IRC Logs for #crux Saturday, 2019-08-24

*** tilman_ has joined #crux00:06
*** jolupa has quit IRC00:06
ryuopop goes the poppler01:20
*** SiFuh has quit IRC01:26
*** SiFuh has joined #crux01:27
*** xor29ah has quit IRC01:38
*** xor29ah has joined #crux01:40
*** SovietPony has quit IRC01:44
*** SovietPony has joined #crux01:46
*** xor29ah has quit IRC01:53
*** xor29ah has joined #crux01:53
*** mavrick61 has quit IRC02:44
*** mavrick61 has joined #crux02:45
*** SiFuh has quit IRC04:25
*** deus_ex has joined #crux04:27
*** SiFuh has joined #crux04:27
*** pedja has quit IRC04:29
Romsterpedja there are a few duplicates https://pastebin.com/vsaRZt6y04:41
*** SiFuh_ has joined #crux05:43
*** SiFuh has quit IRC05:45
*** SiFuh_ has quit IRC05:49
*** SiFuh has joined #crux05:53
*** SiFuh has quit IRC05:59
*** SiFuh has joined #crux06:03
*** brian|lfs has joined #crux06:45
*** brian|lfs has quit IRC06:54
cruxbot[xorg.git/3.5]: mesa3d: updated to 19.1.507:01
cruxbot[opt.git/3.5]: harfbuzz: updated to 2.6.107:01
*** brian|lfs has joined #crux07:18
*** SiFuh has quit IRC08:00
*** SiFuh has joined #crux08:11
pedjaRomster, yes. those were either too old or broke stuff at some point, so I had to fork them.08:30
pedjaand some are just remnants from the original repo I never bothered to remove :)08:31
Romstera ton of footprint errors for new perl version08:32
Romsters/version/versions/08:33
Romsterbut over all it's still going though the ports08:33
pedjaperl is a PITA08:34
Romsteri know right08:34
Romsteri am running a depinst on each port individually08:34
Romsteri'll paste a diff when i am done08:35
pedjaglib-networking and libsoup update broke some stuff, I'll have to check if that's still the case. other then that, I can remove most of the duplicate ports from xfce repo.08:39
pedjawell, after dealing with poppler update, that is :)08:41
pedjawendell is getting 9Gb/s with raid 0 nvmes in his shiny new dev workstation. that's insane.08:42
pedja2x2tb nvme. damn, that's nice :)08:43
Romstersweet08:44
Romsteroh libsoup abi changed? you could of said something08:44
pedjaonly peasants use spinning rust for a system disk these days, apparently08:44
pedjatbh, I forget what was broken, and I was in  a middle of packaging blender deps, so I just reverted to old libsoup version and moved on.08:46
pedjaI meant to get back to it and see what the issue is, but somehow never did :)08:47
Romstersounds like me08:48
pedjaI think it was xfce's screenshot plugin, or something08:49
pedjayeah, too many stuff to keep track of :)08:49
pedjabtw, if you are interested, I updated vulkan in my repo, glslang using system spirv and all :)08:54
Romsteri need to do that08:54
pedja(stole a patch from Fedora for that)08:54
Romsteri remember you said08:54
pedjadidn't test that wine dx12 vulkan thingie, thou08:55
pedjait's a bit of the mess, all vulkan and friends thing, still08:57
Romsteri hope the situation improves08:57
pedjait has to, vulkan has seen a huge industry adoption. sigraph vulkan bof talks are interesting.09:01
pedjathey mentioned dxvk and Proton, amongst other things, which was cool.09:02
Romsterwayland and composite desktops mentioned?09:03
pedjanope :)09:03
pedjathere is a vulkan-based wayland compositor, thou09:03
pedjalike that sway thing09:04
pedjaimho, wayland still has many things to fix/add before it becomes a viable xorg replacement09:05
pedjafor things other then basic desktop use09:06
Romsteri wonder if rtx helps with desktop compositors too09:06
TimB_beautiful window shadows everywhere :P09:07
pedjaray tracing is more of a render/game engines thing09:07
pedjablender will merge optix/rtx support soon-ish, afaik09:08
pedjaidk what the amd gpu equivalent is09:09
