aitor_ | fsmithred, if you are there..., i've just sent you an email related to the live-sdk | 00:33 |
---|---|---|
aitor_ | on the other hand, you were working on a live-config-runit package | 00:34 |
aitor_ | i'm using something like this for the autologin with runit: | 00:35 |
aitor_ | https://paste.debian.net/plain/1175036 | 00:35 |
fsmithred | hi aitor | 00:36 |
aitor_ | hi | 00:37 |
fsmithred | I uploaded a package. Just have to remember where. | 00:39 |
aitor_ | ok | 00:39 |
fsmithred | http://distro.ibiblio.org/refracta/files/extra_packages/live-config-runit_5.2019%2brefracta10u1_all.deb | 00:39 |
fsmithred | that's for the version of live-config in beowulf | 00:40 |
fsmithred | but should work in any version | 00:40 |
aitor_ | thanks | 00:40 |
fsmithred | I will eventually make a devuan package for chimaera | 00:40 |
aitor_ | I also eventually removed the dialogs from the live-sdk in order to simplify the code | 00:41 |
fsmithred | oh, thanks | 00:41 |
aitor_ | suggestions are welcome | 00:41 |
fsmithred | parazyd will be happy about that. He likes to automate builds. | 00:41 |
aitor_ | the code is more readable now | 00:42 |
aitor_ | the dialogs will be located in a separate folder and the user will be able to by-pass them | 00:43 |
fsmithred | cool | 00:44 |
fsmithred | I got the tarball. Thanks. | 00:44 |
aitor_ | hope you like it | 00:44 |
aitor_ | i left glibc-udeb as it is | 00:44 |
fsmithred | ok. What does that mean? | 00:45 |
fsmithred | oh, you had some conflict with that? | 00:45 |
aitor_ | yes | 00:45 |
aitor_ | but i applied a patch to the Makefile of debian-installer | 00:45 |
aitor_ | it was due to an existent symlink in /lib | 00:47 |
aitor_ | more concretly ld-linux.so.2 | 00:47 |
aitor_ | the Makefile uses the find command in /lib and this symlink overrides /lib/i386-linux-gnu/ld-linux.so.2 | 00:49 |
aitor_ | in amd64 this symlink is located in /lib64, so there is an unique ld-linux-x86_64.so.2 in /lib, the one located in /lib/x86_64-linux-gnu | 00:53 |
aitor_ | the patch for the Makefile is include in the extra folder of the live-sdk | 00:54 |
aitor_ | debia-installer uses the source format, no way to apply a quilt patch for that | 00:56 |
aitor_ | time to bed, bye | 00:56 |
fsmithred | good night and thanks. | 00:56 |
aitor_ | see you tomorrow, devuaners :) | 00:57 |
landley | If I wanted to do an sh4 port of devuan, where would I start? | 06:12 |
landley | Hmmm, really two questions. 1) If I wanted to do a musl-libc port of devuan, where would I start, 2) sh4. | 06:13 |
unixbsd | hello | 09:01 |
unixbsd | I try to install with installer. | 09:01 |
unixbsd | Installing grub boot loader [NO]! | 09:01 |
unixbsd | Answer from devuan instlaler: "Installing Grub Boot Loader", please make sure that it works. | 09:02 |
KREYREEN | Referencing https://qubes-os.discourse.group/t/r4-1-x86-64-debian-pv-qube-is-unable-to-use-dgpu-provided-on-pci-00-00-0-through-radeon-on-20089-lenovo-g770/1731 help appreciated.. it's debian-11 running in QubesOS that seems to have userland issue to not use the radeon GPU >.> | 15:49 |
KREYREEN | (debian didn't help x.x) | 15:50 |
danuan | after running update-rc.d , got a load of errors , tracked it down 22 /etc/init.d/ scripts had a filesize of 0 . tracing logs , only thing i found was that server was rebooted earlier in the day . and no reboots after that event | 15:51 |
danuan | plus everything was running so zerooingg did not happend during reboot , as none of those services would have been running , something did it while system was up. and doing a find by date and zero size revealed that /etc/init.d was only one affected | 15:53 |
KREYREEN | danuan, my first guess would be hardware-induced data loss.. did you check the S.M.A.R.T. ? | 15:56 |
