libera/#devuan/ Wednesday, 2020-12-09

tuxd3vbut I believe both contains a spi, to store bootloader00:00
tuxd3v:S00:00
clorti don't think this is nda type stuf00:00
tuxd3vat least some parts of it00:00
clortcan i post the file?00:00
tuxd3vyeah it would be nice to know what is inside :)00:01
clorthttp://0x0.st/ihOX.xml00:01
clortthere is too much information i cannot parse on this webpage. i am lost.00:02
tuxd3vyeah.. Nvidia won the context of hiding things true obscurity :D00:02
clortnano boot flow docu: https://docs.nvidia.com/jetson/l4t/index.html#page/Tegra%2520Linux%2520Driver%2520Package%2520Development%2520Guide%2Fbootflow_jetson_nano.html%2300:03
tuxd3vcontest00:03
onefangI figured out you meant contest by understanding your context.  B-)00:03
clortthis stuff gives me a gross soiled feeling to read.00:05
clortbut u-boot looks familiar - they have a u-boot customization page00:05
clorthttps://docs.nvidia.com/jetson/l4t/index.html#page/Tegra%20Linux%20Driver%20Package%20Development%20Guide/uboot_guide.html00:05
clort"replacing the kernel" looks good00:06
clortbut... what... that makes no sense.  they mean using nvidia kernels?00:07
tuxd3vread the section 'CBoot' of last codument :)00:07
clortit 'chains to u-boot' so after u-boot it's gone yes?00:09
clortit's not some evil hypervisor type thingy is it?00:09
tuxd3v'Supports display and boot logo/bmp splash'00:10
tuxd3vthe nvidia logo should come from there..00:10
tuxd3vtilda in chimaera dies prematurely00:16
tuxd3vhttp://paste.debian.net/1176155/00:16
tuxd3vwhen you start to pull it down and up00:16
tuxd3vsometimes it doesn't survive to one operation00:17
n4diri once had the keyboard shortcuts screwed for tilda. Which made it kinda crash00:17
n4diryou could mv it's config out of the way, to test that00:17
n4diri doubt that is the problem, but it's a quick test.00:18
n4dirtuxd3v: also https://github.com/lanoxx/tilda/issues/348 middle of the page, disable "show on all desktops". As far its me that defeats the whole purpose of tilda00:19
fsmithredtuxd3v, when I start tilda in chimaera, the window is already pulled down. No need for F2, and it does not crash.00:25
fsmithredI'm on amd64 here00:26
tuxd3vsorry ..on a mobile call( but not in the matrix :D )00:31
tuxd3vfsmithred, I am in i38600:31
tuxd3vfsmithred, does you use tilda?00:31
fsmithredno, but it's installed00:32
tuxd3vn4dir, thanks, I don't really know whats going on it just crash00:32
n4dirat the end of that page the dev says a newer version would solve the problem00:32
tuxd3vI never saw it surviving to 3-4 puul up pull down events00:33
n4dirvresion 1.5.300:33
n4diri used it for very long, but not anymore. So no more details i could offer (and am on stable)00:33
fsmithredit's still in Refracta for you00:33
tuxd3vn4dir, they are asking maybe for a coredump00:36
tuxd3vto nalize where the problems comes from.. I didn't read all yet, but will00:36
n4dirtuxd3v: yeah, i don't know of such (besides it exists)00:36
tuxd3vtilda... bad mojo -.-00:40
tuxd3vn4dir, thanks yes they  state that00:41
tuxd3vwe have version v1.5.2-1 on chimaera :(00:42
tuxd3vfsmithred, does you know if its possible to push a package from ceres to chimaera?00:43
fsmithredthat is the normal progression00:43
tuxd3vI believe the consequences could be drastic because of dependencies?00:43
fsmithredI think it happens automatically00:43
fsmithredafter a week or two00:43
tuxd3vho, many thanks00:43
tuxd3vI will wait for it00:44
fsmithredsame version is in ceres00:44
tuxd3vits that I am so badly adicted to tilda, that I  feel myself lost without it00:44
n4dirand with all the right in the world. tilda is pretty cool00:44
n4diri gave up on it as i had low-specs machines for years00:44
tuxd3vceres receives what is in sid right?00:45
tuxd3vprobably they will have to update it00:46
gnarfaceyes00:46
fsmithredyeah00:46
gnarfaceceres == sid00:46
tuxd3vwe don't know how long it will take00:46
fsmithreddid you check for bug reports?00:46
gnarfacemost cases the propogation delays are not more than a couple hours00:46
tuxd3vtil now only the page n4dir showed on github, but the problem seems debugged00:46
n4diron debian i would not hesitate to install a package from unstable on a testing system.00:47
n4dirno idea for devuan.00:47
fsmithredI see two bug reports about crashes at bugs.debian.org00:48
