freenode/#devuan/ Wednesday, 2018-10-24

* DocScrutinizer05 ACKs furrywolf03:04
DocScrutinizer05alas particularly for BT it seems polypaudio is inevitable03:05
DocScrutinizer05the PA-free alternative died a few years ago03:05
furrywolfthere's no way to use bluetooth without pulse?03:12
furrywolfgoogling says https://github.com/Arkq/bluez-alsa works great, and is alive.03:18
gnarfacehmmm.  have you tried it in ascii yet?03:19
furrywolfI don't use bluetooth, so haven't tried it at all...03:20
furrywolfif everything else really does require pulse, we probably should package and distribute that.03:22
gnarfaceyea i think there's a showstopper though03:24
gnarfaceit used to be in debian03:24
gnarfaceit got pulled for a reason03:25
gnarfaceit may not have even been a good reason, i don't know03:25
furrywolf"encourages people not to use pulseaudio" is their idea of a good reason...03:25
gnarfacevery likely, but my point is someone would have put it back by now probably if they hadn't strategically blocked it somehow compatibility-wise03:26
furrywolfalso, anyone else finding packages.debian.org to either be down or to have changed something that breaks google links to it?03:26
gnarfaceseems to be up for me03:28
gnarfaceredirecting to https now, maybe that's new?03:29
furrywolfhttps://packages.debian.org/wheezy/bluez-alsa  is from a google search03:29
furrywolfit gives "03:30
furrywolftwo or more packages specified (bluez-alsa wheezy)03:30
furrywolf"03:30
furrywolfobviously parsing the url wrong...03:30
furrywolfthe only thing I can find about it being dropped is "use pulseaudio".03:31
gnarfacei agree with you that there's clearly an agenda03:31
gnarfacei'm just saying they won't have left it at that.  they'd have made sure to cram jam a monkey wrench in there somewhere to keep people from proving them wrong.03:32
golinuxfurrywolf: Try apulse03:33
golinuxI haven't needed it yet hut many people have to get around PA03:33
gnarfacewill that work with bluetooth audio?  i didn't think apulse was that fancy.03:33
golinuxDunno.  Do some searches on the lists and forum03:33
furrywolfit sounds like there was a period where bluez-alsa did not work with bluez5, so they used that as justification for removing it, and never re-packaged it when it was as expectedly promptly fixed.03:33
furrywolfgolinux:  the topic is making bluetooth audio hardware work without pulse, not making applications work without pulse.  :)03:34
furrywolfand how, as is getting increasingly typical, debian "accidentally" killed a way to not use pulse.03:34
Centurion_Danfurrywolf: do we need to build bluez-alsa for devuan?03:35
Centurion_Dancan somebody build it and prove it works??03:35
gnarfacemy bluetooth dongle stopped working even earlier, so i can't03:35
furrywolfalso, it looks like bluez dropped normal communications methods and started only talking over dbus, which was why bluez-alsa originally broke.03:36
furrywolfCenturion_Dan:  I don't have any bluetooth hardware to test with.03:36
gnarfacei have this one bluetooth dongle that worked in kernel 3.103:36
gnarfacenot a version before or after03:36
gnarfacei tried to file a bug report and the maintainer told me if i could find the bug myself and point it out to him he'd fix it, but only if i also mailed him the hardware (which he could only say he would almost certainly not bother returning)03:38
gnarfaceso i went to radio shack and bought a 12' headphone cable03:39
gnarfaceproblem solved03:39
furrywolfhttps://www.raspberrypi.org/blog/raspbian-stretch/ raspbian packages and uses bluez-alsa themselves to make bluetooth work from rpis without pulse.03:39
furrywolf"From a user point of view, everything should still work exactly as before – the only change is that if you still wish to use PulseAudio for some other reason, you will need to install it yourself."  I like these people!03:40
gnarfaceinteresting03:40
gnarfacei wonder if pulse finally proved too heavy for the rpi cpu03:40