KREYREEN | also i would hexdump these to see what those files are storing.. e.g. could be a malformed header that makes them seem like blank | 15:57 |
KREYREEN | also dmesg might help | 15:57 |
danuan | using a 6 gig ide drive so no smart on that yet | 15:58 |
KREYREEN | i see.. in which case i would move the data in backup (in case there is a hardware damage forcing it to do read/write might cause more data loss) and check it agains something like `badblocks` | 16:00 |
KREYREEN | assuming there being any data that you don't want to lose O.o | 16:00 |
KREYREEN | also syslog might show if there was any cronjob or something that touched these files | 16:00 |
danuan | all happened on exact same date same time , went trough all logs and nothing on that particular date-time | 16:01 |
KREYREEN | might be relevant to health checking https://superuser.com/questions/171195/how-to-check-the-health-of-a-hard-drive | 16:01 |
KREYREEN | danuan, that would point on the hardware failure to me | 16:02 |
KREYREEN | especially when it's on files that are within a same directory where depending on filesystem used these might be on the same block | 16:02 |
KREYREEN | as dataloss usually has these symptoms of leaving blank files behind | 16:03 |
KREYREEN | (at least based on my experience) | 16:03 |
KREYREEN | afaik because the filesystem stores that file hierarchy and points to the data on the specific blocks to be located.. so files/directories stay, but data on them are not reachable O.o | 16:03 |
KREYREEN | and if it has some sensitive data that you don't want to lose and can't exclude hardware damage (that you should have a procedure for if this is commercial use) then i would rip the system from the wall and gave the HDD to the specialist for data restoration | 16:05 |
KREYREEN | ripping it from the wall so that you are not forcing more runtime on the HDD that may result in more data loss or so i was taught to do x.x | 16:05 |
danuan | its just 6 gig system ide drive with backups , important stuff is on zfs | 16:05 |
unixbsd | hello, what is the sources.list of ascii devuan main repository? (only the packages stable, no security)?# | 16:05 |
unixbsd | for pae i386? | 16:06 |
KREYREEN | danuan, then i would try self-diagnostics to verify the current health of the device | 16:06 |
unixbsd | please if you had an template or url for it, thank you very much and in advance looking forward reading you -- kind regards | 16:07 |
fsmithred | unixbsd, deb http://deb.devuan.org/merged ascii main | 16:07 |
KREYREEN | unixbsd, See http://deb.devuan.org/devuan/dists/ ascii has backports, proposed-security, proposed-updates, proposed, security, updates and main available | 16:07 |
danuan | KREYREEN and your question , i can only say i have tested kvm vs xen side by side and kvm came out on top as far as how many things just worked with minimal config . though i never got to gpu passthrough | 16:08 |
fsmithred | ascii ascii-security and ascii-updates are the standard ones to have | 16:08 |
KREYREEN | fsmithred, care to take a look on my issue? ^-^ | 16:08 |
danuan | xen might have been yesterdays sollution , but today kvm seems to get way more attention | 16:08 |
KREYREEN | danuan, ye so far everything but the GPU passthrough works x.x | 16:09 |
unixbsd | is there a repository that works and that is stable? | 16:09 |
KREYREEN | unixbsd, `stable` | 16:09 |
fsmithred | no, don't use 'stable' | 16:09 |
KREYREEN | x.x | 16:09 |
* KREYREEN though it's an alias to current stable~ | 16:09 | |
fsmithred | use codenames, because we are not always in sync with debian | 16:09 |
KREYREEN | ah true sorry x.x | 16:10 |
