| - CandidCadaver (QUIT: Ping timeout: 260 seconds) (~aloo_shu@90.166.98.89) | 00:01 | |
| hramrach | gourd753: You can patch the firmware to remap the keys - see eg. https://git.sr.ht/~gustav/fork-reform/log/gustav/ctrl-super-swap-modes | 00:05 |
|---|---|---|
| hramrach | It does not provide the mapping you want because it does not move the hyper but you get the idea | 00:07 |
| hramrach | this might be more productive long term for key remapping but it's nowhere near generally usable, either https://community.mnt.re/t/another-firmware-for-the-mnt-reform-keyboard-qmk/657 | 00:08 |
| + CandidCadaver (~aloo_shu@90.166.98.89) | 00:11 | |
| gourd753 | Thanks for the links, hramrach. Need to decide if I'm willing to put in the effort on this, I guess. | 00:15 |
| gourd753 | (with the MNT Reform v4 keyboard) | 00:15 |
| gourd753 | because as is, it's a very nicely made object that is just slightly off enough in multiple aspects I go back to the $15 cheap fallback keyboard I'm using because my previous good keyboard died in the last week :( | 00:17 |
| - nsc (QUIT: Ping timeout: 252 seconds) (~nicolas@42-97-142-46.pool.kielnet.net) | 00:26 | |
| + nsc (~nicolas@71-97-142-46.pool.kielnet.net) | 00:28 | |
| - gourd753 (QUIT: Quit: Leaving) (~gourdcapt@user/gourd753) | 00:40 | |
| - paperManu (QUIT: Ping timeout: 252 seconds) (~paperManu@198.16.214.40) | 02:02 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 02:25 | |
| - ericsfraga (QUIT: Ping timeout: 272 seconds) (~user@2.28.216.224) | 02:39 | |
| - sts-q (QUIT: Ping timeout: 260 seconds) (~sts-q@89.58.73.206) | 04:23 | |
| + sts-q (~sts-q@89.58.73.113) | 04:24 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 05:08 | |
| - mlarkin (QUIT: Quit: leaving) (~mlarkin@syn-047-036-115-056.res.spectrum.com) | 06:14 | |
| + mlarkin (~mlarkin@syn-047-036-115-056.res.spectrum.com) | 06:48 | |
| - cobra (QUIT: Ping timeout: 248 seconds) (~cobra@user/Cobra) | 07:05 | |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 10:04 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 10:05 | |
| - amospalla (QUIT: Quit: WeeChat 4.4.2) (~jordi@user/amospalla) | 10:08 | |
| + amospalla (~jordi@user/amospalla) | 10:15 | |
| - Ar|stote|is (QUIT: Ping timeout: 246 seconds) (~linx@149.210.28.180) | 10:17 | |
| + Ar|stote|is (~linx@149.210.17.64) | 10:21 | |
| grimmware | nnng just came to open my pocket and the screen and backlight turned on :D | 10:32 |
| grimmware | I'll plumb in orientation detection and sway socket detection (want it to work across wm restarts), work out a bug with the sysctl uart socket buffering and then I'll write up what I've got in a blog post | 10:33 |
| josch | nice :) | 10:35 |
| - Ar|stote|is (QUIT: Ping timeout: 260 seconds) (~linx@149.210.17.64) | 10:55 | |
| + Ar|stote|is (~linx@149.210.18.40) | 10:59 | |
| * Guest361 -> mjw | 11:21 | |
| ch | yay | 11:35 |
| + paperManu (~paperManu@198.16.214.40) | 11:38 | |
| - Ar|stote|is (QUIT: Ping timeout: 246 seconds) (~linx@149.210.18.40) | 12:44 | |
| + Ar|stote|is (~linx@149.210.16.212) | 12:48 | |
| kensanata | Yesterday I tried to insert an SD card into the slot of my Pocket Reform and managed to miss the correct angle, dropping the card into the slot. I unscrewed that little plate and shook the until until it coughed up the card it had swallowed so all is good. It was quite a surprise. | 12:50 |
| kensanata | Does the EurKEY keyboard layout have a compose key? I guess it does not. Is there a line I can put in my ~/.config/sway/conf.d/input.conf file that would help? Something like input * xkb_options compose:lwin where lwin is actually hyper+rwin or something like that? | 12:55 |
| kensanata | Since I need all the modifiers the keyboard has, it needs to be on some other layer, I fear. | 12:56 |
| + gustav28 (~gustav@c-78-82-52-4.bbcust.telenor.se) | 13:02 | |
| hramrach | you would need to modify the keyboard firmware to produce some other keycode on hyper+rwin | 13:09 |
| kensanata | OK, thanks. | 13:10 |
| hramrach | see eg https://git.sr.ht/~gustav/fork-reform/log/gustav/ctrl-super-swap-modes (for v4 keyboard, the pocket firmware sources are in a different repository) | 13:10 |
| hramrach | but for what you want you would likely just modify the fn matrix without any other code change | 13:12 |
