| plasma41 | eyalroz: I have a feeling redrick would be happy to give you his answer to the last question. | 00:26 |
|---|---|---|
| fsmithred | eyalroz, the dns thing was a wild guess. I looked at the url for the video and it was youtu.be | 00:33 |
| fsmithred | and if you want to change it, change it to anything but your ISP | 00:34 |
| fsmithred | assuming they're screwing with results | 00:34 |
| fsmithred | did you try purging and reinstalling chromium? | 00:36 |
| systemdlete | I downloaded the source to sudo hoping to match the debug logging output with the source code lines. I notice that debian has patches, and I want to know whether/how to apply these patches. The patches are located in a separate subdir (from the "debian" tar file) from the subdir (from the "orig" tar file). | 04:12 |
| systemdlete | Is this just a straightforward matter of using patch(1) to apply these, or is that the wrong way to do this? | 04:12 |
| specing | there is an apt gimme sources command | 04:12 |
| systemdlete | I am not familiar with the exact complexity of the process, sorry. | 04:12 |
| fsmithred | I'm no expert, but you list the patches you want to apply in a file | 04:12 |
| systemdlete | Already have them. | 04:12 |
| fsmithred | I'll remember the name... series (I think) | 04:13 |
| systemdlete | fsmithred: yes, series | 04:13 |
| systemdlete | so I just apply those? | 04:13 |
| fsmithred | they get applied when you build the package | 04:13 |
| systemdlete | ah! | 04:13 |
| systemdlete | Of course. But I don't really need to build it... yet. | 04:13 |
| systemdlete | At least, I hope the source can reveal for me the trouble I am having. | 04:14 |
| systemdlete | I'm trying to be bold-and-brave and figure out why I cannot get sudo to work as the manual pages instruct. I've tried many things, to no avail. | 04:14 |
| fsmithred | what's the problem? | 04:15 |
| systemdlete | I just ran make, but I can see already that patches have not been applied. So I have not done this right? | 04:18 |
| fsmithred | no, not even close | 04:18 |
| fsmithred | you need to read about debian packaging | 04:18 |
| fsmithred | maybe try 'dpkg-buildpackage -us -uc -b' | 04:19 |
| systemdlete | fsmithred: The problem is that I want to run ONE command as root without prompt for password | 04:19 |
| fsmithred | no, run as user | 04:19 |
| systemdlete | aha! | 04:19 |
| systemdlete | thanks | 04:19 |
| fsmithred | install fakeroot | 04:19 |
| fsmithred | you'll need to install a few things | 04:19 |
| systemdlete | sudo can do this, per the man pages | 04:19 |
| fsmithred | sudo do what? | 04:20 |
| systemdlete | I want to run ONE command as root without prompt for password. | 04:20 |
| fsmithred | that's what you're trying to get sudo to do? | 04:20 |
| systemdlete | So I've turned on debug logging in sudo to try to figure out why it is prompting, even though I expressed "NOPASSWD:" in the sudoers file | 04:20 |
| systemdlete | yes, that is it... | 04:21 |
| fsmithred | I have about a dozen commands set up for that | 04:21 |
| systemdlete | So I added a line in visudo: myuser mypc= NOPASSWD: /path/to/program | 04:22 |
| fsmithred | with command aliases | 04:22 |
| fsmithred | mine looks pretty much like that except ALL instead of mypc | 04:23 |
| systemdlete | command aliases are extra at this point; I just want to make this work for now. I can go back and get fancier later | 04:23 |
| fsmithred | no need for just a single command | 04:23 |
| systemdlete | right | 04:23 |
| systemdlete | just a single command | 04:23 |
| systemdlete | for one user on one machine, without prompt for password | 04:23 |
| systemdlete | I mean, I can change mypc to "ALL" and same problem | 04:24 |
| fsmithred | ok | 04:24 |
| systemdlete | this is on ascii, btw, with all updates applied | 04:24 |
| fsmithred | I'm on jessie, but I've got similar on ascii and beowulf | 04:24 |
| systemdlete | I've read through a lot of man pages, and docs I've found around the web. | 04:25 |
| systemdlete | I feel like I should be able to do this one simple thing. | 04:25 |
| fsmithred | here's what I have in /etc/sudoers.d/user_shutdown: | 04:26 |
| fsmithred | user ALL= NOPASSWD: /usr/sbin/pm-suspend, /usr/sbin/pm-hibernate, /sbin/halt, /sbin/reboot | 04:26 |
| fsmithred | works in jessie, ascii, beowulf | 04:26 |
| fsmithred | (worked in wheezy and squeeze, too) | 04:27 |
| fsmithred | you did log out and log in again after making the changes? | 04:27 |
| systemdlete | I tried this in the main sudoers file, using visudo. | 04:27 |
| systemdlete | I opened a new terminal | 04:28 |
| fsmithred | I don't think that's enough | 04:28 |
| systemdlete | how about if I try logging in at a console terminal? | 04:28 |
| fsmithred | that should work | 04:29 |
| systemdlete | Ok, that way I don't need to disconnect from hexchat and all | 04:29 |
| systemdlete | nope. | 04:30 |
| systemdlete | no better from console login | 04:31 |
| fsmithred | didn't work here, either | 04:32 |
| fsmithred | log out and log in | 04:32 |
