| minute | ch: i think checkbox 5 is now complete? https://source.mnt.re/reform/pocket-reform/-/issues/4 | 00:15 |
|---|---|---|
| ch | yes! | 00:15 |
| minute | i read recently in #linux-rockchip that someone almost has rk3588 VOP2 working in uboot | 00:16 |
| - bkeys (QUIT: Ping timeout: 272 seconds) (~Thunderbi@173.16.175.75) | 00:32 | |
| grimmware | minute: does that basically mean having the screen render useful things earlier in boot? | 00:45 |
| grimmware | I need to read into the latest keyboard firmware cos I think it might solve some of the problems I’ve been having when using the accelerometer mod | 00:47 |
| + bkeys (~Thunderbi@173.16.175.75) | 00:48 | |
| ch | i take it to mean that, provided it'll run edk2 | 00:55 |
| ch | but that would be awesome | 00:55 |
| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@47.202.75.129) | 01:02 | |
| + mtm (~textual@47.202.75.129) | 01:05 | |
| - mjw (QUIT: Ping timeout: 248 seconds) (~mjw@gnu.wildebeest.org) | 01:15 | |
| + aperezdc (~aperezdc@2a03:6000:6e61:633::43) | 01:25 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 01:51 | |
| - bkeys (QUIT: Ping timeout: 276 seconds) (~Thunderbi@173.16.175.75) | 02:15 | |
| - nsc (QUIT: Ping timeout: 260 seconds) (~nicolas@i5C74DC72.versanet.de) | 03:13 | |
| + nsc (~nicolas@65-97-142-46.pool.kielnet.net) | 03:15 | |
| - cobra (QUIT: Quit: ZNC 1.8.2 - https://znc.in) (~cobra@user/Cobra) | 04:09 | |
| + jacobk (~quassel@47-186-65-73.dlls.tx.frontiernet.net) | 04:11 | |
| - aloo_shu (QUIT: Ping timeout: 252 seconds) (~aloo_shu@85.51.17.59) | 04:16 | |
| + aloo_shu (~aloo_shu@90.166.98.130) | 04:18 | |
| + cobra (~cobra@user/Cobra) | 04:20 | |
| - MyNetAz (QUIT: Remote host closed the connection) (~MyNetAz@user/MyNetAz) | 04:23 | |
| - paperManu (QUIT: Ping timeout: 248 seconds) (~paperManu@107.159.243.8) | 04:27 | |
| + MyNetAz (~MyNetAz@user/MyNetAz) | 04:34 | |
| - bleb (QUIT: Ping timeout: 252 seconds) (~cm@user/bleb) | 05:40 | |
| + bleb (~cm@user/bleb) | 05:44 | |
| + ajr_ (~ajr@syn-172-091-111-232.res.spectrum.com) | 07:58 | |
| - ajr_ (QUIT: Client Quit) (~ajr@syn-172-091-111-232.res.spectrum.com) | 08:01 | |
| + ajr_ (~ajr@syn-172-091-111-232.res.spectrum.com) | 08:01 | |
| - ajr_ (QUIT: Client Quit) (~ajr@syn-172-091-111-232.res.spectrum.com) | 08:04 | |
| + ajr_ (~ajr@2603:8001:6400:3795::1012) | 08:05 | |
| - ajr_ (QUIT: Client Quit) (~ajr@2603:8001:6400:3795::1012) | 08:07 | |
| + ajr_ (~ajr@syn-172-091-111-232.res.spectrum.com) | 08:07 | |
| - ajr (QUIT: ) (uid609314@user/ajr) | 08:18 | |
| - ajr_ (QUIT: Quit: ZNC 1.9.1 - https://znc.in) (~ajr@syn-172-091-111-232.res.spectrum.com) | 08:21 | |
| + ajr (~ajr@user/ajr) | 08:21 | |
| - ajr (QUIT: Quit: ZNC 1.9.1 - https://znc.in) (~ajr@user/ajr) | 08:40 | |
| - aloo_shu (QUIT: Ping timeout: 252 seconds) (~aloo_shu@90.166.98.130) | 09:45 | |
| + aloo_shu (~aloo_shu@85.51.17.167) | 09:48 | |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@47-186-65-73.dlls.tx.frontiernet.net) | 10:04 | |
| + jacobk_ (~quassel@47-186-65-73.dlls.tx.frontiernet.net) | 10:04 | |
| - n_to (QUIT: Quit: quitidiquit) (~n_to@2a03:4000:6:3662:24b1:57ff:fec6:76c1) | 11:33 | |
| + n_to (~n_to@2a03:4000:6:3662:24b1:57ff:fec6:76c1) | 11:34 | |
| BoostisBetter | minute: I am happy to report that your updated keyboard firmware and after updating psuspend and rmt tools, the psuspend is working perfectly again with the Pocket. The battery ready out on the oled is showing an average, power draw of .300ish amps which is a .200 ish amps improvement. | 12:26 |
| BoostisBetter | minute: is these numbers are true this means that the pocket could last for over 24 hours in psuspend | 12:28 |
| - aperezdc (QUIT: Ping timeout: 265 seconds) (~aperezdc@2a03:6000:6e61:633::43) | 12:29 | |
| BoostisBetter | minute: if these numbers are true this means that the pocket could last for over 24 hours in psuspend | 12:30 |
| + mjw (~mjw@gnu.wildebeest.org) | 12:36 | |
| + gustav28 (~gustav@c-78-82-52-90.bbcust.telenor.se) | 13:02 | |
| - mtm (QUIT: Ping timeout: 252 seconds) (~textual@47.202.75.129) | 13:03 | |
| + mtm (~textual@47.202.75.129) | 13:05 | |
| op_4 | anyone here using the RK3588 and have the misfortune of having to use slack/google meet/zoom? is it powerful enough to get by? | 13:05 |
| BoostisBetter | the RK3588 is more powerful than an x230 i5 | 13:07 |
| BoostisBetter | which runs all of those things without issue | 13:07 |
| BoostisBetter | so I would say that it SHOULD be able to | 13:07 |
| op_4 | sweet! BoostisBetter i think you might have been the tipping point for me here :) | 13:07 |
| op_4 | i have an x230 and it does indeed work | 13:08 |
| BoostisBetter | op_4: no worries, I also have an x230 but I have an i7 version | 13:08 |
| op_4 | BoostisBetter: yeah, tbh me too but it's running qubes so i guess emulating an i5 :p | 13:08 |