TimB_https://preview.redd.it/kdzlo1bbz7i31.jpg?width=640&crop=smart&auto=webp&s=ed3a5af7c6f747efce5baa6e1ef03c0006819935 uhm.. :D09:09
also_uplimefixing it in production09:10
*** SiFuh_ has joined #crux09:13
Romsteromg the electronics technician in me is screaming at that09:14
pedjatemporary hacks tend to become permanent09:15
*** SiFuh has quit IRC09:16
Romstermousepad needs /usr/lib/libpangocairo-1.0.la09:16
Romsternothing else is complaining09:16
pedjalet me check wth is going on :)09:17
Romsteri bumped the version to 0.4.209:17
also_uplimepedja: that reminds me - at my old job, there was a service that needed to be restarted every day or so, because the dev didn't realize sockets counted as fd's, so he never released them. i wrote a tiny script that a tech needed for a couple of customers, to check the amount of used fd's and restart it if necessary. next thing i know, instead of fixing the service, its being shipped in the next release :I09:18
pedjaheh. exactly that09:18
Romsteromg also_uplime09:19
Romsternotlikethis!09:19
also_uplimelol09:19
also_uplimethats not how i imagined my first code for a release version to be09:20
Romsternot releasing resources wen finished is a really derp move09:22
Romsterwen'when09:22
also_uplimeyeah09:23
also_uplimeto be fair, he'd only done web dev at that pont09:23
TimB_aren't resources infinite? especially virtual? :B09:23
also_uplimeup to that point(09:23
also_uplime**09:23
also_uplimei can't type today09:23
pedjaRomster, builds fine here. care to paste a build log?09:26
Romstergod damn it i did a ports -u before i fixed issues but guess what i didn't have git installed so now my diffs are as ugly as fuck09:30
Romsterhttp://dpaste.com/3TY274A09:33
Romstergit driver is a pain in a clean env09:33
Romsterjust look at the Depends on, and forget the rest09:35
TimB_Romster: i use my git repos in a clean env, i install git and pull all the latest updates, then pkg-clean -r and continute my work, and commit inbetween each of them, works fine?09:37
Romsteri forgot to do that09:39
TimB_ah, ok09:39
Romsterfrustrated building everything again09:40
Romstergonna get some food while this compiles again09:43
TimB_have a good one. i am going out for groceries, bbl09:43
pedjaRomster, pull from the repo before you start, I pushed some changes09:46
*** SiFuh_ has quit IRC10:13
*** SiFuh has joined #crux10:13
pedjaTimB_, good job with appindicator ports. q: why are patches lines set to verbose?10:15
Romsteri already started but i'll pull again10:19
TimB_pedja: to scare people :P10:19
TimB_pedja: i stole the patches somewhere else and used them here as they were used over there. maybe slackware? i can't remember10:20
*** jolupa has joined #crux10:20
RomsterTimB_, another handy tip i put them urls at the top of the patch file "patch" will ignore them lines10:21
TimB_which one do you talk about Romster?10:21
*** elderK has joined #crux10:21
Romstergot patch from source, bug report etc10:21
TimB_oh, so you don't forget the source, smart10:22
Romsteryep10:22
Romsterjust look in the patch and go oh that's where i got it from10:22
pedjaTimB_, for a moment I thought they failed to apply :)10:23
TimB_pedja: mission accomplished :B10:23
pedjaheh10:23
TimB_echo "\033[0;31mrunning rm -rf /\033[0m"10:24
Romster======> ERROR: Footprint mismatch found:10:24
RomsterMISSING   drwxr-xr-x      root/root       usr/share/vala/10:24
RomsterMISSING   drwxr-xr-x      root/root       usr/share/vala/vapi/10:24
RomsterMISSING   -rw-r--r--      root/root       usr/share/vala/vapi/libxfce4util-1.0.vapi10:24
Romster=======> ERROR: Building '/usr/ports/xfce-pedja/libxfce4util/libxfce4util#4.14.0-1.pkg.tar.xz' failed.10:24
Romsterdo you need vala? or just -uf the footprint?10:24
*** onodera has joined #crux10:30
pedjarequired, apparently.10:31