| systemdlete | well, logging in from a console is, essentially, doing that: logging out and logging in. Except there's no log out part. | 04:32 |
| fsmithred | Then it should have worked here | 04:33 |
| fsmithred | but it didn't | 04:33 |
| systemdlete | I'll create a 2nd user just for testing. Add that to the sudoers, then log in from console as that new user | 04:33 |
| systemdlete | try it there | 04:34 |
| systemdlete | or, I could try this on my testbox... | 04:34 |
| onefang | I agree with fsmithred, that's the same sort of config I have, on ASCII, and it works fine. | 04:34 |
| systemdlete | shall I conclude that my box is cursed? | 04:35 |
| systemdlete | I mean, there's got to be some explanation | 04:35 |
| onefang | nagios ALL=NOPASSWD: /bin/btrfs device stats * | 04:35 |
| systemdlete | maybe pam or some other security subsystem interfering? | 04:35 |
| fsmithred | doh! | 04:35 |
| systemdlete | doh? | 04:35 |
| systemdlete | what? | 04:35 |
| fsmithred | I'll bet you're trying to run a command that's not in your path | 04:35 |
| fsmithred | no, I'm trying to do that | 04:36 |
| systemdlete | I gave the explicit path in the sudoers, and use the explicit path to invoke in sudo | 04:36 |
| golinux | Maybe this? Haven't read all the scrollback . . . http://forums.debian.net/viewtopic.php?f=16&t=38976 | 04:37 |
| golinux | Just for reference purposes if nothing else. | 04:37 |
| systemdlete | thanks, golinux. I'll take a look at that if we can't figure this out here. | 04:38 |
| fsmithred | OK, it does work here if I make a change and then log in on console without logging out of desktop | 04:39 |
| fsmithred | and I don't need to give the path in the command | 04:39 |
| fsmithred | sudo fdisk -l worked | 04:39 |
| systemdlete | "make a change?" | 04:39 |
| fsmithred | yeah, I added /sbin/fdisk to the list of commands | 04:40 |
| systemdlete | ok, let me try something like that | 04:40 |
| systemdlete | hmmm. still not working for me | 04:42 |
| systemdlete | I added /sbin/fdisk to the list for the same user, logged in from console, no dice | 04:42 |
| systemdlete | still wants passwd | 04:42 |
| fsmithred | your hostname is in /etc/hosts? | 04:43 |
| systemdlete | yes | 04:43 |
| systemdlete | ok, let me try ALL instead... | 04:43 |
| systemdlete | no difference, even if I log in from console | 04:45 |
| fsmithred | try putting it in sudoers.d/something | 04:45 |
| systemdlete | I've never done that. Do I still use visudo? | 04:45 |
| fsmithred | nope | 04:45 |
| systemdlete | just plain editor | 04:45 |
| fsmithred | just edit as root | 04:45 |
| fsmithred | one line | 04:46 |
| systemdlete | omg | 04:47 |
| systemdlete | it worked | 04:47 |
| systemdlete | (this time at least) | 04:47 |
| fsmithred | :) | 04:47 |
| systemdlete | ??? | 04:47 |
| systemdlete | why | 04:47 |
| fsmithred | that's a smile | 04:47 |
| systemdlete | I know. And ??? is perplexed and confused | 04:48 |
| systemdlete | (as well as utterly amazed) | 04:48 |
| systemdlete | I didn't even have to do a console login! | 04:48 |
| fsmithred | I don't know. I used to put my lines in /etc/sudoers and they worked. | 04:48 |
| fsmithred | huh? | 04:49 |
| systemdlete | so, addding it to a include file works. Adding to /etc/sudoers doesn't? | 04:50 |
| systemdlete | that's why I asked "why?" | 04:50 |
| fsmithred | I don't know the answer. It might be hidden in the man page somewhere. | 04:51 |
| systemdlete | checking that now | 04:51 |
| systemdlete | the only thing I see is this: " Note that unlike files included via #include, visudo will not edit the files in a | 04:52 |
| systemdlete | #includedir directory unless one of them contains a syntax error. It is still possible to | 04:52 |
| systemdlete | run visudo with the -f flag to edit the files directly" | 04:52 |
| onefang | Ah yes, I put mine in the include directory, and I don't use visudo. | 04:54 |
| systemdlete | onefang: love it | 04:54 |
| systemdlete | so, I conclude that visudo is necessary to check syntax, but if you don't use visudo, you should, even though there's fewer issues like mine | 04:55 |
| systemdlete | or something even more contorted and ridiculous than that | 04:56 |
| onefang | Meh, if it works, I know the syntax is correct. B-) | 04:56 |
| * systemdlete takes his head and proceeds to bash it against something very, very hard | 04:56 | |
| systemdlete | onefang: indeed | 04:56 |
| systemdlete | In the header of the sudoers file: "Please consider adding local content in /etc/sudoers.d/ instead of directly modifying this file." | 05:00 |
| systemdlete | ok. So, thanks to onefang and fsmithred. | 05:01 |
| fsmithred | yw | 05:01 |
| onefang | Coz /etc/sudoers might get updated by apt. | 05:01 |
| onefang | You are welcome. | 05:02 |
| systemdlete | I don't quite follow the logic of why this works, but... ugh. OK. | 05:02 |
| systemdlete | you guys are great, really | 05:02 |
| systemdlete | I just wish some of these tools were a bit more... idk, consistent? predictable? kind of, what one would expect them to do, rather than some bizarre result if you don't read every last bit of literature ever written about them? | 05:03 |
