| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 00:09 | |
| - Gooberpatrol66 (QUIT: Ping timeout: 276 seconds) (~Gooberpat@user/gooberpatrol66) | 00:09 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 00:24 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-78-82-52-34.bbcust.telenor.se) | 00:25 | |
| - nsc (QUIT: Ping timeout: 265 seconds) (~nicolas@57-97-142-46.pool.kielnet.net) | 01:26 | |
| + nsc (~nicolas@42-97-142-46.pool.kielnet.net) | 01:28 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 02:45 | |
| - paperManu (QUIT: Ping timeout: 255 seconds) (~paperManu@198.16.214.40) | 02:55 | |
| mhoye | josch: yeah, the 25mm one is the one I have, and I like it a lot. | 04:02 |
|---|---|---|
| - op_4 (QUIT: Remote host closed the connection) (~tslil@user/op-4/x-9116473) | 04:05 | |
| + op_4 (~tslil@user/op-4/x-9116473) | 04:05 | |
| - Gooberpatrol_66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 04:19 | |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 04:20 | |
| - sts-q (QUIT: Ping timeout: 245 seconds) (~sts-q@89.58.73.103) | 04:22 | |
| + sts-q (~sts-q@89.58.73.206) | 04:26 | |
| - CandidCadaver (QUIT: Ping timeout: 265 seconds) (~aloo_shu@90.166.98.89) | 04:37 | |
| + CandidCadaver (~aloo_shu@90.166.98.89) | 04:46 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 05:34 | |
| - Gooberpatrol_66 (QUIT: Ping timeout: 252 seconds) (~Gooberpat@user/gooberpatrol66) | 05:35 | |
| xha | i am very confused why the u-boot build is not working. this patch patches something that was never there in the first place: https://source.mnt.re/reform/reform-rk3588-uboot/-/blob/main/0000-dts-makefile.patch?ref_type=heads | 07:49 |
| xha | there is no "CONFIG_ROCKCHIP_RK3588" here: https://gitlab.collabora.com/hardware-enablement/rockchip-3588/u-boot/-/blob/rk3588/arch/arm/dts/Makefile?ref_type=heads | 07:50 |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 07:50 | |
| xha | the patch cannot work... and there where no commits in the collabora repository since the last two weeks that changed anything regarding this device tree Makefile. so i am gussing there was a force-push? | 07:51 |
| xha | this build passed, and i don't understand *how*: https://source.mnt.re/reform/reform-rk3588-uboot/-/jobs/6246 | 07:52 |
| - CandidCadaver (QUIT: Ping timeout: 265 seconds) (~aloo_shu@90.166.98.89) | 08:23 | |
| + CandidCadaver (~aloo_shu@90.166.98.89) | 08:32 | |
| xha | ok, for the rk3588-rock5b branch of u-boot, the patch actually applies. however, that version is super outdated and does not build anymore. | 08:33 |
| - senzilla (QUIT: ) (sid151180@id-151180.hampstead.irccloud.com) | 08:38 | |
| + senzilla (sid151180@id-151180.hampstead.irccloud.com) | 08:39 | |
| + gustav28 (~gustav@c-78-82-52-34.bbcust.telenor.se) | 09:03 | |
| gustav28 | (minute: not yet tested, but here is the code I talked about yesterday: https://git.sr.ht/~gustav/fork-reform/log/gustav/ctrl-super-swap-modes) | 09:47 |
| bluerise | xha: RK3588 support on OpenBSD ain't too bad, but as always the video department is a bit lacking, hah | 09:59 |
| bluerise | xha: so when booting openbsd, which should be easy, you still need the uart | 09:59 |
| bluerise | unless u-boot sets up a simple framebuffer, which I don't think it does | 10:00 |
| bluerise | otherwise, wait a week until mine is here and then I can give some more info :D | 10:06 |
| - cobra (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~cobra@user/Cobra) | 10:10 | |
| xha | bluerise: i actually managed to boot & install openbsd onto an nvme already | 10:16 |
| + mjw (~mjw@gnu.wildebeest.org) | 10:16 | |
| xha | and also boot it from the nvme | 10:17 |
| xha | but yeah, u-boot does not setup a simplefb(4) | 10:17 |
| + cobra (~cobra@user/Cobra) | 10:20 | |
| hramrach | what is hyper and super? When I run evtest on plain USB keyboard the 'windows' key sends LEFT_META or RIGHT_META | 10:22 |
| gustav28 | Mnt labels the LEFT_META key SUPER and HYPER is something only Mnt has, defined as F23 | 10:24 |