gnarfacewell if they made it work with raspbian it might still require systemd somehow though03:41
gnarfaceright?03:41
furrywolfand/or too utter shit.03:41
gnarfaceno, raspbian doesn't give a crap about reliability03:41
furrywolflol03:41
gnarfacethey break their own kernel firmware almost 4 of every 5 updates03:41
gnarfacethey would have had to pull it because it literally cripples performance03:42
gnarfaceand not just a little03:43
gnarfacethey'd be handwaving the problem away otherwise03:43
furrywolfhttps://archive.raspberrypi.org/debian/pool/main/b/bluealsa/  looks like they're packaging it for all archs...  some of them updated this month.03:43
gnarfaceinteresting...03:43
furrywolfbluealsa_0.9_amd64.deb2018-10-11 11:0803:43
gnarfacethere was something else needed03:43
gnarfacea simple pairing app03:43
gnarfacealso absorbed by systemd03:44
gnarfacei was under the impression it could be replicated with a simple python script though03:44
furrywolfanyone here have bluetooth hardware and working (i.e. pulse-free) audio and want to try that package out?03:44
furrywolfalso, do the raspbian people have other packages we should be distributing?  lol03:45
gnarfaceomxplayer03:48
gnarfacebut that's more arm-specific03:49
gnarfacewell, rpi specific03:49
gnarfaceat least03:49
gnarfacenot sure if the bananna pis use the same video chip or not03:49
gnarfaceeven with patches for the video decoding hardware though, vlc is too cpu hungry for a rpi1 if you don't overclock it03:50
gnarfaceso omxplayer is the only really viable player for rpi103:50
furrywolfI can't imagine ever wanting to play video on a pi in the first place.03:51
gnarfaceit works really well at 1080p@6003:54
gnarface... if you're using omxplayer and the video is in a supported encoding and pixel format03:54
gnarfaceit actually makes a nice video player for how low power it is03:54
furrywolfI have no desire for a video player.  lol03:55
gnarfaceoh, heh, well then it would be an exercise in futility for you03:55
gnarfacethe video decoder supports h264 as long as it's yuv420p03:57
gnarfaceand also mpeg2 if you pay like $2.50 extra for the key03:57
gnarface(but it's easy to exceed the bandwidth capabilities of the poor little thing with a mpeg2 stream so i can't imagine how useful that would be)03:58
furrywolfI've been thinking of getting an rpi to log data from my weather station and solar system, to replace a failed nslu2...  no video.  :P04:00
gnarfacei still think you're better off with something from pine64.org04:01
furrywolfI'm not sure what's wrong with the nslu2...  lights do things, but it's not talking to anything.04:01
furrywolfthat pine64 website really sucks.04:08
gnarfacetheir website sucks and their ordering process is a mess04:11
gnarfacefrankly they're just a bunch of kids04:11
gnarfacebut they don't exhibit the overt greed i see from Broadcom04:11
gnarfacetheir hardware is carefully designed and generously priced04:12
gnarfaceif mainline software support is lagging well... it's a sideffect of youthful ambition04:12
gnarfacei'm confident the mainline kernel support will catch up04:12
gnarfacefor a headless server though, the mainline kernel already does all that fine04:13
gnarfacewhere the support is lagging is on stuff mobile users would want... camera, audio, bluetooth, hdmi output, gpu acceleration, etc04:13
furrywolf"The Pine A64 board supports SD, SDHC, and SDXC format microSD card – this means the largest capacity is 256GB. Please note that if a microSD card is formatted as an FAT32 file format, the maximum capacity is 32GB."  seriously, they fucking listen to microsoft's force-people-to-use-exfat crap?  I do not like these people.04:14
gnarfacehmmm04:17
gnarfacei think they just get a lot of android users who don't know shit so they have to warn them about that stuff04:17
gnarfacekeep in mind that 3.11-BSwhatever kernel is the android kernel04:18
gnarfaceso these things boot android too04:18