fsmithredyeah, really. Mixing testing and unstable is pretty safe. Just pin ceres to a lower priority so you only get what you ask for.00:48
fsmithredif it's fixed in git but not in repo yet, I have no idea how long you wait.00:49
fsmithredbackport it.00:49
tuxd3vhttps://bugs.debian.org/cgi-bin/pkgreport.cgi?dist=unstable;package=tilda00:53
n4diri think, but don't know, that as long vst (? -> ups: vte ) is from the repo, compiling tilda itself ain't that much of a problem00:55
n4dirsure would give it a try, if in doubt compile it to ~/bin00:55
tuxd3vn4dir, I understands when you say that now you go for a light desktop reources system... tilda is not low on resources, at least it seems not so low00:59
fsmithredthe github page has instructions: https://github.com/lanoxx/tilda/blob/master/HACKING.md01:02
EHeMAre the Devuan versions of the kbd package modifed from the Debian ones?01:29
EHeMHmm, second thought, nevermind.01:29
EHeMJoy!  Seems kbd got modified inappropriately for Devuan: #79658301:31
EHeMOr perhaps I should point to the changelog at version 2.0.3-2 for kbd.  :-(01:32
clortplease do help fix issues you find01:34
gnarfaceEHeM: please report bugs in forked packages, they'll be taken seriously if legit: http://bugs.devuan.org/01:35
gnarfaceEHeM: (obviously patches would be welcome but i'm sure if it's an obvious mistake they'll jump on it fast either way)01:36
EHeMgnarface: In this case it is a bug in an unforked package, appears it got broken for systemd and no longer works so well for Devuan.01:46
gnarfaceEHeM: well, that sucks but doesn't actually change the advice01:48
EHeMgnarface: I guessed as much, but it seemed a large enough difference to warrant mention before proceeding.01:49
gnarfacefsmithred: ^01:51
gnarfacefsmithred: have you heard about this yet?01:52
fsmithredcatching up01:52
fsmithredanything we modify from debian gets +devuan in the version01:52
fsmithredEHeM, did they remove the init script?02:00
EHeMfsmithred: That is exactly what the changelog says, version 2.0.3-2, "Stop shipping /etc/init.d/kbd and /etc/kbd/* (Closes: #796583)"02:25
fsmithredwe'll add it to the list. Please file a bug report submit@bugs.devuan.org02:27
EHeMfsmithred: Bug #531 (settings on the bug system seem to have gotten interesting).02:52
fsmithred thanks02:54
fsmithredI always use email to interact with the bug system02:54
EHeMExcept Debian managed to get some sort of rather effective spam-fighting measure into place which makes things much easier.   :-(02:57
* EHeM . o O ( Hmm, what did the RPF do to HDMI blanking functionality??? )02:59
danuananyone have any ideas on how to autoexecute a script in singleuser mode . cannot get anything to run after rc1.d/S02single does exec init -t1 S  , runs fine before that as S01 but daemons/processes have stopped yet byt that point.04:36
tuxd3vhello, I need to get the source code of 'libnewt-dev' in chimaera05:42
tuxd3vI updated sources.list with deb-src05:42
tuxd3vbut I can't get the package :/05:42
tuxd3vdoes someone knows I to get it?05:42
tuxd3vthe API changed from beowulf to chimaera, in chimaera is v0.52.21-4+b305:43
masontuxd3v: did you apt update after you added the entry?05:43
tuxd3vshouldn't be major changes since it maintains itself with the v0. but, there are :/05:44
masontuxd3v: And once you did that, how'd you try to get the source?05:44
tuxd3vapt-get source libnewt-dev05:45
masontuxd3v: you said 'apt update' first?05:45
tuxd3vyea, I did05:45
masonHrm.05:45
tuxd3vthere are API changes that I need to follow in :05:45
masontuxd3v: You can always get it from here: https://packages.debian.org/bullseye/libnewt-dev05:46
tuxd3vhttps://gitea.devuan.org/tuxd3v/edev105:46
masonnot altogether sure why the source spec wouldn't have done theright thing.05:46
KREYREEN[cargo-make] INFO - Execute Command: "debootstrap" "--arch=amd64" "--variant=minbase" "--no-merged-usr" "--cache-dir=/home/user/.cache" "beowulf" "/home/user/Repositories/qubes-template-devuan/build"08:01
KREYREENE: Cannot install into target '/home/user/Repositories/qubes-template-devuan/build' mounted with noexec or nodev08:01
KREYREENany idea wha tis this caused by?08:01
KREYREENi can create file in that directory..08:02
* KREYREEN is lazy to read that unreadable shell script that debootstrap is~08:02
* KREYREEN is implementing a handling of devuan TemplateVM in qubesOS08:02