| op_4 | and i had a veyron speedy chromebook (RK3288) and it could just _barely_ squeak by | 13:10 |
| ch | fwupd failing appears to be a permission issue m( | 13:17 |
| minute | BoostisBetter: i also have seen a lower power draw but it's hard to believe :D | 13:18 |
| ch | are we sure its not the leds? :D | 13:18 |
| minute | we are not sure :D but i can do some tests in the lab with the keyboard in isolation | 13:19 |
| minute | old vs new fw | 13:19 |
| BoostisBetter | just happy about the improvement regardless | 13:21 |
| amospalla | are you using imx8mp on reform BoostisBetter? If so, what amp/voltage shows when pseudo suspended? | 13:26 |
| BoostisBetter | amospalla: .300-430-ish amps | 13:36 |
| BoostisBetter | amospalla: also the CPU is obviously cooler. Kind of unbelievable as well. | 13:36 |
| amospalla | I'm upgrading firmware today. | 13:37 |
| amospalla | What is the process to manually upgrade to the latest firmware? (I can not use fwupd). | 13:43 |
| - ericsfraga (QUIT: Remote host closed the connection) (~user@2a00:23cc:b43d:4b00::3ce) | 13:43 | |
| ch | depends | 13:43 |
| ch | oled menu -> X should always work and get you into bootsel mode | 13:44 |
| ch | then picotool load filename.uf2 | 13:44 |
| ch | picotool reset | 13:44 |
| ch | have an external keyboard attached before doing that | 13:44 |
| + bkeys (~Thunderbi@173.16.175.75) | 13:45 | |
| amospalla | Is the latest built artifact a good choice? Or too risky? | 13:48 |
| ch | worked for me yesterday | 13:50 |
| amospalla | Ok, it has an update script. | 13:50 |
| amospalla | oh well, no package named picotool on Debian/Stable :'( | 13:52 |
| amospalla | No luck for me, I'll have to wait till next Debian release. | 13:52 |
| ch | or flash from an sdcard | 13:53 |
| amospalla | true | 13:53 |
| amospalla | or a chroot | 13:54 |
| bkeys | Has anyone got a shipping notification for rk3588? | 14:00 |
| - bkeys (QUIT: Ping timeout: 245 seconds) (~Thunderbi@173.16.175.75) | 14:04 | |
| + bkeys (~Thunderbi@173.16.175.75) | 14:13 | |
| + paperManu (~paperManu@107.159.243.8) | 14:13 | |
| amospalla | I think I upgraded the sysctl firmware, how can I check I actually have the latest one? | 14:21 |
| + ericsfraga (~user@2a00:23cc:b43d:4b00::3ce) | 14:23 | |
| + aperezdc (~aperezdc@2a03:6000:6e61:633::43) | 14:25 | |
| amospalla | nevermind, I have it. I see it is consuming 0.275A, never saw something that low before. | 14:25 |
| josch | maybe a pocket owner can benchmark the battery runtime before and after the change to make sure that these numbers are not fake? | 14:27 |
| amospalla | Comparing it to the values I noted when I got my pocket, it is roughly half the power consumption :S | 14:27 |
| amospalla | josch: yes, I'll do. Will take time, I need a full charge/discharge cycle. | 14:28 |
| BoostisBetter | it is an impressive improvement that is mostly likely due to the idle power consumption efforts. Combined with some loop elements being removed from the keyboard firmware, I think we are just seeing some realistic improvement. Excellent work indeed. | 14:30 |
| BoostisBetter | amospallo: does psuspend only use brightnessctl and NOT dpms? | 14:30 |
| amospalla | brightnessctl + switch into another virtual terminal + setterm blank(from memory) | 14:31 |
| amospalla | Because it has to work system wide. | 14:31 |
| BoostisBetter | ok good | 14:32 |
| BoostisBetter | brightnessctl has worked reliably for me | 14:33 |
| BoostisBetter | I am wondering if we will get working suspend on the RK3588. Quite a few people have that now out in a wild. Do we have reports of suspend working? | 14:33 |
| BoostisBetter | and if so, on what kernel? | 14:33 |
| ch | amospalla: wait, your sysctl wasn't at least 2024-08-26 before? | 14:36 |
| ch | thats the version that brought improved idle power consumption | 14:36 |
| BoostisBetter | ch: if you didn't upgrade then it was at the initial version when the pocket shipped. Minute said mine was way old, for example | 14:37 |
| BoostisBetter | and that version was previous to July as well | 14:37 |
| ch | yes, but upgrading to 2024-08-26B was always a good idea | 14:37 |
| BoostisBetter | yep, although with old non-updated keyboard firmware, it caused the keyboard and sysctrl to crash and be usuable with a power cycle | 14:38 |
| amospalla | ch: I upgraded to the one minute announced on the forums. | 14:38 |
| amospalla | It was on these dates IIRC. | 14:38 |
| ch | thats the one that was announced iirc | 14:38 |
| BoostisBetter | amospalla: you need to update your keyboard firmware as well | 14:38 |
| amospalla | Oh, keyboard firmware too? | 14:38 |
| josch | BoostisBetter: we don't have suspend working with rk3588. I have working suspend with imx8mq on classic reform and kde plasma 6 -- bisecting now to see where it broke... | 14:38 |
