| - S0rin (QUIT: Ping timeout: 260 seconds) (~S0rin@user/s0rin) | 00:34 | |
| + S0rin (~S0rin@user/s0rin) | 00:35 | |
| + klardotsh (~klardotsh@172.56.64.102) | 00:35 | |
| + cwebber (~user@user/cwebber) | 00:39 | |
| - mtm (QUIT: Ping timeout: 252 seconds) (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 02:03 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 03:56 | |
| + mtm (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 04:09 | |
| - natalie (QUIT: Ping timeout: 260 seconds) (~natalie@user/natalie) | 08:39 | |
| + natalie (~natalie@user/natalie) | 09:36 | |
| + qwer (~qwer@78-80-19-34.customers.tmcz.cz) | 10:14 | |
| - mjw (QUIT: Killed (NickServ (GHOST command used by wielaard!~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440))) (~mark@gnu.wildebeest.org) | 11:40 | |
| * wielaard -> mjw | 11:40 | |
| + mark__ (~mark@gnu.wildebeest.org) | 11:40 | |
| C_Keen | hi! I flashed the rescue image v3 to the emmc and turned off the boot switch, now the reform boots into the rescue system, I wanted to switch the boot mode back to sd card momentarily but it seems that reform-boot-config expects not to be run from the recue system itself. changing the /etc/fstab manually to point to mmcblk0p2 and p1 for / and /boot results in /boot mounted from sd card but not | 13:13 |
|---|---|---|
| C_Keen | / | 13:13 |
| C_Keen | so is there a way to boot from sd the root fs without having to flip the switch again? | 13:14 |
| C_Keen | ah I missed updating initramfs | 13:16 |
| * C_Keen -> C-Keen | 13:19 | |
| - qwer (QUIT: Ping timeout: 252 seconds) (~qwer@78-80-19-34.customers.tmcz.cz) | 13:33 | |
| - mtm (QUIT: Ping timeout: 268 seconds) (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 14:04 | |
| Kooda | Has anyone attempted to port the patches to linux 5.19 yet? | 14:51 |
| Kooda | I’ve been asked to give 5.19 + some patches a try, to fix the etnaviv issues | 14:51 |
| - qbit (QUIT: Quit: WeeChat 3.6) (~qbit@h.suah.dev) | 14:54 | |
| josch | Kooda: usually sigrid has the patches ready for the latest kernel version for void but I don't see 5.19 there yet: https://github.com/ftrvxmtrx/void-packages/tree/HEAD/srcpkgs/mntreform2-kernel/files | 14:56 |
| + qbit (~qbit@h.suah.dev) | 14:58 | |
| Kooda | Ok ^^ | 14:59 |
| Kooda | I guess I’ll have to give that a try. | 14:59 |
| Kooda | Btw, any reason we use boot.scr instead of extlinux-like config? I know it’s available on other devices with u-boot | 15:00 |
| josch | Kooda: the uboot version that 99% of reforms out there are running is one where extlinux booting is disabled. The current version of our uboot git has it enabled https://source.mnt.re/reform/reform-boundary-uboot but there is no way to tell who has that installed and who has not. | 15:02 |
| josch | Kooda: so if we would switch from boot.scr to extlinux.conf today, systems would become unbootable. | 15:03 |
| Kooda | I see | 15:03 |
| Kooda | Thanks fot the info! :) | 15:04 |
| Kooda | for* | 15:04 |
| sigrid | there's also 9front that people use on reforms, not only linux | 15:23 |
| sigrid | it knows nothing about extlinux.conf :) | 15:24 |
| Kooda | I would think having a text based format would be beneficial for them too | 15:25 |
| josch | boot.scr is also text-based. I think the advantage of extlinux.conf is, that it's declarative and not turing-complete. | 15:26 |
| sigrid | for us it'd be additional work, i don't know much about extlinux.conf and how it's better than boot.scr | 15:26 |
| Kooda | Ah my bad, I thought boot.scr was binary. | 15:27 |
