systemdlete | phogg: Your suggestions inspired me to look for more efficiency. I took your ideas and applied them to my own style and approach. I manageed to eliminate calls to ls, grep, etc. without resorting to functions (other than the trapped cleanup routine, which sort of needs to be a function) | 00:23 |
---|---|---|
systemdlete | I figured out I could use "echo" rather than "ls" for most of the calls (but not all) since it is really just the file/directory globbing I was after. | 00:24 |
systemdlete | (and it all still works on all of my systems) | 00:24 |
systemdlete | https://paste.debian.net/1292085/ | 00:30 |
systemdlete | nvm last paste; more changes coming... | 01:02 |
systemdlete | https://paste.debian.net/1292090/ | 02:25 |
systemdlete | Now, I've tried to eliminate as much external calls as possible, but there are still some pipelines I think. | 02:25 |
systemdlete | I managed to get rid of every last "ls" and "grep" call! Took me some hard thinking and clever strategies, but I did it. | 02:26 |
systemdlete | phogg: When you get a chance, you might want to look at this updated version of my script. | 02:26 |
* systemdlete is exhausted now and ready for a nap... | 02:30 | |
ted-ious | systemdlete: That's very clever! | 04:08 |
ted-ious | systemdlete: But what is the script for? :) | 04:08 |
systemdlete | ted-ious, it is part of the software that detects other installed operating systems on a multi-boot computer. | 04:12 |
systemdlete | It was failing primarily, I think (per my tracing of them) because of some incompatibility between mdraid and mdadm raid, which are mostly compatible, but the utility programs seemed to be failing. They were calling "dmraid" and I had difficulty finding it in the repos. | 04:13 |
ted-ious | Oh ok. | 04:16 |
systemdlete | Also, the os-prober had earned itself a place in the software hall of shame due to some security issues. The main one, apparently, was the possibility of someone inserting a thumbdrive and getting itself booted into the system (not sure I understand all of that, but...), so I made sure I excluded any removable physical drives in my version. | 04:18 |
systemdlete | EFI will replace the need for grub and its ilk in the long term, but for the time being, I still have older mainboards that only spport the older BIOSes. And, yes, I know that there are libre ROM replacements, but those are a bit scary. I have extra, older mb's here I could test on I suppose. | 04:19 |
systemdlete | Just not a priority atm. | 04:19 |
* systemdlete is tired; steps afk | 04:24 | |
chomwitt | would firefox-esr in Deadalus ask me to update itself ? | 11:53 |
ted-ious | I don't think any devuan or debian firefox tries to update itself. | 12:13 |
ted-ious | That's supposed to be done by apt. | 12:13 |
DFP | Why do Devuan/Debian primarily use names for versions, instead of just numbers? I forget each time which one is which. | 12:49 |
gnarface | DFP: not exactly a support question, but it gets easier with practice and there's a good reason but i can't explain it right now | 13:23 |
gnarface | what's important is that daedalus is the current one | 13:24 |
gnarface | join #devuan-offtopic and i'm sure you can get people to philosophize with you about why names are important and why they're different from a number or a title | 13:26 |
DFP | Alright, thanks. | 13:29 |
concordia | Hello, where can I see the current merge requests for new packages ? | 13:45 |
rrq | it varies depending on the package; generally the built sources are a combination of upstream sources and patches applied at packaging. | 13:51 |
rrq | basically you'll need to pursue that question for any package individually.. perhaps start at https://pkginfo.devuan.org | 13:55 |
rrq | or (with deb-src points in sources.list), you can use "apt-get source $package" to get the source unpacked into a sub-directory. There you typically can inspect the Vcs-Git field (if any) of the debian/control file | 14:00 |
rrq | that brings you to the packaging project | 14:01 |
concordia | rrq, thanks but I'm not a dev, I only wanted to know if there was some ongoing work to add obmenu-generator & urxvt-perls to devuan. If not, I'm not sure I'll be able to do it myself. | 14:07 |
concordia | Even though there is some 3rd party repo for obmenu-generator iirc, but I don't know if that's recommended to use that. Especially considering it is made for Debian. | 14:09 |