rrqKREYREEN wants to be told what "mounted with noexec or nodev" means?08:06
KREYREENrrq, it ain't mounted with noexec nor nodev08:07
KREYREEN /dev/xvdb on /home type ext4 (rw,nosuid,nodev,relatime,discard)08:07
rrqI believe you; maybe flash a mount output line to show the audience?08:07
KREYREEN...08:07
* rrq is too slow08:08
rrqbut doesn't it say nodev?08:08
KREYREENuser@personal:~/Repositories/qubes-template-devuan$ mount | curl -F 'f:1=<-' ix.io08:08
KREYREENhttp://ix.io/2Hgk08:08
KREYREEN...08:08
KREYREENaaaaa i can't reeeaaaad08:08
KREYREENthanku08:08
rrqnw08:08
KREYREENhttps://bin.snopyta.org/?b8631f5f6b58f4f1#9GfxJhqXNbx3BFTHjEssJs8NWLF6yNusuyzFgq4tQnFV11:21
KREYREENany idea why is this failing this way?11:21
KREYREENE: Couldn't find these debs: apt11:21
KREYREENseems like devuan's issue with debootstrap something O.o11:21
rrqperhaps a clear-text paste somewhere would be possible?11:23
gnarfaceuse paste.debian.net and i'll look at it11:23
gnarfaceusually the answer is just repo propagation11:23
gnarfacesometimes it's bad sources.list11:24
KREYREENgnarface, rrq, https://paste.debian.net/1176180/11:24
KREYREENbtw. bin.snopyta.org is using PrivateBin that decrypts the paste in the browser through 256-bit AES and server doesn't know the content11:25
KREYREENalso floss ^-^11:25
rrqyou probably want /merged rather than /devuan11:25
gnarfaceyea11:25
gnarfacei was gonna say that11:25
gnarfaceyour url is wrong for the repo11:25
KREYREENrrq, used merged in beowulf-merged11:25
KREYREENah in the URL11:25
gnarfacecan't verify the rest11:25
KREYREENE: Failed getting release file http://deb.devuan.org/devuan-merged/dists/beowulf/Release11:26
gnarfacethat's with http://deb.devuan.org/merged?11:26
KREYREENno that's with http://deb.devuan.org/devuan-merged O.o checking11:26
KREYREENah it's just merged x.x11:27
KREYREENworks now thanku <311:27
gnarfaceno problem11:27
rrqbtw that bin.snopyta.org seemed to want to install some webassembly or something; not part of my workflow11:28
KREYREENrrq, doesn't require to install anything on my end O.o but it's using webassembly to decrypt the pastes11:49
Hellehrm, isn't a setup without elogind still supposed to be supported ?13:26
HelleI am somehow getting the feeling that something in Devuan's Priorities messes up and hence you end up with debootstrap drawing in a lot of stuff it really shouldn't13:28
gnarfaceHelle: try disabling recommends13:51
gnarfaceHelle: you shouldn't need elogind unless you're using a graphical login13:51
Hellehrm, init gurus, when does sysfs get mounted "these days" ? ie, before the rcS.d/ or only at S09mountall.sh ?15:18
Hellenevermind, being very blind15:25
clorti seem to have elogind16:09
Helleclort: yes, elogind is an option, but I was seeing it pulled in on a freshly debootstrapped setup16:38
clortis it one of those things we'd rather not have?16:38
fsmithredI just did a debootstrap of chimaera and beowulf, and in both cases it pulled in libsystemd016:39
fsmithredelogind is a fork of systemd-logind, so some people don't want it16:39
fsmithredadding --include=libelogind0 causes libsystemd0 to be replaced.16:40
fsmithredHelle, did you say you were getting xorg stuff on debootstrap? I don't see that.16:43
Hellefsmithred: yeah, figured out what was going on16:44
fsmithredwhat was it?16:44
Hellefsmithred: well, mostly, I was adding 3 more packages that I did not expect to impact things, but via some ridiculous Recommends: chains that pulled in Xorg stuff16:45
fsmithredoh, yeah. I can believe it.16:45
fsmithredI don't know of a way to exclude Recommends on a debootstrap.16:45
fsmithredother than in chroot16:45
fsmithredI like to disable those early16:46
asbestodisable what?16:52
masonfsmithred: there's --exclude, but that gets overridden by (I think) recommends, which is why it's insufficient for blocking elogind.16:55
fsmithredI didn't get elogind16:55
fsmithredI included libelogind0 so it would replace libsystemd0. The latter does get pulled in, but it's gone in the final product.16:56
rktaIs the beta installer released yet?18:32
golinuxrkta: A Beowulf 3.1 point release is very close to being released.Chimera beta won't be released until bulleye goes to stable which is still a ways off and Devuan will follow sometime after that.18:40
onefangFor debootstrapping, I debootstrap the bare minimum, disable recommends, then apt install the rest.19:01
