onefang | Even easier is to just accept usrmerge and let it's script do it's thing. | 00:00 |
---|---|---|
rwp | Symlinks have been confusing people since they were invented. Because the "special file" holds a value, which might be anything, and it is only referencing a file by symbolic name. | 00:00 |
onefang | Then the only work you end up doing is arguing about it. lol | 00:00 |
gnu_srs1 | onefang: I'd prefer not to use usrmerge :( | 00:00 |
rrq | yes, I agree, if you start linking you'd just set those sstupid-links and then won't have the ongoing hassle | 00:00 |
gnu_srs1 | Ok, let's start forging packages then :) | 00:01 |
gnu_srs1 | forking** | 00:02 |
onefang | Ah you prefer doing it wrong and doing it the hard way. Gotcha. | 00:03 |
rwp | It's okay onefang if people want to do things differently. | 00:03 |
gnu_srs1 | Anybody knows how many Debian packages are needed to fork? How to find them? | 00:04 |
rrq | or you can just wait until all file references have been corrected to address the actual installation place... like when the kernel guys realize that they need to load firmware at a new place and such... | 00:04 |
rwp | And as I have said before onefang you can change Devuan to usrmerge anytime you want. | 00:04 |
onefang | Yep, my OS build script now installs the usrmerge package with the initial mmdebstrap step. I do it right, I have no problems. Most of the people complaining about it are trying to avoid it. | 00:06 |
gnu_srs1 | rrq: What do you mean by "file references", files moved from /bin to /usr/bin etc. That is the current Debian stupidity! | 00:06 |
onefang | I'm OK with people doing things differently, it's the constant bitching about it that annoys me. | 00:07 |
rrq | gnu_srs1: the full pathname for a file is only a concern at places where full pathname refer to it. | 00:08 |
rwp | To be honest it is the replication of the bitching about it that they should just accept it bothers me. | 00:08 |
rrq | all current problems arise due to many places using full pathname references to some place where the file no longer is | 00:08 |
onefang | Which the symlinks solve. End of. | 00:09 |
rrq | a change like stupid-links should have been executed by changing all those references first, and then moved the installation place | 00:09 |
gnu_srs1 | Lets start using PATH as much as possible then?? | 00:10 |
rrq | that would have been ideal | 00:10 |
rrq | even kernel firware loading could have had an environment variable for the fireware root directory | 00:12 |
rrq | (though it already took some years to get firmware collated into a dedicated root path) | 00:14 |
rrq | (since programming for a device probably takes a mind set or skill different from such system-wide concerns) | 00:15 |
rwp | It's a basic concept disconnect that the kernel thinks it owns / and so thinks fw files in /lib/ is correct and the userland thinks having everything in subdirectory is nice so thinks files in /usr/lib/ is correct. The two concepts never were in conflict before but now forcing them into one or the other requires symlinks. | 00:28 |
rwp | Really if Ken Thompson was truly a god he would have known in the 1970s that 50 years in the future people would be doing this UsrMerge thing and he would have put the OS in /os instead of in / initially and avoided all of this kerfuffle. :-) | 00:30 |
onefang | So we need to replace /usr merge with /os merge now? B-) | 00:34 |
onefang | That's an improvement, one less letter to type. | 00:35 |
rwp | If the OS is required by the forces-that-be to be in a subdirectory then /os would make more sense to me than /usr does. | 00:36 |
rrq | the saddest part to me is that the Debian package maintainers nowadays seem to see themselves as product people rather than librarians, and now Debian is understood as that Gnome Desktop with a swirl rather than a (vast) library as a union of collections of (FOSS) software. | 00:51 |
brocashelm | libraries are useful to ensure software works without installing the whole thing (as other distros like slackware and arch tend to do) | 00:56 |
rrq | btw: given that libc6 Pre-Depends on debconf and debconf Pre-Depends on libc6, how can those be installed? | 01:27 |
onefang | Use mmdebstrap or debootstrap, that'll install them. | 01:29 |