fsmithred | that last fact makes it more likely to work. Devuan is 99% pure Debian. | 14:11 |
concordia | fsmithred, alright then I'll give it a try. | 14:11 |
fsmithred | we don't generally add packages, we just fork the ones that require systemd. | 14:12 |
fsmithred | I've seen obmenu-generator mentioned on the forum multiple times. There might be some useful information there. | 14:12 |
fsmithred | dev1galaxy.org | 14:13 |
concordia | I'll have a look, thanks. I have an other unrelated question. I use openrc but I noticed wpa_supplicant init file is missing. I added dhcpcd to default runlevel. wired networks works fine (and starts automatically) but not wpa_supplicant. | 14:16 |
concordia | I have to manually issue : "wpa_supplicant -Dnl80211 -iwlan0 -c/etc/wpa_supplicant/wpa_supplicant.conf" | 14:16 |
concordia | On gentoo, /etc/init.d/wpa_supplicant is provided ( https://dpaste.com/228ZQV358.txt ) but maybe things are done differenly on Devuan ? | 14:19 |
yann-kaelig | Hello | 14:58 |
yann-kaelig | Is that true ? Alsa-firmware are not packaged for Debian ? | 14:58 |
buZz | likely they are in firmware-linux or something, i'd guess | 15:01 |
buZz | there -is- a debian package called alsa-firmware-loaders though | 15:01 |
yann-kaelig | buZz: Hi. Well, That the first things I have done, no alsa firmwares at all. The one I'm looking is multiface_firmware_rev11.bin | 15:06 |
yann-kaelig | apt-file search output nothing | 15:06 |
buZz | i think its in firmware-linux-nonfree. | 15:08 |
buZz | its mentioned on https://wiki.debian.org/Firmware/List | 15:08 |
buZz | but not sure | 15:09 |
buZz | if your driver is already reporting looking for it, you could just plonk the file in /lib/firmware somewhere ;) | 15:09 |
yann-kaelig | What does that mean : Download is performed unsandboxed as root as file '/home/ykaelig/waveform_64bit_v12.5.11.deb' couldn't be accessed by user '_apt'. - pkgAcquire::Run (13: Permission denied) | 15:34 |
fsmithred | yann-kaelig, I don't know the full meaning of those words, but I get the same message whenever I do apt download as root. And I've been ignoring that message for about 10 years. | 15:36 |
yann-kaelig | really weird. This softwar ework as expected on Windows but I get Illegal Instruction on Linux platform because it has been built with more strict CPU flags on it | 15:36 |
fsmithred | yann-kaelig, if you added a third-party repo to sources.list be very careful about what gets pulled in. | 15:39 |
yann-kaelig | No third-party repo. An official .deb from Traktion DAW | 15:40 |
yann-kaelig | So much issues and complicated things if you want to do simple audio processing on Linux. Everythign is so smooth on my Offline windows10 I have cleaned from all the garbage | 15:40 |
fsmithred | By third-party I mean not devuan repo. | 15:41 |
yann-kaelig | Nothing like that | 15:41 |
yann-kaelig | Few days ago I was trying to install Spleeter from Deezer, on linux it's a pain and headbreaker, on windows everythign is in one standalone exe with an opensource GUI | 15:43 |
yann-kaelig | Anyway. Thx | 15:44 |
yann-kaelig | Ok, So I have installed wavefrome12 with the new packages dependencies, but now when I removed the software and run autoremove not all installed packaged are removed | 15:51 |
yann-kaelig | The following NEW packages will be installed: | 15:51 |
yann-kaelig | ffmpeg lame libavdevice59 libcdio-cdda2 libcdio-paranoia2 libid3tag0 libimlib2 liblo7 libportmidi0 waveform12 xjadeo | 15:52 |
yann-kaelig | The following packages were automatically installed and are no longer required: | 15:52 |
yann-kaelig | lame libid3tag0 libimlib2 liblo7 libportmidi0 xjadeo | 15:52 |
yann-kaelig | ? | 15:52 |
yann-kaelig | why ffmpeg libavdevice59 libcdio-cdda2 libcdio-paranoia2 are not removed ? | 15:53 |
yann-kaelig | Why I lost my previous state before waveform12 installation ? | 15:54 |
ballsxd | hey guys, upon trying to update my packages i get the following error: "Errors were encountered while processing: | 15:57 |
ballsxd | linux-image-6.4.0-4-amd64 | 15:57 |
ballsxd | mpd | 15:57 |
ballsxd | E: Sub-process /usr/bin/dpkg returned an error code (1)" I found that removing these packages with either dpkg or apt didnt work but manually deleting them from /var/lib/dpkg/info does work. is it recommended to do so? | 15:57 |