| * furrywolf wonders why systemdlete is bashing his head against poettering's head | 05:03 | |
| systemdlete | Ah, I know: "obvious." That's the word | 05:03 |
| systemdlete | furrywolf: Because (if you didn't already know this) when I finally stop, it will feel much better. | 05:04 |
| systemdlete | (not sure about Mr. P) | 05:04 |
| onefang | Just remember that all of these tools are written by random people, for random reasons. | 05:06 |
| systemdlete | The man page should say, somewhere, preferably in the section about included files, that things work differently in the external files than in the main sudoers file | 05:06 |
| systemdlete | So I should expect random results sometimes. LOL! | 05:06 |
| fsmithred | most of these things have been around for decades | 05:07 |
| fsmithred | the *.d thing is fairly recent | 05:07 |
| fsmithred | give the docs a few more years to catch up | 05:08 |
| systemdlete | https://www.sudosatirical.com/articles/systemd-set-to-declare-independance/ | 05:10 |
| systemdlete | I'm not the only fool who has fallen into the jaws of sudo, see: https://serverfault.com/questions/883434/sudo-nopasswd-not-working-on-a-specific-command | 05:12 |
| systemdlete | and here, http://toroid.org/sudoers-syntax, apparently it is acknowledged that not everything is quite what one would expect: " if you take the EBNF in the manpage seriously enough, you will discover that the implementation doesn't follow it." | 05:14 |
| systemdlete | I post these as declarations that I am not (quite) crazy. (Yet.) | 05:15 |
| systemdlete | I may have found the culprit, here: https://help.ubuntu.com/community/Sudoers which says "The sudoers file is read in one pass so when multiple entries match for a user, they are applied in order. Where there are conflicting values, the last match is used (which is not necessarily the most specific match)." | 05:27 |
| fsmithred | oh | 05:28 |
| systemdlete | I had my user spec before the sudo group rule! | 05:28 |
| fsmithred | you're in there more than once? | 05:28 |
| systemdlete | not exactly. My user is part of the sudo group | 05:28 |
| fsmithred | ah, makes sense | 05:29 |
| systemdlete | so the first match is on my user spec, but it ignores that when it matches the sudo group spec | 05:29 |
| systemdlete | sense, finally. | 05:29 |
| systemdlete | but even then, not what I'd expect. I'd think the first rule that matches would be the effective rule. | 05:29 |
| fsmithred | I think the guy in the serverfault thread didn't use the right line in his sudoers | 05:29 |
| fsmithred | nah, the last | 05:30 |
| systemdlete | why the last? | 05:31 |
| onefang | So the stuff in the includes gets to override it. | 05:31 |
| systemdlete | ah, true. | 05:31 |
| systemdlete | which is what ultimately happened in my case. | 05:31 |
| systemdlete | but includes are recent | 05:31 |
| systemdlete | but anyway, this is good to know. Ties it all together, finally. Thanks again | 05:32 |
| fsmithred | I learned something | 05:32 |
| systemdlete | from the sudoers man page (which I did read, and I remember reading this part, but it just didn't quite hold in my brain): "When multiple entries match for a user, they are applied in order. Where there are multiple matches, the last match is used (which is not necessarily the most specific match)." | 05:34 |
| fsmithred | like I said - hidden in the man page | 05:35 |
| systemdlete | I guess at the point I was reading that section, I had not yet observed that the user in question was part of the sudo users group, or that a sudo users group existed (I know I should have known that sudo group existed, but it just wasn't in my main cache at that moment) | 05:36 |
| systemdlete | yes. | 05:37 |
| systemdlete | hidden in plain sight. For anyone who actually READS. And I really need to improve my man page reading skills. I do this a lot it seems. | 05:37 |
| * systemdlete kicks himself in the head, twice, just to be sure | 05:38 | |
| onefang | Twice with each foot? | 05:38 |
| systemdlete | no, feet go in mouth. | 05:38 |
| systemdlete | (I feel so dumb...) | 05:38 |
| systemdlete | I actually absorbed the BNF descriptions more than the plain text. | 05:40 |
| onefang | lol | 05:40 |
| systemdlete | Next time I'm in the forest, I will be sure to look at the trees... in great detail. | 05:41 |
| slvr | get better at cache eviction. | 05:41 |
| systemdlete | slvr: How? | 05:44 |
| systemdlete | Read each man page twice from beginning to end, without falling asleep I mean? | 05:44 |
| fsmithred | sounds painful | 05:45 |
| fsmithred | uh-oh, it's past my bedtime | 05:45 |
| fsmithred | g'night | 05:45 |
| onefang | Good night. | 05:46 |
| furrywolf | cyas | 05:46 |
| systemdlete | night fsmithred | 05:47 |
| systemdlete | ok, time for a system test here. Thanks to all again | 05:48 |