rrq | and according to https://www.debian.org/doc/debian-policy/ch-relationships.html, "Pre-Depends does not permit circular dependencies to be broken." | 01:29 |
rrq | "If a circular dependency is encountered while attempting to honor Pre-Depends, the installation will be aborted." | 01:30 |
onefang | Those two things are kinda important to the system, so would get installed at the beginning, before there's a full system to screw up. Something needs to be installed first, then second, ... then eventually you have a working system. | 01:33 |
onefang | But it's bound to be that a bunch of those early needed packages might depend on each other. Even in circles. | 01:33 |
onefang | So I'm not surprised things are broken at early system build until enough of it is built. | 01:34 |
onefang | But might also be better off asking Debian about that. | 01:35 |
rrq | are you saying that Pre-Depends means something else for some packages? | 01:36 |
onefang | I'm saying that the very first few packages that need to be installed might have screwed up dependencies coz they are the FIRST needed to be installed. And it's up to Debian to reconcile that with their policy. | 01:37 |
onefang | Living next door to a major construction site, they start at the bottom and build upwards, but the bottom is just dirt and gravel. Can't build the foundations until the dirt and gravel is in the proper place. | 01:38 |
onefang | Though some of that dirt is ending up on my floor. Gotta make a mess when you start building things. High rise apartment blocks or operating systems. | 01:40 |
onefang | The amount of juggling of that dirt and gravel is surprising, but I guess that sort of juggling is needed. Debian is juggling things to get those first base packages installed. After that is done, then they are in a better place to follow the rules, or notice they broke their own rules in getting things up and running. | 01:42 |
al1r4d | wait, what "https://dev1galaxy.org/viewtopic.php?pid=47027#p47027"????? blocked for security reason???? | 01:45 |
rwp | gnu_srs1, I forgot but newer coreutils added an ln -r,--relative option to have the command convert full paths to relative paths. I should have remembered it earlier. That makes generating them much easier for everyone. | 01:45 |
onefang | al1r4d: Maybe someone accidentally typed a password into a forum post? | 01:46 |
rrq | it's not that the forum post is blocked; it's rather that that OP's ISP intercepts DNS requests with bogus responses because they don't like the OP attempting to resolve some certain QFDN | 01:59 |
rrq | FQDN | 01:59 |
onefang | Or like one of my web pages is now blocked by Google, and thus Firefox, for totally bogus reasons? | 02:01 |
rrq | this one is not about you, though | 02:01 |
onefang | I'm just saying there's more than one way to get this problem. | 02:02 |
rrq | indeed, Internet is getting crowded by middlemen | 02:03 |
rrq | (or is it "with" rather than "by"?) | 02:03 |
onefang | Why not both? | 02:04 |
rrq | inclusiveness FTW :) | 02:04 |
onefang | https://dev1galaxy.org/viewtopic.php?pid=47027#p47027 works for me, so yeah, probably a network issue closer to al1r4d | 02:05 |
rrq | I'm technically unhappy with Pre-Depends in some cases being used wrongly so that installalling requires an outside logic that patches around that breakage. | 02:07 |
rrq | and especially as it concerns the most fundamental building pieces. | 02:08 |
onefang | Fair enough. The ones you mentioned are outside the logic anyway, coz they are part of the logic which hasn't been installed yet. | 02:09 |
onefang | As for that forum thread, the various apt-panopticons can connect to the repo.jing.rocks package mirror fine. | 02:10 |
rrq | well the installing system is fine and dandy, but the system installed into runs into that quirk, and the installing system must thus "know fiddly details" about that | 02:11 |
onefang | Ah the quirks get in the way of upgrades of one of those packages? | 02:11 |
onefang | But if this is Debian doing something wrong they can easily fix, tell Debian. I don't think we fork either package. | 02:13 |
rrq | my interest is more academic than that; I'm not in a grief, but rather reviewing the bootstrap | 02:13 |
rrq | guess I had mistakenly hoped that the most fundamental building blocks had had more technical rigidity | 02:17 |