| hramrach | also try evtest in case it does already send a different key | 13:13 |
| - Ar|stote|is (QUIT: Ping timeout: 252 seconds) (~linx@149.210.16.212) | 13:23 | |
| + Ar|stote|is (~linx@149.210.16.106) | 13:27 | |
| minute | josch: security bug report > /etc/apt/sources.list.d/mntre.sources hat Permissions 666 aka -rw-rw-rw-. Sollte denke ich besser 644 sein. | 14:05 |
| minute | josch: reported by axel beckert | 14:05 |
| minute | kensanata: yes, sway supports xkb options | 14:06 |
| josch | minute: thanks! fixed in reform-system-image git | 14:08 |
| minute | josch: cool, that was fast | 14:08 |
| josch | minute: i can add a maintainer script snippet to reform-tools which does the chmod on the next reform-tools upgrade if it finds the file to exist | 14:08 |
| minute | josch: thanks, i just wanted to ask what's the best approach, and that sounds very good | 14:09 |
| josch | and add a check to reform-check | 14:09 |
| josch | i'm preparing a release right now | 14:09 |
| minute | awesome, thanks a ton | 14:11 |
| josch | no worries, this is a big blunder by me indeed | 14:12 |
| josch | i'll add another check to the mkimage.sh script | 14:12 |
| josch | it already checks for files owned by non-root users | 14:12 |
| josch | it should also check for files writable by non-root users | 14:12 |
| minute | cool | 14:16 |
| minute | i wonder if factorio would run on the rk3588 | 14:36 |
| gsora | poor rk3588 | 14:40 |
| gsora | crushed by the load of factories :P | 14:40 |
| ryukazou | The most important part is could rk3588 handle the growth of the factory. | 14:59 |
| ryukazou | And my spaghetti conveyor belt | 15:01 |
| + spew (~spew@155.133.15.121) | 15:05 | |
| josch | ryukazou: main bus ftw ;) | 15:18 |
| josch | minute: new reform-tools version pushed and reform-debian-package pipeline is running | 15:34 |
| josch | minute: once reform-tools 1.57 is in the MNT repos, please test and if it works you can spread the news | 15:35 |
| minute | josch: great, thank you | 15:49 |
| hramrach | gustav28: Rebased the port to v3 https://source.mnt.re/hramrach/reform/-/commits/gustav/ctrl-super-swap-modes One thing that comes to mind is that the swap only lasts for as long as the firmware is running, and is unswapped when powered off. That will likely be annoying when not having the keyboard always powered as would often be the case with an external keyboard. | 15:54 |
| minute | hramrach: i wonder if we have any options to save config values to flash on rp2040? | 16:01 |
| minute | https://github.com/raspberrypi/pico-examples/blob/master/flash/program/flash_program.c | 16:02 |
| minute | turns out we can | 16:02 |
| minute | https://www.makermatrix.com/blog/read-and-write-data-with-the-pi-pico-onboard-flash/ | 16:02 |
| minute | so we can make a settings struct with some CRC and store it, or default if CRC no good | 16:03 |
| minute | could also make a little notepad in the keyboard OLED and save files to flash ;) | 16:03 |
| + mark_ (~mjw@gnu.wildebeest.org) | 16:04 | |
| gustav28 | minimalist typewriter if powered via power bank ;) | 16:08 |
| gustav28 | I had this idea of a p(ersistent)config.h/.c that handles settings using key/value pairs with strings as keys | 16:09 |
| gustav28 | I also looked into macOS detection for the ALT-SUPER switch, seems possible but could be confusing to a new user | 16:09 |
| gustav28 | hramrach: cool! I have a mnt source account now so I will open a MR for my changes to keyboard4 this evening | 16:10 |
| + cobra (~cobra@user/Cobra) | 16:23 | |
| hramrach | gustav28: also consider this: https://source.mnt.re/reform/reform/-/merge_requests/72 | 16:24 |
| hramrach | I haven't tested any of this yet | 16:25 |
| ch | minute: i was wondering if we'll need to preserve some config region in the rp2040 flash (and then: how) | 17:51 |
| ch | minute: was mostly pondering this in the context of hardware-specific sysctl fw | 17:51 |
| hramrach | well, see above for the flash writing demo | 17:54 |
| hramrach | but also the hardware-specific sysctl would only need to read the config, not write it - unless the hardware changes as well | 17:55 |
| hramrach | if it's teh screen type then it could potentially get replaced, and then the sysctl would need to be updated | 17:56 |
| hramrach | also maybe the configuration does not need to be stored, a kernel/u-boot driver could tell sysctl which screen is attached based on device tree | 17:56 |