| ham5urg | I have installed Devuan-Beowulf/Testing. Libfreerdp is a package I depend and as stated here https://pkginfo.devuan.org/cgi-bin/d1pkgweb-query?search=freerdp&release=beowulf it's version is 2. After installation I took a look and found /usr/lib/x86_64-linux-gnu/libfreerdp-core.so.1.1.0. Just to be sure I installed Testing, I check sources.list which shows: deb http://pkgmaster.devuan.org/merged beowulf main contrib | 08:43 |
| ham5urg | Just curious what happened? | 08:44 |
| Jjp137 | from what I can tell, you need to install freerdp2-related packages if you want to actually use version 2 | 08:54 |
| Jjp137 | there's similar examples of that in Debian but I'm drawing a blank right now... | 08:57 |
| Jjp137 | oh right it's similar to like SDL 1.2 and SDL 2 if you heard of those libraries | 08:57 |
| TauronGuy | Beowulf...what a cool name | 08:59 |
| TauronGuy | is it a bad idea to install to DE's during install? | 09:01 |
| TauronGuy | or will there be conflict | 09:01 |
| Jjp137 | do you mean two* DEs? if so, they may conflict if they use different policykit backends, which is detailed in the release notes | 09:04 |
| Jjp137 | this is assuming ascii of course | 09:04 |
| TauronGuy | yes two lol | 09:04 |
| TauronGuy | yes in ascii | 09:04 |
| Jjp137 | you can get around it by installing the other one manually afterwards | 09:05 |
| Jjp137 | just pick your favorite one for now :) | 09:05 |
| TauronGuy | so can I do KDE/Cinn can coexist | 09:05 |
| TauronGuy | but they are both my fav : | 09:05 |
| TauronGuy | :/ | 09:05 |
| TauronGuy | you're right better to just one DE | 09:05 |
| TauronGuy | i can install another one later | 09:06 |
| Jjp137 | yea | 09:06 |
| ham5urg | I installed Guacamole (compiled from original source code, downloaded via git) and I'm just curious why https://pkginfo.devuan.org/cgi-bin/d1pkgweb-query?search=freerdp&release=beowulf shows freerdp2 as the only option but a look at the filesystem shows /usr/lib/x86_64-linux-gnu/libfreerdp-core.so.1.1.0 | 09:08 |
| Jjp137 | it's a leftover from ascii, which only has 1.1 | 09:08 |
| Jjp137 | (well, unless you had ascii-backports enabled) | 09:08 |
| ham5urg | Is it possible to update pkginfo.devuan.org? | 09:09 |
| Jjp137 | uh that's generated from a script iirc but I don't think there's anything wrong with its results since freerdp2 is indeed the only option on beowulf | 09:10 |
| Jjp137 | did you upgrade from ascii? | 09:10 |
| eyalroz | fsmithred: Changing my ISP didn't work. | 09:11 |
| eyalroz | fsmithred: My DNS I mean. | 09:11 |
| ham5urg | No, I debootstrapped a fresh beowulf.# cat /etc/apt/sources.list | 09:11 |
| ham5urg | deb http://pkgmaster.devuan.org/merged beowulf main contrib | 09:11 |
| Jjp137 | oh hmm | 09:11 |
| TauronGuy | i can't find the beowulf download | 09:12 |
| eyalroz | fsmithred: Purging and re-installing chromium didn't work either. | 09:13 |
| Jjp137 | the most common way to have a beowulf install right now is to change your sources.list and then dist-upgrade | 09:13 |
| eyalroz | And for everyone on this channel - I'm trying to resolve this problem: https://unix.stackexchange.com/q/524317/34868 | 09:13 |
| eyalroz | on beowulf, and any ideas are welcome. | 09:13 |
| pardis | ham5urg: The output of `dpkg -S /usr/lib/x86_64-linux-gnu/libfreerdp-core.so.1.1.0` would be helpful here | 09:14 |
| ham5urg | pardis: libfreerdp-core1.1:amd64: /usr/lib/x86_64-linux-gnu/libfreerdp-core.so.1.1.0 | 09:20 |
| xinomilo_ | eyalroz, no trouble here with chromium. rarely using it, but just tested it with yt and it works. did you try running it from terminal? | 09:21 |
| pardis | ham5urg: Interesting, and what does `apt-cache policy libfreerdp-core1.1` say? | 09:21 |
| pardis | Might need to pastebin that one, it'll probably be several lines | 09:21 |
| xinomilo_ | ham5urg, there's no such package in beowulf. probably leftover from ascii. try purging it if nothing complains | 09:23 |
| ham5urg | pardis: https://paste.debian.net/1087691/ | 09:25 |
| pardis | ok, so that package isn't available from an APT repo on your system | 09:26 |
| pardis | At least that's consistent with what your package search showed | 09:26 |
| ham5urg | https://paste.debian.net/1087692/ | 09:26 |
| pardis | But this means you've either installed it without using APT, or changed your sources.list since installing it | 09:27 |
| pardis | Since that apt-cache policy output confirms that APT does not see it in a repo | 09:27 |
| TauronGuy | whats better lightdm or sddm | 09:27 |
| golinux | TauronGuy: There are no beowulf isos yet. | 09:36 |
| golinux | Official ones at least. | 09:36 |
| golinux | TauronGuy: The DM will possibly depend on which backend your desktop of choice uses. | 09:38 |
| golinux | Read the ascii release notes to get a feel for it. Things are a little more flexible in beowulf but best to pick the most compatible one. | 09:40 |
| TauronGuy | what about KDE | 09:41 |
| TauronGuy | which one would work best | 09:41 |
| ham5urg | pardis: I've installed the packages as mentioned here: https://guacamole.apache.org/doc/gug/installing-guacamole.html#optional-dependencies | 09:42 |
| ham5urg | # apt update | 09:42 |
| ham5urg | Hit:1 http://pkgmaster.devuan.org/merged beowulf InRelease | 09:42 |
