| josch | minute: unfortunately it seems that quite a bit of work remains to be done before atuin can be in debian: https://paste.debian.net/1320720/ | 00:05 |
|---|---|---|
| - mjw (QUIT: Ping timeout: 260 seconds) (~mjw@gnu.wildebeest.org) | 00:35 | |
| josch | uff... problems keep getting more with atuin... | 00:35 |
| josch | error: package `atuin-common v18.3.0 (/<<PKGBUILDDIR>>)` cannot be built because it requires rustc 1.77 or newer, while the currently active rustc version is 1.75.0 | 00:35 |
| josch | linux 6.9 is now in Debian kernel packaging git master https://salsa.debian.org/kernel-team/linux/-/commit/badef23070beb796b2c94722c581f035524f90f6 | 00:46 |
| josch | it includes the new debian/rules clean-generated target which I'll be using in reform-debian-packages | 00:47 |
| josch | minute: running reform-setup-encrypted-nvme (which also calls reform-migrate and reform-boot-config) in qemu on the gitlab CI works but instead of 6 minutes (on my a311d) the job takes 75 minutes... | 00:52 |
| josch | minute: is this worth the benefit of having some testing of the system image and reform-tools? | 00:52 |
| * bluerise_ -> bluerise | 00:54 | |
| + mgd (~mgd@host-89-240-113-218.as13285.net) | 01:20 | |
| - mgd (PART: !!unknown attribute: msg!!) (~mgd@host-89-240-113-218.as13285.net) | 01:20 | |
| - jjbliss (QUIT: Remote host closed the connection) (~jjbliss@1464766-static.elnsmiaa.metronetinc.net) | 02:49 | |
| + jjbliss (~jjbliss@1464766-static.elnsmiaa.metronetinc.net) | 02:53 | |
| - chomwitt (QUIT: Ping timeout: 268 seconds) (~chomwitt@2a02:587:7a19:9600:1ac0:4dff:fedb:a3f1) | 03:08 | |
| midfavila | noam, Zaba, Twodisbetter: jumping in with partial logs but wrt: the reform vs framework thing, its like the difference between free software and open source, from my perspective | 03:45 |
| midfavila | or freedom-enforcing/restrictive vs non-enforcing/permissive licenses | 03:46 |
| midfavila | the latter is a problem because it doesnt require you to not screw people over | 03:46 |
| midfavila | in doing so you tacitly endorse the status quo | 03:46 |
| midfavila | i.e. proprietary systems, and the unjust power structures they help in perpetuating | 03:46 |
| midfavila | maybe im reading too far into it for some views, but that the reform's schematics are available is a huge draw for me both as a consumer and a hobbyist | 03:47 |
| midfavila | its not just that theyre available, that theyre convenient from an engineering standpoint | 03:47 |
| midfavila | its that that convenience is a) putting minute et al on a more equal, democratic playing field with the end users, and it also just makes the reform and similar easier to use | 03:48 |
| midfavila | like when it comes to hardware, if schematics, source code, etc are available, i'll almost always choose that over a system where it isnt, because beyond my weird ideological opinions, its just less headache inducing | 03:48 |
| midfavila | if i want to port software to the reform i just pull up some specs, roll up my sleeves, and get cracking. i dont have to jump through weird legal hoops or look for leaked docs or just cope with the fact that the manufacturer wont let me use the hardware as i please because muh property rights | 03:49 |
| midfavila | in the real world that can be as simple as not needing to conatminate my kernel with blobs during build time. its a pain in the ass to have to deal with those because they arent maintained for older or more obscure software, rendering otherwise useful hardware e-waste, and they can introduce serious bugs that cant be fixed by the community | 03:50 |
| midfavila | if a modern gluebook is a knife twelve inches in your back, the framework pulls it out halfway | 03:50 |
| midfavila | thats not progress | 03:50 |
| midfavila | thats like, doing slightly better than the average | 03:50 |
| midfavila | like wow you'll sell me a battery and a screen great job round of applause everyone | 03:51 |