unixbsd | http://auto.pkgmaster stuffs is offline | 16:10 |
fsmithred | KREYREEN, qubes and radeon? I don't know anything about it. | 16:10 |
KREYREEN | x.x | 16:10 |
KREYREEN | fsmithred, info on diagnosing radeon on devuan would be appreciated as this seems as userland issue O.o | 16:10 |
KREYREEN | basically debian has dGPU on PCI available, but refusing to use it | 16:11 |
fsmithred | unixbsd, if deb.devuan.org gives you problems, pick a nearby mirror from the list | 16:11 |
fsmithred | at devuan.org | 16:11 |
fsmithred | KREYREEN, maybe a bios setting for order of vid cards? | 16:12 |
danuan | KREYREEN if you need to trick to xen maybe try AMD proprietary drivers ? and i would look in to kvm passthrough of that particular card , see what that comunity is saying as it might be applied | 16:12 |
danuan | trick - stick | 16:13 |
fsmithred | gotta go. afk. | 16:13 |
KREYREEN | fsmithred, this is on xen so BIOS should provide the vid card | 16:14 |
KREYREEN | danuan, i never got the AMD proprietary to work on this dGPU | 16:14 |
* KREYREEN noticed something in settings and is trying reboot of the VM | 16:14 | |
Junicchi | it might be me or the updates but | 16:14 |
Junicchi | the whole sound system on my machine is broken now | 16:14 |
danuan | i remeber reading that it is a hit and miss for passthrough , and major hacks with extracting vga bios and passing it to domus to initialise as files for some videocards | 16:15 |
Junicchi | vga passtrough huh | 16:15 |
Junicchi | fuck | 16:18 |
Junicchi | s/vga/pci/g | 16:18 |
KREYREEN | still no idea x.x | 16:35 |
danuan | KREYREEN just a though , isnt that lenovo a laptop? and would that mean you are trying to passthrough a primary and only gpu that main system uses ? | 16:43 |
danuan | as far as i know that means you need to make dom0 a headless system somehow | 16:43 |
unixbsd | is there a webpage with all infos about sources.list? similar to the classical one: https://wiki.debian.org/SourcesList (old good webpage) | 18:10 |
unixbsd | the only one I know is: https://www.devuan.org/os/packages but not a wiki. | 18:13 |
golinux | That's the only page | 18:13 |
KREYREEN | danuan, it is laptop and it has iGPU and dGPU | 18:13 |
golinux | We need more people involved to maintain a wiki. | 18:14 |
golinux | Last effort died a slow death. New effort hasn't launched. And no one to write content so for now the wensite is it. | 18:15 |
golinux | website | 18:16 |
KREYREEN | o.o | 18:19 |
KREYREEN | devuan has a wiki? | 18:19 |
KREYREEN | https://wiki.devuan.org/ returns "Welcome to nginx!" lol | 18:21 |
specing | Should be "Welcome to systemd-nginx" | 18:26 |
KREYREEN | system dicked would return 404 | 18:26 |
DHE | 500 user tried to use a non-default setting | 18:27 |
KREYREEN | default sucks~ | 18:32 |
golinux | KREYREEN: https://friendsofdevuan.org/doku.php | 18:42 |
golinux | It is so out of date Beowulf isn't even mentioned. | 18:42 |
golinux | And no one liked the doku wiki format. | 18:43 |
KREYREEN | like if you want people contributing then i would make an official wiki on wiki.devuan.org and mentioned it on devuan.org.. | 18:43 |
KREYREEN | would be personally interested to contrib :3 | 18:43 |
golinux | As I said above . . . New effort hasn't launched | 18:44 |
golinux | There will be an announcement if it ever happens. You are welcome to contribute | 18:44 |
mason | KREYREEN: It's on my task list, but I've diverted to something less obnoxious. There was a thread on dng if you want to throw in ideas. | 18:44 |
KREYREEN | link it then :p | 18:45 |
KREYREEN | would be currently interested in makeing devuan image for qubesOS tbh | 18:45 |
fsmithred | go for it | 18:46 |
KREYREEN | abstracting it atm | 18:46 |