| hramrach | but whatever they are I suppose I would want to swap hyper and super to make the corner key usable by the OS | 10:24 |
| gustav28 | in their documentation here https://mntre.com/documentation/reform-keyboard-v4-manual.pdf | 10:24 |
| gustav28 | this is for full reform/standalone keyboard, not sure about the pocket | 10:25 |
| hramrach | it's the same for posket as far as the left bottom keys are concerned | 10:26 |
| gustav28 | ok. my changes above is for swapping ctrl-hyper (so that ctrl is bottom left) and/or swapping super-alt (so that super matches Mac keyboards), independently | 10:29 |
| hramrach | the layer change key is so much used that as far as I am concerned it can be on the side of the case, no need to have it on the keyboard at all | 10:29 |
| hramrach | not so easy on the pocket because there are fewer keys on the keyboard, and the layer switch is needed to type at all | 10:30 |
| hramrach | If this is merged adding a third swap mode is not that difficult, anyway | 10:36 |
| hramrach | hm, maybe not so easy because it's swapping a key that does send a keycode with one that does not | 10:37 |
| gustav28 | pocket reform keyboard firmware seems to be entirely seperate (in MNT Pocket Reform repo), so my changes would not apply to it I think | 10:43 |
| gustav28 | what's the layer key? | 10:43 |
| gustav28 | (would be nice with a unified firmware though! in principle :) | 10:44 |
| hramrach | I suppose it's the hyper that does the layer switch to the additional functions | 12:01 |
| gustav28 | oh yes, layers as in additional functions. that's hyper on the keyboard4 as well | 12:10 |
| hramrach | also I think that the switch functions should come last in the menu (or just before 'x'), status is probably more useful than switch | 12:18 |
| gustav28 | Good point! | 12:27 |
| hramrach | about the unified firmware .. thre is separate firmware even for the v2/v3 keyboard | 12:40 |
| hramrach | tried to port it to the v3 firmware as well https://source.mnt.re/hramrach/reform/-/tree/gustav/ctrl-super-swap-modes?ref_type=heads | 12:43 |
| + paperManu (~paperManu@198.16.214.40) | 12:45 | |
| hramrach | Also if you swap hyper and ctrl then hyper-a is very easy to press accidentally | 12:50 |
| grimmware | minute: yeah what size *are* the pocket reform trackballs? I don't think it's in the handbook | 13:00 |
| ch | PREF-TBBALL-10R01 POM Ball Black 15mm | 13:22 |
| ch | is in the handbook bom | 13:22 |
| grimmware | ah my bad, I did search it but I looked under the wrong sections | 13:25 |
| hramrach | Looks like the v4 keyboard uses a different chip from v2/v3. It does not mean that the keyboard logic could be shared but it would need some adaptation laeyr because the API is somewhat different sometimes | 13:25 |
| hramrach | on the other hand, the v4 and pocket key matrix is different so the sharing would be more difficult | 13:26 |
| grimmware | mhoye's supplier doesn't do the balls in 15mm, I've just bought a random one on ebay | 13:32 |
| minute | grimmware: 15mm | 13:58 |
| - aperezdc (QUIT: Ping timeout: 260 seconds) (~aperezdc@2a03:6000:6e61:633::43) | 14:25 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 15:22 | |
| bluerise | xha: cool, you got a dmesg? | 15:45 |
| bluerise | RK3588 <3 | 15:45 |
| + aperezdc (~aperezdc@2a03:6000:6e61:633::43) | 16:21 | |
| xha | bluerise: here you go https://xha.li/paste/pocket_rk3588_dmesg.txt | 16:45 |
| xha | lots of "not configured" lol | 16:45 |
| svp | the milk-v mars cm does in fact start at boot, but due to a missing device tree it wont do much on its own, it tries to boot from nvme and defaults to hdmi for output. imaging the nvme drops me to a shell, basically what i was expecting | 17:37 |
| minute | svp: oh wow ok. in classic reform or pocket reform? | 17:42 |
| svp | classic reform | 17:43 |
| minute | svp: nice, and you get graphics on hdmi? | 17:43 |