| BoostisBetter | or you will have issues of the sysctrl crashing | 14:38 |
| amospalla | I only upgraded the sysctl. | 14:38 |
| amospalla | Well, I just ran the included script in ./scripts | 14:38 |
| ch | the current fw discussion was always about the keyboard firmware :) | 14:38 |
| BoostisBetter | amospalla: this is the one you want for your keyboard | 14:39 |
| BoostisBetter | https://source.mnt.re/reform/pocket-reform/-/jobs/7256/artifacts/download?file_type=archive | 14:39 |
| BoostisBetter | it has the keyboard firmware already built with an install script already there explaining how to upgrade it | 14:39 |
| BoostisBetter | you will need another usb keyboard though | 14:39 |
| BoostisBetter | but my limited time running this version has been VERY good | 14:40 |
| BoostisBetter | and the best thing is you can recover the usb hub and the keybaord independently with it | 14:40 |
| amospalla | Whatever, I upgraded the sysctl to the latest build artifact and the power consumption apparently improved (if the oled values are correct). | 14:41 |
| ch | try plugging a charger and then unplugging it to see if theres a difference | 14:42 |
| josch | $ git bisect bad v6.12 | 14:43 |
| josch | Segmentation fault (core dumped) | 14:43 |
| josch | yay | 14:43 |
| BoostisBetter | josch: good times | 14:44 |
| amospalla | ch: yes, plug/unplug, it still consumes less power than before. | 14:45 |
| amospalla | when plugged, voltage gets higher, and amps get negative and around 1.4A | 14:46 |
| BoostisBetter | how much power the pd pulls has to do with the stater of your battery. Max discharge is around -1.9 amps | 14:46 |
| BoostisBetter | close it gets to 100% the more it tapers off, and at 100% it will just be the charger supplying operating requirements. Love that. The batteries are nice and cool once they are charged. | 14:47 |
| BoostisBetter | amospalla: when you upgrade the sysctrl or keyboard make sure you reboot the system. This ensures that the file system is correct and not locked. | 14:50 |
| amospalla | Should the new keyboard firmware make the computer consume less energy? | 14:50 |
| amospalla | If so, I should wait to test the time it gets the battery to deplete with the latest sysctl firmware, and then test again with the latest keyboard firmware. | 14:51 |
| BoostisBetter | yes the new keyboard is the key to a much better overall netted power consumption. However with the new sysctrl I saw slight improvements as well. | 14:53 |
| - mjw (QUIT: Ping timeout: 260 seconds) (~mjw@gnu.wildebeest.org) | 14:53 | |
| * Guest4113 -> mjw | 14:55 | |
| amospalla | But I have not applied the keyboard firmware (from usb-preempt branch), but the sysctl firmware (from main branch). | 14:55 |
| josch | Bisecting: 85361 revisions left to test after this (roughly 16 steps) | 14:57 |
| josch | second time was a charm | 14:57 |
| josch | of course it would not segfault inside gdb... | 14:58 |
| - bkeys (QUIT: Quit: With every step we take, danger will follow closely) (~Thunderbi@173.16.175.75) | 15:01 | |
| josch | when bisecting linux and you re-use your current .config with another kernel version -- do you do anything other than just pressing and holding the enter key as all the differing config options scroll by? | 15:02 |
| + bkeys (~Thunderbi@173.16.175.75) | 15:02 | |
| minute | amospalla: looking forward to your results! | 15:04 |
| BoostisBetter | amospalla: ditto | 15:04 |
| Zaba | josch: make olddefconfig perhaps? | 15:12 |
| josch | Zaba: thank you! I'll try that out at the next bisection step in ~6.5 hours :) | 15:13 |
| - bkeys (QUIT: Ping timeout: 272 seconds) (~Thunderbi@173.16.175.75) | 15:15 | |
| josch | ./include/trace/events/ext4.h:1891:46: internal compiler error: Segmentation fault | 15:26 |
| josch | this is not my lucky day... | 15:26 |
| BoostisBetter | ahh the joys of debugging work. | 15:27 |
| * leonardo- -> leonardo | 15:55 | |
| BoostisBetter | in other news, I already have an idea for something I am going to work on to improve teh Next when it comes out. has to do with the touchpad. | 15:57 |
| BoostisBetter | any guesses? I will hint that it is something Windows already has. | 16:03 |
| BoostisBetter | Crap, just had the USB hub just crash again. System was still running but no USB port was active. Keyboard and trackball unresponsive and the lpc was also unresponsive and inaccessible. | 16:11 |
| BoostisBetter | Had to power cycle... Bummer. | 16:12 |
| - jacobk_ (QUIT: Ping timeout: 260 seconds) (~quassel@47-186-65-73.dlls.tx.frontiernet.net) | 16:16 | |
| ch | josch: lol. is that with the newest git from unstable? (i think jonathan uploaded a new git yesterday) | 16:17 |
| BoostisBetter | amospalla: seems like it is evtest and trying to get a handle on the keyboard. I guess the hook changes and psuspend spams the system waiting for the keyboard to awake. This causes the whole usb hub to take a giant dump. | 16:17 |
