| chartreuse | Yes, just was meaning in the context why the dtb exists, and that probably isn't the best option to grab for getting the machine name (imx8mq-mnt-reform2) | 00:00 |
|---|---|---|
| chartreuse | Especially if someone adds a custom dtb | 00:00 |
| josch | the -hdmi variant exists not because of the non-free blob but because unconditionally using the -hdmi variant would drive the panel in a non-optimal way | 00:02 |
| josch | the -hdmi variant is not supposed to exist which is also why upstream kernel doesn't have it | 00:02 |
| jackhill | anyone chainloading grub with uboot's EFI support? | 00:02 |
| josch | Lucas Stach wanted to fix this problem for good but alas the free time... | 00:03 |
| chartreuse | For driving the panel in a non-optimal way, I was always getting those messages about the clock rate being wrong when I wasn't using the hdmi one. I only swapped at the moment to test some things out I think | 00:04 |
| josch | chartreuse: details are in /usr/share/flash-kernel/db/all.db above the entry for the HDMI version | 00:04 |
| chartreuse | Ah okay | 00:06 |
| josch | jackhill: yes, ex-parrot experimented with that | 00:06 |
| josch | jackhill: see logs from 2023-08-03 | 00:07 |
| josch | jackhill: and bkeys wanted to make it work | 00:07 |
| chartreuse | Yeah I'll probably swap back to plain for now but will look into that | 00:07 |
| josch | chartreuse: what do you think about this solution: https://source.mnt.re/reform/reform-boundary-uboot/-/jobs/4886/artifacts/browse | 00:07 |
| chartreuse | That could work, maybe a little wasteful making a duplicate copy of that build | 00:08 |
| chartreuse | Just think a dtb independent solution would be nice if someone has their own custom dtbs | 00:08 |
| josch | chartreuse: it's 1.01 MB of waste because gitlab CI does not support symlinks | 00:09 |
| josch | chartreuse: +1000 -- find somebody who makes this happen :) | 00:09 |
| josch | Lucas Stach had plans of making it happen in 2022 ;) | 00:10 |
| chartreuse | Hmm, jack detect isn't working, must have not done a change to the -hdmi version I did to plain, or did something else wrong | 00:12 |
| chartreuse | make a blob free hdmi happen? Or a way of configuring so the display is DCSS with HDMI as well? | 00:13 |
| josch | the second | 00:14 |
| josch | blobs are in there anyways | 00:14 |
| chartreuse | Swapping back to single, I wonder if that performant thing is why my CPU temps seem a couple degrees higher than before | 00:14 |
| josch | Lucas planned to drive them with the eLCDIF controller instead of DCSS. | 00:14 |
| josch | chartreuse: can you confirm that the u-boot flash.bin filename is correct like this? Then i'm tagging this with another release. | 00:15 |
| chartreuse | Sure I'm on single again but I'll swap back after confirming this one is correct too | 00:16 |
| chartreuse | E: Failed to fetch https://source.mnt.re/reform/reform-boundary-uboot/-/jobs/4745/artifacts/raw/imx8mq-mnt-reform2.bin | 00:18 |
| chartreuse | Using reform-check with the patch you gave before | 00:19 |
| josch | are you sure you applied that correctly? it's missing the -flash part | 00:20 |
| chartreuse | https://paste.debian.net/1323675/ | 00:20 |
| josch | oh wait | 00:21 |
| chartreuse | Doesn't have -flash in it | 00:21 |
| josch | apparently i fail at copypasting | 00:21 |
| chartreuse | With -flash it does work though | 00:21 |
| josch | oooooh | 00:21 |
| josch | found it, sorr | 00:21 |
| josch | thank you for checking | 00:22 |
| chartreuse | I'll reboot into dual display, though that might not work as 4745 is an older artifact since my uboot is a year old | 00:22 |
| chartreuse | I'll be updating that after though | 00:22 |
| chartreuse | Yeah, E: Failed to fetch https://source.mnt.re/reform/reform-boundary-uboot/-/jobs/4745/artifacts/raw/imx8mq-mnt-reform2-hdmi-flash.bin 404 Not Found [IP: 49.13.21.45 443] | 00:25 |
| chartreuse | But that's likely due to my older uboot? Or that it's picking an older artifact or such | 00:25 |
| josch | chartreuse: no, that's because i need to release a new reform-tools which has the next u-boot release tag in it | 00:26 |
