we be in terms of ability to install stuff?
2126[19:50:41] <dpkg> Firmware is software to operate electronic devices, usually contained in EPROM or flash memory. Some Linux kernel drivers require firmware to be provided from userspace, notably for <WiFi> devices. Most firmware files are not part of a Debian release as they do not conform to the <DFSG>; some are available via <contrib> and <non-free> packages, ask me about <search>. See also <installer firmware>. replaced-url
2127[19:50:58] <jhutchins> Darn, I thought it had the package name.
2128[19:51:29] <jim> jhutchins, tried a netinstall with firmware, didn't find the wireless... but she installed some packages manually, some firmware packages, then we saw an interface in /sys/class/net
2129[19:51:31] <jhutchins> jennie: If you have non-free enabled, try apt search "nonfree"
2141[19:54:52] <jhutchins> nmcli is in network-manager, or you can use static config.
2142[19:54:55] <jhutchins> !interfaces
2143[19:54:56] <dpkg> Your network configuration is in the file /etc/network/interfaces ; "man 5 interfaces" for documentation, "zless /usr/share/doc/ifupdown/examples/network-interfaces.gz" for example configurations. Start and stop your networking with ifup -a and ifdown -a respectively. replaced-url
2144[19:55:07] <jim> ok, we need to know if the installer works on that image
2164[19:59:58] <jmcnaught> jennie: your sources.list is incomplete because you did not configure a network mirror during installation. See replaced-url
2165[20:00:28] <jhutchins> Network Firmware is typically in non-free,
2166[20:00:49] <jennie> jhutchins why cant i install it from usb like i did yesterday?
2192[20:05:02] <jim> sney, add main to the cdrom line
2193[20:05:31] <jim> she;s paused
2194[20:05:46] <sney> jim: it is there, at the end. are we looking at a different paste? replaced-url
2195[20:05:47] <SerpentSpeech> ah there is no network at all yet?
2196[20:06:18] <jim> correct, right now jennie has no net on her debian machine
2197[20:06:35] <SerpentSpeech> no cable to use?
2198[20:06:42] <sney> this laptop only has wifi iirc
2199[20:06:48] <SerpentSpeech> aight
2200[20:06:49] <jim> no port to plug it into
2201[20:07:16] <SerpentSpeech> interesting project this :)
2202[20:07:44] <sney> you can get some catch-22 situations installing debian without a network connection, but it usually doesn't require *this* much going around in circles
2203[20:07:46] <jennie> i call wlo1 after ls /sys/class/net
2204[20:07:47] <jim> it's been a horrible blocking problem since yesterday or day before
2205[20:08:14] <jennie> i can see wl01 after i installed firmware for wifif
2232[20:13:13] <jmcnaught> Maybe sorting out /etc/network/interfaces and getting a connection should come first, then edit sources.list.
2233[20:13:19] <sney> jennie: put this in /etc/network/interfaces, replacing ESSID and PASSWORD with your actual wifi network and password, then 'ifup wl01' replaced-url
2234[20:13:30] <jim> jennie, on cdrom, and, not 'free' but 'main'
2235[20:13:33] <SerpentSpeech> Akuw btw did you install debian in english by your self? Installer prompt says choose langauge if i remember right. Perhaps you wanted system in english?
2238[20:13:54] <sney> jim: again, it already has main. if you add another main to the end of a line that already says main, it will change nothing. (and apt will probably warn about it)
2239[20:14:23] <jim> right... so the sources are good already
2240[20:14:56] <jim> and, time to install wpasupplicant
2243[20:15:53] <sney> there is no judd bot so the , commands will do nothing
2244[20:16:38] <sney> wireless-tools doesn't depend on wpasupplicant but it may have been installed automatically otherwise, depending on tasks chosen, task-laptop for instance
2277[20:23:27] <jim> jennie, well, network-manager, then we won't need those 4 lines any more
2278[20:23:33] <sney> that's iwlist, which is in wireless-tools.
2279[20:23:45] <sney> !sbin
2280[20:23:45] <dpkg> Some binaries, particularly system utilities are installed to /sbin or /usr/sbin rather than /bin or /usr/bin. The former directories are not in a standard user's PATH. Try calling them directly, e.g. /sbin/ifconfig, or running them as root. See also <buster su>.