specing | Honestly the systemd hate kinda subsided | 18:46 |
KREYREEN | doubt | 18:47 |
KREYREEN | new gen likes systemd because they are stupid imho :p | 18:47 |
golinux | Humans are stupid (but that's for offtopic) | 18:48 |
KREYREEN | meh.. the qubes backend is such a garbage x.x | 18:50 |
fsmithred | wiki discussion starts here: https://lists.dyne.org/lurker/message/20200904.045101.f4c51de2.en.html | 18:50 |
specing | Is there any development on a systemd competitor? Something that implements similar functionality in a proper way? | 18:53 |
fsmithred | s6 | 18:54 |
fsmithred | at least for init and service management | 18:54 |
fsmithred | not for network, mounting, and whatever else systemd wants to do | 18:55 |
mason | specing: Look at launchd, SMF. Those are what systemd wanted to be, and both are free software and hence can be useful with the free kernels/systems available. | 18:58 |
mason | But honestly, sysvinit works fine. | 19:00 |
danuan | just wait till systemd gets a registry and deletes /etc | 19:01 |
danuan | the linux can be free of this old unix crap | 19:02 |
Helle | .... you joke, but gconf -> GSettings & dconf | 19:05 |
mason | KREYREEN: Oh, forgot to mention, there's a moribund #devuan-wiki that will one day spring into living colour once a wiki is up. Also, FWIW, current thrust is twiki/foswiki, possibly using custom apparmor profiles to address some of their code structuring issues. | 19:05 |
mason | But I'm working on a repository collider first. | 19:06 |
danuan | Helle yes it seems modern day linux distributions have moved in direction of windows , where it seems very hard to build a minimal server without xwindows interface being a dependancy for one thing or another. | 19:16 |
Helle | danuan: I mean I am actually not too angry about X libs being pulled in | 19:17 |
Helle | so far I've been without X on the server though | 19:17 |
Helle | hell, the main thing that draws in X is the fact I install a full version of vim (which has an X dependency just for clipboards....) | 19:17 |
danuan | even with debian/devuan you have to be picky and careful about installing things , or it pull in a shitload | 19:18 |
Helle | danuan: I mean disabling recommends is important | 19:18 |
suavedandy | fsmithred: So I followed these instructions: https://dev1galaxy.org/viewtopic.php?id=3258 | 19:34 |
suavedandy | And the installer still asks what file system I should format the partition in. | 19:35 |
fsmithred | see the config file to disable that | 19:35 |
fsmithred | no_format=yes (I think) | 19:36 |
suavedandy | fsmithred: Alright, will search for it. | 19:38 |
fsmithred | 55:#no_format="yes" | 19:38 |
fsmithred | uncomment that | 19:39 |
IanJ | Is there a way to --no-install-recommends by default? | 19:54 |
xinomilo | create file in /etc/apt/apt.conf.d/ with 'APT::Install-Recommends "false";' | 19:57 |
IanJ | Thanks xinomilo :) | 19:58 |
IanJ | It was starting to wear having to type that each time I installed something :) | 20:01 |
fsmithred | aptitude -R ... | 20:06 |
IanJ | Don't have aptitude installed, never had a need for it. | 20:14 |
IanJ | Some of this stuff has crazy dependancies. | 20:15 |
IanJ | I didn't install cpp, but if I try to remove it, it wants to uninstall xorg. | 20:16 |
xinomilo | xorg is a metapackage | 20:17 |
IanJ | Maybe it is, but what has removing a c pre processor got to do with xorg? | 20:18 |
n4dir | aptitude why cpp | 20:20 |
n4dir | will probably only lead that far | 20:20 |
fsmithred | apt rdepends cpp | 20:21 |
fsmithred | will tell you what depends on cpp | 20:21 |
fsmithred | g++, build-essential | 20:22 |
fsmithred | gcc | 20:22 |
danuan | aptitude is nice for managing things from an interface , took me a while to switch from dselect as i was too used to that | 20:23 |