| chartreuse | Ah okay | 00:26 |
| chartreuse | But the filename seems correct even with the wrong tag | 00:26 |
| josch | chartreuse: i would be greatful though if you could test that once it's in the repo :) | 00:26 |
| chartreuse | No worries, just let me know when. | 00:27 |
| josch | yes, if i fix the job id, the download works | 00:27 |
| josch | chartreuse: uff we were so productive today -- makes me very happy! :) | 00:27 |
| chartreuse | Heh, fixing tiny bugs that caused very confusing symptoms | 00:27 |
| chartreuse | Mainly me using an old system image, at least it wasn't my older v2 install, that probably would have been worse. I decided back then to start fresh on a v3 rather than upgrade since i had a custom kernel built on v2 | 00:28 |
| chartreuse | I wonder if reform-display-config could be sped up greatly by calling flash-kernel instead of update-initramfs -u. Since it's just a dtb change rather than needing to build a new initram-fs img. Unless I'm missing something | 00:32 |
| josch | chartreuse: i hope that i'm not missing any corner cases, but i think you are right | 00:35 |
| josch | but then i also wonder how often people switch between modes usually and whether that's worth the risk of potentially breaking something? | 00:36 |
| chartreuse | It does seem to work at least for me and only takes seconds rather than over a minute with an initramfs rebuild | 00:37 |
| josch | chartreuse: are you switching display mode often? | 00:38 |
| chartreuse | I'm not sure really if it would or wouldn't break anything. The update-initramfs really only seems to be being used for the post hooks that call flash-kernel. There's no module differences between single and dual. | 00:39 |
| chartreuse | I've never even actually used the hdmi output once | 00:39 |
| josch | ah okay | 00:39 |
| chartreuse | I set to dual recently to maybe test it out and see if how it was just in case | 00:39 |
| josch | then i rather leave things as they are | 00:39 |
| josch | no need to experiment with something for which there is no immediate need by anybody | 00:39 |
| chartreuse | But chaning the call to just flash-kernel does work for me, I'll continue testing it I guess. It only affects people with imx8mq ones and no one is complaining it seems | 00:40 |
| josch | chartreuse: if you tell me that you plan to switch the display mode often, then i'll make the change | 00:41 |
| chartreuse | I don't but if I ever have to do a presentation it'd be more a nice to have. I just don't want to break someone else without testing it more I guess | 00:41 |
| josch | it might've just me being young and stupid in january 2022 | 00:43 |
| josch | +been | 00:43 |
| josch | but right now i think just calling flash-kernel should be enough | 00:43 |
| chartreuse | I will say it is working just fine and I've just swapped back and forth a few times and confirmed with /proc/device-tree/model each time | 00:44 |
| chartreuse | And perhaps it being fast would encourage people that do use HDMI to stick on single unless they need it if there is performance benefits to the internal | 00:45 |
| josch | personally i have never noticed it | 00:45 |
| josch | but i mostly hang out on the terminal | 00:45 |
| josch | i needed hdmi mode for slides -- but those were created with latex, so no animations either :D | 00:45 |
| - vagrantc (QUIT: Ping timeout: 260 seconds) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 00:46 | |
| josch | chartreuse: i should go to bed now | 00:46 |
| josch | i'll finish reform-tools tomorrow | 00:46 |
| chartreuse | No worries, thanks for all the help | 00:47 |
| josch | minute: i could wait with the next reform-tools release until after your a311d nvme experiments and then ship with the new bananapi u-boot? | 00:47 |
| josch | chartreuse: thank you for your help as well! really cool :) | 00:48 |
| chartreuse | I do think that single maybe is faster in firefox when scrolling with a lot of images, but it's hard to really judge | 00:48 |
| chartreuse | Not as fast as when I used to be able to force hardware acceleration on (though it was glitchy) | 00:48 |
