libera/#devuan/ Wednesday, 2024-06-05

fatsothe change to lightdm for some reason screwed up my scaling config00:15
gnarfacefatso: here's what i have for mesa on current stable http://paste.debian.net/1319210/ (if the paste is still there, i put that up before you disconnected)00:16
fatsoor maybe it was because i tried to use an x11 session? because it was screwed up on x11 first00:16
gnarfacewhat were using to scale?00:16
gnarfacedid you make that xorg.conf change?00:16
fatsoit's still there, thank you00:16
gnarfaceyou probably don't need both architectures unless you're using Wine00:17
fatsoi didn't but i configured sddm to reflect KDE's configuration, so now my SDDM login screen is exactly like my lock screen00:17
fatsoand I didn't make the xorg.conf change00:18
gnarfacewhat are you scaling, the entire desktop or just fonts?00:19
gnarfaceif it's the whole view, this may be something you can do with xrandr either on command-line or in the xorg.conf too00:19
gnarfaceif it's just fonts, i guess that's toolkit specific00:21
fatsoit's the whole view00:24
gnarfaceyea, that's a xrandr thing i think, which means you can also specify it in xorg.conf with your intel driver if you want, or you can add it to a startup script with the console binary00:25
gnarfacei dunno if the kde config will conflict but you can likely do it instead of00:25
fatsoI'd rather not mess with it again haha00:37
nemohttps://news.ycombinator.com/item?id=40578414  where does devuan stand on this?02:32
nemowill /var/tmp get erased prior to reboots?02:33
rrqwhat program/script would erase /var/tmp ?03:39
buZzrm03:41
rrqyou mean that one of the rc6.d/K* files would have `rm /var/tmp/*` ?03:42
buZzoh, like that03:42
buZzcould be?03:42
rrqnone of mine though... and it depends on init system and setup I guess03:42
rrqaccording to FHS 3.0, /var/tmp contains "Temporary files to be preserved between reboots."03:46
rrq(but who bothers about FHS nowadays?)03:47
buZzwell, systemd03:47
buZz:P03:47
buZzon this debian machine here its completely stuffed with systemd files03:48
buZzwell, 10MB of it03:48
XenguyI didn't know about /var/tmp, I always created my own directory off the / , to persist across reboots since /tmp flushes everything on reboot03:51
rwpThe only totally safe time to remove things from /tmp and /var/tmp is at initial single user boot time before going multiuser and starting up sshd and gettys and such.  That's why we usually clear /tmp at boot time.  It's the safe time to do it.03:54
rwpBut /var/tmp is supposed to preserve files across reboots.  So that one is more problematic.  Can't clear it at boot time.03:54
XenguyMaybe /var/tmp is a feature and not a bug?03:54
rwpAnd there will be endless discussion about how to clear a temporary directory during the normal multiuser run time of the system in the presence of hostile agents operating on the system.03:55
rwpI went looking for the tmpreaper README and found a copy of it here: https://fossies.org/linux/tmpreaper/debian/README.security03:58
rwpIt's good reading for the background of the problem of it on a multiuser system with malicious agents actively operating on it.03:59
XenguyFair enough, but it seems like if one has hostile users on a system the main battle has been lost?04:04
rrqmmm at most universities there are at least 2 students with accounts that will want to "break in"... which possibly counts as "hostile"04:06
XenguySure they exist, but the main mission is to detect and ban such users, no?04:07
rrqor maybe find them and employ them :)04:10
XenguyFriends and enemies I suppose : -)04:11
fatsokde says my thunderbolt subsystem is not available. Should I see it even when nothing is connected to it:05:34
fatso?05:34
fatsoI'm on a thinkpad T480, know to have thunderbolt issues, but successfully managed to update its firmware05:35
fatso*known05:36
gnarfacefatso: dunno, did some quick searches though and i found one suggestion to disable "Thunderbolt BIOS Assist Mode" in the BIOS so fwupd can see it06:25
gnarfacehttps://www.reddit.com/r/thinkpad/comments/12tf6xv/psa_t480_thunderbolt_controller_v23_is_now_on/06:25
fatsoI did it and updated the firmware06:28
fatsotried changing that setting to "enabled" and it resulted in thunderbolt not even being recognized by "lspci"06:29
gnarfacei don't have one, but they're popular so someone around here certainly knows...06:38
onefangAm I doing something wrong, or does /var/cache/apt NOT like living in a symlink?  I keep getting an extra / added between the path and the package when it's trying to find it during install.09:26
amarsh04hmm... package hd-idle is still broken in Devuan - I had reported a bug against it in Debian but it hasn't been fixed10:57
cousin_luigiamarsh04: Which one are you talking about?11:06
cousin_luigiamarsh04: 1.05 is seemingly abandoned, debian testing has switched to the 1.21 go reimplementation11:06
amarsh04Debian bug 1069637, version 1.21 doesn't appear to work with Devuan11:08
cousin_luigiamarsh04: Is it about the configuration file in /etc/default ?12:56
amarsh04if one installs the later version, it default to hanging as sane defaults (even disabling by default) are not provided13:00
cousin_luigiamarsh04: I'm sorry, are you running Daedalus?13:08
amarsh04cousin_luigi - ceres13:12
cousin_luigiamarsh04: Possibly #devuan-dev then?13:14
amarsh04thanks13:20
rwpXenguy, rrq, I describe it like that, malicious local agents, because on a personal laptop where there is only one user and it only operates on the coffee table on a protected LAN then there is little worry about an attacker breaking the system.  Trying to keep it in perspective.21:55
rwpBut for an OS distribution which will be run in all types of environments then there will be all types of complaints about security of the default configuration.  The default configuration should be secure by default.21:55
rwpThat's harder to do in the absolutely general case than it at first appears.21:56
rwpMeanwhile...  I am running tmpreaper on my systems and I have configured it to expire files from both /tmp and /var/tmp after they age sufficiently to make them old files that should be removed automatically.21:56
Xenguyrwp, Interesting, first I've heard of tmpreaper22:08
rwpRemoving files from random users in a tmp directory is somewhat tricky-hard.  I think the process should become the uid of the owner of the file before attempting the unlink action.  That seems like it would have the best chance at avoiding being gamed.22:12
rwpIt's never just a single file though.  It's a possible entire deep tree of files.  And that's also tricky.22:13
rwpThat's where the new openat*() set of library functions have now taken everything over.22:13

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