| hramrach | the only problem with that is when rebooting sysctl (most likley during upgrade) | 17:57 |
| minute | josch: how about licensing like this? https://source.mnt.re/reform/reform-handbook/#licenses | 18:02 |
| jfred | speaking of minimalist typewriters, imagine if the keyboard controller could show a basic UI on the primary display and not just on the OLED... | 18:20 |
| jfred | probably not wired up for that :) | 18:20 |
| hramrach | hmm, dodgy. Generally dual licensing is used to resolve problems like these, and if there are few contributors to the stuff it's likely OK. However, sounds like something that could use talking to an actual lawyer that specializes in this. | 18:25 |
| minute | hramrach: yeah, this is only to be able to ship it as a .deb | 18:27 |
| hramrach | for example, would it be clearer to also relicense the schematics as dual OHL/GPL to make sure new renders can be included in the handbook without hassle? | 18:27 |
| minute | so any updates would come through this repo... | 18:27 |
| minute | btw factorio works on rk3588 reform https://mastodon.social/@mntmn/113386088280317754 (short video in reply) | 18:34 |
| hramrach | I see, the OHL is incompatible with GPL because it has extra obligations related to hardware designs that are not part of GPL so relicensing the schematics would be counter-productive | 18:38 |
| hramrach | Then the situation is indeed quite dodgy | 18:39 |
| josch | minute: this means, that the artwork/graphics can either be re-used under the terms of GPL3+, CERN-OHL-S 2 *or* CC-BY-SA 4.0? | 18:48 |
| minute | josch: yeah well i guess CERN-OHL and GPL are not super useful for graphics but yeah, if you want! | 18:49 |
| gustav28 | minute: cool thing factorio works so well! | 18:50 |
| minute | gustav28: yeah, i'm surprised! | 18:51 |
| josch | minute: let me write a debian/copyright for you and then you just tell me whether that's correct or not | 18:51 |
| sigrid | I was gonna ask if DayZ is playable on rk3588 but I'm going to find out soon enough myself | 18:56 |
| hramrach | FWIW if you want to know what Linux thinks is the size of your screen install libdrm-tests, and run modeprint $(cat /sys/class/graphics/fb0/name | sed -e 's/drmfb$//') | 18:57 |
| minute | josch: cool | 19:05 |
| minute | sigrid: ah nice, it looks like we sent out your module today | 19:06 |
| - mjw (QUIT: Killed (mercury.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:14 | |
| * mark_ -> mjw | 19:14 | |
| + Guest7761 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:14 | |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 19:15 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 19:16 | |
| sigrid | minute: \o/ yep, can't wait | 19:20 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 19:30 | |
| + staticbunny (~textual@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 19:32 | |
| staticbunny | I think there is a bug with the keymap for ~ on the pocket. Is this a known issue? | 19:36 |
| staticbunny | it comes out as ";10;27~" or ";2;27~" depending on using control + ESC or Menu + Shift + ESC | 19:38 |
| staticbunny | are the keymaps defined in usb_hid_keys.h or is their an easier way to edit them? | 19:39 |
| josch | minute: is this correct? https://paste.debian.net/1333727/ | 19:41 |
| minute | staticbunny: hmm the right combo should be hyper+shift+esc | 19:42 |
| josch | errrr s/oru/or/ :) | 19:43 |
| staticbunny | @minute yeah sorry i should have said hyper+shift+esc for my second commands. that produces ";10;27~" | 19:43 |
| minute | josch: haha yes if that is fixed then LGTM | 19:44 |
| josch | minute: i hope so -- i'm going to mail ftp-master now and put you in cc | 19:44 |
| staticbunny | im seeing something on the community site about being able to map things with a .xkb file. I wonder if that got borked. I just dont know where to look to see where its bad. | 19:45 |
| minute | staticbunny: in which application does it produce that? | 19:45 |
| minute | staticbunny: for me it produces ~ | 19:45 |
| staticbunny | foot in sway | 19:45 |
| minute | staticbunny: which shell? bash? | 19:45 |
| staticbunny | zsh | 19:45 |
| staticbunny | trying bash | 19:46 |
| staticbunny | same behavior | 19:47 |