2281[20:23:59] <jennie> I tried iwlist wlo1 scannig but it didnt work
2300[20:27:48] <jim> also, now you can install the DE you want
2301[20:28:01] *** debhelper sets mode: +l 1185
2302[20:28:10] <sney> !tasksel
2303[20:28:10] <dpkg> tasksel is a program to select and install "tasks": a collection of packages for a system to perform a specific task (e.g. desktop environment, web server). Ask me about <standard task>, <server tasks> and <laptop task>. replaced-url
2304[20:28:51] <jim> would tasksel not have DEs that are not on dvd 1?
2305[20:29:12] <jennie> yes, worked network-manager
2306[20:29:29] <jennie> oh its asking for cdrom to insert in
2308[20:29:49] <sney> jennie: you still need to fix sources.list, I thought you saw the message from dpkg above. here it is again.
2309[20:29:52] <sney> !buster sources.list
2310[20:29:53] <dpkg> A suitable /etc/apt/sources.list for "Buster" has the lines: "deb replaced-url
2311[20:30:00] <jim> hit enter if the usb is plugged in
2312[20:30:14] <sney> jennie: ^ make sure your sources.list has these lines, and you can comment or remove the cdrom now, since you have a network connection.
2313[20:31:26] <jennie> yes, thanks sney, jim hitting enter after inserting cdrom is still asking for cdrom, i am chaging sourcles
2314[20:31:52] <jim> also comment out line 5
2315[20:33:17] <jackal> hello, i have this weird question - can you install debian from within already installed distro? sort of like you download an installer, open it and install debian like so, w/o any cd or iso images
2316[20:34:01] <jim> jackal, there's something called deb-bootstrap
2317[20:34:03] <sney> jackal: there is a tool called debootstrap that can do that, for installing in a chroot etc
2318[20:34:09] <sney> !debootstrap
2319[20:34:09] <dpkg> debootstrap can create a basic Debian system from scratch, without apt/dpkg. Useful for installing in a <chroot>. It is key to installing Debian GNU/Linux from a Unix/Linux system. replaced-url
2320[20:35:10] <jim> jennie, what DE did you want?
2321[20:36:02] <jackal> i dont think debootstrap is what i need
2323[20:36:22] <jim> jackal, it does the thing you asked for
2324[20:36:45] <sney> jackal: well, there isn't really a gui installer that will run in an existing OS.
2325[20:37:07] *** Quits: _aeris_ (~aeris@replaced-ip) (Remote host closed the connection)
2326[20:37:09] <jim> jackal, basically, it installs the minimal debian by installing packages (hence a "bootstrapping" of debian, wherever you want
2327[20:37:14] <jim> )
2328[20:38:12] <jackal> jim, i asked if there's an installer i can use to install it from within an os, to replace that os afterwards
2329[20:38:16] <jim> you would need to install deb-bootstrap on the machine you want to put debian on
2330[20:38:19] <jackal> sort like windows, i think
2331[20:38:21] <sney> debootstrap is the tool used by the debian installer, so the resulting system is the same, even if it doesn't have that familiar red-and-blue menu. it's also used by packaging tools like pbuilder.
2387[21:27:07] <mixxit> it just says link disconnected i tried to disable r8169 and use r8168-dkms as suggested in various posts but i still have no connectivity
2391[21:29:15] *** Quits: wyatt8740 (~wyatt8740@replaced-ip) (Remote host closed the connection)
2392[21:30:42] <sney> mixxit: as you can see from the topic, unstable support is in #debian-next on OFTC (note, right now you are on freenode) - that channel is more capable of debugging pre-release issues.
2393[21:30:57] <mixxit> oh sorry sney i didnt notice, its been a while since ive been in irc!