rrq | possibly there's a file in /etc/grub.d/ with '#!/usr/bin/sh' as it's first line? | 16:00 |
rrq | ballsxd: ^ | 16:01 |
yann-kaelig | Ok so apt-cache rdepends --installed ffmpeg : | 16:01 |
yann-kaelig | Reverse Depends: imagemagick-6.q16 < ffmpeg:i386 | 16:01 |
yann-kaelig | Why ? Just why is this software taking the right on ffmpeg. I never asked for that | 16:02 |
yann-kaelig | Is was working as expected before without ffmpeg, why is he trying to change his own state ? | 16:02 |
rrq | how did you get ffmpeg installed? | 16:03 |
yann-kaelig | from waveform12 | 16:03 |
yann-kaelig | as a NEW packages | 16:03 |
rrq | is waveform12 known by apt? | 16:04 |
yann-kaelig | I do not understand your question | 16:04 |
rrq | try: apt-cache policy waveform12 | 16:05 |
yann-kaelig | The package waveform12 has been removed | 16:05 |
yann-kaelig | with after that a autoremove | 16:06 |
rrq | doesn't matter. if it's known by apt, "policy" will tell you | 16:06 |
yann-kaelig | It was a standalone .deb and I installed the package with apt install -f <package.deb> | 16:07 |
yann-kaelig | N: Unable to locate package waveform12 | 16:07 |
rrq | yes, maybe you are right that the autoremove logic imagines ffmpeg being expected by imagemagick-6.q16 and therefore not a candidate for autoremove | 16:10 |
yann-kaelig | Yes, because when I run a remove on ffmpeg then next after that following packages were automatically installed and are no longer required: | 16:11 |
yann-kaelig | libavdevice59 libcdio-cdda2 libcdio-paranoia2 | 16:11 |
yann-kaelig | these one that were installed for waveform12 | 16:11 |
yann-kaelig | Ok, so now back to my previous state, everybody is happy, autoremove bad logic ^^ | 16:14 |
yann-kaelig | Time for a new package manager | 16:14 |
fsmithred | all the other ones have copied apt | 16:15 |
yann-kaelig | But also time for a new distro logic. This old logic where everyone is part of the system doesn't work. Time to redesign the system | 16:17 |
fsmithred | not sure what you mean, but the ultimate in configurability is probably gentoo. | 16:17 |
fsmithred | if you want more like windows there are things like appimage | 16:18 |
fsmithred | every app can have its own libraries | 16:19 |
yann-kaelig | The same issue occur on Gentoo on another level. Opensuse is working on something that I was studying. A core ( main ) minimal system where other software does not interfere with it. | 16:23 |
yann-kaelig | It could be also manager with the help of file system level | 16:24 |
ballsxd | rrq: there are actually many files in there starting with a shebang | 16:25 |
ballsxd | 00_header 10_linux 25_bli 30_uefi-firmware 41_custom | 16:25 |
ballsxd | 05_debian_theme 20_linux_xen 30_os-prober 40_custom README | 16:25 |
ballsxd | however reading any of them doesnt make me any smarter | 16:26 |
yann-kaelig | Well, CU later. Have a nice day/night | 16:26 |
[-_-] | Hiiii | 16:44 |
[-_-] | http://0x0.st/HOZF.txt | 16:44 |
[-_-] | see the dh_usrlocal | 16:44 |
[-_-] | what is the cause of this error? | 16:45 |
buZz | is it a directory? | 16:49 |
buZz | try ls -l home/atom/HWD/deb/st/my-st-debian/st-custom/debian/st-custom/usr/local/bin | 16:50 |
buZz | eh, add a / before | 16:50 |
buZz | might be getting confused by the multiples of st-custom ? | 16:51 |
[-_-] | home/atom/HWD/deb/st/my-st-debian/st-custom/debian/st-custom/usr/local/bin is a directory | 16:55 |
[-_-] | home/atom/HWD/deb/st/my-st-debian/st-custom/debian/st-custom/usr/local/bin/st is not | 16:55 |
[-_-] | so close yet so far | 17:05 |
buZz | [-_-]: well, 'st' would be a file there, i guess? | 17:40 |
buZz | i never really liked using /usr/local/ for anything at all | 17:40 |
buZz | but i guess its the norm, so meh | 17:40 |
[-_-] | it has been fixed | 17:41 |
[-_-] | the problem is with prefix != PREFIX | 17:41 |
[-_-] | https://wiki.debian.org/Packaging/Intro?action=show&redirect=IntroDebianPackaging | 17:41 |
[-_-] | here | 17:41 |
buZz | ah :D | 17:43 |
ksx4system | image for Raspberry Pi that I'm using does not contain "menu-config" script, where do I download it? | 18:40 |
[-_-] | guys do you all use sudo ? | 20:15 |
[-_-] | how do you turn off or restart computer as non-root user? | 20:15 |
rwp | As a non-root user I tap the power button and let it shutdown. | 20:17 |
[-_-] | how do you do it from inside the system? | 20:17 |
rwp | Or for a reboot I will Control-Alt-F1 to get the vt console and then Control-Alt-Delete to trigger "shutdown -r now". | 20:17 |