| sigrid | it is, the source is text | 15:27 |
| josch | the resulting boot.scr has a small binary header but otherwise is readable text | 15:27 |
| + mtm (~mtm@c-73-27-62-116.hsd1.fl.comcast.net) | 15:32 | |
| - Nulo (QUIT: Ping timeout: 268 seconds) (~Nulo@user/nulo) | 17:24 | |
| cinap_lenrek | boot.scr is binary | 17:33 |
| cinap_lenrek | yes, its text, but wrapped in some braindead uboot header with a checksum | 17:34 |
| cinap_lenrek | so you can't just open it in a text editor and modify it (????? WHY? ????) | 17:34 |
| + Nulo (~Nulo@user/nulo) | 17:34 | |
| cinap_lenrek | wonders of uboot | 17:34 |
| cinap_lenrek | so what we do now is actually just jank ANOTHER text file into memory (our plan9.ini essentially)) from boot.scr to a known fixed address and thats it | 17:35 |
| cinap_lenrek | now that file you can just modify like a normal person | 17:35 |
| minute | :3 | 18:02 |
| cinap_lenrek | sorry minute for the noise. | 18:06 |
| minute | meanwhile https://shop.mntmn.com/products/mnt-reform-ls1028a-module-preorder | 18:07 |
| cinap_lenrek | \o/ | 18:07 |
| minute | need 70 people with surplus cash to make this real though | 18:08 |
| Kooda | Oooh | 18:10 |
| C-Keen | I need to save up for the pocket reform :) | 18:11 |
| C-Keen | 30% more power usage is quite something | 18:11 |
| cinap_lenrek | fuck it | 18:11 |
| cinap_lenrek | i'll get one | 18:11 |
| sigrid | :O | 18:15 |
| - cwebber (QUIT: Ping timeout: 252 seconds) (~user@user/cwebber) | 18:15 | |
| minute | cinap_lenrek: nice | 18:16 |
| Kooda | I wonder how much hotter it gets. My Reform already goes above 60°C sometimes and it’s quite uncomfortable for my hands ^^' | 18:18 |
| minute | in my experience it doesn't really get much hotter | 18:18 |
| Kooda | nice :) | 18:19 |
| sknebel | PCIe over HDMI is an interesting trick. I guess that means the second lane is not easily available internally? (since it probably goes directly card-edge->socket on the mainboard?) | 18:29 |
| cinap_lenrek | minute: überweisung is raus! viel glück das es klappt! | 18:29 |
| minute | cinap_lenrek: danke!!!! | 18:30 |
| minute | sknebel: yep | 18:30 |
| sknebel | also a bit surprised by the price difference for the 16GB version - does that need a much fancier board (more layers etc?)? or are memory chips that bad if you buy small amounts? | 18:33 |
| minute | the chips are just a lot more expensive | 18:38 |
| C-Keen | it's a bad time for independent HW projects :/ | 18:39 |
| sknebel | yeah, just noticed that too looking some up on mouser - of course its double size chips, not more of them (and either way they make a bigger part then I thought - holy heck is the difference between "I want a stick of desktop RAM" and "I want individual chips" large) | 18:39 |
| Kooda | I still haven’t made the Reform version of my keyboard because µCs are so hard to come by :( | 18:45 |
| Kooda | I guess I should switch to RP2040 too… | 18:45 |
| C-Keen | or an esp32 :) | 18:46 |
| Kooda | Equally overkill hehe | 18:47 |
| C-Keen | but not affected by the chip shortage | 18:47 |
| minute | Kooda: i now have a working keyboard on rp2040, although for the pocket version. but you can copy from that | 18:47 |
| Kooda | I should look into that yeah, because the rp doesn’t have program memory. | 18:48 |
| C-Keen | does anyone know whether technomancy got further with his atreus PCB for the reform? | 18:48 |
| minute | i also recently brought up the OLED on it | 18:48 |
| Kooda | nice | 18:48 |
| minute | so it's not far from being useful for big reform as well | 18:48 |