rrq | probably comes back to the best operative rule: ignore me :) | 02:18 |
onefang | lol | 02:19 |
onefang | So for all my tests repo.jing.rocks rocks. | 02:21 |
onefang | But I only just stopped holidaying and weekending today. | 02:21 |
golinux | OT . . . How can you holiday and weekend? Thought you've been in bed for weeks? | 02:22 |
* golinux is confused as to onefang's state | 02:23 | |
onefang | I was bed ridden for weeks, at the beginning of December. Then holiday season kicked in just as I was recovering. Then I did my usual "Monday is still the weekend somewhere" thing yesterday. Later this month is my birthday, so think I'll squeeze in some work now. | 02:24 |
onefang | Work includes looking at that forum post and poking at that mirror to see how well it is right now. | 02:25 |
golinux | Happy hear you're doing better . . . | 02:26 |
onefang | Thanks. | 02:26 |
onefang | Damn, hyperreal was next on my mirror work list. lol | 02:28 |
onefang | Welcome back hyperreal You are next on my mirror work list. B-) | 02:32 |
e54 | I installed Devuan, now i have STD's , What now | 02:36 |
e54 | Oh think golinux gave em to me :( | 02:41 |
e54 | she not even cute | 02:41 |
Xenguy | WTF are you talking about? | 02:41 |
onefang | golinux gave you Several Tidy Devuan's? | 02:43 |
Xenguy | .oO( Read The Fine Manual ...) | 02:43 |
e54 | onefang: Yes | 02:43 |
brocashelm | devuan cures systemSTDs | 02:44 |
onefang | lol | 02:44 |
e54 | Doe's not make sense,but yes | 02:44 |
e54 | How is the gang tonight? | 02:45 |
Xenguy | OT AF | 02:45 |
rustyaxe | systemdos | 02:45 |
rustyaxe | can someone remind me how to prevent a package from ever being installed by apt? | 02:46 |
e54 | Xenguy: I dont need no stinking manual :) | 02:46 |
n4dir | rustyaxe: i think it is done with pinning, and i think it is described in the section: Prevent/Selective insallation ... | 02:51 |
n4dir | https://wiki.debian.org/AptConfiguration | 02:51 |
rwp | rustyaxe, You can add a pinning file in the /etc/apt/preferences.d/ directory that blocks it. | 02:51 |
rwp | What n4dir said. | 02:51 |
rustyaxe | Prevent/selective installation from a third-party repository? | 02:52 |
n4dir | mhh. wait a second. I think there is just preference. You will need to figure out what pinning priority is needed to not install it. | 02:53 |
rustyaxe | not seeing a way to just outright block a package? | 02:53 |
n4dir | i think it is something - (minus), no? | 02:53 |
rustyaxe | theres a way | 02:53 |
rustyaxe | I've done it before its just been long time | 02:53 |
brocashelm | through apt pinning mainly | 02:53 |
brocashelm | set priority in the negative range | 02:53 |
n4dir | i was sure it was on this devuan in /etc/apt/... but perhaps it was on refracta. But if brocashelm knows, then all is good | 02:54 |
brocashelm | Pin-Priority: -1 | 02:54 |
n4dir | brocashelm: if you use -2, it won't install it two times in a row, right? :-) | 02:54 |
brocashelm | it goes in /etc/apt/preferences.d | 02:54 |
brocashelm | n4dir: lol | 02:54 |
rwp | In earlier days before Devuan we pinned to block systemd. Here is an example: https://unix.stackexchange.com/questions/316744/systemd-apt-pinned-to-1-and-installed-in-upgrade-from-debian-8-to-debian-9 | 02:55 |
e54 | Don't use -3. It will install windows 7 | 02:55 |
rustyaxe | n4dir: rwp: thanks | 02:56 |
brocashelm | i'm still paranoid, so systemd is pinned nonetheless | 02:57 |
* rustyaxe adds brltty orca and gpm | 02:57 | |
brocashelm | in the wildcard range | 02:57 |
n4dir | i for one would still always have a close look. As i pretty much never do pinning, i don't really trust myself. | 02:57 |
rustyaxe | brltty caused me a 3 hour drive recently | 02:57 |
rwp | It won't hurt anything to be safe. | 02:57 |
brocashelm | things like libsystemd0, not installed at all | 02:57 |
brocashelm | elogind is the bandage for now | 02:57 |
rustyaxe | Damn thing blocked the serial interface to the card readers that unlock the doors :O | 02:58 |
n4dir | looks like "apt-mark hold packagename" would also prevent it from being installed. If the manpage is correct. probably is | 02:58 |
rustyaxe | eh a file with them is perfect, i can drop it into my "template" that has the crap i need to move in to a machine | 02:58 |