| midfavila | i still have to lock myself into your form factor and you wont give me schematics to fab my own low-level parts of tinker with things or whatever | 03:51 |
| midfavila | the reform might be (by default anyway) slower and less performant than the competition, but the quality and not-making-me-want-to-sudoku-myself-factor is leagues beyond almost anything else on the market | 03:52 |
| midfavila | tl;dr framework drools reform rules | 03:52 |
| midfavila | blogpost over | 03:52 |
| midfavila | thanks for coming to my ted talk | 03:52 |
| dozens | man that guy ted must have talked so much, he's like famous for it | 03:56 |
| noam | midfavila: sudoku yourself?? | 05:08 |
| midfavila | character rerolling. i say it sardonically but there are times where ive been tinkering with something and ive hit a wall because of licensing and it makes me want to throw things | 05:09 |
| midfavila | like for example matrox still makes graphics cards and afaik they dont require proprietary firmware. quad DP and single slot. ideal for professional workstations... but the X11 driver is proprietary and only they distribute it and it hasnt been updated in years | 05:10 |
| midfavila | if the matrox x11 driver was open source i would sell my workstation's WX5100 immediately | 05:11 |
| josch | midfavila: wrt performance you might be interested in the upcoming rk3588 upgrade for the reform :) | 07:13 |
| josch | minute: source.mnt.re CI is doing "500 Internal Server Error" again because no more space for build artifacts -- i'm going to delete some from the non-main branches to fix this again, okay? | 07:19 |
| josch | https://source.mnt.re/reform/reform-system-image/-/jobs/4570 | 07:19 |
| pandora | did anyone manage to install miniconda or so on an MNT? I tried the aarch64 installer but it always quiets with "Illegal instructions". I suspect that the cpu on the pocket is not supporting all opcodes? Maybe I am just doing it wrong | 07:32 |
| pandora | oh ... actually. The build for the rpi4 works. I don't know why the generic aarch64 doesn't | 07:40 |
| pandora | @ephase“11:08 PM <ephase> I don't think so because I have CPU and Memory on my waybar, all become red when full “ - could u share ur config how u got cpu and mem in ur waybar? Wanna do the same :) | 07:48 |
| ex-parrot | josch: 75 minutes doesn't sound too bad for a thing if you can just ignore it and let it run | 07:58 |
| minute | midfavila: thanks for the speech! | 08:58 |
| josch | minute: i also sent you some messages but the important question is whether 75 minutes is just too horrible or not? Maybe this job should be optional and only trigger on scheduled runs or something? | 09:01 |
| minute | josch: 75 mins are indeed horrible because they block the CI... so optional is good. ooc what's the main issue, qemu can't use kvm in ci? | 09:03 |
| josch | minute: i have yet to find a hosting provider which hands out virtual machines which also allow kvm | 09:05 |
| josch | minute: most of them give you a virtualization environment with kvm disabled for reasons beyond my knowledge | 09:05 |
| josch | but it's probably a good reason because i do not know of one where you get it | 09:05 |
| - digitalrayne (QUIT: Quit: じゃまたね) (~digitalra@vps-446f4f39.vps.ovh.ca) | 09:52 | |
| + digitalrayne (~digitalra@vps-446f4f39.vps.ovh.ca) | 09:53 | |
| violet | friendy refresh on trying to get an account activated on the mnt gitlab. username is artemis / artemis@artemis.sh | 10:30 |
| josch | minute: ^ | 10:33 |
| violet | logging off now becaue its bed time but ill be around again tomorrow | 10:37 |
| + mjw (~mjw@gnu.wildebeest.org) | 10:45 | |
| - nybble (QUIT: Ping timeout: 256 seconds) (~sprang@user/csprng) | 10:46 | |
| josch | minute: i found an old system image for LS1028A on my disk with a kernel from March this year: 6.7.9-reform2-2+reform20240327T174618Z | 10:47 |