fsmithredonefang, that's what I do, too. Also, I discovered today that you can't do a debootstrap of chimaera, but using 'testing' works.19:17
fsmithredalso, I plan on making chimaera isos after beowulf point-release is done19:17
* onefang heads back to bed.19:28
* clort frolics with the power of kdenlive20:21
unixbsdhi20:46
unixbsdhow to get on devuan / gui-free X-free a way that it puts internet on wireless whatever the dongle wifi, unplug replug, bit like BSD does?20:47
rwpunixbsd, There are multiple non-graphical ways.  Both by using manager software, and by running the low level commands yourself.20:58
rwpunixbsd, Easiest might be wicd and wicd-curses which I will recommend as being very easy.20:59
asbestoI'm using directly wpa_supplicant with a config file and a shell script for dhcp and stuff. Quite simple20:59
unixbsdare they already some bash script or whatever that do put ifup/dhclient/wpa key on whatever wlanX ?=20:59
asbestoor you can use... let me search. there's a fantastic tool by a friend of mine20:59
unixbsdcan you give you script?20:59
asbestolet me search the link for ya20:59
unixbsdwould be cool20:59
rwpEasily scripting this is basically the same as using wicd-curses or conman or other manager.  (shrug)21:00
rwpFor dedicated non-mobile systems that connect to wifi I use /etc/network/interfaces in the normal way with wpa_supplicant.21:01
rwpThat's a very reliable and solid configuration.21:01
Hellewicd does require DBus and things as does iwd21:03
HelleI should package eiwd at some point (iwd, without DBus)21:03
rwpdbus?  Why oh why...21:04
Hellewpa_supplicant or the way it integrates with /etc/networks is otoh still passable21:04
Hellerwp: well, that is why eiwd exists21:04
asbestohttps://github.com/KatolaZ/setnet21:04
asbestohere it is21:04
asbestoIDK if this is the last version or what21:04
rwpI am an emacs user and Emacs for goodness sake also pulls in dbus.  But I don't use it.21:04
asbestobecause I remember the author using HIS own git repository instead of github but i can't find it21:04
asbestoanyway there is it21:04
asbestoit's just a shell script, very nice and simple21:04
asbestoand it works :)21:05
HelleBut (e)iwd replaces wpa_supplicant and has some more modern options and eiwd, while not having a commandline client, works fine with just their config files (it only used dbus for the client)21:05
Hellebut is currently not packaged and I should fix that21:05
Hellealso means probably trying to convince Debian to include it21:05
asbestoyou mean setnet?21:06
Helleno, eiwd21:06
asbestowell it's just a shell script21:06
asbestoah ok :D21:06
asbestofrom setnet.sh I just moved to having a config. file for wpasupplicant, and that is what I'm using now21:07
Hellewpa_supplicant however as mentioned doesn't support some more recent wifi developments annoyingly21:07
asbestojust because I'm lazy and too '70 UNIX oriented :)21:07
asbestoah21:07
asbestoas I read somewhere, wpa_supplicant "is a can of worms"21:07
rwpunixbsd, Are you intending it to be a mobile client?  Connecting to many different random Access Points?  Or a local non-mobile one always connecting to your local one?21:07
Helleyes, that is why people are hesitant to implement new code on it21:07
asbestoI just have a bunch of AP configured into a config file, and it works good when I'm on the move.21:08
asbestobut sometime I prefer using single files just for one access point at a time, to avoid any delay / problems21:08
asbestoanyway setnet is great :)21:09
Helleeh, if it's a wrapper around wpa_supplicant it's a bit whatever, there is integration for that in /etc/network scripts21:10
antonio_nmcli is the answer21:13
rwpanswer to the question of how to cause yourself pain and suffering?21:13
Hellenmcli is networkmanager, no thanks21:14
HelleDependencies all over and bad design for anything but simple use cases21:15
rwpMultiple times now across several years and multiple releases I have had Network Manager down the main network connection during upgrades and then forget to bring the network back up again.21:15
Hellewicd is already a lot lighter, and iwd seems to be doing even better, but I've yet to try it in production21:16
rwpToo many drives to the datacenter to get on the console.21:16
rwpFool me once.  Shame on me.  Fool me twice.  Shame on you.21:16