| BoostisBetter | systemd actually automatically restarted the service, and this is when evtest first failed and the awaiting keybaord messages happen. | 16:18 |
| BoostisBetter | I think assuming that the hub is going to give the keyboard the same address is perhaps not prudent here and is the source of the issue | 16:18 |
| amospalla | check your dmesg, I suspect that keyboard disapears, evtest stops and psuspend stops with it. Then systemd restarts the failed service, and because there is no keyboard it waits forever for it to exist. | 16:19 |
| BoostisBetter | That is possible, because the keyboard DID disappear. | 16:21 |
| BoostisBetter | minute: I wonder if the hub could automatically detected if it is disconnected and reset to connect again? | 16:21 |
| antti | Hello Folks | 16:22 |
| antti | I just got a mnt-reform v4 keyboard. I noticed that there isn't an insert key and I want to change the left ctrl for the hyper key. | 16:22 |
| antti | I also I want to disable the rgb backlight and only enabled when I need it. | 16:22 |
| antti | Based in the following links | 16:22 |
| antti | https://shop.mntre.com/products/mnt-reform-keyboard-40 | 16:22 |
| antti | With version 4.0, we are now using the ARM-based controller RP2040 as it is faster and better | 16:22 |
| antti | documented than the ATmega32U4. | 16:22 |
| antti | This links said that it is easy to update via USB but there aren't instructions. | 16:22 |
| antti | https://mntre.com/docs-reform.html#keyboard | 16:23 |
| antti | MNT Reform Standalone Keyboard Manual (V4) | 16:23 |
| antti | https://mntre.com/documentation/reform-keyboard-v4-manual.pdf | 16:23 |
| antti | this manual has only a few details about it v4 keyboard | 16:23 |
| antti | where can I find the official way to modify some keys ? | 16:23 |
| antti | I found an old link in https://community.mnt.re "Another firmware for the MNT Reform keyboard: QMK | 16:23 |
| antti | from 2021 that he mnt keyboard supports qmk | 16:23 |
| antti | https://community.mnt.re/t/another-firmware-for-the-mnt-reform-keyboard-qmk/657/2 | 16:23 |
| antti | is this the way how can I change a few keys ? | 16:23 |
| ch | in the pdf you linked it says the source and instructions are at https://source.mnt.re/reform/reform/-/tree/master/reform2-keyboard4-fw?ref_type=heads | 16:23 |
| ch | in qmk both mntre and mntre_v3 are for processor=atmega32u4, so that wont work with v4 | 16:25 |
| ch | but maybe you can contribute a config to qmk thats using rp2040 | 16:26 |
| + bkeys (~Thunderbi@173.16.175.75) | 16:27 | |
| antti | thanks for the replies . I saw the gitlab repo but how do you which file do you need to modfy for the keyboard and then swap the keys | 16:27 |
| antti | https://source.mnt.re/reform/reform/-/raw/master/reform2-keyboard4-fw/README.md | 16:27 |
| ch | https://source.mnt.re/reform/reform/-/blob/master/reform2-keyboard4-fw/src/usb_hid_keys.h?ref_type=heads looks relevant | 16:28 |
| + jacobk (~quassel@47-186-65-73.dlls.tx.frontiernet.net) | 16:31 | |
| minute | BoostisBetter: keyboard disappeared with new fw? does it come back after usb reset from oled menu? | 16:32 |
| ch | https://bugs.debian.org/1092015 unstable users might want to hold off upgrading for a bit until its clear if this affects arm64 too or whats going on | 16:36 |
| ch | ('libwlroots-0.18: WLR compositors's fail to run due to "undefined symbol: wlr_output_is_drm" | 16:36 |
| ch | ') | 16:36 |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@47-186-65-73.dlls.tx.frontiernet.net) | 16:39 | |
| BoostisBetter | minute: the lpc does not respond or give me a menu. It as if someone just unplugged the whole hub. | 16:41 |
| - bkeys (QUIT: Ping timeout: 260 seconds) (~Thunderbi@173.16.175.75) | 16:41 | |
| amospalla | BoostisBetter: does your pocket lose keyboard only when it is pseudo-suspended? Never ever in other situation? | 16:41 |
| BoostisBetter | amospalla: in this instance it has resumed 3 minutes earlier from psuspend. Was working fine. Then the hub disconnectes and psuspend went into waiting for keyboard status | 16:42 |
| BoostisBetter | But before, yes it seemed to do that while psuspend was active. | 16:43 |
| amospalla | Battery 0: Discharging, 96%, 25:36:00 remaining | 16:47 |
| antti | https://source.mnt.re/reform/reform/-/blob/master/reform2-keyboard4-fw/src/usb_hid_keys.h?ref_type=heads looks relevant -> interesting I can see a lot of keys here. these should be the keys that I need to check in order to swap it. but how do you know what is the main file for that ? | 16:48 |
| BoostisBetter | amospalla: with psuspend active right. That is what I was seeing as well. Massive. | 16:48 |
| minute | BoostisBetter: the menu is rendered by the keyboard itself. so you couldn't open that, yes? | 16:48 |