| josch | good night _o/ | 00:48 |
| chartreuse | gnight | 00:48 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:20) | 01:12 | |
| + reform30678 (~thciprian@161.97.200.147) | 01:18 | |
| - reform30678 (QUIT: Client Quit) (~thciprian@161.97.200.147) | 01:20 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 01:26 | |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@2601:380:837e:24e0:b413:1241:5e3:eebd) | 01:28 | |
| + vagrantc_ (~vagrant@2600:3c01:e000:21:7:77:0:50) | 01:32 | |
| - vagrantc (QUIT: Ping timeout: 248 seconds) (~vagrant@2600:3c01:e000:21:7:77:0:20) | 01:34 | |
| * vagrantc_ -> vagrantc | 01:34 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 01:44 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:02 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:05 | |
| + jacobk (~quassel@2603:3013:103:ef00:b413:1241:5e3:eebd) | 02:13 | |
| + robin (~robin@user/terpri) | 02:17 | |
| - mhoye (QUIT: Quit: Lost terminal) (~mhoye@li319-32.members.linode.com) | 02:29 | |
| - jacobk (QUIT: Ping timeout: 276 seconds) (~quassel@2603:3013:103:ef00:b413:1241:5e3:eebd) | 02:47 | |
| - chomwitt (QUIT: Ping timeout: 265 seconds) (~chomwitt@2a02:587:7a22:1700:1ac0:4dff:fedb:a3f1) | 03:32 | |
| + chomwitt (~chomwitt@2a02:587:7a1d:e400:1ac0:4dff:fedb:a3f1) | 03:34 | |
| - nsc (QUIT: Ping timeout: 255 seconds) (~nicolas@i5C74DCE7.versanet.de) | 03:52 | |
| + nsc (~nicolas@216-98-142-46.pool.kielnet.net) | 03:54 | |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@45.134.140.153) | 04:12 | |
| - jn (QUIT: Ping timeout: 265 seconds) (~quassel@user/jn/x-3390946) | 04:38 | |
| + jn (~quassel@2a0a:a549:89c2:0:20d:b9ff:fe49:15fc) | 04:38 | |
| - jn (QUIT: Changing host) (~quassel@2a0a:a549:89c2:0:20d:b9ff:fe49:15fc) | 04:38 | |
| + jn (~quassel@user/jn/x-3390946) | 04:38 | |
| - chomwitt (QUIT: Ping timeout: 248 seconds) (~chomwitt@2a02:587:7a1d:e400:1ac0:4dff:fedb:a3f1) | 04:47 | |
| + jacobk (~quassel@2601:380:837e:24e0:b413:1241:5e3:eebd) | 05:12 | |
| - cobra (QUIT: Ping timeout: 265 seconds) (~cobra@user/Cobra) | 05:20 | |
| + cobra (~cobra@user/Cobra) | 05:20 | |
| violet | chartreuse: josch: i dont really understand what is the purpose of switching between the two device trees on imx8mq anyway? | 05:44 |
| violet | i always boot dual-display | 05:45 |
| violet | i use hdmi a lot, because i use hdmi when im at my desk and internal display when im in bed | 05:47 |
| violet | i would definitely not want to reboot for that | 05:48 |
| josch | violet: this is the explanation: https://sources.debian.org/src/flash-kernel/3.108/db/all.db/#L1192 | 07:11 |
| josch | minute: do you remember a deeper reason for not defaulting to the hdmi dtb? | 07:12 |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@2601:380:837e:24e0:b413:1241:5e3:eebd) | 07:37 | |
| violet | interesting, i havent noticed performance problems with it | 07:44 |
| violet | meaning, my browser is slower than the rest of the display stack, and anime seems to play fine | 07:45 |
| josch | violet: you watch videos in your browser on imx8mq and it works for you? | 08:23 |
| josch | especially anime was always a pain on imx8mq because everything is h265 encoded and the hardware decoder produced lots of nasty artifacts when trying to decode h265... | 08:23 |
| + Jonas___ (~Jonas@82-65-231-86.subs.proxad.net) | 09:10 | |
| violet | yeah for anime i pre-transcode it to h264 | 09:47 |
| violet | but actually, watching youtube in the browser works fine enough for me | 09:48 |
| violet | my trick is that i use mobile youtube | 09:48 |
| violet | and i limit the resolution to 480p | 09:48 |
| violet | this works ok but has some framedrops. its fine for videos where that doesnt matter like a tutorial or talk or something | 09:49 |
| violet | for smooth playback i will use mobile youtube to search for a video but then download to watch outside the browser | 09:49 |
| violet | https://m.youtube.com/?persist_app=1&app=m <- if you click this link you will persistently switch your youtube to load mobile youtube and auto redirect normal youtube links to mobile | 09:52 |