pedjaI have 0.44 installed10:32
Romsterthen vala needs to be listed on libxfce4util10:34
pedjait is. now :)10:34
Romsterhmm actually that's gobject-introspection not vala sorry...10:37
Romsterlibxfce4ui does need vala10:40
Romsterxfce4-settings missing garcon10:41
Romsteri'll just go though this entire thing again and make a new diff10:42
pedjaui depends on util, and that one requires vapigen10:49
*** stoffepojken has quit IRC10:54
*** stoffepojken has joined #crux10:56
TimB_stenur: trying bluez now with my bt headphones. works pretty well over a distance as well. not sure what to expect but i would say it's pretty decent11:31
TimB_stenur: if you don't use consolekit to manager your session, did you add your user to the plugdev group to manage your bt controller?12:00
Romstergst-plugins-base gst-plugins-base-32 gtk gtk-32 gtk3 gtksourceview libglade librsvg librsvg-bin librsvg-compat libwnck-gtk3 libxfcegui4 pango-compat vte3 vte3-ng12:33
Romsterall need rebuilding because of libtool argh12:33
Romsteri hate libtool12:33
also_uplimeim not sure why they thought it was a good idea to make warnings red12:35
Romsterso colour blind people can't see them?12:36
TimB_just to catch attention?12:37
Romsterit got also_uplime's attention :)12:41
TimB_so, it works? :P12:43
*** SiFuh has quit IRC13:02
*** jolupa has quit IRC13:07
*** SiFuh has joined #crux13:32
*** elderK has quit IRC13:46
*** raph_ael has quit IRC14:08
*** jolupa has joined #crux14:39
*** raph_ael has joined #crux14:48
*** onodera has quit IRC14:51
*** onodera has joined #crux14:55
stenurTimB_: no, for now i have an entry in sudoers.d/$HOSTNAME for that.  (For group "media", which can also mount some sticks etc.)14:56
stenurI do not even have that group.14:57
TimB_stenur: i see.. i am thinking about making a little readme about bluetooth on CRUX for bluez14:57
stenurShame that bluetoothctl has no manual.14:57
TimB_to just attach it to the port, people should take notice of readme files.. i hope..14:57
stenurI do.14:58
TimB_stenur: it's well explained in gentoo wiki, or arch one.. not sure14:58
TimB_i would basically just copy their instructions for kernel configuration and some more noteworthy stuff to get it going14:58
*** jolupa has quit IRC14:59
stenurAh, kernel configuration, ok.14:59
TimB_consolekit does take care without further ado, i figured it would be good to have instructions for a more puristic approach as well15:00
stenurbluetoothctl itself seems not been fully developed; the alias name overwrites the device reported one, it cannot be used for eg "connect"15:00
stenurI have written aliases to not mess around with device ids.15:01
TimB_ah15:01
stenurI do not know about consolekit, TimB_.15:01
stenurI have read an article someone about "seats" springing into existence.15:02
stenurThat's it.  And it was quite some time ago.15:02
TimB_https://wiki.gentoo.org/wiki/Bluetooth#Device_pairing15:02
stenurFunnily, this Philips monitor paired without entering anything, automatic.15:03
TimB_my stuff does the same15:03
stenurTimB_: yes, though i do not know what the agent does; i do not have this enabled (only short, does not matter here).15:03
TimB_sure, i dunno either, lightdm wants it i believe, and sessions just run smooth since, so why not?15:04
stenurI now have an alias ptv-pair 'printf "power on\nconnect 00:F4:8D:33:38:FA\n" | sudo bluetoothctl'15:04
stenurwhat is "lightdm" btw.  I did not find it nowhere?15:04
TimB_it's a loginmanager15:05
stenurAnd the reverse, too.  (Alias)15:05
stenurAha.  Ok.15:05
TimB_did i push it to contrib yet? not sure15:05
TimB_i did15:05
stenurTo my shame, i do not track contrib.git yet at all. :-)15:06
stenurBut will now, for bluez.15:06
stenurHm. Actually i am lying.  I do track contrib via HTTPUP, but not via git.15:08
