IRC Logs for #crux Monday, 2017-03-27

*** chinarulezzz has joined #crux00:11
*** timcowchip has quit IRC01:53
*** moth has quit IRC02:14
*** emmett1 has joined #crux02:22
*** cippp has quit IRC02:24
*** _________mavric6 has quit IRC02:29
*** timcowchip has joined #crux02:29
*** _________mavric6 has joined #crux02:30
brian|lfsso Rufus can't even make a dam UEFI windows 10 install media04:00
jaegersure it can, I've used it many times for that04:07
*** chinarulezzz has quit IRC04:07
*** chinarulezzz has joined #crux04:08
brian|lfsQuestion why don't we have gfdisk on CRUX04:08
brian|lfsseeing we all have EFI now or msot of do I think it should be added to core?04:09
jaegerparted is there04:09
jaegerwell, in opt04:09
brian|lfsyes parted is there but for install be nice to have gfdisk04:12
jaegerI suppose it's preference, though parted supports gpt fine04:13
*** tilman has joined #crux04:36
*** nwe has quit IRC04:37
*** nwe has joined #crux04:37
*** abenz_ has joined #crux05:21
*** abenz_ has joined #crux05:57
cruxbot[opt.git/3.3]: freerdp: 20170320 -> 2017032406:02
brian|lfsseeing we all have EFI now or msot of do I think it should be added to core?06:16
brian|lfsseeing we all have EFI now or msot of do I think it should be added to core?06:16
brian|lfs 06:16
brian|lfs ;;06:16
cruxbot[opt.git/3.3]: freerdp: use git-describe version scheme06:34
*** chinarulezzz has quit IRC06:44
timcowchipI'm trying to view an ipcam d0 I need xorg-xf86-video-v4l?07:28
ryu0jaeger: if you ship a recent enough util-linux, it has adequate GPT support now. i don't see much point in gptfdisk anymore.07:29
ryu0jaeger: nor parted really. i only ever needed it for obscure labels sometimes.07:29
timcowchipdoesn't gparted need parted?07:31
ryu0yes it does, but i thought the crux install media didn't have a GUI present.07:31
ryu0all i was getting at is there's not much reason left to keep parted in core.07:32
ryu0it's making more and more sense for it to be in opt, if anywhere.07:32
timcowchipoh, install media....nvm07:32
ryu0err nevermind, it appears it's not in core anymore.07:32
ryu0util-linux provides enough features with fdisk.07:33
ryu0hm. curious they still include lilo in the base. i would consider syslinux to be a better BIOS bootloader now.07:33
ryu0it's actually still developed and doesn't require you to reinstall it every time...07:34
ryu0syslinux also comes with a EFI bootloader now too...07:35
ryu0timcowchip: ipcam? is that a network stream?07:37
timcowchipits a amcrest cam on my front porch07:38
timcowchipits has a 2 ip's07:38
timcowchipone for wifi one for wired07:39
timcowchipI can view it in google-chrome with an extension07:39
ryu0i've only heard of V4L being used on local devices.07:39
timcowchipbut not on crux07:40
timcowchipoh ok07:40
ryu0i don't know if it even has a network backend.07:40
timcowchipso not what I'm looking for07:40
timcowchipI remember using it for webcams07:40
frinnstryu0: lilo is stupidly easy07:47
*** timcowchip has quit IRC07:59
*** onodera has joined #crux08:15
ryu0frinnst: so that's why you still use it?08:32
ryu0frinnst: i found it a hassle to have to reinstall it every time your kernel is rebuilt.08:33
*** abenz_ has joined #crux08:41
*** onodera_ has joined #crux09:03
*** onodera has quit IRC09:03
*** onodera_ has quit IRC09:03
*** onodera has joined #crux09:03
frinnstryu0: i use "make install" when building the kernel - copies the image to /boot/vmlinuz, renames the old vmlinuz to vmlinuz.old and runs lilo (if installed)09:25
frinnstit usually use it for VMs and such09:38
frinnstdoesnt work well with btrfs and i'll use uefi when i buy a m209:38
frinnstspeaking of m2, i should buy one now09:38
*** john_cephalopoda has joined #crux09:51
dbrookeyou can hook the kernel 'make install' for UEFI etc., I boot with the kernel stub and customised a script I found on the 'net -
frinnstthe only thing it does if you dont have lilo installed is copy the image10:02
*** emmett1 has quit IRC10:05
abenz_which one frinnst ?10:15