| josch | minute: that one seems to work fine. I then upgraded only the kernel to 6.8.12-1+reform20240605T104210Z | 10:48 |
| josch | and that produces lock-ups. | 10:48 |
| josch | I have an LED on my SSD which will turn on once i press the power button and then stay on (it will blink when data is being transferred) | 10:48 |
| josch | that LED turns off when the lock-up happens | 10:48 |
| josch | great, i'm trying to uninstall linux 6.8 so that i can boot with linux 6.7 again but it locks up before "apt remove" can complete XD | 11:04 |
| josch | good news is of course that the issue is quite reproducible so if somebody has too much free-time on their hands, it would be not too hard to bisect the kernel to find the commit that introduced this regression | 11:06 |
| josch | oh no wait, good news! | 11:10 |
| josch | i just had a lock-up with kernel 6.7 | 11:10 |
| josch | so maybe it's something about my system | 11:10 |
| josch | huh... is this a thermal issue? | 11:11 |
| josch | maybe not... /sys/devices/virtual/thermal/thermal_zone?/temp says 51000 | 11:13 |
| josch | and now i put a 120 mm fan on top of the heat sink and the problem remains... | 11:32 |
| - mjw (QUIT: Ping timeout: 264 seconds) (~mjw@gnu.wildebeest.org) | 11:59 | |
| josch | minute: do you think i can hit the merge button on https://source.mnt.re/reform/reform-tools/-/merge_requests/72 so that the pcie related fixes in the initramfs for imx8mp can be deployed to users? | 12:10 |
| josch | i tested big parts of the MR but not all of it as my LS1028A is a bit broken it seems :( | 12:10 |
| josch | so i didn't manage to have reform-setup-encrypted-nvme finish | 12:10 |
| josch | but the "dmesg -n 7" in the initramfs hook works as intended | 12:10 |
| * Guest2692 -> mjw | 12:56 | |
| + ericsfra` (~user@ra-vpn234.ra-vpn.ucl.ac.uk) | 13:00 | |
| - ericsfraga (QUIT: Ping timeout: 252 seconds) (~user@2.28.216.209) | 13:02 | |
| - ericsfra` (QUIT: Ping timeout: 264 seconds) (~user@ra-vpn234.ra-vpn.ucl.ac.uk) | 13:09 | |
| pandora | Got another question about the pocket reform battery… when I query the battery state with upower it always reports the same “Energy”. That makes it impossible to get a “TimetoEmpty”. The energy rate and percentage works tho. Is the energy reported by the driver? | 13:49 |
| minute | mjw: does the sd card in your reform contain important data or should i override it with latest ls1028a image? | 14:58 |
| mjw | minute, override it! :) | 14:58 |
| minute | mjw: ok cool | 14:58 |
| minute | josch: merged! sorry i accidentally didn't uncheck the "delete source branch" box | 14:58 |
| + ericsfraga (~user@2.28.216.209) | 15:00 | |
| minute | josch: maybe your ls1028a needs the bodge wire for jtag reset | 15:05 |
| minute | josch: i can do that (after my holidays) | 15:05 |
| minute | there's only 3 people who have ls1028a without that fix, josch, jacqueline and cinap_lenrek | 15:05 |
| minute | jacqueline did the fix herself | 15:06 |
| minute | josch: the loglevel fix is in that reform-tools MR, right? | 15:08 |
| minute | i applied this fix now manually on mjw's sd card | 15:09 |
| mjw | ACTION feels special, thanks | 15:09 |
| minute | :D | 15:09 |
| minute | everything looks good so far | 15:11 |
| mjw | ACTION also feels a little embarrassed, here we have the most open, modular, fixable hardware and instead of fixing it myself I sent it in to get it fixed/upgraded... | 15:11 |
| minute | i'll leave it turned on for a while during lunch (which is now) to see if i also get the hang that josch gets... (but i don't expect it) | 15:12 |
| minute | core_cluster-virtual-0 | 15:12 |
| minute | Adapter: Virtual device | 15:12 |
| minute | temp1: +46.0°C | 15:12 |
| minute | mjw: it's good to check in on the health of our devices personally from time to time :D | 15:13 |
| minute | > 15:48:11 up 45 min, 3 users, load average: 0.09, 0.06, 0.01 | 15:48 |
| minute | no hang | 15:48 |
| minute | > temp1: +53.0°C | 15:48 |