| ham5urg | ... | 09:42 |
| ham5urg | # apt install libfreerdp-dev | 09:43 |
| ham5urg | ... | 09:43 |
| ham5urg | See here: https://paste.debian.net/1087693/ | 09:43 |
| ham5urg | Cite from the pastebin: "libfreerdp-dev is already the newest version (1.1.0~git20140921.1.440916e+dfsg1-13+deb9u3)." | 09:44 |
| eyalroz | xinomilo_: Running chromium from the terminal, you mean? | 09:44 |
| eyalroz | No, but sure, why not. | 09:44 |
| golinux | TauronGuy: https://files.devuan.org/devuan_ascii/Release_notes.txt | 09:45 |
| xinomilo_ | ham5urg: https://pkginfo.devuan.org/cgi-bin/d1pkgweb-query?search=freerdp2-dev&release=beowulf | 09:47 |
| xinomilo_ | libfreerdp-dev is from ascii. let it go.. | 09:47 |
| ham5urg | I don't have packages from ascii. Never touched ascii. | 09:48 |
| xinomilo_ | well, you got it from someplace | 09:48 |
| ham5urg | Yes, that's for sure. But neither did I installed it manually nor did I edited sources.list. | 09:49 |
| ham5urg | # cat /etc/apt/sources.list | 09:50 |
| ham5urg | deb http://pkgmaster.devuan.org/merged beowulf main contrib | 09:50 |
| ham5urg | # apt install libfreerdp-dev | 09:50 |
| ham5urg | Reading package lists... Done | 09:50 |
| ham5urg | Building dependency tree | 09:50 |
| ham5urg | Reading state information... Done | 09:50 |
| ham5urg | libfreerdp-dev is already the newest version (1.1.0~git20140921.1.440916e+dfsg1-13+deb9u3). | 09:50 |
| xinomilo_ | # apt purge libfreerdp-dev | 09:51 |
| xinomilo_ | if nothing complains remove it. check dpkg|apt.logs if you want to see how it installed originally | 09:51 |
| xinomilo_ | eyalroz, yes, and check terminal output | 09:52 |
| eyalroz | xinomilo_: Two errors which look harmless: | 09:58 |
| eyalroz | ERROR:browser_dm_token_storage_linux.cc(101)] Error: /etc/machine-id contains 0 characters (32 were expected). | 09:58 |
| eyalroz | ERROR:sandbox_linux.cc(364)] InitializeSandbox() called with multiple threads in process gpu-process. | 09:58 |
| xinomilo_ | yes minor. is hardware acceleration enabled in settings? settings -> advanced settings -> system , use hardware acceleration.. | 10:00 |
| xinomilo_ | i would also suggest running a new chromium profile to check | 10:01 |
| TauronGuy | anyone know how to install brave on devuan | 10:29 |
| eyalroz | xinomilo_: Disabling hardware acceleration makes the video work. But - without audio (which is otherwise working) | 11:25 |
| eyalroz | So you're on to something! | 11:25 |
| eyalroz | xinomilo_: Also, a new profile exhibits the same behavior. | 11:34 |
| eyalroz | I should mention I do see the audio's "peaks" in its stream on pavucontrol, but - no audio comes out. I am Getting audiop with Firefox and indepdent playing of the videos. | 11:34 |
| xinomilo_ | Firefox plays videos with hardware acceleration on? | 11:36 |
| xinomilo_ | and what does this say from chromium https://www.youtube.com/html5 ? | 11:37 |
| eyalroz | xinomilo_: FF plays videos with its default settings. | 11:54 |
| eyalroz | chromium says. on that URL, all [v]'s | 11:54 |
| r3boot | eyalroz: the /etc/machine-id issue is an easy fix btw; uuidgen | sudo tee /etc/machine-id | 12:01 |
| eyalroz | r3boot: 1. So should that not be generated by default to avoid getting errors? 2. I don't have uuidgen on my system, which package is it from? | 12:31 |
| eyalroz | r3boot: 2. Never mind, it's uuid-runtime | 12:32 |
| r3boot | And 1) and this is my uninformed opinion, since I've been out of distro maintenance for a looooong time. /etc/machine-id is afaik mostly used on systemd-based systems, which use the content of that file as a unique machine identifier (and to be fair, most old unix systems have a similar thing, mostly based on the mac address of the first network card on their bus). This in turn caused things like browsers to use that file in part to ... | 12:34 |
| r3boot | ... generate a unique fingerprint (I think/assume. I have no proof of this) | 12:34 |
| r3boot | Back in the day it was mostly useful for licensing btw | 12:34 |
| fsmithred | ham5urg, how did you install beowulf? | 12:57 |
| fsmithred | eyalroz, same question - how did you install beowulf? (you probably told me already and I forgot) | 13:28 |
| fsmithred | I'm wondering how you got /etc/machine-id | 13:28 |
| ham5urg | fsmithred: I used https://www.ibiblio.org/refracta/ what is looking like a mistake. The people on this channel are right, even as I have beowulf in my sources.list, my system is not a beowulf-devuan. | 13:45 |
| fsmithred | how is it not beowulf? | 13:46 |
| fsmithred | there are newer refracta isos at refracta.org. | 13:47 |
| ham5urg | # apt-cache policy libfreerdp-dev | 13:48 |
| ham5urg | libfreerdp-dev: | 13:48 |
| ham5urg | Installed: 1.1.0~git20140921.1.440916e+dfsg1-13+deb9u3 | 13:48 |
| ham5urg | Candidate: 1.1.0~git20140921.1.440916e+dfsg1-13+deb9u3 | 13:48 |
| ham5urg | Version table: | 13:48 |
| ham5urg | *** 1.1.0~git20140921.1.440916e+dfsg1-13+deb9u3 100 | 13:48 |
| ham5urg | 100 /var/lib/dpkg/status | 13:48 |
| fsmithred | that package does not exist in beowulf | 13:48 |