furrywolfthe only thing they're warning you about is that they intentionally crippled it to comply with microsoft's corruption of the sd card standard.04:18
gnarfaceoh, i see what you're saying04:19
gnarfacei'm not sure that's actually the case though04:19
furrywolfnormal linux doesn't care in the slightest what filesystem you put on the card...  the 32gb-fat32-on-sd limit is, however, required for official licensing of the sd card spec.04:20
gnarfacei'm reading that as a warning about the limitation of fat32 not the hardware04:20
furrywolfno, there's no technical reason for it.04:20
furrywolfit's a legal reason.04:20
gnarfacebut it's only a problem if you're using fat32 to being with...04:21
gnarfacewhat would you be doing with fat32 on a headless linux logging server?04:21
furrywolfif you officially license the sd spec, you're required to reject sd cards over 32gb with fat32 filesystems on them.04:21
gnarfacereally?04:21
furrywolfyep04:21
gnarfaceand that's not a limit of fat3204:21
furrywolfyep04:21
gnarfacei thought the natural limit for fat32 was lower than 32GB actually04:22
gnarfaceanyway04:22
gnarfacei'm assuming the sd card license was already tied to the boards by the manufacturer04:22
gnarfacekeep in mind these guys aren't manufacturing this stuff either04:22
gnarfaceit's just rebranded android hardware04:22
furrywolfmost devices made before >32gb cards were a thing will happily use 64gb or 128gb cards with fat32 systems on them anyway...  then when the big cards come out, devices that wanted to officially use them had to intentionally add checks to reject them.04:22
gnarfaceso it may have been a choice the manufacturer made that they had to keep with in order to keep the cheap hardware04:23
gnarface(in order to re-use an existing design as the base instead of having to design something from scratch like they're sopine boards)04:23
gnarface*their04:23
furrywolfit was due to microsoft's influence on the sd spec...  they wanted to discourage use of non-microsoft operating systems, by making devices use filesystems that linux etc couldn't read.04:24
gnarfaceheh, seems like they wasted a lot of effort on that.  someone should have told them fat32 was already supported04:24
furrywolfthere is still no legal-in-the-US support for >32gb sd cards on linux.04:24
gnarfacei'm gonna have to test that04:25
furrywolfoh, it works fine.04:25
furrywolfit's just moved to a non-US-based FUSE module, not the kernel.04:25
gnarfacei am actually not sure i have any sd or microsd cards larger than 32GB here04:26
gnarfacebut i've never put fat32 on them anyway04:27
furrywolfyou must not own any consumer electronics?  :P04:27
furrywolffor example, a digital camera, a mp3 player,...04:28
gnarfacei guess not04:28
gnarfacei mean i have several smaller cards just none above 32GB04:28
furrywolfI mean, that all consumer electronics only work with fat32...04:32
furrywolfif you never put fat32 on them, you must not be using any of them in consumer electronics.04:33
gnarfacehmmm04:42
gnarfaceyou're right, i think the wii does use encrypted files on fat3204:43
gnarfacei'm pretty sure the wii-u reformats the entire thing though once you put it in04:43
gnarfacewhat i meant is that, as a general rule, i don't use fat32 on them in linux04:47
gnarfacei've not tested fat32's limits in this context04:48
furrywolfI do, because I add or remove files from these devices from linux.  heh.04:48
furrywolfi.e. put mp3s on the mp3 player, or download images from my camera.04:48
golinux(Some people just don't get that this is a devuan help channel)04:52
furrywolfI add and remove files from my fat32-formatted cards from devuan.  :P04:53
gnarfacesorry04:53
nacelleits all washing out anyways now that MS has released the patents around that to OIN04:55
golinux#debianfork04:58
* man_in_shack flails09:08
blergHow can I get my laptop to suspend when I close the lid? Debian wiki says a systemd service replaced pm-utils. Do pm-utils still work?12:13
djphin devuan, they should12:14
blergdjph: Cool. (I am using Devuan)12:14