| BoostisBetter | minute: that is correct. The keyboard was totally unresponsive. | 16:48 |
| minute | BoostisBetter: ok so the keyboard main loop hung or crashed, hmm. | 16:48 |
| BoostisBetter | minute: I tried plugging another keyboard in as well and the hub did nothing with it. | 16:48 |
| BoostisBetter | minute: so the sysctrl was likely also crashed. | 16:49 |
| BoostisBetter | minute: or like I speculated, the USB hub connecting all of it took a dump | 16:49 |
| BoostisBetter | minute: lastly what amospalla is reporting is what I am seeing as well. This means that psuspend is almost better than a real suspend. | 16:51 |
| minute | BoostisBetter: hub and sysctl are not really related to each other. but interesting that the new keyboard was not detected as well | 16:51 |
| amospalla | my script is a wannabe suspend, can't really compare | 16:52 |
| BoostisBetter | minute: hmm, if we can just keep the keyboard stabke then psuspend will be super useful for the Mplus crowd not upgrading. | 16:52 |
| BoostisBetter | minute: So I just enabled sway idle and left my pocket alone. I come back to it and normally wiggling the track ball will wake it and I'll get the screen and the keyboard backlight will come back on. But now it's acting the exact same way. So the keyboard is disconnected. The OLED menu is not visible and I can't, it doesn't have any way to wake the machine. I'm going to plug a keyboard in and see | 16:54 |
| BoostisBetter | if that's recognized. But my guess is the hub will be down as well. | 16:54 |
| BoostisBetter | minute: the hub is still active and I am back in | 16:56 |
| BoostisBetter | minute: I should note that this time psuspend was disabled and not running. This just happeed while the pocket was on and plugged in. It is possible that the Librem 5 charger I am using is doing something weird to the sysctrl or something. | 16:57 |
| BoostisBetter | minute: looking at the dmesg logs from journalctl it seems that there is a usb error: usb 1-1.3: device descriptor read/6, error -32 | 17:00 |
| BoostisBetter | then the hub is listed as device 4, resets and is listed as device 5 | 17:00 |
| BoostisBetter | I this this relisting of the hub is what is causing the keyboard to disconnect | 17:01 |
| BoostisBetter | xwayland, on another tangent seems to be causing the etnaviv gpu to crash. I need to figure out which applicants I am using that is using xwayland. | 17:02 |
| minute | BoostisBetter: it can also be the other way around, that the keyboard hangs and disturbs the hub | 17:02 |
| BoostisBetter | minute: lsusb reads out the way it should and shows both keyboard and sysctrl | 17:03 |
| minute | BoostisBetter: ok but keyboard responds not to hyper+enter yes? and holding hyper and rolling trackball does not change the backlight? | 17:04 |
| BoostisBetter | minute: I think the sysctrl is still running as waybar seems to be reporting a battery precentage that is correct. I guess it really is isolated to just the keyboard this one time. | 17:04 |
| BoostisBetter | minute: yes that is correct | 17:04 |
| BoostisBetter | I am using a System76 keyboard with it now | 17:04 |
| BoostisBetter | no mouse | 17:04 |
| minute | BoostisBetter: ok, so the keyboard rp2040 has crashed or is hung | 17:04 |
| BoostisBetter | minute: ok, that seems probable | 17:05 |
| BoostisBetter | minute: is the only way to get it back to power cycle? | 17:05 |
| minute | i wonder how to reproduce that | 17:05 |
| minute | BoostisBetter: you can remove the keyboard lid and press the reset button of the keyboard. but not very comfy | 17:06 |
| minute | we could look into hw watchdog of the rp2040, but that's also a bandaid | 17:06 |
| BoostisBetter | minute: amospalla: I am wondering if rmt tools are maybe something here. Swayidle is using rmt leds to turn the backlight off on the keyboard. When resume is activated by moving the mouse, then rmt leds is setting the leds to 0 0 50 | 17:07 |
| minute | if we had a good reproducer, that would be best | 17:07 |
| minute | i'm currently in a train and will see how my pocket behaves | 17:08 |
| BoostisBetter | reseting the keyboard brought keyboard and oled back | 17:09 |
| amospalla | BoostisBetter: you can find if psuspend has some interaction. | 17:11 |
| BoostisBetter | it didn't this go around because it was systemctl stopped | 17:11 |
| BoostisBetter | but rmt leds were still doing things | 17:11 |
| + jacobk (~quassel@2603:8080:b200:7b02:77cb:6304:f9db:dda1) | 17:12 | |
| amospalla | disable everything, use it for a couple of days, enable only one thing and use it for a couple of days, until you enable everything. If psuspend is triggering something you will locate what is this way. | 17:12 |