| staticbunny | let me try it in a text editor outside the terminal to confirm its not something i did | 19:47 |
| minute | staticbunny: hmmmmm do you have a recent kbd firmware? | 19:47 |
| minute | staticbunny: for me it also works in zsh | 19:47 |
| minute | staticbunny: which trackball button does "wheel mode" for you? | 19:48 |
| staticbunny | does that get updated when doing the firmware update in the repo? | 19:48 |
| minute | staticbunny: currently there isn't a _really nice_ keyboard fw update script like there is for the system controller | 19:51 |
| minute | staticbunny: do you still have the original imx8mplus cpu module? | 19:51 |
| staticbunny | yeah | 19:51 |
| staticbunny | new module is on the way ;-p | 19:51 |
| minute | ok cool | 19:52 |
| josch | okay, mail to ftp-master send | 19:52 |
| minute | staticbunny: you can download the latest kbd firmware build file from here https://source.mnt.re/reform/pocket-reform/-/jobs/6332/artifacts/browse/pocket-reform-keyboard-fw/pocket-hid/build/ | 19:52 |
| josch | lets hope and see :) | 19:52 |
| staticbunny | sure thing | 19:52 |
| minute | staticbunny: and this is the update script, but you have to change the path because it assumes there is a "build/" folder there https://source.mnt.re/reform/pocket-reform/-/blob/main/pocket-reform-keyboard-fw/pocket-hid/flash.sh?ref_type=heads | 19:53 |
| staticbunny | got it, i'll give it a shot | 19:53 |
| minute | cool. you can also use lsusb to confirm that "bus 1" is correct | 19:53 |
| minute | on the rk3588 it is bus 3... | 19:54 |
| minute | josch: little bug report for reform-migrate: bnys tried it on a system that is preinstalled on emmc, to migrate to nvme. no sd card inserted. but i think he didn't specify --emmc. the script doesn't detect that there is no sd card though. it fails near the end with /dev/mmcblk1p1: not a block device. it would be cool if it would check this prerequisite at the beginning | 20:14 |
| minute | josch: ah, with `--emmc` he gets `E: writing to eMMC not supported on MNT Reform 2 with RCORE RK3588 Module`. but it should be, also on pocket with rk3588 | 20:17 |
| staticbunny | also before i forget, any recommendations on an LTE m.2 modem? | 20:35 |
| staticbunny | 5G ones are expensive and the only one with wifi i found is LTE-M/2G | 20:36 |
| josch | minute: the git commit which introduced EMMC_USE=false in the machines.conf for RCORE says that it was done that way because the rk3588 can be soft-bricked by wrong flashing and i think i remember that you told me that we'll change that once you found a way to unbrick accidentally soft-bricked boards? | 20:43 |
| amospalla | Is there an easy way to modify kernel boot paremeters on pocket? | 20:44 |
| minute | josch: it should be fine because the uboot on emmc will prefer the sd card if inserted. we just shouldn't enable EMMC_BOOT yet | 20:46 |
| minute | josch: bnys just tested it by manually changing EMMC_USE to true and it went fine! | 20:46 |
| josch | nice :) | 20:46 |
| minute | amospalla: yes, make extlinux/extlinx.conf on /boot (preferably on an sd card so you can easily recover) | 20:46 |
| amospalla | thank you! | 20:46 |
| minute | amospalla: here's a random example http://dump.mntmn.com/extlinux.conf.txt | 20:47 |
| minute | amospalla: but that's not directly for pocket, so you need to totally adjust it | 20:47 |
| amospalla | I want to try a kernel parameter to set font to 16x32 on boot | 20:47 |
| minute | amospalla: there's a merge request for uboot from hramrach that i merged today | 20:47 |
| minute | amospalla: it does exactly that | 20:47 |
| minute | amospalla: which cpu? | 20:48 |
| amospalla | oh, ok. Stock. | 20:48 |
| minute | amospalla: if you are not afraid to flash this uboot, you can use it: https://source.mnt.re/reform/reform-imx8mp-uboot/-/jobs/6442/artifacts/file/imx8mp-mnt-pocket-reform-flash.bin | 20:49 |
| minute | amospalla: if you brick uboot on emmc you have to recover using a slightly complicated method | 20:49 |
| amospalla | I was trying to do something with the pocket reform font size issue, but I guess this already solves the issue. | 20:49 |
| minute | yeah | 20:50 |
| amospalla | I am afraid of bricking actually, will wait. | 21:19 |
| - staticbunny (QUIT: Ping timeout: 252 seconds) (~textual@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 21:39 | |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 21:40 | |