| sknebel | you were planning to redesign the touchbad/trackball to use RP2040 too, right? I forgot to check what's on the board I screwed in yesterday :D | 18:51 |
| Kooda | everything is still using atmega32u(2|4) | 18:54 |
| Kooda | as far as I know | 18:54 |
| Kooda | (oh and the nxp chip as the system controler) | 18:54 |
| C-Keen | the LPC | 18:55 |
| josch | LS1028A order placed -- lets make this happen! | 18:58 |
| C-Keen | <3 | 18:58 |
| minute | josch: thanks!! | 18:59 |
| - mjw (QUIT: Killed (NickServ (GHOST command used by mark__!~mark@gnu.wildebeest.org))) (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 19:00 | |
| * mark__ -> mjw | 19:00 | |
| + wielaard (~mjw_@2001:1c06:2488:1400:9e5c:8eff:fe8f:a440) | 19:00 | |
| bkeys | josch: wait we can order the LS1028A for the Reform now? | 19:22 |
| C-Keen | bkeys: yes | 19:26 |
| bkeys | Link? | 19:26 |
| bkeys | Sadly right now I am not in a spot where I will have the money for it, that and my Reform is still out getting repaired | 19:26 |
| cinap_lenrek | minute → meanwhile https://shop.mntmn.com/products/mnt-reform-ls1028a-module-preorder | 19:26 |
| minute | bkeys: deadline to order is end of the year | 19:28 |
| bkeys | Alright, I will probably be able to cough it up by then | 19:28 |
| bkeys | minute: As someone who likely has this in his hands right now; would you say it's an upgrade? | 19:31 |
| bkeys | In terms of CPU performance, ignoring RAM | 19:32 |
| minute | it depends on the workload. it's def. an upgrade for emacs and javascript | 19:33 |
| minute | for compiling, not really, because on imx8mq it scales well to 4 cores | 19:33 |
| bkeys | Mmhm, and GPU performance? | 19:33 |
| minute | i would say similar or slightly worse | 19:33 |
| bkeys | Oh and does the LS1028A get any UEFI? | 19:34 |
| minute | oh, this i don't know. | 19:34 |
| bkeys | It seems like for my use case the soquartz SoM would be a better fit since I believe it has better GPU | 19:34 |
| - klardotsh (QUIT: Quit: nyaa~) (~klardotsh@172.56.64.102) | 19:39 | |
| minute | bkeys: possibly yes. i need to still get that to work though. | 19:41 |
| Kooda | Isn’t someone already using the CM4 adapter? | 19:42 |
| minute | i did, but i couldn't bring up mipi on the rpi yet | 19:47 |
| minute | and soquartz didn't work at all so far, also didn't work for jacqueline | 19:47 |
| Kooda | damn | 19:48 |
| bkeys | CM4 would have vulkan, wouldn't it? | 19:48 |
| minute | yes | 19:48 |
| Kooda | SOQuartz has too, iirc | 19:48 |
| minute | rcm4 bringup completion is in my plan in a few weeks | 19:49 |
| bkeys | Having GPU acceleration and UEFI would be awesome and CM4 should have UEFI | 19:49 |
| minute | cm4 has some drawbacks though, only 1x pcie and only usb2 | 19:50 |
| * jnerula__ -> jnerula | 20:34 | |
| josch | minute: I'm currently stuffing my reform with 3d printed parts to hold additonal hardware -- is the design of the new heatsink already finished? will it be bigger or placed differently compared to the current one? And where do you place the eDP adapter board? | 21:47 |
| josch | the LS1028A heatsink that is | 21:47 |
| + doctorhoo (~hanno@194-18-252-127-no2005.tbcn.telia.com) | 21:50 | |
| minute | josch: that's not yet finalized! interested in where you are stuffing things btw | 21:51 |
| minute | the edp adapter is tiny though | 21:51 |
| josch | minute: i'll post about it on the forum once i'm done -- but it'll be no big deal, in the worst case i just re-print the parts with some modifications next year :) | 21:51 |