| BoostisBetter | amospalla, right now I have it sleeping as I am doing a battery test like you. 92%, enabled psuspend. Going to see how much drains in 4 hours. Once I resume, I will disable all things that use rmt in psuspend, and try it like that. I will also remove all instances of rmt in my sway config to make swayidle or something else isn't using it as well. | 17:18 |
| BoostisBetter | amospalla: the good news, is that I think we are real close and this is just something minor. | 17:18 |
| BoostisBetter | minute: I don't mind trying to figure this out with your help. I can do whatever you need. | 17:19 |
| BoostisBetter | minute: going to dinner now, but I'll be back in about 3 hours. | 17:19 |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@2603:8080:b200:7b02:77cb:6304:f9db:dda1) | 17:24 | |
| minute | indeed trackball moves slower on usb-preempt fw. | 17:25 |
| minute | i now edited .config/sway/config.d/input and added: | 17:26 |
| minute | accel_profile "adaptive" | 17:27 |
| minute | pointer_accel 1.0 | 17:27 |
| minute | that's much better! | 17:27 |
| [tj] | https://groups.google.com/g/linux-sunxi/c/OT3Or42xFWk | 17:39 |
| minute | hmm, rk3588 pocket reform headphone audio sounds bitcrushed in some frequencies | 17:41 |
| minute | i wonder if there's a clocking/mode mistake | 17:41 |
| ch | minute: btw did you recently test ethernet on rk3588/pocket? | 17:44 |
| minute | [tj]: thx for that link! | 17:45 |
| minute | ch: hmm not sure, do you have issues with it? | 17:45 |
| ch | i think i saw it not working, will retest with a more controlled environment later | 17:45 |
| ch | congress was not the best testbed :) | 17:46 |
| minute | yeah :D | 17:48 |
| + ajr (~ajr@user/ajr) | 18:13 | |
| + piroko (~piroko@104.225.216.16) | 18:29 | |
| grimmware | minute: BoostisBetter's problems sound pretty familiar to me, I'm going to re-read what they've written tomorrow as I may have been triggering the same thing with my accelerometer mod when plugged in to a power supply | 18:41 |
| grimmware | if it *is* the same thing I can tell you some more about the conditions under which I've triggered it | 18:44 |
| ch | grimmware: in case this only happens when connected to a charger, can you try my usbpd branch? | 18:44 |
| grimmware | ch: is this a separate branch or the stuff you merged to main a while ago? | 18:45 |
| ch | https://source.mnt.re/zeha/pocket-reform/-/commits/zeha-usbpd | 18:46 |
| ch | not merged yet | 18:46 |
| ch | cant use usb devices on the charging port with it at the moment | 18:46 |
| - ericsfraga (QUIT: Quit: ERC 5.6.1-git (IRC client for GNU Emacs 31.0.50)) (~user@2a00:23cc:b43d:4b00::3ce) | 18:48 | |
| grimmware | hmm, I'd have to merge our branches | 18:48 |
| ch | yeah probably | 18:49 |
| grimmware | I'll give it a look tomorrow as the merge might not be too nasty, we're both in the spi_com code for example but I wouldn't be surprised if we've managed to avoid stepping on each other's toes because I didn't touch the pd stuff | 18:50 |
| minute | BoostisBetter: shut down pocket in previous train. pulled out of sleeve in new train. keyboard not responding. maybe there's just a bug in my new branch | 18:58 |
| + mark_ (~mjw@gnu.wildebeest.org) | 19:02 | |
| - mark_ (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 19:17 | |
| + bkeys (~Thunderbi@173.16.175.75) | 19:26 | |
| + jacobk (~quassel@2603:8080:b200:7b02:77cb:6304:f9db:dda1) | 19:35 | |
| - bkeys (QUIT: Ping timeout: 265 seconds) (~Thunderbi@173.16.175.75) | 19:38 | |
| + bkeys (~Thunderbi@173.16.175.75) | 19:40 | |
| BoostisBetter | not sure, but similar things were happening before the branch as well. I can confirm that resetting the keyboard controller manually always brings the keyboard and oled back up | 19:52 |
| + mark_ (~mjw@gnu.wildebeest.org) | 19:53 | |
| BoostisBetter | minute: I am really curious if the sysctrl firmware (which has changed) is maybe polling the keyboard or something in a way that is contributing to this. Seems kind of illusive. | 19:54 |
| - bkeys (QUIT: Ping timeout: 244 seconds) (~Thunderbi@173.16.175.75) | 19:56 | |
| BoostisBetter | minute: came back from dinner, and the keyboard has disconnected again as well. It seems to be happening all the time now. From 17:12 to 19:49 the keyboard was fine. Then it disconnected again. | 19:56 |
| BoostisBetter | ch: I would be willing to try your branch, can you link me to the artifacts to download? | 20:08 |
| BoostisBetter | I will say that the latest disconnect I had happened while only on battery power. BUT, plugging into power while in psuspend or swayidle does tend to lead to crashes on the sysctrl. | 20:09 |
| ch | BoostisBetter: on https://source.mnt.re/zeha/pocket-reform/-/commits/zeha-usbpd click the green tick -> in the bottom right corner there is a download button -> then click build:archive | 20:09 |
| ch | BoostisBetter: this click-path should work for most things | 20:10 |
| BoostisBetter | thanks, got it. | 20:12 |