fsmithred | oh, I just do aptitude on command line. I don't like the interface because it reminds of dselect. | 20:23 |
fsmithred | I had a really hard time with that, coming to debian from suse. | 20:24 |
IanJ | https://82.46.16.105/V/QBY8H7 | 20:24 |
IanJ | I just don't understand why a pre-processor, something you would use to compile. Would be needed by compiled binaries of any kind? Maybe I'm missing something... | 20:25 |
n4dir | i sure heard and experienced weird stuff as soon metapackages are involved, but not that. Sounds weird indeed | 20:29 |
IanJ | *nods* How packge management works is some kind of voodoo. I have no idea how it all works I just know sometimes it has some strange ideas :) | 20:30 |
n4dir | if i simulate a removal of cpp it wants to remove xorg too. Not that info would be of any use for you, but yeah, confirmed | 20:31 |
danuan | do it enough times and make a procedure list of what to install and what to skip, i do just apt for xless servers , but aptitude when doing a full blown desktop (otherwise it takes forever ) | 20:35 |
n4dir | sounds to me like a different question/problem | 20:37 |
IanJ | n4dir: who would be responsible for defining the dependencies, the package maintainer of cpp on debian? | 20:39 |
n4dir | now as far i can tell it is the dependency listed in the file debian/rules ? For sure a file in the debian folder of the source code | 20:41 |
n4dir | else apt-cache show <packagename> lists those dependencies | 20:41 |
n4dir | and for xorg it sure doesn't mention cpp | 20:41 |
suavedandy | fsmithred: Found it with a text search. Have no idea how to find a specific line in Nano. | 20:42 |
suavedandy | I wanted to clarify. Do you type the same partition name instead of pressing ENTER when the installer asks you for the home partition? | 20:43 |
fsmithred | it's on line 55 as indicated by the sed output I posted. Not very far to scroll down in a text editor. | 20:43 |
fsmithred | I don't remember how to do the btrfs install | 20:44 |
fsmithred | whatever I wrote, whatever the other guy wrote | 20:44 |
suavedandy | fsmithred: mount -o subvolid=258 "$home_part" /target ; check | 20:47 |
suavedandy | That's one of the lines I had to change. | 20:47 |
fsmithred | that looks a little familiar | 20:47 |
suavedandy | The only difference is "-o subvolid=258" part. | 20:48 |
fsmithred | and there's a different subvol for root? | 20:49 |
suavedandy | The home partition mounting doesn't happen when you skip choosing the home partition, right? As far as I remember from checking the config, it was inserted into some form of construction. | 20:49 |
suavedandy | Yes. | 20:49 |
n4dir | IanJ: i misunderstood your question, obviously. Yes, the maintainer of a package. | 20:50 |
fsmithred | right | 20:50 |
suavedandy | An if statement, perhaps? | 20:50 |
fsmithred | both subvols are part of /dev/mapper/storage? | 20:50 |
suavedandy | Yes. | 20:50 |
fsmithred | might need to disable a test to put the same thing in for both | 20:50 |
IanJ | n4dir: I will contact them, see if they can take a look at it. | 20:50 |
fsmithred | hang on while I look | 20:51 |
n4dir | IanJ: well, the problem is that cpp is not listed as a dependency of xorg, and the other way around neither. | 20:51 |
suavedandy | fsmithred: The home subvolume still needs to be mounted seperately tho. | 20:51 |
n4dir | I fooled with apt rdepends, apt-rdepends and apt-cache depends, but sure can't make any sense of it (i hardly use those tools) | 20:51 |
suavedandy | Because of "-o subvolid=…" | 20:52 |
fsmithred | you took care of that by editing the mount commands, right? | 20:52 |
suavedandy | Yes. | 20:52 |