abenz_the 960 pro?10:15
*** snux has joined #crux10:35
frinnsti only found pro in 512gb and up10:37
frinnsttoo expensive10:37
*** blueness has quit IRC10:51
*** snux has quit IRC11:09
*** onodera has quit IRC11:31
*** onodera has joined #crux11:31
*** abenz__ has joined #crux11:42
*** abenz_ has quit IRC11:45
*** chinarulezzz has joined #crux12:08
*** blueness has joined #crux12:25
*** emmett1 has joined #crux12:28
*** onodera has quit IRC13:43
*** emmett1_ has joined #crux14:10
*** blueness has quit IRC14:14
*** onodera has joined #crux14:37
cruxbot[contrib.git/3.3]: containerd: updated to version 977c51115:41
cruxbot[contrib.git/3.3]: runc: updated to version a01dafd15:41
cruxbot[contrib.git/3.3]: docker: updated to version 17.03.0-ce15:41
*** lounge has joined #crux15:51
jaegerfrinnst: if you want to go nvme, the intel 600p line is pretty cheap15:59
jaegerI bought one recently. Haven't had it for a long time but it's fast16:01
*** emmett1_ has quit IRC16:01
*** SiFuh_ has quit IRC16:04
*** emmett1 has quit IRC16:07
*** SiFuh has joined #crux16:08
frinnst960 evo was about the same16:10
jaegerOh, nice. I didn't know they had come down closer. That's good to hear16:11
*** chinarulezzz has quit IRC16:20
pedjastupid question: NVME SSD's are plugged where exactly?PCI-Express and/or M.2 slot?16:22
pedja(I am way behind a new tech :) )16:23
pedjathe idea behind them is that the system sees them more like RAM than the traditional spinning rust?16:26
abenznot really16:26
abenzSATA had the speed at 6gbps16:26
abenzbut SSD tech allows for faster speeds16:26
abenzso alternative methods of connecting the NVME based storage to system were needed16:27
pedjaNVMe over M.2 or U.2 (using PCI Express 3.0 ×4 link) 32gbps (Wikipedia).damn, that's fast :)16:28
abenzso you have a few ways, one of which is using an adaptor that plugs into your PCIE slot, into this adapter you plug in the nvme16:29
abenzlike this:
abenzbut recent motherboards offered the slot directly so you can plug it right in16:29
jaegerMany motherboards now have m.2 slots that support both AHCI SATA and NVMe16:29
pedjahow is Linux support for all that?16:33
jaegerseems fine16:33
pedjaah, yes, you have one :)16:34
pedjathanks abenz, jaeger16:34
jaegerpedja: also, hopefully those warnings you mentioned are gone from docker daemon startup now16:36
pedjaheh.docker is whining little bitch :)16:37
jaegerit can be, yeah16:37
pedjasince you were busy, I forked your docker ports and updated to 17.03.0 a while back.16:43
pedjaonly difference is that I pass the whole commit ID to make.It was the only way to shut it up :)16:44
*** chinarulezzz has joined #crux16:44
pedjaalpine port does something similar, iirc16:44
jaegerI did it with the short commit ids and I didn't see the warnings but I wasn't sure what they looked like before16:45
pedjaif 'docker info' shows N/A for containerd/runc version, daemon will complain,'s harmless, afaict16:47
jaegeroh. hrmm. Maybe I have to switch to the full commit? I'll test it16:47
pedjathey should really fix that :)16:48
pedjawell, once the containerd or runc (I forgot) hit 1.0, it will become easier16:49
pedjabigger problem for me is the warning that JFS doesn't support d_type16:50
*** chinarulezzz has quit IRC16:50
*** chinarulezzz has joined #crux16:50
pedjawhich will become an error in a few release (what was supposed to be .16)16:51
jaeger'docker info' does show "<short> (expected: <long>)" but seems to work fine16:53
pedjait's a sanity check of sorts.16:54
jaegerusing the long commit does change the docker info output slightly. I'll update them to use that, though I'm going to leave the port version the short commit16:56
pedjabut no matter what I try, I can't get it to recognize that the version of tini is correct.a bit annoying16:56
pedjathat's what I did16:56
pedjainterestingly enough, dynamic version of tini completely breaks it.static one works fine.16:58
pedjaalpine containers don't work with either :)16:59