2415[22:03:42] <Vatum> kernel 4.19.0-11 and 4.19.0-13 (amd64) breaks lcd backlight ( /sys/class/backlight empty ) on radeon systems (and maybe on others). the last kernel that works is 4.19.0-9 for me
2416[22:03:52] *** Quits: bitblit (~bitblit@replaced-ip) (Remote host closed the connection)
2456[22:35:11] <Gigglebyte> Are there any other browsers that are still written for 32-bit besides Chromium and Firefox? The reason I ask is that my laptop died, took it to the shop, and someone gave me a really old HP Pavilion dv4000 that only runs 32-bit. It has Linux Mint running on it, and I just installed Chromium. I can't find a working link for Chrome, but Google discontinued support for 32-bit back in 2017.
2457[22:35:39] <Gigglebyte> Anyone know of an older link to Chrome? And is it reasonably safe?
2483[22:54:28] <n4dir> delpan: i vaguley remember "aptitude purge ~c" and as i only remember it vaguely a quick web search gave this: replaced-url
2484[22:54:59] <n4dir> take care with it. If it goes south, i don't wanna be the culprit. Hertzog has quite a reputation, of course
2485[22:54:59] <jhutchins> I doubt if there's any automatic cleanup of those files, lots of things write to them.
2486[22:55:15] <jhutchins> delpan: What is your actual goal?
2487[22:56:53] <delpan> jhutchins: well, to remove cron:D i usually remove it (with apt-get --purge, or dpkg --purge) but only just noticed it leaves them 2 dir's behind and they are empty.
2488[22:57:07] <delpan> so wondered if some known dependency of other systool or something
2489[22:57:26] <jmcnaught> delpan: other packages put files in those directories too, so that they can add cron jobs.
2490[22:58:10] <delpan> the postrm script should contain what to remove when uninstalling, correct? or does it use other instructions
2491[22:58:25] <delpan> jmcnaught: i understand but they were empty
2497[23:05:14] <zabriski> Hey, I broke my xorg, er uh, window manager I guess. I have three monitors and 2 gpus, they're normally all plugged into one and I wanted to see about setting up my center monitor to be on my second gpu for better gaming performance. My google-fu wasn't strong enough to get anything to output to xscreen 1, only 0 would display. After I switched
2498[23:05:14] <zabriski> everything back and it appeared to be working, I saw discover had updates including nvidia drivers. I updated, restarted and now I can't make any window fullscreen on my secondary monitors, only the center, and tiling triggers on the edges of the center monitor instead of the outsides of the secondary ones
2499[23:05:54] <jhutchins> jmcnaught: He left. Removing cron will gain him experience points!
2500[23:06:09] <zabriski> I also can't invoke the alt+space searchbar on anything other than the center monitor
2501[23:06:19] <jhutchins> I wonder if he thinks that's necessary to install an alternative.
2502[23:07:30] *** Quits: fflori (~fflori@replaced-ip) (Remote host closed the connection)
2503[23:08:32] <zabriski> it seems like there's a 800x600 xscreen overlapping the upper left corner of the left monitor that I think might be breaking stuff. When I login the kde splash screen displays another smaller instance overlapping on top of the correct resolution one
2508[23:14:17] <jmcnaught> zabriski: by default there is no /etc/X11/xorg.conf or files in /etc/X11/xorg.conf.d/ so if these exist you could move them somewhere else for now to get back to Xorg defaults (which autodetects everything when there is no xorg.conf)
2511[23:22:02] <zabriski> I can't see who messaged to just remove xorg.conf and let it regenerate but that fixed it, thanks for the help. Sorry my google-fu is bad and I gotta ask dumb questions in here, everything I was reading was telling me to run various commands to remake it, which kept keeping the bad setting
2513[23:23:14] <jhutchins> zabriski: X has been around a long time, and there's plenty of documentation on the web that's been obsolete and no longer true. Of course, no dates or versions on the docs.
2518[23:25:34] <zabriski> also are mouse acceleration settings stored in xorg.conf on debian? after letting it regen it seems like I have the mild acceleration that comes default again
2531[23:43:51] <robobox> for all you know you could be getting docs for X11R5
2532[23:44:38] <zabriski> jhutchins: if you ever need to know for some reason I dug through my notes and it's /usr/share/X11/xorg.conf.d/50-mouse-acceleration.conf that handles mouse accel, you have to make the file yourself and put in the settings for raw input to truly disable acceleration