| minute | svp: where's the bootloader stored on that one, emmc or sdcard? | 17:43 |
| minute | svp: soo with the internal hdmi adapter it could probably be used with the internal display then... i figure usb doesn't work? | 17:44 |
| svp | opensbi is, surprisingly enough, on a onboard eeprom, im booting straight off of nvme | 17:45 |
| svp | (i had to settle for the emmc-less version) | 17:45 |
| minute | ah interesting. and what kind of shell are you getting? | 17:46 |
| svp | the internal adapter is not going to help, it only has one hdmi output and it's assigned to the outside hdmi port. it does have 4-lane dsi which is what i'm using. also, linux shell, it does manage to boot all the way to gdm | 17:52 |
| svp | had to adjust the kernel cmdline because it was looking for a sd card. i cant seem to get that to work so i went straight for the nvme since it was being detected | 17:53 |
| hramrach | hm, looks like Mars does not support virtualization | 17:54 |
| svp | and yes the hdmi framebuffer happily shows up, boot console and all | 17:54 |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 17:54 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 17:55 | |
| hramrach | and up to 8GB RAM, not great but usable | 17:56 |
| svp | until radxa gets their act together this is the only cm4 risc-v solution i know of | 17:57 |
| hramrach | way better than loading a riscv core into a fpga | 17:58 |
| hramrach | the SPI NOR is great for bootloader, and it's very cheap as well. I don't understand why so few boards have one | 18:01 |
| hramrach | or is it SPI, they do not specify the bus in the board description | 18:01 |
| hramrach | yes, it shows SPI in the block diagram | 18:02 |
| svp | on a related note the orange pi cm5 also starts but i can only use it with the serial console so that isnt helpful. even less helpful: the fact it runs uart at 1500000 bps | 18:05 |
| minute | yeah, rk3588 uart is always 1.5 mbaud | 18:09 |
| svp | hm, usb also isnt playing nice on mars | 18:13 |
| minute | svp: could be usb hub | 18:16 |
| minute | svp: you probably need to find the gpio line for the usb reset and toggle that | 18:16 |
| hramrach | getting riscv in reform would be nice | 18:17 |
| svp | unsure, even the port with the direct connection to the som isnt responsive. | 18:17 |
| minute | svp: that goes through another usb hub on the rcm4 which is on the same reset line iirc | 18:18 |
| svp | huh! hmm. | 18:18 |
| hramrach | hm but have too many computers already, should stop looking at SoMs, there is no point for me | 18:18 |
| minute | it's GPIO10 on the rpi cm4 | 18:19 |
| minute | checking https://milkv.io/docs/mars/compute-module/hardware | 18:21 |
| grimmware | is this the first time that someone other than minute has tried a unique cm4 module in the adapter? | 18:21 |
| grimmware | or a unique SoM at all? | 18:22 |
| minute | svp: it's GPIO52 | 18:23 |
| minute | grimmware: no, someone else tried soquartz | 18:23 |
| minute | but they couldn't get the DSI to work | 18:23 |
| grimmware | it's fuckin sweet either way | 18:23 |
| minute | yess | 18:23 |
| grimmware | can't imagine any other ecosystem where this would happen | 18:23 |
| minute | svp: do you have gpioset / gpioget etc? | 18:24 |
| svp | i also have a bigtreetech cb1 laying around but i dont think thats gonna surprise me, jamming it in | 18:24 |
| svp | no i dont have it in the vendor system image, i'll see if i can get that installed... | 18:25 |
| minute | svp: for sdcard to work on the milkv you might need to add no-1-8-v; to the sd card node in dts, as sdcard is fixed to 3v3 in reform | 18:26 |
| minute | and dsi display is a lot trickier, but essentially also adding a bunch of dts nodes | 18:27 |
| + Gooberpatrol_66 (~Gooberpat@user/gooberpatrol66) | 18:38 | |
| svp | gpioset 0 52=1 seems to do the trick | 18:39 |
| - Gooberpatrol66 (QUIT: Ping timeout: 246 seconds) (~Gooberpat@user/gooberpatrol66) | 18:39 | |
| svp | sure enough, blindly logging in from the reform itself happily lists a new session in who | 18:40 |