fsmithred | looks like home_dev gets tested for ending in a digit, which is not fatal, and also checks that it's not the same partition as the esp (efi partition) | 20:54 |
fsmithred | so you should be good to put the same logical partition for both | 20:54 |
fsmithred | dev/mapper/storage | 20:54 |
fsmithred | n4dir, apt-cache depends will tell you the deps for the package you name | 20:55 |
fsmithred | rdepends will tell you what packages depend on the one you name. You have to think backward when you read the output.. | 20:56 |
n4dir | yeah, that i got. But as the man page of apt-rdepends mentioned it, i gave it a try | 20:56 |
n4dir | apt rdepends and apt-rdepends differ quite a bit too | 20:56 |
n4dir | mainly apt rdepends gives me apps i haven't installed | 20:57 |
fsmithred | I don't have apt-rdepends | 20:57 |
fsmithred | how did you get it? | 20:58 |
n4dir | try playstore or ebay | 20:58 |
fsmithred | it's in the repo | 20:58 |
n4dir | ha ha. sorry. i thought you were kidding. just install | 20:58 |
fsmithred | I've never heard of it | 20:58 |
n4dir | i never heard of apt rdpedends. | 20:58 |
fsmithred | apt-cache rdepends has been around for a long time | 20:58 |
n4dir | but was puzzled as it listed xdm. checked, not installed, installed apt-rdepends, that looks more like what i got | 20:58 |
n4dir | still no clue bout the relation of cpp and xorg though | 20:59 |
fsmithred | yeah, I saw xdm and wondered about that | 20:59 |
n4dir | this install is 3 months old? So as you see: i only need it if i fool around. | 20:59 |
n4dir | most of the time i get away with aptitude why | 20:59 |
n4dir | well: i hardly de-install programs | 21:00 |
suavedandy | Still wondering if BtrFS's compression works with encrypted volumes. After all, the encrypted volume appears as random data… probably… | 21:39 |
mason | suavedandy: It should work anywhere/everywhere. | 21:49 |
suavedandy | Okay. I have no idea how encryption works on GNU/Linux then… | 21:51 |
fsmithred | normally you encrypt a partition and then open it and create a filesystem on it. (single filesystem or lvm with multiple logical volumes.) | 21:56 |
suavedandy | fsmithred: Turns out you can just mount one top subvolume with two subsubvolumes. | 22:00 |
sixwheeledbeast | It depends how you set it up i suppose | 22:00 |
suavedandy | So THAT'S why everyone was saying you must have a top volume. | 22:00 |
fsmithred | sorry, I don't understand | 22:00 |
suavedandy | You see, nested subvolumes get mounted with the top volume. | 22:01 |
fsmithred | since btrfs is a filesystem, you probably want to put that inside an encrypted partition. | 22:01 |
suavedandy | So you can switch subvolumes around that way kinda. | 22:02 |
fsmithred | https://btrfs.wiki.kernel.org/index.php/FAQ#Does_btrfs_support_encryption.3F | 22:02 |
suavedandy | Or maybe not. | 22:03 |
suavedandy | Anyway. | 22:03 |
suavedandy | https://help.ubuntu.com/community/btrfs | 22:03 |
suavedandy | Here. | 22:03 |
suavedandy | It says they have a toplevel subvolume in their layout. | 22:04 |
fsmithred | there's nothing about encryption on that page | 22:04 |
suavedandy | It prob works. I remember installing Ubuntu with BtrFS and encryption just fine. | 22:05 |
suavedandy | It has automatic subvolumes like on OpenSUSE. But it's simpler, as you can see. | 22:06 |
fsmithred | the kernel.org page explains the different ways of encrypting with btrfs | 22:07 |
suavedandy | The @ subvolume is mounted to / using the kernel boot option rootflags=subvol=@ | 22:07 |
suavedandy | The @home subvolume (if it is used), is mounted via the mount option subvol=@home in fstab. | 22:07 |
suavedandy | fsmithred: Not talking about the encryption. Talking about the subvolume layout. | 22:08 |