blergNow to figure out how to get a display manager to play nice with Stumpwm.12:15
ejrwew lads, after about half a year of not using devuan, i am back now, and more happy with it than ever :)12:29
jyrihi13:41
pilgrimmPardon if this is a stupid question, does anyone know if the Wacom digitizer on Lenovo's Yoga 720s work with Devuan?13:42
pilgrimmUsing my other casual drawing pc that has plain deb on it because my main workhorse crapped out, but I want to go back to devuan :<13:45
jyrianyone aware that ceres has totally broken GTK support on applications? I mean that even $ reportbug13:46
jyrizsh: segmentation fault  reportbug13:46
jyristrace points to GTK problems and any program that uses GTK refuses to launch13:46
jyripilgrim: don't know without trying :)13:47
gnarfacepilgrimm: the kernel is the same13:49
gnarfacehardware support shouldn't be different unless it somehow requires systemd13:49
gnarfacejyri: has this been happening for a long time?13:50
pilgrimmgnarface: ah so it's a kernel thing then. i thought it was a package thing but then again that actually checks out from my experiences13:50
pilgrimmI'm using sysvinit on debian and it works fine with my pen so I highly doubt it's dependent on systemd. Thank you for the response, now I'm confident enough to put devuan on here :)13:52
jyrignarface: I just noticed it today13:53
gnarfacepilgrimm: it probably does also need some or another xorg package and maybe some config snippet, but that usually comes along with a standard xorg install13:53
gnarfaceat least these, i'm guessing: xserver-xorg-input-wacom, libwacom-bin, libwacom-common, libwacom213:54
gnarfaceexcept where systemd was removed, the packages and dependences are all the same as on debian13:55
gnarfacejyri: i care, but i can't do anything about it.  i'll avoid updating today.  thanks for the warning :)13:55
gnarfacejyri: (i do assume this type of thing will get fixed upstream)13:56
jyrignarface: ok :) I do apt update ; apt -y upgrade ; apt -y autoremove ; apt clean  every morning when I come to the office (this is my workstation)13:56
gnarfaceoh boy13:56
gnarfaceceres on a workstation13:56
jyri:D13:56
gnarfacethat takes a special type of bravery13:56
jyriI like to live on the edge13:56
jyriit has only once bitten me in the ass before (then it was worse since it removed Xorg)13:57
jyribut that one was easy to downgrade, now I'm not sure which package caused this since it wasn't any that has gtk on the name of it13:59
gnarfacewell, it's worth checking the debian bug tracker to see if anyone else has noticed it14:00
sicknesshi all17:13
sicknesswhat is the best/fastest way to install ceres? install devuan ascii then change sources.list then do a dist-upgrade? (that is the advice I've found on the forums)17:15
jyrisickness: that's it17:16
jyribut, don't install today, I think that I run in an bug that messes up gtk17:17
sicknessjyri: ah, I was just curious to test it, nothing production, so even if it explodes it's ok :)17:22
golinuxjyri: https://igurublog.wordpress.com/2012/11/05/gnome-et-al-rotting-in-threes/17:33
golinuxThe gnome/GTK thing has been coming for a long time17:34
golinuxsickness et al:  Be aware that there are some packages that are newer in ascii than in beowulf or ceres.17:35
golinuxOur workflow is changing to now modify packages in ceres and let them move down to older releases.17:36
pankeriniWow, I didn't imagine the GTK/gnome problem lasted this long17:37
sicknessah, ok17:40
golinuxYeah, it's been a long time coming . . .17:49
golinuxBottom line is that the gnome sickness goes hand in hand with the systemd sickness17:50
premobosshi18:01
premobossdevuan ascii 64 bit. already did update/upgrade. Amule is quite instable, it crash often (at least 4 times today).18:02
premobossPC with 4GiB Ram ,a nd 200G of free space, fs is ext4.18:03
enycpremoboss: hrrrm silyl question, does the machine memtest86+ ok?  i'd then look in kernel log (dmesg) to see what sort of crashes...18:06
premobossok, i go for kernel log.18:07