[-_-] | that can't be done on latest stable because of some bug in Xorg | 20:18 |
rwp | To reboot a remote computer from say an ssh connection then root is required. Either su or sudo to run shutdown. | 20:18 |
rwp | Hmm... New bugs. I hate it when people just keep messing with things! | 20:18 |
rwp | I noticed a few new (serious) bugs in my latest Daedalus Stable install but I did not notice that one. Let me go double check it. | 20:19 |
rwp | Just tested it (ctrl-alt-F1 then ctrl-alt-delete) on Daedalus Stable and it rebooted okay no problem. | 20:21 |
rwp | I have installed and am using xserver-xorg-legacy though so that might be why it is working for me. | 20:22 |
[-_-] | what is that legacy for? | 20:24 |
rwp | And they quit before I could respond. :-( | 20:30 |
rwp | For pedantic completeness I will put this here. https://manpages.debian.org/bookworm/xserver-xorg-legacy/Xorg.wrap.1.en.html | 20:33 |
Xenguy | I still maintain that this type of sudo usage is an (unfortunate) Ubuntu'ism | 20:36 |
ksx4system | any ideas why strongSwan started ignoring private key in letsencrypt certs? it just doesn't see it | 22:53 |
ksx4system | eg. ipsec listcerts does not contain "has private key" | 22:53 |
ksx4system | worked just fine before I've upgraded to daedalus | 22:53 |
gnarface | ksx4system: not sure if it's related but they moved the location of something and i had to move my winehq certs too | 23:18 |
gnarface | .... so it might be related | 23:18 |
gnarface | yea, instead of one big keyring somewhere they're now supposed to be in separate *.gpg files in /etc/apt/keyrings/ | 23:19 |
gnarface | i didn't write down the command to extract the key from the bigger file but you should be able to find it somewhere near the reference to the migration to /etc/apt/keyrings/ | 23:20 |
gnarface | (i also had to add it to my sources.list for the winehq entry like this: deb [arch=amd64,i386 signed-by=/etc/apt/keyrings/winehq.gpg] http://dl.winehq.org/wine-builds/debian/ bookworm main) | 23:20 |
ksx4system | gnarface, it tries to read key from correct location and logs that it failed | 23:21 |
gnarface | hmm, dunno then | 23:21 |
ksx4system | 2023-09-16T23:06:08.652936+02:00 datura charon: 00[CFG] loading private key from '/etc/ipsec.d/private/key.pem' failed | 23:21 |
ksx4system | verbose as hell | 23:21 |
gnarface | hmm. this was the only related change i can recall on the switch to daedalus | 23:22 |
ksx4system | it looks like newer strongSwan version just refuses to work with Letsencrypt certs | 23:22 |
gnarface | could have been just a key format change perhaps? i've seen that once or twice in the past | 23:22 |
ksx4system | gnarface, I've used whatever certbot generated | 23:23 |
ksx4system | I don't even know what format is it ;) looks like SSL certificate used to look tho | 23:23 |
gnarface | i've never used certbot, i just remember that i had to recreate keys by hand here and there over the years as a response to various emergency security updates to openssl and gnutls | 23:24 |
gnarface | sometimes the old ones would be incompatible without actually looking visually different in format | 23:24 |
gnarface | doesn't happen very often | 23:25 |
ksx4system | I guess something like this happened | 23:25 |
gnarface | i think those were mostly for much older keys than daedalus though | 23:25 |
ksx4system | gnarface, brand new SSL keys generated on daedalus (I had to add another subdomain) | 23:25 |
gnarface | hmmm.... | 23:26 |
ksx4system | gnarface, the key format is the same, I've checked Letsencrypt docs and strongSwan docs | 23:36 |
* ksx4system would have a Fortigate long time ago if he could afford it | 23:37 | |
gnarface | huh | 23:37 |
gnarface | well i wonder if they just decided to ignore letsencrypt... seems like that would have been news though | 23:37 |
ksx4system | gnarface, SSL certificate itself is just fine btw, I've verified that with website hosted on lighttpd | 23:38 |
ksx4system | Chrome says this certificate is perfect | 23:38 |
ksx4system | gnarface, apparently letsencrypt switched from RSA to ECDSA | 23:42 |
ksx4system | key format is the same | 23:42 |
ksx4system | algorithm is not ;) | 23:42 |
gnarface | ah, there we go | 23:43 |
gnarface | i knew it had to be something like that | 23:57 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!