| josch | i'm putting a huawei e3372 usb lte modem next to the batteries on the side with the sd-card slot. Between that and the ethernet port there is the enough space for my usb uart adapter. I'm putting the adapter directly into the case and then i can plug a micro-usb cable into my reform to connect to it via serial. The other side then carries the laird antenna (thank you!!) as intended as well as a mini-usb | 21:55 |
| josch | hub. Next to the trackball there are then additional usb ports for some more hardware like bluetooth. | 21:55 |
| minute | josch: cool! so far i think the heatsink will be similar/same place and size | 21:55 |
| minute | cool @ internal usb! | 21:55 |
| josch | i love that there is so much free volume to throw additional hardware in :) | 21:56 |
| minute | just put 2x battery board 2.1 into my reform for daily testing | 21:56 |
| minute | (with protection for each cell) | 21:56 |
| minute | had some problems to get it going because i ordered milliohm resistors from mouser but they were actually megaohms... | 21:56 |
| josch | minute: will battery board 2.1 come with some protection against the contacts touching each other and producing a short circuit? | 21:57 |
| Kooda | josch: ooh I might steal the usb uart idea :D | 21:57 |
| Kooda | woops at resistors x) | 21:57 |
| josch | Kooda: the alternative idea was to have the usb uart adapter on the outside and connect to GND/RX/TX via a 2.5 mm audio jack | 21:58 |
| Kooda | josch: yeah I was planing on doing that | 21:58 |
| minute | josch: i am thinking to include one of the 3d printed or tape protections. but also the new (clone) battery clips that we have now are not so close together anymore | 21:58 |
| josch | cool! | 22:01 |
| mtm | minute: the resistors were only off nine orders of magnitude, that's not a problem, right? ;-) | 22:22 |
| khm | the main cm4 drawback is you need a time machine to acquire one | 22:39 |
| minute | khm: well, that's the other issue, yeah | 22:54 |
| vkoskiv_ | Trying to think of cool things to do with the OLED | 22:57 |
| minute | bongo cat | 22:57 |
| vkoskiv_ | I'd like to do some more keyboard and possibly LPC fw work in general | 22:57 |
| vkoskiv_ | Bongo cat! | 22:57 |
| vkoskiv_ | I want that as an option in the menu | 22:58 |
| minute | you could write a display driver for it and use it as a second screen... | 22:58 |
| minute | vkoskiv_: jacqueline did it already (bongo cat) | 22:58 |
| vkoskiv_ | That'd be a fun project, I haven't written a linux driver | 22:58 |
| minute | a simple framebuffer driver is surprisingly straight forward. but the downside is that in this case you would need to send raw HID packets to update the display. i guess it would work easier from userspace | 22:59 |
| vkoskiv_ | Trying to understand how the current oled logic works. Is the MNT logo integrated into the font? | 23:00 |
| vkoskiv_ | The one shown on power-on | 23:00 |
| minute | yes. | 23:00 |
| minute | i have some tools/examples to dump pixels to the display if you need | 23:01 |
| vkoskiv_ | Ah :D I didn't bother trying to blit out the font data to see | 23:01 |
| vkoskiv_ | I saw the oled string writer script, messed around with that. | 23:01 |
| minute | vkoskiv_: https://source.mnt.re/reform/reform/-/tree/master/reform2-keyboard-fw/kbdgfx-demo | 23:02 |
| minute | ah ok | 23:02 |
| vkoskiv_ | Thanks, I missed this, will give it a go. | 23:02 |
| vkoskiv_ | Even just a linux program to draw some fun stats on there would be neat, I think. | 23:02 |