rustyaxe | my bashrc etc | 02:58 |
e54 | are you guys on bullseye/bookworm/trixie now? | 03:00 |
rwp | I never considered "holding" a package to prevent it from being installed. But the man page does support that idea. I have only used "hold" to keep installed packages from upgrading or removing. | 03:01 |
rwp | e54, Yes. | 03:01 |
n4dir | same here. i was pretty astonished | 03:01 |
e54 | rwp great answer. | 03:02 |
rwp | It makes sense since it prevents a new package from upgrading an existing. A new package. Which is basically just the same as if there is nothing installed then don't install it. | 03:02 |
rwp | e54, This is the Devuan support channel so of course we are talking about Devuan with 4 Chimaera (Bullseye) being OldStable, 5 Daedalus (Bookworm) being Stable, 6 Excalibur (Trixie) being Testing, and Ceres (Sid) Unstable. | 03:04 |
e54 | rwp ok | 03:05 |
rwp | Devuan is an overlay of all package that are not explicitly banned, and that is only systemd, with anything that hard depends upon systemd being banned by implication and most of those having been forked to allow installation without systemd. | 03:05 |
rwp | All of the rest of the tens of thousands of packages are installed from Debian verbatim. It is Debian. | 03:06 |
e54 | That is understood | 03:06 |
Xenguy | .oO( Debian the way it might have been ... ) | 03:07 |
rwp | The Universal Operating System! | 03:07 |
e54 | Not entirely true, but ok | 03:07 |
Xenguy | Universal, will always love that | 03:07 |
e54 | I have seen packages here before debian | 03:08 |
rwp | That's Devuan now. Debian has unfortunately decided to actively shed that vision. But this is getting OT. Happy to continue it in #devuan-offtopic though. | 03:08 |
n4dir | i wouldn't be too happy, but am on my way to bed anyway. | 03:09 |
rwp | Don't go to bed angry n4dir. Stay awake and plot revenge! :-) Good night! | 03:10 |
n4dir | ha. | 03:10 |
Xenguy | .oO( revenge of the dispossessed ... ) | 03:11 |
e54 | Why? | 03:15 |
e54 | What is the topic? | 03:16 |
e54 | Dead air? | 03:16 |
e54 | How about alpine linux? | 03:20 |
e54 | Thaey cool | 03:20 |
rwp | Let's please take this to #devuan-offtopic where OT is On Topic there. :-) | 03:21 |
joerg | thanks | 03:34 |
e54 | Is Devuan an OS copy of people that know what there doing? | 04:53 |
mason | It's a copy of Debian, not of people. | 04:54 |
e54 | So they know what they are doing? You dont? | 04:55 |
mason | Do you have a support question? | 04:55 |
e54 | Yes, Devuan is like a project of retartds. excse my engkish | 04:57 |
mason | e54: That's not a question. Do you have a support question? | 04:58 |
e54 | why is Devuan a thing? Just fuckin up the ecosystem? | 05:00 |
mason | That's close enough to a question. Debian wants users to have a choice of init systems. | 05:00 |
mason | Devuan* I mean. | 05:03 |
e54 | so debian has choice? | 05:04 |
mason | Yes, but not as easily or as straightforward as with Devuan. | 05:04 |
e54 | so devuan is like debian and not mainstream | 05:06 |
e54 | like debian that sucks more | 05:06 |
mason | Do you really think so? | 05:07 |
joerg | please take it #devuan-offtopic! | 05:08 |
mason | That's a good idea. | 05:08 |
joerg | last call | 05:08 |
e54 | Lol, this channel is sooo backed up | 05:10 |
e54 | joerg: let me take this to off topic | 05:11 |
e54 | joerg: this channel is dead | 05:12 |
fluffywolf | lol, I was just thinking that, but I can't do that here. | 05:14 |
Xenguy | It's a thing | 05:15 |
movedMacBookToDv | hello there ya'll | 09:28 |
movedMacBookToDv | its been a month since i moved to dv1 but i cant seem to figure out how to resolve an ethernet issue | 09:29 |
movedMacBookToDv | the issue seems to happen everytime i make the machine sleep or hibernate - afterwhich the ethernet doesnt work and i have to reboot | 09:30 |
brocashelm | what is the issue with the ethernet? can't connect with eth0? | 09:30 |
movedMacBookToDv | it seems related to this - https://wiki.debian.org/InstallingDebianOn/Apple/MacBookPro/Early-2015-13-inch#Immediate_wakeup_after_sleep | 09:30 |
movedMacBookToDv | the machine doesnt even identify eth0 exists or the device exists | 09:30 |
brocashelm | i see, it's on a mac | 09:31 |