| minute | i used this image https://source.mnt.re/reform/reform-system-image/-/jobs/4529/artifacts/file/reform-system-ls1028a.img.gz | 15:48 |
| minute | josch: so might be something about your module or setup | 15:48 |
| minute | (system is not yet on SATA though... this is running from SD) | 15:49 |
| minute | ah, poweroff doesn't cleanly work, there's a (known) issue where it OOPSes | 15:49 |
| minute | weird that it does > [ 2777.405956] reform2_lpc: removing ... | 15:50 |
| minute | so it removes lpc driver instead of doing powerdown?! | 15:50 |
| + mark_ (~mjw@gnu.wildebeest.org) | 16:01 | |
| + chomwitt (~chomwitt@2a02:587:7a19:9600:1ac0:4dff:fedb:a3f1) | 16:37 | |
| noam | Hmm, how loud do the CES speakers get? | 16:44 |
| noam | has anyone modded the Reforrm to be, uh, even louder? :P | 16:44 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 17:13 | |
| - ericsfraga (QUIT: Ping timeout: 268 seconds) (~user@2.28.216.209) | 17:20 | |
| josch | minute: oh thank you for all your replies! Let me go through them. :) | 17:29 |
| josch | minute: I actually changed the settings of the reform-tools project such that merge requests would have their source branch deleted by default -- this is to stop clutter from accumulating which at some point has to be cleaned up. If a branch gets merged, I do not think there is much reason to keep the branch, no? Because its contents can then also be found in the "main" branch. And the local git of the | 17:31 |
| josch | user who prepared the MR will still have the branch on their PC. | 17:31 |
| josch | minute: i thought the bodge wire for jtag reset was only to allow it to *boot*. Is the bodge wire also required to assure continuous operation? If yes, is there a plausible reason why I didn't experience this problem until now? As you know I had it on for many, many hours in the past -- not the least to test reform-tools and reform-system-image stuff but also to do all the PCIE graphics card foo. And it | 17:33 |
| josch | never died by itself... | 17:33 |
| josch | minute: the loglevel fix is now part of the initramfs. If it's just in reform-hw-setup, then it's too late to allow entering of the luks passphrase during very early boot. | 17:34 |
| josch | minute: yes, the CI results of job 4529 is exactly what I used. But I also had the issue with earlier images. Notably images from March with which I hadn't had these issues back then. | 17:35 |
| josch | And yes, my system was also running from SD. reform-setup-encrypted-nvme never managed to finish | 17:35 |
| josch | minute: I documented the LS1028A powerof issue here 2 days ago so that it does not get forgotten: https://source.mnt.re/reform/reform-debian-packages/-/issues/9 | 17:36 |
| josch | minute: thanks a lot for all your feedback. It indeed sounds to be a problem of my specific unit, which I'm very relieved about. Phew... I was really worried I'd have to do another round of kernel bisecting. XD | 17:37 |
| + ericsfraga (~user@2.28.216.209) | 18:30 | |
| - ericsfraga (QUIT: Remote host closed the connection) (~user@2.28.216.209) | 18:33 | |
| * mjw -> Guest7064 | 18:43 | |
| - Guest7064 (QUIT: Killed (copper.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:43 | |
| * mark_ -> mjw | 18:43 | |
| + Guest7064 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 18:43 | |
| + nybble (~sprang@user/csprng) | 18:54 | |
| + ephase (~ephase@2a01:e0a:2a:5300:8af3:6216:8fce:7058) | 20:22 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 20:24 | |
| - ephase (QUIT: Quit: WeeChat 4.3.1) (~ephase@2a01:e0a:2a:5300:8af3:6216:8fce:7058) | 21:32 | |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-64-230.tukw.qwest.net) | 22:26 | |
| + colinsane (~colinunin@97-113-64-230.tukw.qwest.net) | 22:30 | |
| - colinsane (QUIT: Remote host closed the connection) (~colinunin@97-113-64-230.tukw.qwest.net) | 23:29 | |
| + colinsane (~colinunin@97-113-64-230.tukw.qwest.net) | 23:34 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!