TimB_they are the same files? what do you mean? :D15:08
stenurAnd lightdm there is. It was not in your github repo, that much is plain.15:08
TimB_it has been, for the longest time, but it is deleted over there15:09
stenurYeah, i am tracking some .git repos for half a decade or longer, even when i did not have had the opportunity to use CRUX beyond ISO install image.15:09
TimB_i still have lightlocker in my repo, that i wanted to commit to contrib at some point because it makes use of lightdm perfectly :)15:09
stenurNeed to throw those away.15:09
stenurDesktop manager, aha.  I do not have a desktop :-)15:10
TimB_like slim, gdm or whatever15:10
TimB_lightdm works very stable and fast, plus it makes use of pam which enables me to unlock my laptop with my yubikey15:11
stenurHaha.  No, not me, luckily this is still possible.  Sorry, TimB_.15:11
stenurThat is cool, however.15:11
TimB_haha whatever floats your boat right?15:11
TimB_it's a nice feature not having to type passwords for anything when in uni15:12
stenurAt the end of 2017 i shortly had a Debian, and i could not even move the mouse without causing swap orgies.15:13
TimB_what hardware is this?15:13
stenurAnd ... to my shame ... i could not even find the button which opens a terminal.15:13
stenurYes, Yubikey/alike really is a good thing i consider to have at some time.15:13
stenurThat was an old MacBook Air.  But it was in a VirtualBox VM.15:14
TimB_ok15:14
*** raph_ael has quit IRC15:15
*** SiFuh has quit IRC15:15
*** SiFuh has joined #crux15:16
*** raph_ael has joined #crux15:20
*** SiFuh has quit IRC15:22
*** onodera has quit IRC15:23
*** SiFuh has joined #crux15:26
*** SiFuh has quit IRC15:38
*** SiFuh has joined #crux15:39
*** SiFuh has quit IRC15:45
*** SiFuh has joined #crux15:50
*** SiFuh has quit IRC16:11
*** SiFuh has joined #crux16:12
*** SiFuh has quit IRC16:17
*** SiFuh has joined #crux16:22
stenurWhat is that "regtest" of libsndfile doing? It is enabled by not passing --disable-sqlite, sqlite is only used for it?16:25
stenurYou get a footprint mismatch if you have sqlite installed otherwisse.16:26
TimB_stenur: is it producing missing or rather just new files? there is a configure option in prt-get.conf for that i think16:28
stenurPardon?16:28
TimB_the footprint mismatch, is it about MISSING or NEW files?16:29
*** SiFuh has quit IRC16:29
stenurNEW, a -regtest program is build, because i have sqlite.16:29
stenurI wonder why libsndfile does not say --disable-sqlite, because that is only used for that, and what is _that_ -regtest program doing??#16:30
TimB_PKGMK_IGNORE_NEW="yes" it's in pkgmk-conf16:30
TimB_no idea16:30
stenurI did -if now. Just wondering.16:30
stenurTimB_: your pulseaudio does not work here, it cannot open module-alsa-card.so: "libalsa-util.so: undefined symbol: pa_alsa_ucm_device_update_available"16:54
stenurYep, src/modules/alsa/alsa-ucm.c does not provide this without HAVE_ALSA_UCM!16:59
TimB_stenur: iirc, you need to deploy an aplay.conf or what it's called to use alsa api over pulse17:00
stenurNah, "+void pa_alsa_ucm_device_update_available(pa_alsa_ucm_device *device) {}" in src/modules/alsa/alsa-ucm.c here. Let me try.17:02
stenurOk, it still does not work here, but at least that error goes away.17:12
stenurhttps://pastebin.com/raw/k2tJ0VUQ17:12
stenurDo not ask me what alsa-ucm is, though.  But here it is not.17:13
TimB_dunno17:13
TimB_are you trying out pa now?17:14
TimB_did you switch the programs settings to use pulse output as well?17:14
stenurYes, i wanted to have a log.17:14
stenurWhich programs?  I was trying "ogg123 -d pulse /x/p/ub/media/schulze_cyborg.ogg" yet only, it says "ERROR: Cannot open device pulse."17:15
stenurEven though i recompiled libao and it now has the pulse module.17:15