| fsmithred | if you clicked on Download link at my ibiblio site, it brings you to sourceforge | 13:48 |
| ham5urg | # cat /etc/apt/sources.list | 13:48 |
| fsmithred | those images are devuan jessie | 13:48 |
| ham5urg | deb http://pkgmaster.devuan.org/merged beowulf main contrib | 13:48 |
| fsmithred | name of iso you used? | 13:49 |
| ham5urg | # lsb_release -a | 13:49 |
| ham5urg | No LSB modules are available. | 13:49 |
| ham5urg | Distributor ID:Devuan | 13:49 |
| ham5urg | Description:Devuan GNU/Linux beowulf/ceres | 13:49 |
| ham5urg | Release:10 | 13:49 |
| ham5urg | Codename:n/a | 13:49 |
| fsmithred | so I'm guessing it was a refracta10-something | 13:49 |
| ham5urg | I deleted the iso already so I'm not sure... | 13:50 |
| xinomilo_ | when did you install beowulf? | 13:50 |
| xinomilo_ | was it before 02/2018? | 13:50 |
| fsmithred | couldn't have been | 13:51 |
| fsmithred | lsb shows it was a beowulf build | 13:51 |
| ham5urg | no, just some weeks ago | 13:51 |
| xinomilo_ | that package used to be in buster/beowulf till 24/2/2018 : https://tracker.debian.org/news/935520/freerdp-removed-from-testing/ | 13:51 |
| ham5urg | May the refracta version was from before that date | 13:51 |
| ham5urg | Maybe* | 13:52 |
| fsmithred | isos were uploaded april 15 | 13:52 |
| fsmithred | made a couple days before that, I think | 13:52 |
| ham5urg | I don't know, but it does not matter to much for me, as it is just a testing machine. | 13:52 |
| ham5urg | But thanks for investigating | 13:53 |
| fsmithred | well I'd like to know what's going on | 13:53 |
| fsmithred | I'm the one who made that iso | 13:53 |
| fsmithred | amd64 or i386? | 13:53 |
| ham5urg | amd64 | 13:53 |
| ham5urg | Its strange, as sources.list points to a beowulf repo | 13:54 |
| fsmithred | you got a no-X iso, right? | 13:54 |
| ham5urg | Yes | 13:54 |
| fsmithred | ij | 13:54 |
| fsmithred | ok | 13:54 |
| fsmithred | I'm booting the iso in vbox to have a look | 13:55 |
| fsmithred | my sources.list had deb.devuan.org, not pkgmaster | 13:58 |
| fsmithred | and nothing freerdp installed | 13:58 |
| fsmithred | where'd you get it? | 13:59 |
| fsmithred | aptitude why libfreerdp-dev | 13:59 |
| fsmithred | ham5urg, I think you need to install freerdp2-dev since that's what's available in beowulf | 14:05 |
| ham5urg | I installed all packages as named here: https://guacamole.apache.org/doc/gug/installing-guacamole.html#required-dependencies | 14:06 |
| ham5urg | I believe I booted with refracta and installed/used debootstrap | 14:07 |
| ham5urg | So my system was a blank devuan on harddisk. | 14:08 |
| fsmithred | oh, so you started with a pure devuan | 14:08 |
| ham5urg | And after that I started to install all packages I needed. | 14:08 |
| fsmithred | that would explain why you have pkgmaster in sources.list | 14:08 |
| fsmithred | ok, here's my theory - there were problems with the repo in april and may | 14:08 |
| xinomilo_ | instructions in url are wrong/outdated : "Debian / Ubuntu packagelibfreerdp-dev" | 14:09 |
| ham5urg | I failed to install beowulf with any other distributions | 14:09 |
| fsmithred | some things were not updating | 14:09 |
| fsmithred | so the old package may have still been hanging around and you were able to install it | 14:09 |
| fsmithred | you should be able to use the newer package | 14:09 |
| fsmithred | have you done update/upgrade in the last week? | 14:10 |
| fsmithred | if so, everything should be ok | 14:10 |
| xinomilo_ | actually instructions are correct, for debian stretch/devuan ascii. not for buster/beowulf | 14:14 |
| eyalroz | fsmithred: I clean-installed ASCII a few days ago, but kept my home folder from a previous installation; then dist-upgrade'ed to beowulf | 14:37 |
| fsmithred | eyalroz, I can reproduce the error about /etc/machine-id, but it has no effect on function. I can still play the video. | 15:08 |
| eyalroz | fsmithred: Maybe I should try installing an independent chromium binary and see what that does. | 15:22 |
| fsmithred | maybe we should compare package lists | 15:23 |
| eyalroz | fsmithred: Want a pastebin of dpkg --get-selections? | 15:28 |
| eyalroz | anyway, stand-alone chromium build complains about a missing sandbox. running it without a sandbox, I get video and no audio - even with hw acceleration enabled. | 15:30 |
| eyalroz | Speaking of audio, by the way | 15:30 |
| fsmithred | weird | 15:30 |
| eyalroz | pulseaudio is not running in the background by default, for some reason | 15:30 |
| fsmithred | oh | 15:30 |
| eyalroz | and I need to run it manually to be able to use pavucontrol | 15:30 |
| fsmithred | pulseaudio is nowhere to be seen here | 15:30 |
| eyalroz | ... which shows me a "dummy output" | 15:30 |
| eyalroz | "here" = Devuan ? | 15:31 |
| fsmithred | except maybe on a picture at the post-office | 15:31 |
| fsmithred | no here means my machines | 15:31 |
| fsmithred | I never install it | 15:31 |
| eyalroz | so what do you use to set the volume? | 15:31 |
| eyalroz | (e.g. on Cinnamon) | 15:31 |