movedMacBookToDv | probably because ethernet connects to this machine via a thunderbolt adapter | 09:31 |
movedMacBookToDv | if i remove the adapter and then put it back in other thunderbolt port - then also the OS fails to recognize it a new device | 09:31 |
brocashelm | i might be mistaken, but i think that script should still work with sysvinit (my /lib/systemd/system-sleep has a couple of ootb scripts from packages for support, despite no systemd pid running) | 09:33 |
movedMacBookToDv | true... brocashelm .. i use the same machine... as on the debian wiki for the issue with wifi - i have it with ethernet.. i tried rmmod thunderbolt and then doing modprobe thunderbolt .. still nothing works excepts reboot | 09:33 |
brocashelm | hmmm, and it looks like it was last revised in 2017? so, are you using the latest daedalus with 6.1 kernel? | 09:34 |
movedMacBookToDv | actually the issue is not immediate wake up after sleep but the next issue wifi after sleep | 09:34 |
movedMacBookToDv | yes.. daedalus and 6.1 kernel | 09:34 |
movedMacBookToDv | Linux dvn 6.1.0-17-amd64 #1 SMP PREEMPT_DYNAMIC Debian 6.1.69-1 (2023-12-30) x86_64 GNU/Linux | 09:35 |
movedMacBookToDv | to be exact | 09:35 |
brocashelm | is your issue exactly like this? https://forums.debian.net/viewtopic.php?t=153697 | 09:36 |
brocashelm | well, doesn't specify if it's a mac, but the error messages? | 09:36 |
brocashelm | regarding the wifi not working on wake up | 09:37 |
movedMacBookToDv | similar except for wifi .. this happens on ethernet which works with thunderbolt adapter - a physical adapter | 09:37 |
movedMacBookToDv | i have no issue with wifi | 09:37 |
brocashelm | oh, ok | 09:37 |
brocashelm | had them mixed up, then | 09:37 |
movedMacBookToDv | https://wiki.debian.org/InstallingDebianOn/Apple/MacBookPro/Early-2015-13-inch#WiFi_won.27t_work_after_hibernation - this is the right link | 09:38 |
movedMacBookToDv | next subsection on same page | 09:38 |
movedMacBookToDv | sorry.. for wrong link | 09:38 |
movedMacBookToDv | this is where the modprobe playing did not work when i was troubleshooting | 09:38 |
brocashelm | hate to have to link to archwiki, but this might also provide some clues: https://wiki.archlinux.org/title/Laptop/Apple | 09:39 |
brocashelm | what exact model is your macbook? | 09:39 |
brocashelm | mid-2010s or so? | 09:40 |
movedMacBookToDv | the same model that i linked in debian wiki | 09:40 |
movedMacBookToDv | exactly same machine | 09:40 |
brocashelm | apple macbookpro from early 2015? | 09:41 |
brocashelm | hmm | 09:41 |
movedMacBookToDv | true | 09:41 |
movedMacBookToDv | macbookpro12,1 to be exact model name from 2015 | 09:41 |
movedMacBookToDv | a1502 is the model | 09:41 |
movedMacBookToDv | from your arch wiki link - https://wiki.archlinux.org/title/Power_management#USB_autosuspend - i think this could be the issue that i am facing | 09:43 |
brocashelm | i wonder if there's anything from that archwiki that could point you in the right direction? | 09:43 |
movedMacBookToDv | i feel so the same | 09:43 |
brocashelm | yeah | 09:44 |
brocashelm | some clues might be in there | 09:44 |
brocashelm | wouldn't hurt to try | 09:44 |
brocashelm | sorry, i'm not much help here; never really owned a macbook | 09:44 |
movedMacBookToDv | i can understand | 09:46 |
movedMacBookToDv | it is this adapter- | 09:46 |
movedMacBookToDv | https://4get.ca/proxy?i=https%3A%2F%2Fvmart.pk%2Fwp-content%2Fuploads%2F2019%2F04%2Fproducts-Apple_Thunderbolt_to_Gigabit_Ethernet_Adapter_main.jpg - | 09:46 |
brocashelm | checking my grub config, i see i've always had usbcore.autosuspend=-1 as part of GRUB_CMDLINE_LINUX_DEFAULT | 09:47 |
GoatAvenger | movedMacBookToDv, what kind of ports are available on that device | 09:47 |
movedMacBookToDv | or this - https://arstechnica.com/gadgets/2012/06/hands-on-apples-thunderbolt-gigabit-ethernet-adapter/ | 09:47 |
brocashelm | usbcore.autosuspend=-1 disabled autosuspend for all usb devices | 09:47 |
brocashelm | *disables | 09:47 |
movedMacBookToDv | 3.5mm jack, 2 thunderbolt port, and 2 USB 3, 1 HDMI , 1 SD card and of course magsafe charging port | 09:48 |
brocashelm | ok, that makes sense. i think that line i just mentioned for your grub could do the trick | 09:48 |