| josch | svp: wow this is really cool news! it's really cool you are trying this out :) | 18:52 |
| - jacobk (QUIT: Quit: http://quassel-irc.org - Chat comfortably. Anywhere.) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 18:57 | |
| minute | svp: nice! | 19:16 |
| svp | hm, i think figuring out DTBs and DTSs is beyond my skill level, but it doesnt seem too far away in terms of arms' reach | 19:27 |
| svp | still fighting against the motherboard usb hub as port J5 doesnt feel like outputting power when the hub is operational but sometimes does when it isnt, but i am already accepting my destiny as it's definitely something i did a while back. perhaps it's getting close to the time for a motherboard upgrade anyway? ^^; | 19:33 |
| + erle (~erle@user/erle) | 19:44 | |
| ch | hramrach, josch: i just downloaded the pocket rk3588 image from https://source.mnt.re/reform/reform-system-image/-/jobs/6260/artifacts/browse and that still shows the setfont problems (and doesnt actually change the font). should that work alreadY/ | 20:03 |
| ch | nice to see that rcore in idle draws about the same or maybe less power than the imx8mp | 20:21 |
| ch | i think i'll flash a keyboard fw with shorter idle timeout for the kbd backlight | 20:23 |
| ch | just too much power draw | 20:23 |
| ch | powertop thinks the gpu is running at 300*G*Hz. guess thats not right :> | 20:29 |
| grimmware | got lid closed detection working aww yiss | 20:34 |
| ch | nice | 20:35 |
| grimmware | the issue was that I was still including the one axis that totally doesn't matter (i.e. along the hinge) in the angle calculations which was obviously throwing it off when the thing wasn't totally level | 20:37 |
| ch | i think this is better than what i ever saw on imx8: # ON BAT chg=0 mps_flt=00 input=5000mV@56mA charge=0mA discharge=275mA p=2.12W ttempty=328min | 20:37 |
| svp | since it's way over my bar i'll just add a line in rc.local to ready the hub >_> helps little due to no display but it's a start. if anyone more knowledged wants to give a try to baking the dtb id be glad to give it a try though | 20:38 |
| grimmware | ch: what command? I can run it on my pocket with imx8mp for comparison if you like | 20:38 |
| grimmware | might not be useful if it's not an equivalent workload tho | 20:39 |
| grimmware | man I'm looking forward to getting my rcore on tuesday | 20:40 |
| ch | yeah repro is hard. you can look at the Amps figure on the battery status on the oled | 20:41 |
| ch | (probably only correct when on battery) | 20:41 |
| grimmware | eh, yeah difficult to tell becasue I'd be eyeballing a figure that keeps changing and trying to match that up to like, load average or something | 20:43 |
| josch | ch: this problem seems to be known to hramrach: https://source.mnt.re/reform/reform-system-image/-/merge_requests/112#note_11220 | 21:00 |
| ch | josch: well that seems to be about console-setup, which runs later. but the initramfs stuff already doesnt work at all? | 21:01 |
| josch | ch: i'm afraid this is something you pocket owners have to figure out :/ | 21:02 |
| minute | my input to this would be: break=top or break=modules into initramfs (over serial for example, with extlinux.conf that has keep_bootcon) and try to load the font manually and try to understand what exactly fails, maybe even with a homebrew setfont executable with debugging | 21:06 |
| ch | yeah its just weird because it was working already | 21:07 |
| josch | ch: it worked on your systems but the vanilla system might not have some stuff installed that you do have installed maybe? | 21:09 |
| ch | it worked on my system, and with the current packages it also fails on my system | 21:09 |
| ch | and it also fails on the image | 21:10 |
| josch | nice, sounds some regression because $foo got upgraded to a new version :/ | 21:10 |
| ch | wondering how it worked though | 21:13 |