| fsmithred | either alsamixer in a terminal or volumeicon-alsa on the panel | 15:32 |
| eyalroz | When I use alsamixer-gui, it has no effect on anything | 15:32 |
| fsmithred | I think you'd need to get rid of PA for that to work | 15:32 |
| fsmithred | probably same with volumeicon-also | 15:33 |
| fsmithred | alsa | 15:33 |
| fsmithred | how do you run chromium in a sandbox? Is that built into the browser? | 15:33 |
| eyalroz | ... let me rephrase that. It starts working again if I kill pulseaudio | 15:33 |
| fsmithred | oh | 15:34 |
| fsmithred | can you keep it killed, or does it keep coming back? | 15:34 |
| eyalroz | Maybe doing so will have a positive effect on chromium? let's see... | 15:34 |
| eyalroz | fsmithred: Yay! | 15:35 |
| eyalroz | That worked. | 15:35 |
| fsmithred | cool | 15:35 |
| fsmithred | you have audio and video now, right? | 15:35 |
| eyalroz | fsmithred: Video with no acceleration, audio when pulseaudio doesn't run. | 15:36 |
| fsmithred | I have no advice on PA except to get rid of it. It's too complicated for me to figure out. | 15:37 |
| fsmithred | and when I need a sound server, I use jack | 15:37 |
| eyalroz | fsmithred: So, how do you add volumeicon-alsa to the (Cinnamon) panel? | 15:37 |
| fsmithred | I think you add it to startup applications, wherever that is | 15:38 |
| eyalroz | let me try that | 15:39 |
| fsmithred | it's /usr/bin/volumeicon | 15:39 |
| eyalroz | Hmm, In the "startup applications" dialog, it shows me "PulseAudio", but that's not actually started. | 15:40 |
| fsmithred | make the entry manually | 15:40 |
| eyalroz | How come the default audio configuration on Debian is so f'ed up? | 15:40 |
| fsmithred | maybe they can't figure out how to use PA, either | 15:40 |
| fsmithred | or maybe it's as bad as many people say | 15:41 |
| eyalroz | fsmithred: Well, if they can't figure it out they should configure something else. Like putting the volume icon on people's taskbar and with a decently-painted icon | 15:43 |
| eyalroz | plus, the volumeicon starts the ncurses mixer rather than the (atrocious-looking) GUI mixer | 15:43 |
| fsmithred | Refracta has a little black speaker icon. Is that enough? | 15:43 |
| eyalroz | Better than nothing, that's for sure :-) | 15:43 |
| fsmithred | I just get a volume slider | 15:44 |
| eyalroz | I used to run mint before switching to Devuan on principle | 15:44 |
| fsmithred | right-click for settings | 15:44 |
| eyalroz | So I'm kind of spoiled | 15:44 |
| fsmithred | yeah, debian and devuan are more d-i-y | 15:44 |
| fsmithred | oh, Open mixer gives you alsamixer. | 15:45 |
| eyalroz | fsmithred: But that's silly. I mean, of course, not every whim should be catered for, but the default configuration of everything should be reasonable; and for audio, it sort of isn't | 15:45 |
| eyalroz | yeah, ncurses alsamixer! How 1990s :-) | 15:46 |
| fsmithred | it'll work when your mouse dies | 15:46 |
| gnarface | there is a alsamixer-gui too | 15:46 |
| fsmithred | you can change it if you can find a descent mixer | 15:46 |
| gnarface | it's just the same thing in gtk | 15:46 |
| Hurgotron | eyalroz: works better than Pulseaudio... | 15:46 |
| eyalroz | Hurgotron: pulseaudio worked fine on ASCII | 15:47 |
| eyalroz | for me anyway | 15:47 |
| fsmithred | for some reason I thought alsamixer-gui was gone | 15:47 |
| Hurgotron | eyalroz: oh? What mixer do you use for pulseaudio-ascii? | 15:48 |
| fsmithred | alsamixergui (no hyphen) | 15:48 |
| eyalroz | Hurgotron: pavucontrol, it's very nice GUI-wise | 15:48 |
| fsmithred | pavucontrol I think it's called | 15:48 |
| eyalroz | You can't put it in your taskbar though. At least, I've not been able to | 15:48 |
| eyalroz | I wonder if I can make volumeicon start alsamixergui | 15:49 |
| fsmithred | yes, right click to change preferences | 15:49 |
| eyalroz | fsmithred: Oh yeah, works :-) ... another thing which should have been configured by default IMHO. | 15:51 |
| eyalroz | Speaking of which... do Debian people - I mean, integrators, no package maintainers - speak to you guys (= Devuan people)? Or is there sort of a disconnect? | 15:52 |
| fsmithred | mostly disconnect | 15:53 |
| fsmithred | recently (last few months) there's been collaboration on other init systems | 15:53 |
| fsmithred | and on maintaining sysvinit | 15:54 |
| eyalroz | Nice | 15:54 |
| fsmithred | yeah | 15:55 |
| fsmithred | we do have control over the task- packages for the desktops. | 15:55 |
| fsmithred | so you could make a case to the devs for removing PA, I think. | 15:55 |
| eyalroz | Does that mean there's a chance they'll make systemd a configurable choice during installation? Or is it just on openrc, upstart etc? | 15:55 |
| fsmithred | not sure exactly what pulls it in | 15:55 |
| fsmithred | well, someone recently posted that he's running sid without systemd | 15:56 |
| fsmithred | I don't know if debian will give you a choice of inits during install | 15:57 |
| eyalroz | fsmithred: I don't mind pulseaudio, myself; I just care about whatever installs by default have a working default configuration | 15:57 |