brocashelm | it will be told to the kernel at boot time, so no need to have to manually do it afterwards or reboot to get it working again | 09:48 |
movedMacBookToDv | but where do i put that line for grub.. | 09:49 |
movedMacBookToDv | probably on a file on /boot/efi ? | 09:49 |
brocashelm | your grub file is in /etc/default | 09:49 |
brocashelm | look for the file grub | 09:49 |
movedMacBookToDv | oh | 09:49 |
brocashelm | look for GRUB_CMDLINE_LINUX_DEFAULT="" | 09:49 |
brocashelm | usually has quiet or splash in there, somewhere | 09:49 |
GoatAvenger | /etc/default/grub ^ | 09:50 |
brocashelm | try that, and then run sudo update-grub | 09:50 |
brocashelm | reboot | 09:50 |
brocashelm | then hibernate and retest | 09:50 |
movedMacBookToDv | it has quiet for me | 09:51 |
brocashelm | you could add it alongside or replace it | 09:51 |
brocashelm | example of what i use: GRUB_CMDLINE_LINUX_DEFAULT="intel_pstate=disable usbcore.autosuspend=-1" | 09:52 |
brocashelm | so just add a space and include the usbcore.autosuspend=-1 next to it | 09:52 |
movedMacBookToDv | i c | 09:53 |
GoatAvenger | could also check out h-node.org for usb-to-ethernet hardware | 09:53 |
GoatAvenger | the DUB-E100 is cheap and available on ebay | 09:54 |
GoatAvenger | better driver support, perhaps | 09:54 |
movedMacBookToDv | i have the driver supported already .. but just that it doesnt load after sleep/hibernate | 09:55 |
movedMacBookToDv | i am currently using it to connect and chat.. | 09:55 |
movedMacBookToDv | added - GRUB_CMDLINE_LINUX_DEFAULT="quiet intel_pstate=disable usbcore.autosuspend=-1" | 09:57 |
movedMacBookToDv | ran update-grub | 09:57 |
GoatAvenger | I found the following as well --> https://www.kernel.org/doc/html/latest/admin-guide/thunderbolt.html | 09:58 |
GoatAvenger | if you scroll to the bottom to the section (Forcing power) | 09:58 |
GoatAvenger | seems you could check for the presence of "/sys/bus/wmi/devices/86CCFD48-205E-4A77-9C48-2021CBEDE341/force_power" | 09:59 |
GoatAvenger | and if it's there, try (echo 1 > /sys/bus/wmi/devices/86CCFD48-205E-4A77-9C48-2021CBEDE341/force_power) after a resume from sleep where the adapter is unresponsive | 09:59 |
movedMacBookToDv | i see.. | 09:59 |
GoatAvenger | just a guess, though | 09:59 |
movedMacBookToDv | if grub update thing doesnt work.. i'll trying rmmod thunderbolt-net and adding it back again | 10:00 |
movedMacBookToDv | let me sleep for a bit and get back | 10:00 |
GoatAvenger | good luck.. | 10:00 |
movedMacBookToDv | or actually i can remove the thunderbolt adapter and then reconnect | 10:00 |
movedMacBookToDv | if it doesnt work.. i need to reboot.. | 10:01 |
movedMacBookToDv | either way | 10:01 |
movedMacBookToDv | brb | 10:01 |
movedMacBookPro | sadness.. grub update or even rmmod thunderbolt-net followed by modprobe thunderbolt-net didnt work | 10:10 |
brocashelm | damn :( | 10:10 |
brocashelm | maybe then look into that DUB-E100 goatavenger linked you to? | 10:11 |
movedMacBookPro | true.. | 10:12 |
movedMacBookPro | i cant keep rebooting every day morning... | 10:12 |
GoatAvenger | and if it's there, try (echo 1 > /sys/bus/wmi/devices/86CCFD48-205E-4A77-9C48-2021CBEDE341/force_power) after a resume from sleep where the adapter is unresponsive | 10:12 |
GoatAvenger | as root, can try that too | 10:12 |
GoatAvenger | if no-go, then it's a kernel/driver bug, most likely | 10:12 |
movedMacBookPro | this directory is empty - /sys/bus/wmi/devices/ | 10:13 |
GoatAvenger | ok, no-go there.. then | 10:13 |
brocashelm | btw, the line i pasted earlier: the intel_pstate=disable was specific to my preference for acpi for power management, and because i use older intel processors | 10:13 |
brocashelm | you just needed the usbcore.autosuspend=-1 to try it | 10:14 |
brocashelm | but your macbook is from 2015, so might be fine | 10:14 |
movedMacBookPro | oh .. i c | 10:15 |
brocashelm | it starts with sandy bridge and newer | 10:15 |
brocashelm | i prefer using acpi_cpufreq driver instead | 10:15 |
movedMacBookPro | wow.. its been a long time on ubuntu .. and i just relealized that ubuntu pastebin needs a login/signup .. sigh | 10:18 |
movedMacBookPro | i think sandybridge is 3rd gen.. and this is 5th gen cpu | 10:19 |
movedMacBookPro | not sure | 10:19 |
movedMacBookPro | any other pastebin alternatives | 10:19 |