| ch | minute: btw, on rcore the hdmi when hdmi output off seems to confuse my display. it goes into a cycle of "no signal" and "off" | 21:14 |
| josch | if there was a version of the system image where this was known to work, you could bisect | 21:14 |
| minute | ch: hmm | 21:14 |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.28.180) | 21:46 | |
| + Ar|stote|is (~linx@149.210.28.180) | 21:47 | |
| minute | aha, this is not yet in the sway i'm running https://github.com/swaywm/sway/pull/8156 | 21:48 |
| minute | maybe, with a lot of luck, this could fix chromium being crashy on rk3588 (https://gitlab.freedesktop.org/mesa/mesa/-/issues/11244) | 21:49 |
| minute | (`wayland-info | grep syncobj` doesn't yield anything) | 21:49 |
| hramrach | ch: set the font in flash-kernel. See https://source.mnt.re/reform/reform-rk3588-uboot/-/merge_requests/5 | 21:51 |
| - S0rin (QUIT: Read error: Connection reset by peer) (~S0rin@user/s0rin) | 21:52 | |
| ch | that all seems like piling workarounds on top of workarounds | 21:53 |
| hramrach | we have a pile of workarounds now | 21:54 |
| ch | minute: maybe its in sway from experimental (1.10 got tagged 4 mins ago, seems a bit fresh :) | 21:54 |
| hramrach | setting the font in u-boot would be nice but we don't have a safe way to upgrade u-boot so passing parameters in flash-kernel is the next best option | 21:55 |
| minute | ch: i think it isn't in 1.10 unfortunately, but it's in wlroots that we have... so i'll try to build sway master from source | 21:56 |
| ch | arch wiki claims "The kernel chooses the one to use based on its evaluation of the screen resolution." | 21:57 |
| ch | would be good if the autodetect would work | 21:57 |
| hramrach | the resolution information is not available over mipi | 21:57 |
| ch | screen rotation seems to have a similar problem. the attached hdmi display comes up rotated, i guess because the internal display has to be rotated | 21:57 |
| ch | isnt there already display specific driver code which knows this? | 21:58 |
| hramrach | and there is no way to rotate each separately for the console | 21:58 |
| hramrach | in wayland yes, in kernel no | 21:59 |
| hramrach | well, not sure about wyland, x11 can sure do it | 21:59 |
| ch | in wayland it works | 22:00 |
| ch | but i mean, what happened to "the kernel should be in charge" and kms and all that stuff | 22:00 |
| minute | maybe we are missing some info on the panel structure? | 22:00 |
| hramrach | it would have to come from dtb if the mipi cannot provide any info | 22:01 |
| hramrach | http://devicetree.org/schemas/display/panel/panel-common.yaml# | 22:09 |
| hramrach | hm, broken URL | 22:09 |
| minute | hramrach: not necessarily, we can provide drm structures in the kernel driver | 22:11 |
| minute | of the panel. like it's physical size | 22:11 |
| hramrach | https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/devicetree/bindings/display/panel/panel-common.yaml | 22:12 |
| minute | but that's already known to drm, not sure if accurate > physical_width: 95 mm, physical_height: 151 mm | 22:12 |
| hramrach | that drm structure needs to be populaated from something | 22:12 |
| minute | well it is populated by the driver... | 22:12 |
| minute | hramrach: https://github.com/torvalds/linux/blob/master/drivers/gpu/drm/panel/panel-jdi-lt070me05000.c#L294 | 22:13 |
| hramrach | the size sunds about right | 22:13 |
| hramrach | ugh, there is specific driver for that one panel | 22:14 |
| minute | btw, current sway master doesn't work on rk3588 | 22:14 |
| minute | hramrach: of course... | 22:14 |
| minute | hramrach: also, we patch that driver | 22:14 |
| minute | because the standard driver is not working correctly | 22:14 |
| minute | also it needs tweaking depending on quirks of each DSI host | 22:15 |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-78-82-52-34.bbcust.telenor.se) | 22:15 | |
| minute | unfortunately struct drm_display_info doesn't have any rotation attribute. | 22:16 |