stenurStarted PA via pulseaudio --start --exit-idle-time=-1 --log-target=syslog.17:15
TimB_mh, k17:15
TimB_no idea about ogg123 but that should work i suppose17:16
TimB_i do use pavucontrol17:16
stenurEven as root, no chance.17:16
stenurWell, i look into that this evening.  It is not as big as i feared, and maybe as a shared system service it would be ok.17:17
stenurProblem here is that ALSA stutters if i plug in HDMI. Whatever that is.17:18
stenurSo i thought maybe get rid of xmix and bluealsa and use pulseaudio, if you all use that.17:19
TimB_'it just works'17:26
*** SiFuh has joined #crux18:46
*** zimmer_Dl has joined #crux19:27
*** zimmer_Dl has quit IRC19:28
*** SiFuh has quit IRC19:29
*** SiFuh has joined #crux19:30
*** zimmer_Dl has joined #crux19:42
*** zimmer_Dl has quit IRC19:44
*** Workster has quit IRC19:47
*** SiFuh has quit IRC19:49
*** SiFuh has joined #crux19:50
*** zimmer_Dl has joined #crux19:55
*** SiFuh has quit IRC19:56
*** zimmer_Dl has quit IRC19:56
*** zimmer_Dl has joined #crux20:06
*** zimmer_Dl has quit IRC20:08
*** zimmer_Dl has joined #crux20:14
*** zimmer_Dl has quit IRC20:14
*** zimmer_Dl has joined #crux20:34
*** zimmer_Dl has quit IRC20:34
*** zimmer_Dl has joined #crux20:40
*** zimmer_Dl has quit IRC20:42
*** zimmer_Dl has joined #crux20:55
*** zimmer_Dl has quit IRC20:57
*** SiFuh has joined #crux21:01
*** SiFuh has quit IRC21:06
*** zimmer_Dl has joined #crux21:08
*** zimmer_Dl has quit IRC21:11
*** zimmer_Dl has joined #crux21:16
*** zimmer_Dl has quit IRC21:18
*** monarrk has joined #crux21:27
*** zimmer_Dl has joined #crux21:31
*** zimmer_Dl has quit IRC21:34
*** monarrk has quit IRC21:35
*** zimmer_Dl has joined #crux21:39
*** zimmer_Dl has quit IRC21:41
*** SiFuh has joined #crux21:41
stenurPulseAudio is a real resource hog. Sound is fine, but over bluetooth. I have crystal clear bluez5-alsa, but total mess via PA.21:52
stenurIt seems compressed or so.  /usr/share/sounds/alsa/Front_Right.wav sounds like Micky Mouse.21:52
*** zimmer_Dl has joined #crux21:53
TimB_stenur i tried my bt headset for over an hour today, sounded great as long i was in range21:54
stenurHigh fidelity playback (A2DP sink), this should be the right thing. autodetect_mtu=1 for bluetooth discover also does not change.21:54
stenurTimB_: lucky you!21:54
stenurHmm.  Out of ideas for the moment.21:54
TimB_maybe, i was suprised how well it worked. i remember it being tricky when i left it, over a year ago21:54
TimB_i did use blueman btw, but i don't think that's needed21:55
stenurI installed ncpamixer to get through the sink jungle.21:56
*** zimmer_Dl has quit IRC21:56
TimB_i don't have a sink jungle?21:56
TimB_and it doesn't really hog much resources for me either, it did, i think with flat_volumes enabled21:57
TimB_i did some more changes today, some upped the resource usage again, but not much and it sounds sharper21:57
stenurNee, it is just that pactl list is a bit overwhelming, for example.21:57
stenur 21:57
stenurOh, almost 10 percent CPU, even if idle 1.7 percent!21:58
TimB_what? for me, under usage, 1,3%21:59
TimB_with pulseeffects on top of it21:59
stenurReally???21:59
TimB_yes21:59
*** SiFuh has quit IRC21:59
TimB_https://i.imgur.com/9dcWgYu.png22:00
stenurNo, 9.0-9.6 percent why ogg123 sends and PA goes to laptop builtin speaker.22:00
TimB_0.0% on my laptop idle, reported by htop22:01
stenurMysterious.22:01
stenurOk, 0.7 percent idle now.  With bluetooth disabled.22:02
stenurAnd enabled, too.  Whatever.22:03
TimB_http://dpaste.com/0F1R01N client.conf http://dpaste.com/12J6GK7 default.pa and http://dpaste.com/0MECDTH daemon.conf22:04
TimB_all in /etc/pulse22:04