movedMacBookPro1 | for wifi - removing brcmfmac and then re-injecting it as per the debian wiki works .. but am not sure why doing the same for thunderbolt or ethernet doesnt work - https://bin.mha.fi/?862da35601612048#4VcoQvTykVhJuSuszF4uxyUreEinaACPsTtFogd43G8p | 10:23 |
movedMacBookPro1 | i think i was mentioned for pastebin .. but then i had to reconnect.. i received a notification that i was mentioned .. but the not the actual chat message lol | 10:23 |
movedMacBookPro | lol rmmod and then modprobe bcm5974 - the ethernet driver neither disconnects the connection nor reconnects it.. | 10:28 |
movedMacBookPro | i feel am terrible at this debugging thing | 10:28 |
rrq | does it help with: rmmod -f | 10:29 |
brocashelm | you could use dpaste.org or termbin for pastebin alternatives | 10:33 |
movedMacBookPro | i thought so.. so its tg3 and not bcm594 from https://bin.mha.fi/?862da35601612048#4VcoQvTykVhJuSuszF4uxyUreEinaACPsTtFogd43G8p | 10:36 |
movedMacBookPro | removing rmmod and then modprobe tg3 stops n/w connection and then reconnects as expected.. | 10:37 |
rrq | so adding tg3 to /usr/lib/pm-utils/defaults (SLEEP_MODULES) might work | 10:39 |
movedMacBookPro | but that doesnt work after physical removal adapter followed by its reconnection or after the sleep/hibernate when left connected - so driver removal and | 10:39 |
movedMacBookPro | let me try that pm-utils thing.. later i'll remove that grub thing.. as it doesnt work | 10:40 |
rrq | (SUSPEND_MODULES actually) | 10:40 |
movedMacBookPro | i just have shell script - https://bin.mha.fi/?fbd915544fbcc046#HDQuyMbPko1radSNTq17TJpTbFLoiFX25UdN7cL4i6Sc | 10:43 |
movedMacBookPro | so its sleep.d directory ... i have a hunch that this might work.. so do i create script like this https://wiki.debian.org/InstallingDebianOn/Apple/MacBookPro/Early-2015-13-inch#WiFi_won.27t_work_after_hibernation in /usr/lib/pm-utils/sleep.d - rrq ?? | 10:45 |
rrq | so just making it SUSPEND_MODULES="tg3" doesn't work ? (and uncommenting the line) | 10:47 |
movedMacBookPro | wow termbin seems nice than dpaste or bin.mha - thanks brocashelm for this suggestion and previous debugging hints | 10:48 |
brocashelm | cheers | 10:49 |
brocashelm | how's the progress? | 10:49 |
movedMacBookPro | but where do i add SUSPEND_MODULES=tg3 ? | 10:49 |
rrq | edit /usr/lib/pm-utils/defaults .. findel the SUSPEND_MODULES= line and change it | 10:49 |
movedMacBookPro | actually.. i dont wanna suspend tg3 .. am totally fine leaving the thunderbolt adapter / ethernet adapter powered up when the machine is sleeping | 10:49 |
rrq | no that will unload the module and reload it | 10:50 |
movedMacBookPro | sadness.. no such file as pm-utils/defaults | 10:50 |
rrq | at suspend and hibernation | 10:50 |
rrq | ok. that idea relies on you using pm-utils | 10:51 |
movedMacBookPro | root@dvn:/usr/lib/pm-utils# tree | 10:51 |
movedMacBookPro | . | 10:51 |
movedMacBookPro | └── sleep.d | 10:51 |
movedMacBookPro | └── 60_wpa_supplicant | 10:51 |
movedMacBookPro | 2 directories, 1 file | 10:51 |
rrq | apprently you use something else | 10:52 |
movedMacBookPro | oh.. | 10:53 |
movedMacBookPro | brocashelm - the progress is like... i might have to sleep it off.. and see if debugging this tomorrow helps | 10:53 |
Nietz | good to hear from you again @movedMacBookPro | 16:37 |
Nietz | Do you have any Ethernet adapters that would rely on external power? I'm thinking trying an Ethernet adapter through an external expansion hub or docking station might be helpful. Then we could ensure a stable power supply to the device while we attempt to reproduce it. | 16:39 |
Nietz | check out something like this: Belkin Thunderbolt 3 Dock Core With Thunderbolt 3 Cable - Usb C Hub - 7-In-1 Docking Station For Macs & Windows, 60W Upstream Charging, With Gigabit Ethernet, Displayport & Audio Ports https://a.co/d/idmYIqz | 16:53 |
Nietz | I usually try to isolate hardware issues first. | 16:53 |
joerg | duh, I wrote my own lstree command, wasn't aware there's a tree(1) already - *me scratches head* | 19:05 |
rwp | What's wrong with "find -ls"? | 19:12 |
rwp | Or simply "find" if one just wants the names. | 19:13 |
unixman_home | Huh ... https://termbin.com/hte3 | 19:34 |
rwp | https://manpages.debian.org/bookworm/tree/tree.1.en.html | 20:03 |