fsmithredpremoboss, did you see this bug report? https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=84606118:08
premobossfsmithred, no i go to read now.18:09
premobossah... then is that.18:09
fsmithredlooks like it didn't get fixed. Last post says use the jessie version18:10
premoboss2 time ago when he drashed, i got a pop ul windows wint error similar to that bug... it was like a "debug" window, you can do "close" or "continue" but if i choose "continue", it iterate in endless loop so i was forced to chose "close"18:11
premobossso, we must wait for bugfix, ok.18:12
fsmithredmaybe it's fixed in sid/ceres. That report is 18 months old. Check the changelog.18:15
nemogolinux: heh. Redhat GnomeOS (in comments on that link) - yeah prettymuch18:18
nemogolinux: I wonder when we'll get SystemDKernel18:18
nemothey kinda already tried to inject themselves in there once already and were beaten back18:19
nemobut that was when Linus was not busy removing himself from power18:19
MinceRgkh already has write access to Linus' branch18:26
MinceRso it's not long until systemd-kerneld18:26
* nemo sighs18:27
nemowelp. BSD is still an option18:27
pankeriniYear of the Hurd Desktop soon18:28
golinuxSo is forking the kernel18:29
KatolaZyeah, sure....18:29
KatolaZ:D18:29
MinceRhurd is dead18:29
MinceRand who will maintain the kernel fork?18:30
MinceRhttps://www.gnu.org/software/hurd/open_issues/systemd.html18:30
nemoproblem with forking is the ever growing burden of integrating useful stuff from the other side of fork, and maintaining enough people. it's about the community vote, and so far at least the community is unfortunately like... maybe 5% anti-systemd to 95% "I have a desktop, who cares"18:30
KatolaZMinceR: the hobbits18:30
nemoI kinda wonder what the % is amongst active linux/FOSS contributors18:30
MinceRit seems to me that the "modern" side isn't interested in implementing useful stuff anyway18:30
MinceRso i wouldn't worry about that18:30
nemoMinceR: well... the corporate coders will presumably implement anything they are assigned to do.18:31
nemoMinceR: heh '<pochu> let's create our own init system, make it better than anyone else<pochu> let's create our own init system, make it better than anyone else'18:32
nemoMinceR: «With hookers, and blackjack»18:32
KatolaZnemo: the percentage of what?18:32
nemohttps://www.youtube.com/watch?v=e35AQK014tI18:33
nemoKatolaZ: I was just wondering if literally everything in the entire linux ecosystem could become systemd dependent, or if the % of those who actually help build stuff under linux in userspace regular apps, desktops or system might push back18:34
nemoif anti-systemd sentiment is stronger in that group, it'd be harder for that to happen18:34
KatolaZstill don't get the point, but I guess it's OK18:35
KatolaZ:)18:35
KatolaZthere is no reason why everything at the application leve should depend on systemd18:36
KatolaZthere are some applications which might need some of the facilities made available by systemd18:36
KatolaZe.g., session management18:36
KatolaZbut none of those is *necessary*18:36
KatolaZ(meaning that there are always other ways of obtaining the same result without using the facility made available by systemd)18:37
MinceRit's kind of like that most mono applications don't need the "non-core" parts not covered in microsoft's useless "promise"18:37
MinceRand yet all major mono applications use them anyway18:37
KatolaZit's a bit different MinceR18:37
MinceRin the "modern" world, necessity is not a factor18:37
KatolaZwhy would an editor need to know what init system is currently running?18:38
MinceRso that it can depend on systemd18:38
KatolaZnon sequitur18:38
MinceRbusiness as usual18:38
KatolaZbut which business?18:38
MinceRred hat's18:39
KatolaZMinceR: red hat does not develop everything in the FOSS world...18:39
MinceRindeed, they just aim to destroy everything in the FOSS world18:40