fsmithred | if you wanted to use luks, make one big partition on the drive, encrypt it, open it, and then do the btrfs setup, then install the system. | 22:09 |
suavedandy | I'm saying that mounting two subvolumes is unnecessary. | 22:09 |
miskatonic | i better stick to good old ext* | 22:10 |
fsmithred | miskatonic, me too. | 22:10 |
suavedandy | miskatonic: The only reason why we have so much fiddling is because Debian/Devuan isn't Ubuntu and doesn't provide installation on BtrFS root in its installer. | 22:11 |
fsmithred | suavedandy, are you saying that you don't need to have /home in fstab? | 22:11 |
suavedandy | Yep. | 22:11 |
fsmithred | ok, I guess you can find out pretty fast if that works or not | 22:11 |
fsmithred | add #, save and reboot. | 22:12 |
suavedandy | fsmithred: No, I was very wrong. | 22:35 |
suavedandy | I read the instructions wrong. | 22:35 |
suavedandy | The Ubuntu's article just stated that level 5 subvolumes are mounted first. It didn't say only one subvolume needs to be mounted. | 22:36 |
suavedandy | The BtrFS installation sure requires some hard brain flexing. | 22:46 |
fsmithred | and what do you get for that extra work? | 23:01 |
mason | suavedandy: Why did you move off of ZFS? | 23:16 |
mason | I might have asked this previously, but I'm not remembering. | 23:17 |
suavedandy | Well, I wasn't very successful with it. | 23:21 |
fsmithred | mason, have you got a minute or two to meet me in video chat? I want to test new chromium properly. | 23:27 |
suavedandy | fsmithred: That article seems to be quite helpful in terms of setting BtrFS. https://mutschler.eu/linux/install-guides/ubuntu-btrfs/ | 23:27 |
suavedandy | It's for Ubuntu but, I mean, Debian-based. | 23:28 |
suavedandy | So it applies. | 23:28 |
suavedandy | At least partly. | 23:29 |
fsmithred | btrfs inside luks is what I assumed would work. | 23:29 |
mason | fsmithred: Sure, yes. Devuan meetingspace? | 23:31 |
fsmithred | yeah | 23:31 |
fsmithred | on my way | 23:31 |
suavedandy | There's no other way of encrypting BtrFS. Unlike ZFS which has native encryption. | 23:31 |
mason | I still use LUKS with ZFS. I don't like the native encryption. | 23:32 |
suavedandy | fsmithred: So apparently every entry in my fstab is mounted with UUID. | 23:40 |
suavedandy | Is that something required for GPT? | 23:40 |
Helle | no, but it is common | 23:44 |
mason | suavedandy: I much prefer human-readable GPT labels. /dev/disk/by-partlabel/foo | 23:44 |
suavedandy | "/dev/disk/by-partlabel/Linux\x20filesystem" | 23:48 |
suavedandy | Now thats pretty | 23:48 |
fsmithred | yeah, uuid is standard now, and it's good if you have multiple disks, in case they come up in a different order. | 23:51 |
fsmithred | but uuid sucks to read | 23:51 |
fsmithred | I use /dev/mapper names if they exist | 23:51 |
fsmithred | and I have used labels in the past. They're nice. | 23:51 |
suavedandy | fsmithred: Can't you just /dev/device_name? | 23:54 |
fsmithred | not on an encrypted partition | 23:54 |
fsmithred | sorry | 23:54 |
fsmithred | encrypted filesystem | 23:54 |
suavedandy | I was referring to ESP. | 23:54 |
fsmithred | not sure how I do that. I'll check. | 23:55 |
fsmithred | on one laptop /dev/sda1 is mounted to /boot/efi, on the other laptop the esp is not mounted. | 23:56 |
fsmithred | since both are laptops and only have one drive, there's no chance of being out of order | 23:56 |
fsmithred | desktop is bios boot | 23:56 |
suavedandy | And what about a desktop with one drive? | 23:57 |
fsmithred | I use dev names in that case. Also in my VM tests. | 23:58 |
fsmithred | least work involved | 23:58 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!