rwp | unixman_home, You will need to install the "tree" package if you want to have it installed locally. | 20:04 |
joerg | unixman_home: that's strange. anyway https://linux.die.net/man/1/tree | 20:05 |
unixman_home | rwp, I figured out it was not installed by default after I looked for it. Thanks. :) | 20:07 |
joerg | rwp: my (homegrown) lstree does a recursive ls -l | 20:07 |
rwp | The tree command originated on the MS-DOS side of the family tree. Then people implemented a version for Linux too. | 20:07 |
rwp | joerg, I have several commands that I use that I wrote for myself that are re-implementations of other commands and I still use mine because I just like them better. It's okay that you have your own lstree that you wrote. That's cool. And you can do more fun things with it and similar to it because you wrote it. | 20:08 |
joerg | https://termbin.com/zosx | 20:09 |
rwp | Look how compact and concise that is! Recursive shell scripting. FTW! :-) | 20:11 |
rwp | joerg, It took me a moment to realize that lstree is actually different from the tree command as ls lists all of the tree components of the pathname given. Whereas the tree command is more like the find command in that it is browsing down directory tree paths. | 20:20 |
rwp | I am still impressed with your use of recursion in the lstree command. That's cool. | 20:21 |
joerg | :-)# | 20:58 |
msiism | I've set up a Devuan workstation at work today and would like this machine not to show the beautified version of the GRUB menu. However, I couldn't seem to find a way to make that happen. | 22:54 |
mason | msiism: Look at /boot/default/grub - you can set lots of options there. | 22:55 |
msiism | I've had a look at /etc/default/grub. But that didn't give me any clues. There wa sno GRUB_THEME line I could comment out. | 22:55 |
mason | msiism: Do you know what precisely you want to turn off? | 22:55 |
msiism | I did swtich GRUB to console mode at one point, which did the job, sort of. | 22:56 |
msiism | I'd like to see the "traditional" boot menu, without the background. | 22:56 |
msiism | Without the Devuan-themed background, that is. | 22:57 |
golinux | It is part of the default clearlooks-phenix sapphire theme | 22:57 |
mason | msiism: Do you have "splash" in GRUB_CMDLINE_LINUX_DEFAULT? If so, removing that and running update-grub might do it. | 22:58 |
mason | As in, remove "splash", rather than the whole line. | 22:58 |
golinux | which is found here: /usr/share/desktop-base/ | 22:59 |
msiism | Yeah, okay. I'll test that tomorrow. Thanks. | 22:59 |
msiism | golinux: Good to know. | 22:59 |
golinux | fsr does the magic of setting it up but will not be around much this week | 23:00 |
msiism | Yeah, I guess, removing "splash" from that line will probably do it. | 23:01 |
mason | msiism: don't forget the "update-grub" as that file is used to generate the grub config - it's not itself part of the live grub configuration | 23:01 |
msiism | Yeah, I know. | 23:02 |
rwp | msiism, The Ludite that I am I always set GRUB_TERMINAL=console (Legacy BIOS boot only) GRUB_DISABLE_SUBMENU=y GRUB_CMDLINE_LINUX_DEFAULT="fsck.repair=yes" | 23:20 |
rwp | AFAIK it is not possible to have a simple text grub boot on UEFI as UEFI requires a graphical console. | 23:21 |
rrq | no, text is fine | 23:21 |
rwp | Agreed. Text/Plain too not this Text/HTML that many things decide is text! | 23:22 |
rrq | could also use syslinux but that would probably be a bit of constant "fight" against installing grub (idk) | 23:24 |
rwp | Has anyone got a uboot recipe for amd64 booting? | 23:26 |
msiism | rwp: I actually switched over to "console" mode for one boot today. There's nothing wrong with it, I guess. It just looks different. | 23:36 |
rwp | You said you wanted "not to show the beautified version of the GRUB menu" and that would seem to be closer to it. | 23:41 |
rwp | I just want a simple and easy to understand boot screen. The fancier things get the harder it is to use. | 23:42 |
rwp | msiism, You said GRUB but as for the rest of the boot flow perhaps the setting you were looking for was setting FANCYTTY=no in /etc/lsb-base-logging.sh ?? See use in /lib/lsb/init-functions file. | 23:46 |
msiism | I'll keep it in mind. | 23:50 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!