jaegeris tini or docker-init needed in general? I've never used it17:00
pedjait stops the containers gracefully instead of killing them17:01
jaegerAh, the PID 1 reaping problem17:03
pedjasome people use tini, some dumb-init.they both work fine, for most containers17:05
pedjaI've set tini as default, and I might try dumb-init for alpine, maybe it works better17:06
cruxbot[contrib.git/3.3]: containerd: switched make command to full commit hash17:08
cruxbot[contrib.git/3.3]: runc: switched make command to full commit hash17:08
pedjaif you are looking for GUI, portainer and Weave Scope are pretty nice17:09
jaegerI don't use one so far17:10
pedjaI played with them, to see what's the all fuss about, and to visualize what the hell is going on17:12
jaegerI'll take a look at adding tini to the build, maybe17:13
pedjasince I am researching how to make temporary PXE boot server using Docker, they might come in handy17:14
pedjaif you do, I would suggest static build, I had some issues with dynamic17:14
jaegerstatic build is required, I imagine17:14
pedjashared doesn't depend on much, iirc, glibc and few other libs, but for some reason Docker prefers static build17:16
pedjaat least when I test it17:16
pedjaand yes, you'll have to use a specific commit of it17:17
pedjaotherwise subtle breakage happens17:18
pedjafinicky beast, this Docker17:18
pedjain theory, you should be able to use any container runtime you want in the future. now, *that* would be interesting17:24
jaegerdoes tini have a way to pass the commit via cmake or make or do I need to keep the .git dir?17:28
jaegerI don't see an easy one searching through the files17:28
jaegerIt does work keeping the .git dir, just curious17:29
pedjathat's how I did it17:36
pedjait might be wrong way to do it, since 'docker info' lists N/A as init version, though17:39
pedjabut I am out of ideas17:40
pedjaand it's an exact version that it requires, so it shouldn't a problem, but who knows17:41
jaeger is what I used, but the source tarball is tarred from the exact commit17:42
jaegerI don't have strong feelings about it, most of my usage is /bin/bash which doesn't have the problem17:42
jaegerinit version: 949e6fa17:43
otwieraczAm I the only one with such problem?17:43
jaegerotwieracz: do you have the appropriate dependencies installed?17:43
otwieraczWell, yes.17:43
jaegerfreetype and freetype-32 are both in the list17:43
otwieraczI was using depinst at least.17:43
jaegermight want to double check it or even reinstall freetype-32 if it's broken somehow17:45
otwieraczoh wait!17:46
otwieraczI don't have compat-32 enabled!17:46
otwieraczI've got a feeling like „why there was only one or two dependencies installed while there was complete list…”17:47
*** tsaop has joined #crux18:13
otwieraczwell, this time db-32 failed.18:21
*** john_cephalopoda has left #crux ("Trees are hiding until the final day.")18:28
*** timcowchip has joined #crux18:33
*** tsaop has quit IRC18:38
*** timcowchip has quit IRC18:47
*** timcowchip has joined #crux19:14
dbrookeI have a port whose upstream source now contains UTF-8 characters in some of the file path/name and because pkgmk sets LC_ALL=POSIX bsdtar fails the extraction, any thoughts?19:42
*** blueness has joined #crux19:43
*** chinarulezzz has quit IRC19:51
*** chinarulezzz has joined #crux19:51
*** cippp has joined #crux20:13
*** lounge has quit IRC20:19
*** blueness has quit IRC20:20
*** blueness has joined #crux20:44
*** blueness has quit IRC20:51
*** blueness has joined #crux20:57
*** blueness has quit IRC21:05
*** cippp has quit IRC21:18
*** cippp has joined #crux21:20
Worksterdbrooke, override unpack_source() in your Pkgfile is a option22:07
*** timcowchip has quit IRC22:07
*** blueness has joined #crux22:20
*** just_fun has quit IRC23:14
*** just_fun has joined #crux23:28
*** ryu0 has quit IRC23:44
*** blueness has quit IRC23:52
*** blueness has joined #crux23:52
*** onodera has quit IRC23:56

Generated by 2.14.0 by Marius Gedminas - find it at!