libera/#devuan/ Saturday, 2019-09-07

drawkulaoverwriting the wrong partitions or drives with a clicky coloured tool would not feel better than making mistakes with dd00:09
golinuxEven as clicky-loving as I am, I wouldn't do that either00:13
* drawkula still gets adrenaline tsunamis from dd-ing to partitions or disks...01:06
* golinux gets sweaty palms and palpitations01:07
drawkulafor critical commands there is only one way to do them...01:08
drawkulalook at them long enough before hitting enter...01:09
drawkulamaybe start typing those lines with a # to avoid accidentally launching them01:09
golinuxTriple check over several hours01:10
drawkula:-)01:10
drawkulasome really focussed minutes should do...01:10
golinuxSometimes I unplug unmounted drives01:11
golinuxI am old and move slowly01:11
specinguse the links in /dev/disk/by-id, not /dev/sd* directly01:11
drawkulacomparing /proc/partitions to the expected drive/parition size can give hints...01:12
drawkulai often sit too long in front of screens ... overtired ... hit shappens! bit i try my best to minimise it01:13
rdavdrawkula: do a "tail -f /var/log/messages" and pop the disk out and then back in and you will see the correct /dev/sdX the kernel has given01:49
drawkulanah... usb plugs are to *hity01:51
rdavalso can use the bash auto-complete to check out the devs that are available for dd to use ie "dd if=foo of=/dev/sd[leave this blank]"  and hit tab to see the /dev/sdX available01:51
rdavI use hotplug sata in a box for this01:52
drawkulaheyyy... I'm using unix for +7- 3 decades now01:52
drawkula7 -> /    (shitty shift key)01:52
* drawkula ____( Trust me, I know what I'm doing. )01:52
rdavhmm my first unix was in .... 1980?01:52
rdavawesome!01:52
rdavI know the feeling after dd wrong disk some yrs back01:53
rdavso I take these extra steps to double check01:53
drawkuladd HAS to feel dangerous... that's a security feature101:55
gnarfaceif anyone has trouble starting Xorg after upgrading ceres today, make sure you're using the glvnd versions of the various pacakges rather than the non-glvnd ones (dunno why, dunno why it keeps happening, either)02:05
gnarfacesorry of the various nvidia driver related packages02:05
gnarfaceto be clear02:05
gnarfacenot the first time that has sideswiped me, though i'm not clear if the non-glvnd ones sometimes worked and i didn't notice, or if the part i keep missing is it switching them out during upgrades02:06
gnarfacei *thought* they were both working before but i could be wrong02:06
gnarfacei also purged the old version of nvidia-persistenced which doesn't appear to have an updated version in the repo yet, though i'm not sure if that step was necessary now02:07
Xenguy_Sendmail?  Surely you jest = )03:10
XenguyI always thought postfix should have been the default MTA, but they keey picking Exim for some reason03:11
Xenguyer, keep03:11
hightower3hm strange, so on ppc63le ascii I had kernel 4.9. Then I upgraded to unstable which also added kernel 5.2. Now, when grub chooses 5.2 and boots it, a couple lines below it says 'starting kernel via...' and this hangs with 1 cpu core going to 100%, whereas with 4.9 it works.03:24
lifebookhey! some news about 4.19 beowulf vs 4.9 ascii kernels combat in computer with low memory 256 MB and monothread CPU10:43
lifebookwith 4.19 the system is not usable because it keeps on swapping10:44
lifebookjust tested with 5.2 (buster-backports from debian) and good surprise10:45
lifebookthe system may swap but does not keep on swapping10:46
lifebook5.2 behaves like 4.910:46
lifebookthere may be something wrong in memory management from 4.1910:48
lifebookmaybe in mm_percpu_wq10:48
lifebookthe memory is allocated but not freed10:49
lifebookas 5.2 kernel has better protection against spectre n friends10:50
lifebooktime to get on without 4.9 ascii kernel10:51
lifebookif anyone else got trouble with ressource limited system yet wants memory management10:55
lifebookgo get 5.210:55
lifebook:)10:55
lifebookit may also be related to rcu_gp (grace period)10:56
lifebookthanks to all11:01
XenguyIs everyone else on Ascii getting a huge update this morning, including ssh etc.?16:05
yeti32/36 packages upgraded on amd64/i38616:16
yetifor my last 2 ascii VMs16:19
fsmithred75 packages upgraded five hours ago. I just checked and there are another 41.16:34
fsmithred82 available on another ascii, and this is the one I've been using to test the update notifier. Previous dist-upgrade was three days ago.16:47
ErRandir61 updates here16:54
XenguyThanks folks17:13
systemdlete2My laptop hangs when it exits xfce to lightdm (slim is no longer installed)21:43
systemdlete2boots to lightdm without issue and I can log in without issue21:43
systemdlete2(othr than display being sideways)21:43
systemdlete2xfce display is landscape21:44
systemdlete2also, I cannot get cinnamon to launch; it just hangs.21:45
sky-kunHello folks.21:52
sky-kunAnyone here have good knowledge with nvidia cuda and blender? I didn't get it to work on my devuan ascii. I have the nvidia drivers version 390 installed, but in blender no cuda support is detected.21:54
hightower3Hey not particularly important, but just repeating the report from yesterday, which is that on Linux 5.2.0-2-powerpc64le (the kernel from debian unstable), my install hangs at message "Booting Linux via __start() @ 0x000000000002000000" and once CPU core goes to 100% (I notice this because running in a VM)22:46
hightower3Kernel 4.9 works fine22:46
hightower3s/once/one/g22:47
xrogaansystemdlete: have you checked your logs?23:06

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