| hramrach | the panel-common does, though | 22:17 |
| minute | interesting, i wonder where that goes in the kernel | 22:20 |
| minute | also, i think we use fbcon:rotate_all | 22:20 |
| minute | but there's also fbcon:rotate | 22:20 |
| + gustav28 (~gustav@c-78-82-52-34.bbcust.telenor.se) | 22:20 | |
| minute | IIRC rotate didn't work at first so i used rotate_all... maybe that can be looked at again | 22:20 |
| ch | cmdline says fbcon=rotate:3 | 22:22 |
| gustav28 | crap, I think my keyboard4 just died :( | 22:24 |
| gustav28 | I tried running ./flash-direct.sh, got response no device in BOOTSEL. then I entered the firmware update mode of the keyboard and tried again, got somewhat different output. then I unplugged the keyboard and all my devices on the usb hub it was connected to stopped working. after replugging them they all work, except the keyboard. | 22:24 |
| minute | ch: ah sorry | 22:25 |
| gustav28 | (terminal output https://paste.sr.ht/~gustav/fce42ecbebf8c9a2c9c554210cd27b991e1bb0d8) | 22:25 |
| josch | hramrach, ch: about setting fbcon=font:TER16x32 for the kernel cmdline: setting this in boot.scr will be tricky because u-boot knows that it runs on imx8mplus but not that it runs on the pocket, or the classic reform or the reform next. It just has ${board}=nitrogen8mp | 22:25 |
| minute | gustav28: hmm probably just borked the firmware... maybe it'll come back after a reboot, or you can try the programming switch | 22:26 |
| hramrach | I did not manage to get the display resolution from anywhere so I cannot expect the console-setup to pick the right font size | 22:26 |
| gustav28 | minute: will try reboot. any idea of why it happened? installed picotool via brew, maybe mismatch with pico-sdk? programming switch will make it visible to picotool again, hopefully? | 22:28 |
| ch | maybe just wonky usb at that moment? | 22:28 |
| hramrach | but serrting it in u-boot would not work because it's shared between pocket and big reform, indeed | 22:28 |
| minute | gustav28: yes | 22:34 |
| minute | gustav28: i've never tried flashing on a mac, too | 22:34 |
| minute | hramrach: how do you expect to get the resolution? there are several ways to get it, but depends on your context | 22:35 |
| minute | hramrach: for example, `cat /sys/class/drm/card0-DSI-1/modes` | 22:35 |
| hramrach | minute: that does not have physical size, though | 22:37 |
| gustav28 | minute: ok! just to confirm, the programming switch is inside the case, so I need to open it up? | 22:39 |
| minute | hramrach: ah you mean the physical size, to determine dpi? | 22:39 |
| minute | gustav28: yes | 22:39 |
| minute | hramrach: uboot is not shared between big reform and pocket, there are special builds for each, because i set the fbcon rotation only for pocket variants in uboot. | 22:40 |
| hramrach | so that should work | 22:41 |
| minute | i.e. https://source.mnt.re/reform/reform-rk3588-uboot/-/blob/main/mnt-pocket-reform-rk3588_defconfig?ref_type=heads#L137 | 22:41 |
| minute | so one could set a bigger font there, for example | 22:41 |
| hramrach | I suppose you can merge the font setting then | 22:42 |
| minute | yeah, i just need to test it. i'll make a note for tomorrow | 22:42 |
| gustav28 | aww yea it works! and my fw changes work as intended, just one minor bug (the menu doesn't close or react when switching mode) | 22:44 |
| gustav28 | thanks for the help minute | 22:44 |
| gustav28 | it flashed fine with the programming switch | 22:44 |
| minute | gustav28: ah, great to hear | 22:44 |
| ch | rebooting seems to not work every n-th time | 22:44 |
| minute | ch: are you on the very newest kernel? | 22:45 |
| minute | ch: or, what do you mean specifically? | 22:46 |
| ch | 6.11.4-mnt-reform-arm64 #1 SMP Debian 6.11.4-1+reform20241025T201754Z (2024-10-25) | 22:46 |
| ch | i think the panel turns off, and then something is stuck and it doesnt boot (panel stays black) | 22:46 |
| ch | cant attach a serial console atm | 22:47 |