TimB_daemon.conf nice level is too low, should go back to -11, it doesn't make much of a difference for me, else you should look at resample-method, flat-volumes, default-sample-format and the rest22:05
TimB_load-module module-udev-detect tsched=0 in default.pa might be helpful22:06
*** zimmer_Dl has joined #crux22:07
stenurOh wow, wait. That sched thing i try right away..22:07
TimB_i do not have a user config22:09
*** zimmer_Dl has quit IRC22:09
stenurNo. That wasn't it. I will look at your pastes.22:09
TimB_like i said, it was acting weird for me without further configuration too. and i do suspect flat-volumes to be the bigges culprit22:09
*** zimmer_Dl has joined #crux22:15
*** zimmer_Dl has quit IRC22:16
stenurHA!  sample-rate makes it sound normal. But have clicks for each output device, but at least all equal now!  Thanks, didn't even look at that yet.22:18
TimB_cool22:18
*** zimmer_Dl has joined #crux22:23
stenurYes, cool, thanks.  96000/48000 even better.  4.0 percent CPU left, it likely had to resample a lot!22:23
stenurI still have some clicks though.22:26
stenurHaha, and mind you, when i open firefox in the other account and start a video then playback speed halfs.22:26
*** zimmer_Dl has quit IRC22:27
ryuostenur: i once had to disable dmix because it was too much for the ARM cpu to handle.22:27
ryuoit fixed all the audio problems.22:27
stenurryuo: ALSA is no longer running here!22:28
stenurOr: is /etc/asound stuff still checked now?22:28
ryuoit is.22:28
stenurWait.22:28
ryuoit has an impact on anything still talking to ALSA.22:29
stenurNo, does not change anything. Firefox started directly, not via "apulse firefox". Start it in other account with own pulseaudio, speed halfs.22:30
ryuoerm... PA has to use ALSA or another thing to output sound.22:31
ryuoif it uses ALSA, asound configuration will still have some impact.22:31
stenurDoes not restore it seems, need to stop.  And wait a few seconds.  Quite a few seconds.  Still.22:31
stenuro-ho-hoo!  Even worse!  Quarters!22:33
*** zimmer_Dl has joined #crux22:33
stenurMaybe i should only start one PA as a system service?22:34
*** zimmer_Dl has quit IRC22:35
stenurHe does not seem to like that the way it is compiled, somehow.22:37
TimB_pa system service is discouraged by upstream22:37
stenurYes. What the heck. I go back to ALSA. Thank TimB_.22:38
stenurThanks TimB_ i wanted to say.22:38
TimB_really? ^^22:39
stenurWhat really? Back to ALSA? Yes, in process. That bluez5-alsa worked just fine.22:40
TimB_ok22:42
*** zimmer_Dl has joined #crux22:42
stenurYeah i mean, almost four hours is enough. Won't come to bed before hm six in the morning. And want to start on a working system on Monday.22:43
*** zimmer_Dl has quit IRC22:43
TimB_but didn't it work just now?22:46
stenurNo, i had some "click hangs".  And when the firefox in the browser account wanted to play sound, music speed slowed down by half and more.22:47
TimB_wtf22:47
stenurYes.  Do you use multiple user accounts?22:48
TimB_no22:48
*** zimmer_Dl has joined #crux22:50
*** zimmer_Dl has quit IRC22:52
*** zimmer_Dl has joined #crux22:58
*** zimmer_Dl has quit IRC22:59
*** zimmer_Dl has joined #crux23:06
*** zimmer_Dl has quit IRC23:08
stenurA nice sunday! Ciao.23:15
*** john_cephalopoda has quit IRC23:17
*** john_cephalopoda has joined #crux23:19
TimB_to you too23:20
*** zimmer_Dl has joined #crux23:21
*** zimmer_Dl has quit IRC23:22
*** zimmer_Dl has joined #crux23:28
*** zimmer_Dl has quit IRC23:30
*** jolupa has joined #crux23:31
*** zimmer_Dl has joined #crux23:42
*** zimmer_Dl has quit IRC23:44
*** jolupa has quit IRC23:45
*** SiFuh has joined #crux23:51
*** zimmer_Dl has joined #crux23:51
*** zimmer_Dl has quit IRC23:53
*** SiFuh has quit IRC23:55
*** junland has quit IRC23:58

Generated by irclog2html.py 2.14.0 by Marius Gedminas - find it at mg.pov.lt!