| BoostisBetter | Can you paraphrase what your branch is bringing to the table? | 20:12 |
| BoostisBetter | It mainly improves PD selection? | 20:12 |
| BoostisBetter | also I pulled from pipeline #3297 right? | 20:13 |
| ch | mostly PD stuff, yes. 3297 yes | 20:13 |
| ch | let me just check sth | 20:13 |
| ch | ok, this doesnt work as is on your hardware | 20:14 |
| ch | if you want to try it, try it from https://source.mnt.re/zeha/pocket-reform/-/commit/63b9f2282549db90f585933907368cd9c7f270c6 | 20:15 |
| BoostisBetter | I just installed it form that pipeline though, seems to work. I should take it from what you just linked though? | 20:20 |
| ch | yes | 20:20 |
| BoostisBetter | roger, installing it now | 20:21 |
| BoostisBetter | installed and rebooting | 20:25 |
| BoostisBetter | regarding the xwayland clients, found codium, brave, and betterbird all using it. They all are able to use wayland though. Made some changes and they are all happily using wayland. Something though like lxpolkit, pasystray, and gsm-settings are all dependent on xwayland. | 20:43 |
| BoostisBetter | I doubt they are crashing though, so I am not worried. Most likely it was Brave. | 20:43 |
| BoostisBetter | ch: so far now problems using your branch. I do see faster pd charging negogitations though | 20:45 |
| BoostisBetter | should more chargers work with this? Like maybe the Framework charger? | 20:45 |
| BoostisBetter | ch: YES!!! The Framework charger works with it now. How awesome is that! | 20:46 |
| ch | yes, more chargers should work | 20:46 |
| ch | and you should see better charging behaviour | 20:46 |
| ch | as said, usb devices mostly dont work on the charging usbc | 20:46 |
| BoostisBetter | looks like it is trickle charging compared to the others, which I like. I prefer low and slow as it is just better for the batteries. | 20:47 |
| BoostisBetter | I rarely plug anything in, and mostly have the pd port used for charging anyway | 20:47 |
| ch | the charger ic should switch automatically to faster charging once its safe to do so | 20:49 |
| ch | depending on the charging brick and the supported voltages that might not work without a hw mod, but if so it never did | 20:49 |
| BoostisBetter | seems to be working well enough. | 20:55 |
| BoostisBetter | Really appreciate it so far. | 20:55 |
| BoostisBetter | in other news I am trying to force LibreOffice to use wayland, but I can't seem to get soffice to use it. | 20:55 |
| BoostisBetter | I only ever use Writer | 20:56 |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@2603:8080:b200:7b02:77cb:6304:f9db:dda1) | 21:07 | |
| josch | bisecting imx8mq is not going well... kernel 6.6 is stuck in the initramfs... :/ | 21:23 |
| BoostisBetter | josch: bummer! Appreciate the efforts! | 21:30 |
| BoostisBetter | minute: seems the keyboard is exclusively the issue. Did another psuspend, and I had to reset the keyboard to get the keyboard working again. not trying to stress just reporting some hopefully useful information. | 21:35 |
| + Ar|stote|is (~linx@149.210.3.168) | 21:38 | |
| antti | Hello there. I'm trying to figure out how to switch the left ctrl vs the hyper key in mnt-reform v4 keyboard. I cloned the gitlab repo and ../reform2-keyboard4-fw/README.md there is only an example of build.sh | 22:03 |
| antti | Which tools do I need to flash it ? How can I make a copy of the current version ? in this new v4 model is the "usb flashing mode" the current way to push the flash on it? How do you know which file modify . somebody here point me to a file called usb_hid_keys.h but I don't see how can I use it . | 22:03 |
| - mjw (QUIT: Killed (iridium.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2486:a800:7602:5eff:dc71:a72c) | 22:06 | |
| * mark_ -> mjw | 22:06 | |
| + Guest1720 (~mjw@2001:1c06:2486:a800:7602:5eff:dc71:a72c) | 22:06 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-78-82-52-90.bbcust.telenor.se) | 22:15 | |
| BoostisBetter | hi antti: asking on the community forum is a great way to get help as well. Asking here MIGHT be quicker but only if the relevant people are on and are free. At least on the forum you give everyone a chance to help as well. | 22:16 |
| antti | BoostisBetter: oh I see. thanks for let me know. I will post it there. | 22:43 |
| BoostisBetter | no worries. Folks here are really awesome as well. Just might miss them depending on when you ask questions or say something. | 22:47 |
| BoostisBetter | IRC is fun like that. | 22:47 |
| josch | antti: also, not many folks are with keyboard v4 yet | 22:48 |
| antti | oh that explains why there isn't a qmk support | 22:53 |
| antti | well I more than happy to help with the documentation. this is my first keyboard that it has openhardware :P | 22:54 |