| hramrach | gustav28: you should return a specific value to close the menu | 22:47 |
| minute | ch: kernel seems new ;3 | 22:47 |
| vkoskiv | I think I asked about this a while back, but is it normal to have 935 packages held back in apt still? | 22:48 |
| ch | on unstable? not really | 22:48 |
| hramrach | there is some qt breakage, or was a copuple days ago :/ | 22:49 |
| vkoskiv | Tried to scan through the full output of `apt -s full-upgrade` but I'm not convinced that it's a safe upgrade, I got tons and tons of output. | 22:49 |
| vkoskiv | I don't know how to work out why those 935 packages are being held back, beyond googling and getting handwavey answers :D | 22:49 |
| ch | logs? | 22:50 |
| vkoskiv | Full output of `apt -s full-upgrade` (2.2MB): https://mister-muffin.de/p/x0hV.txt | 22:55 |
| josch | vkoskiv: you are using repos that are not from Debian. It might help to first disable tham and/or remove the packages coming from them. | 22:57 |
| ch | (KEYMAP=y and sleep makes setfont work, meh) | 22:57 |
| vkoskiv | I have deb & deb-src from debian, then deb from reform-debian-repo | 22:57 |
| josch | vkoskiv: can you upload the output of "apt-cache policy"? | 22:59 |
| vkoskiv | https://mister-muffin.de/p/FZZu.txt | 23:00 |
| gustav28 | hramrach: yes, thanks - I made it so the menu closes after changing mode and force pushed to my branch, if you're interested. it's not ideal as you may want to toggle both modes before exiting the menu but without closing the menu it's not clear to the user if anything was toggled at all. ideally I would want a little toggle icon switching in the | 23:00 |
| gustav28 | menu :) | 23:00 |
| ch | freecad -> unsatisfied dependency on qtbase-abi-5-15-13 | 23:00 |
| ch | guess that doesnt help | 23:00 |
| ch | qt/kde is still unhappy in unstable indeed | 23:01 |
| vkoskiv | ch: How did you work that out? 118 lines containing 'freecad' :D | 23:02 |
| josch | vkoskiv: and maybe next this handy one-liner: python3 -c 'import apt_pkg;apt_pkg.init();cache=apt_pkg.Cache(None);print("\n".join(sorted([f"{pkg.name:<44} {pkg.current_ver.ver_str}" for pkg in cache.packages if pkg.current_ver for pfile, _ in pkg.current_ver.file_list if not pfile.not_source and not pfile.site.endswith(".debian.org")])))' | 23:03 |
| gustav28 | minute: the keyboard fw flashes fine now even without the programming switch - must have been the usb hub they keyboard was connected to that messed up the communication (now I have it directly connected to the computer) | 23:03 |
| josch | vkoskiv: other than that, I'd try running "apt install" with the stuff that it wants to remove that it really should not to figure out why it wants to remove it | 23:03 |
| ch | vkoskiv: the lines after "875 upgraded, 433 newly installed, ..." are not really useful in this case | 23:04 |
| josch | like, why does it want to remove systemsettings or freecad | 23:04 |
| ch | vkoskiv: i picked a package that seemed meaningful from the list of stuff apt wanted to remove, and then checked https://qa.debian.org/dose/debcheck/unstable_main/1730005201/arm64.html | 23:04 |
| ch | (link from https://qa.debian.org/dose/debcheck/unstable_main/index.html ) | 23:05 |
| vkoskiv | josch: https://mister-muffin.de/p/OzG4.txt | 23:05 |
| josch | vkoskiv: okay, thank you | 23:06 |
| josch | yeah, there seems to be a qt issue right now... | 23:06 |
| vkoskiv | So it's expected to have these held back, then? | 23:07 |
| josch | the uninstallable ones, yes | 23:08 |
| ch | right now yes, three days ago not | 23:08 |
| + S0rin (~S0rin@user/s0rin) | 23:31 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-78-82-52-34.bbcust.telenor.se) | 23:37 | |
| + gourd753 (~gourdcapt@user/gourd753) | 23:40 | |
| gourd753 | Question: what can I do with the hyper/super keys on the MNT keyboard? | 23:42 |
| gourd753 | as it is, Hyper mostly just causes me to fail to do stuff that would rely on CTRL by muscle memory | 23:42 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!