* rwp laughs21:16
Hellebut eiwd would allow for a dbus-less install for things like embedded systems21:16
Helleso replacing wpa_supplicant with something, well modenr21:16
Hellemodern21:16
unixbsdI am back... ok let s go to https://github.com/KatolaZ/setnet21:19
unixbsdit lookks prettty nice this setnet21:21
unixbsdit does use just shell21:21
unixbsdI wonder if it would work with those enxXXXXX stuffs21:21
unixbsdI dont know yet, I will test it first.21:22
unixbsdhopefully hte network layer can be better with that thing setnet21:22
antonio_asbesto has a script called petacci.sh which does the same thing21:23
asbestoantonio you're a fucking troll <321:27
antonio_asbesto: is that true or not?21:28
Helleanyone know what the Devuan plans for network device naming rules are ? just sticking with persistent-net rules as Debian has had since Lenny ?21:39
Hellefrom what I can find online (but it's not authorative), yes and old style (eth0/wlan0 type) naming scheme21:42
Hellegood to know, that makes my life easier writing this21:44
unixbsdman, that debootstrap is completely broken on devuan. I have no idea what is going on with your server.21:49
unixbsdI works very fine on debian.21:49
golinuxHella Devuan uses eth0 etc21:56
n4dirfailed to fetch deb.devuan.org in sources.list yesterday. Same today. Error message says: "unable to fetch mirrors.dotsrc.org"21:58
n4diranything i can do about it besides waiting?21:58
fsmithredunixbsd, I did three debootstrap installs today. Two chimaera and one beowulf. What problem are you having?21:59
unixbsdthey need l�ikely to host the devuan packages and mirror on a debian server21:59
fsmithredwhat do you mean?22:00
unixbsddebian has lot of money to get those machines running ...22:00
HelleI've just ran a bunch of Debootstraps with both Beowulf and Chimaera and uuuuh, it's my code that is leaving the result broken, not Devuan :P22:00
fsmithredyou don't need to worry about debian servers22:00
fsmithredwhat url did you use?22:00
unixbsdthe debian servers are heavy ones, excellent machines to serve packages.22:00
Helle...22:00
unixbsddevuan is smaller community in comparison.22:00
fsmithredwas deb.devuan.org too slow for you?22:00
Helledon't mix sources and then complain22:01
unixbsdslow and broken while fetzching missing22:01
fsmithredI don't think you can mix sources on a debootstrap22:01
fsmithredpick a mirror near you from the list22:01
fsmithredor one you know is fast22:01
n4dirwhats even more weird that i can ping both URL's. hmmm ...22:01
fsmithredif packages are missing, the mirror is probably in the middle of an update.22:02
n4dirso just wait?22:02
fsmithredcheck the mirror checker22:02
unixbsdI can just tell you that 2-3 last years ago it worked fine.22:02
unixbsdsince lockdown and first covid, your devuan servers de.devuan.org is failing regularly.22:02
fsmithreduh, except whenever it stopped updating22:03
unixbsdI tried abroad from different countries.22:03
fsmithreddon't use country codes, just pick a server22:03
fsmithredthe country codes just send you to the same round-robin as deb.devuan.org22:03
fsmithredcrap-shoot22:03
fsmithredapt-panopticon (the mirror checker): https://borta.devuan.dev/apt-panopticon/results/Report-web.html22:04
fsmithredsame thing, northern hemisphere: https://sledjhamr.org/apt-panopticon/results/Report-web.html22:05
n4dirso i stick to deb.devuan.org and wait?22:06
fsmithredn4dir, or pick a different mirror22:07
n4dirno need to get excited with old horses. i will wait22:07
n4dirthanks for the info and the link22:07
fsmithredthat works22:07
unixbsdwhy not asking debian to host or to mirror your packages? there is no shame at all, at least we would be sure that it would work fine.22:11
fsmithredamprolla is getting some testing and patching lately to address the problems22:11
fsmithredyou think they would cooperate?22:11
unixbsdwhy not?22:12
unixbsddebian aim no harm22:12
fsmithredwe can't get them to keep init scripts in packages22:12
unixbsdwait... I may try to ask.22:13
unixbsdman, seems 5 bucks per months22:18
aitor_oops, you were talking about network managers and dbus mitigation...23:03
aitor_i've just sent an email to DNG23:03
tuxd3vin chimaera a package version is updated frequently, or once in chimaera that package will for sure be in next stable?23:59

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