| josch | antti: if v4 works like the versions before, then the file you might want to modify is src/matrix.h You will see comments like // Top row, // Second row, // Third row | 22:55 |
| josch | antti: since you want to swap the first and second keys in the last row (right?) you should probably look under // Sixth row | 22:56 |
| josch | and then swap KEY_HYPER with KEY_LEFTMETA | 22:56 |
| josch | and then README.md tells you how to build it | 22:58 |
| josch | the script flash-direct.sh has the picotool invocations that let you flash it | 22:58 |
| josch | i ordered keyboard v4 but don't have one yet, so i cannot give you help from my own experience -- just reading the stuff in git | 22:59 |
| antti | I see. thanks for let me know all this. I was lost when I read usb_hid_keys.h | 23:00 |
| antti | I'm trying to switch keys four row first key (ctrl) for six row first key (hyper) . this is the first modification. thne I would like to see how can I map the insert key and then the pause key becasue when I'm playing 0ad I need to hit the menu with the rodent | 23:01 |
| josch | antti: note, that unless you want to change the hyper key, you can also use this method which avoids having to change your firmware: https://community.mnt.re/t/remapping-keyboard-without-firmware-modification-adding-sysrq-key/687 | 23:06 |
| josch | antti: it sounds like you want to do this change: https://paste.debian.net/hidden/b5498744/ | 23:08 |
| antti | A moderator must manually approve your new account before you can access this forum. --> step one completed.I need to wait for it. | 23:11 |
| josch | antti: one sec... | 23:13 |
| josch | antti: welcome to the forum :) | 23:13 |
| antti | sweet. | 23:13 |
| antti | let me check | 23:13 |
| antti | I was using with Xorg key mapping but I prefer to avoid this because I have multiple hosts and sometimes I forget to push the settings. if it is in the keyboard I just need to plug and play :P | 23:14 |
| josch | antti: ah, it's an external keyboard v4? | 23:15 |
| antti | yes I just order the keyboard standalone. it took like more than 3 months but I finally have it. | 23:15 |
| BoostisBetter | yeah MNT is a small company doing assembly in house. Good quality from excellent people, all worth the wait. | 23:16 |
| antti | is there a dark mode for https://community.mnt.re hidden somewhere ? | 23:16 |
| antti | BoostisBetter: sweet. I saw a talk about mnt in the last ccc and it looks like they are working hard to have a working arm64 alternative. | 23:17 |
| josch | antti: Preferences → Interface → Color Scheme | 23:18 |
| antti | I found the dark mode. wow is amazing. | 23:18 |
| antti | thanks josch . | 23:19 |
| antti | it has a strong magenta. | 23:19 |
| BoostisBetter | yeah Team MNT love magenta | 23:20 |
| antti | +1 | 23:20 |
| antti | oh that explains why the default led is magenta in the v4 standalone keyboard | 23:21 |
| antti | https://community.mnt.re/t/mnt-reform-keyboard-4-0-standalone-firmware-backup-change-howto/2891 | 23:32 |
| antti | that is the post Folks. let me know if I need to include something else | 23:33 |
| antti | thanks | 23:33 |
| vkoskiv | On my earlier keyboard at least, circle + x kicks it into flashing mode. Not sure there is a way to recover the firmware already on there though. | 23:34 |
| josch | probably via "picotool dump" | 23:37 |
| BoostisBetter | has anyone tried getting openBSD running on a MNT Reform or Pocket? | 23:37 |
| josch | ah no, "picotool save" | 23:37 |
| josch | the man page looks wrong... i should fix this... | 23:38 |
| antti | which cpu/soc are you using in the mnt reform or pocket ? | 23:39 |
| josch | My main computer is a MNT Reform with A311D. | 23:40 |
| vkoskiv | Same here | 23:40 |
| josch | I don't understand all the hype about the rk3588 -- 3.6 GB RAM should be enough for everything, no? ;) | 23:41 |
| ch | no :D | 23:41 |
| ch | this is not kansas^i386-land anymore | 23:41 |
| josch | hey i can *address* more than that easily! :D | 23:41 |
| ch | hahaha | 23:42 |
| ch | PAE wants to have a word with you | 23:42 |
| BoostisBetter | josch: the hype is 32gb RAM and a CPU that is more powerful than a x230 i5 basied thinkpad | 23:42 |
| BoostisBetter | I can DO A LOT with my X230 still | 23:42 |
| BoostisBetter | nice to think the same about a Pocket or Next. | 23:43 |
| antti | I can't find A311D in obsd pages. rk3588 is here > https://www.openbsd.org/74.html | 23:44 |
| ^alex | BoostisBetter, iirc openbsd didn't initialize the screen with the default compute module | 23:46 |
| BoostisBetter | gotcha. Would love it if OpenBSD worked. One day perhaps. | 23:52 |
| antti | well you will have more changes than waiting for obsd support than any other Os (even in the linux world) | 23:56 |
| antti | well you will have more changes waiting for obsd support than any other Os (even in the linux world) . | 23:58 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!