| hramrach | amospalla: you can also adjust the commandline in flash-kernel | 21:50 |
| josch | minute: in my rush to fix this problem i forgot to "git add" the maintainer script in the last reform-tools release m( | 21:51 |
| hramrach | echo 'LINUX_KERNEL_CMDLINE="fbcon=font:TER16x32"' >> /etc/default/flash-kernel ; update-initramfs -u | 21:51 |
| josch | fixing this up now... | 21:51 |
| hramrach | but if you put something overly broken there you might need to boot from SD card to edit the file again | 21:52 |
| hramrach | not nearly as complicated as recovering from bad u-boot | 21:53 |
| josch | hramrach: should that not be "sudo flash-kernel" instead of "update-initramfs -u"? | 21:53 |
| hramrach | it does do the flash kernel part as well | 21:54 |
| hramrach | I don't know how to use flash-kernel but noticed it's integrated into initramfs through some hook | 21:55 |
| + staticbunny (~textual@76-223-253-78.lightspeed.frokca.sbcglobal.net) | 22:02 | |
| minute | hramrach: thanks, good hint, i always forget about this method | 22:04 |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-78-82-52-4.bbcust.telenor.se) | 22:15 | |
| hramrach | /buffer 1 | 22:25 |
| hramrach | found a rp2040-zero, looks like I can do some flash memory torturing without breaking my keyboard | 22:31 |
| hramrach | cool, it's programmed by mass storage write | 22:34 |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.16.106) | 22:48 | |
| staticbunny | Anyone ever used the T99W175 5G modem? | 22:49 |
| + Ar|stote|is (~linx@149.210.16.106) | 22:49 | |
| minute | staticbunny: unclear what the chip is on that one | 22:51 |
| staticbunny | @minute i think its the snapdragon X55 | 22:53 |
| josch | does somebody remember the environment variable tricks to make opengl based programs maybe work on platforms like a311d that do really have a high enough opengl version? | 22:57 |
| - sts-q (QUIT: Remote host closed the connection) (~sts-q@89.58.73.113) | 23:01 | |
| josch | no luck with MESA_GL_VERSION_OVERRIDE=3.2 MESA_GLSL_VERSION_OVERRIDE=320 :( | 23:06 |
| minute | josch: 3.3, 330 | 23:07 |
| minute | josch: in extreme cases, 4.2, 420, or 4.5, 450 | 23:07 |
| minute | josch: for factorio i used 3.3 | 23:08 |
| minute | josch: a while ago it was also still possible to do PAN_MESA_DEBUG=gl3 | 23:08 |
| + gustav28 (~gustav@c-78-82-52-4.bbcust.telenor.se) | 23:09 | |
| josch | i get it to start with LIBGL_ALWAYS_SOFTWARE=1 MESA_GL_VERSION_OVERRIDE=4.2 | 23:11 |
| josch | but it is unplayably slow of course | 23:11 |
| josch | "it" could be factorio but it could of course be anything else | 23:12 |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-78-82-52-4.bbcust.telenor.se) | 23:21 | |
| kop316 | minute: I just saw the mastodon post. If I have one, can I just paypal/get an invoice for the extra card cost? I ask because actually sending it back from the US will probably cost just as much | 23:28 |
| minute | kop316: you can just keep it then | 23:30 |
| kop316 | minute: you sure? I don't mind doing that really. or if you want the card back, I can send it. I just wasn't sure if the issue was you needed the card back, or if it was you lost the money. | 23:31 |
| minute | kop316: it was our fault for not checking thoroughly, just in case for whoever it's convenient they could send the card back, but if not, no problem | 23:32 |
| kop316 | understood. Sorry that happened..... | 23:32 |
| minute | yeah, they could have at least told us ^^ | 23:34 |
| staticbunny | You got screwed over by the LCD mfg as well recently too right? | 23:34 |
| josch | there is a new item to that story: the cables... | 23:35 |
| staticbunny | sigh | 23:35 |
| minute | well asiaRF didn't screw us over, they just did unconventional packaging | 23:36 |
| staticbunny | well hopefully it doesnt all come out of your pocket. That sounds like their bad | 23:36 |
| staticbunny | seems like the majority of stories i hear about sourcing hardware is a money pit unless you have someone in shenzhen that can be oversight | 23:38 |
| minute | well, 90%+ of our business there went smoothly so far | 23:46 |
| minute | it's a bit of an unusual amount of trouble there atm | 23:46 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!