| vkoskiv_ | Like a waybar-esque thing where there are a few modules you can choose from and rearrange, and it updates every N seconds | 23:03 |
| minute | one could probably run Xvfb and then grab its contents with a slightly modified tool in a loop | 23:03 |
| minute | vkoskiv_: yeah, a pluggable stats thing would be great. | 23:03 |
| minute | vkoskiv_: graphic EQ | 23:03 |
| minute | network/cpu/ram graph | 23:04 |
| vkoskiv_ | The sine demo is super cool :D | 23:11 |
| vkoskiv_ | 60fps, looks like. | 23:11 |
| minute | haha yeah | 23:13 |
| khm | I just want a cartoon face that looks progressively more distressed as system load increases | 23:14 |
| khm | like the dude at the bottom of the wolf3d ui | 23:14 |
| minute | hehe | 23:14 |
| khm | has anyone had any luck using nxp's provided driver instead of etnaviv? apparently it provides newer gl versions | 23:15 |
| minute | khm: works only with NXP kernel | 23:15 |
| minute | so you would need to tweak/port our dts over to that | 23:16 |
| minute | they do some things differently than in mainline | 23:16 |
| vkoskiv_ | minute: Do you have time in the near future to go over the MRs for the reform repo? | 23:25 |
| vkoskiv_ | I think it would make sense to ignore draw commands coming from the SoM when any of the menus are open. That way anything drawn from linux won't override the important stuff shown | 23:26 |
| vkoskiv_ | I'll try and implement taht. | 23:26 |
| minute | vkoskiv_: yes, i just noticed that i have to take care of some MRs | 23:27 |
| minute | i'm writing it in my list | 23:27 |
| sigrid | I would be very interested in spi for lpc one in particular | 23:28 |
| vkoskiv_ | The battery status stuff? | 23:28 |
| sigrid | yeah | 23:28 |
| vkoskiv_ | I'd love that in the mainline fw too, yeah. | 23:28 |
| vkoskiv_ | That's a big one, getting proper battery monitoring and stuff | 23:29 |
| vkoskiv_ | Dang, I'll just leave this sine wave demo running as I do other stuff :D | 23:29 |
| sigrid | once it gets merged it will make sense to write an spi driver for 9front to expose that info | 23:29 |
| minute | yes, i will merge it | 23:29 |
| sigrid | also to do proper power off | 23:29 |
| minute | i just need to tweak some of the code, tiny nits | 23:29 |
| vkoskiv_ | If you have any comments on my lil PR that disables the circle+1 when the SoM is on, let me know. I can tweak that stuff. | 23:30 |
| minute | sure thing | 23:31 |
| minute | it's in my todo list now to take care of these | 23:31 |
| vkoskiv_ | Cool, thank | 23:33 |
| vkoskiv_ | I just copypasted a few more draw_sine() calls and it looks pretty. | 23:34 |
| minute | haha | 23:36 |
| vkoskiv_ | Hah, it seems like I managed to crash the kbd somehow | 23:50 |
| vkoskiv_ | With the kbdgfx program | 23:50 |
| vkoskiv_ | I got the bootup animation and now the keyboard doesn't keyboard | 23:50 |
| vkoskiv_ | I'll click reset... | 23:51 |
| vkoskiv_ | Nope, reset button doesn't bring it back either | 23:52 |
| vkoskiv_ | Can I just yank the UART and USB cables while it's running or should I power cycle the machine | 23:53 |
| vkoskiv_ | A dfu-programmer command from the kbd fw flash script got it back | 23:56 |
| vkoskiv_ | (I tried messing around with the programming switch too, accidentally put it into dfu mode) | 23:56 |
| vkoskiv_ | Looks like udev left /dev/hidraw0 in there as inactive, and now the oled lives at /dev/hidraw4 | 23:57 |
| vkoskiv_ | Not sure if that's intended | 23:57 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!