nemoKatolaZ: there's no reason why it *has* to - but it could be harder and harder to avoid, what if gtk3 did *everything* over dbus and relied on a systemd specific implementation18:43
KatolaZnemo: it's already like that :)18:43
nemoKatolaZ: can't predict the future but what it *feels* like is an all consuming glob trying to shove everything into process 0 or the kernel18:43
KatolaZwelcome to 201818:43
nemoKatolaZ: can't be totally like that since you guys are still able to shim it out18:44
KatolaZuh?18:44
KatolaZgnome is not installable cleanly in devuan18:44
nemoKatolaZ: hm? devuan has gtk3 apps18:44
nemoand ones using dbus even18:44
fsmithrednemo, I have a no-dbus build with most of the apps that normally are included in Refracta18:44
nemoKatolaZ: but yes gnome3 is not installable18:44
KatolaZI know nemo18:44
KatolaZ:)18:44
nemoI was just trying to imagine rot spreading18:44
KatolaZeven if in principle it's just a matter of a wrong dep on the debian package18:45
KatolaZsince it should be possible for GNOME to work with libpam-systemd18:45
nemoIt's like someone looked at Microsoft Windows with its HTML parsing and font rendering and widget drawing in ring 0 and though "man, that's a damn good idea.18:45
KatolaZwhich is provided by elogind18:45
nemo"18:45
nemolinux kernel - now with wayland18:46
nemoblazing fast UIs!18:46
KatolaZthat's not gonna happen nemo18:46
nemoDid you guys ever read the MS justification for why they did that? I thought it was pretty funny18:46
nemoit boiled down to - we aren't fast enough to compete w/ OS with real security measures otherwise, and, we're gonna be super careful18:47
nemolast year there was a priv escalation in the UI code...18:47
KatolaZnemo: I stopped caring about M$ more or less at the same time I got the first PC18:47
KatolaZ:)18:47
nemostill need to use it for work unfortunately18:47
nemobut major win on my part, just last month I got our first *OFFICIAL* linux server at the office18:48
nemountil that point they were adamant that they could only support Windows18:48
nemothis allows for a possible transition once Win 7 is EOL'd18:48
KatolaZthe few times I had to use windoes, I just used cygwin under it18:48
nemoonly issue is Oracle which irritatingly charges more for the Linux builds even though the performance is better - I think the logic is if you're using linux you must be a big operation and therefore they can extract more from you18:49
nemoso convincing them to relicense the db and other oracle junk under linux might be a sticking point18:49
nemobut who knows18:49
nemoBTW, that linux server is Ubuntu 14.04 - I am hoping to convince them to allow me to transition it to Devuan Ascii before next spring EOL of ubuntu 14.04 but they might just force me to upgrade to 16.04 instead18:53
UsLhello vuans! So, my system changes sometimes and the values in /etc/fancontrol settings becomes outdated. hwmon02 is suddenly hwmon01 and so on. Is there a way to solve that or make the fans go 100% when settings are outdated?19:03
KatolaZUsL: what do you mean by "my system changes sometimes"?19:48
UsLDevice path of hwmon2 has changed19:56
UsLDevice name of hwmon2 has changed19:56
UsLConfiguration appears to be outdated, please run pwmconfig again19:56
UsLso in this instance    hwmon0 is acpitz19:57
UsL   hwmon1 is dell_smm19:57
UsL changed places19:57
UsLit can happen now and then.. Have not pin pointed the reason19:58
terraAnyone able to properly boot from devuan iso with grub2 ?22:02
terraI'm able to boot but installer is trying to find a cdrom device.22:02
golinuxComment out the cdrom line in /etc/fstab22:09
golinuxterra: ^^^22:09
terrareally?22:11
negevhi22:23
negevrunning beowulf i'm seeing consistently high cpu usage indicated in top, with no apparent processes listed using anywhere near what the total is indicated as22:24
negevanyone else seeing this?22:24
negev4.18.0-2-amd64 on a xen vps22:24
golinuxterra: Yeah, misread your question22:25
leonegev: try atop instead of top23:04

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!