| eyalroz | fsmithred: It doesn't give you that choice now, but it could be a way which would theoretically allow merging Devuan back into Debian | 15:57 |
| fsmithred | ok, when the beowulf beta isos come out, please test install cinnamon | 15:57 |
| eyalroz | (or maybe I'm being naive here) | 15:57 |
| eyalroz | fsmithred: I'll do that. | 15:58 |
| fsmithred | we already have code in devuan installer to let you choose init system | 15:58 |
| fsmithred | so it's possible that will migrate to debian | 15:58 |
| fsmithred | it wouldn't be the first thing that's done that | 15:58 |
| fsmithred | there's been collaboration on elogind | 15:58 |
| fsmithred | it's possible that devuan will be obsolete in a few years | 15:59 |
| eyalroz | What, in your opinion, would be a sufficient set of conditions to "fold" or "merge" Devuan? | 16:00 |
| fsmithred | it's also possible that they will think up new shit that we don't want | 16:00 |
| fsmithred | if our reason for existence goes away | 16:00 |
| fsmithred | devuan was formed because debian forced systemd on everyone | 16:00 |
| fsmithred | if we're going to keep talking about this, we should move to #debianfork | 16:01 |
| eyalroz | so, will proper support for no-systemd in all packages, plus init system selection during installation, be enough? | 16:01 |
| fsmithred | don't know. It's not my call. | 16:01 |
| eyalroz | Ok, nm, I'll re-ask there. | 16:01 |
| KatolaZ | eyalroz: what is your point? | 16:02 |
| KatolaZ | are you a debian developer? | 16:02 |
| fsmithred | he's trying to get me to clarify my statement that devuan might be obsolete in a few years | 16:02 |
| KatolaZ | wishful thinking is cheap :) | 16:02 |
| fsmithred | well, I did also say they might come up with new shit | 16:03 |
| KatolaZ | there is a lot of resistance against anything that promotes multiple choice and/or alterntive choices to systemd | 16:03 |
| KatolaZ | elogind has remained in experimental for ages | 16:03 |
| KatolaZ | because the systemd team does not want it | 16:03 |
| eyalroz | KatolaZ: I'm not a distribution developer/maintainer/activist | 16:04 |
| eyalroz | KatolaZ: https://eyalroz.github.io/ | 16:04 |
| krzych | hi, what you use for virtualization? qemu does not want to work... How to install virtualbox on devuan ascii? | 16:05 |
| eyalroz | KatolaZ: So you're saying the systemd team has some political influence within Debian... hmm. | 16:05 |
| fsmithred | krzych, vbox is in ascii-backports | 16:06 |
| fsmithred | and qemu does work for other people. I only use it to boot isos. | 16:06 |
| KatolaZ | eyalroz: I am not saying anything | 16:06 |
| KatolaZ | I am reporting facts | 16:07 |
| KatolaZ | which you can access as well by reading debian-devel | 16:07 |
| KatolaZ | or debian-init-diversity | 16:07 |
| eyalroz | KatolaZ: Fair enough... ;-) I'm not on those lists. I'm in over my head enough on other reading material | 16:07 |
| KatolaZ | so am I | 16:08 |
| KatolaZ | :) | 16:08 |
| r3boot | .. what I still dont understand, is why instead of writing a patchset for gnome which rips out logind support and replaces it with a pam stack, that people ripped the logind part out of systemd, rebranded + patched that to elogind, and are satisfied with that. But I'll take this to #debianfork, dont worry :) | 16:10 |
| krzych | fsmithred: tnx | 16:11 |
| fsmithred | yw | 16:12 |
| fsmithred | lol eyalroz I just read your description of working on a personal project. Too true. | 16:12 |
| eyalroz | fsmithred: You mean the "scratching an itch" bit? | 16:13 |
| fsmithred | yeah | 16:13 |
| eyalroz | Most of my time spent working on bidi support in Thunderbird went in to writing HTML email, which I'm personally against | 16:15 |
| eyalroz | and with an HTML editing component in Mozilla which is a steaming pile of garbage | 16:16 |
| fsmithred | on a debootstrap install of beowulf made from a refracta10 live iso, lsb_release -a shows the Release is 10. | 16:20 |
| fsmithred | How does it know that? I didn't create /etc/lsb-release or put 10 in there anywhere | 16:20 |
| fsmithred | it says devuan (exactly the same output as ham5urg got earlier | 16:21 |
| fsmithred | ok, lsb-release hasn't been devuanized yet | 16:23 |
| fsmithred | ham5urg, I just noticed that my beowulf debootstrap install has a couple of ascii lines in sources.list. If you have those, you should comment them out and run another update. | 17:14 |
| ham5urg | which one? | 17:18 |
| fsmithred | which one what? | 17:18 |
| fsmithred | any ascii lines in any beowulf sources.list should be removed | 17:19 |
| ham5urg | I've done so after installation I guess. I'm not sure. | 18:04 |
| unixman | No Pale Moon package for Devuan? :( Well, I can just get the statically compiled binary then. | 23:02 |
| onefang | I use the Debian PPA for Pale Moon, works fine for ASCII. http://download.opensuse.org/repositories/home:/stevenpusser/Debian_9.0/ | 23:09 |
| unixman | I got the static build. It's working fine. | 23:16 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!