| violet | i think you can switch back with https://youtube.com/?persist_app=1&app=d or https://youtube.com/?persist_app=1&app=desktop | 09:52 |
| violet | the video controls are slightly weird like this because its designed for touch screen. but it loads so much faster | 09:53 |
| violet | and then i use ublock origin to block the recommended videos bar on the right so i dont have to see that | 09:53 |
| josch | violet: using a youtube proxy like invidious would also be an option for you, no? | 10:16 |
| violet | i find them hit-or-miss for whether they're actually working | 10:16 |
| josch | violet: what is your ublock rule to hide the recommended videos? i only managed to make it a gray box but not to make it disappear | 10:18 |
| violet | unfortunately for me it's the same, just an empty box | 10:18 |
| josch | ah okay | 10:19 |
| violet | for the desktop version of the website there is a great browser extension "DF YouTube (Distraction Free)" which can filter out recommendation bar, end-cards, home page stuff, comments, and so on. however you like it, with a list of checkboxes to choose what you want and dont want. but it does nothing for mobile youtube | 10:22 |
| josch | thank you for the recommendation | 10:23 |
| josch | i usually just copypaste the youtube url into my terminal to use it with mpv XD | 10:23 |
| violet | yeah i love that mpv can do that | 10:23 |
| vkoskiv | Ehm, I'll take my 35 simultaneously open YouTube tabs, thank you very much | 10:23 |
| josch | vkoskiv: but you are also on the a311d and not the imx8mq ;) | 10:24 |
| vkoskiv | I mostly hoard youtube tabs on my desktop, 4GB isn't enough for much of that :D | 10:32 |
| josch | 3.6 GB :) | 10:33 |
| - _rm (QUIT: Changing host) (johan@halin.me) | 10:44 | |
| + mjw (~mjw@gnu.wildebeest.org) | 10:50 | |
| violet | i think just add 64 gigs of swap, surely there are no problems with this (sarcasm) | 11:03 |
| josch | kidding aside, the machine is my daily driver with 8 GB of swap on nvme | 11:16 |
| violet | yeah 8GB swap works well for me too | 11:22 |
| violet | im am curious to see how the imx8mp feels. im working on setting up my pocket right now | 11:23 |
| violet | im doing gentoo on it though because ive been wanting to do that anyway | 11:25 |
| violet | https://source.mnt.re/vimja/mnt-reform-overlay already has the kernel packaged right now which i think is neat. i hope to package reform-tools as well though because there's a lot in there that would be useful on a gentoo system as well | 11:27 |
| josch | violet: it'd be useful to find out how we can improve sharing reform-tools across multiple distros | 11:28 |
| josch | violet: right now it is super debian-centric | 11:28 |
| josch | but ideally, the debian-parts should be factored out into the packaging and the source itself should be distribution agnostic | 11:28 |
| violet | yeah i noticed that while looking at it earlier. i'll share what my first draft of packaging it looks like once i have it | 11:29 |
| minute | josch: the hdmi dtb makes the internal display run at a lower freq and might be slightly lower performance because of lcdif vs dcss | 11:39 |
| josch | minute: but what does "lower performance" mean in practice? | 11:40 |
| josch | minute: i have used the hdmi dtb for nearly a year and didn't notice a problem | 11:40 |
| - mjw (QUIT: Ping timeout: 248 seconds) (~mjw@gnu.wildebeest.org) | 11:53 | |
| * Guest3536 -> mjw | 12:01 | |
| + reform23345 (~andypiper@89.36.117.58) | 13:01 | |
| * reform23345 -> andypiper | 13:01 | |
| andypiper | channel | 13:04 |
| - andypiper (QUIT: Client Quit) (~andypiper@89.36.117.58) | 13:05 | |
| - q66 (QUIT: Quit: WeeChat 4.3.5) (~q66@q66.moe) | 13:17 | |
| + q66 (~q66@q66.moe) | 13:18 | |
| + chomwitt (~chomwitt@athedsl-13986.home.otenet.gr) | 13:23 | |
| grimmware | Shipping notification! | 13:24 |
| grimmware | Looks like I’m getting a new computer for *checks date* my partner’s birthday | 13:25 |
| gsora | perfect gift! | 13:25 |
| josch | exactly, then you need to buy a second one for yourself ;) | 13:28 |
| - chomwitt (QUIT: Ping timeout: 248 seconds) (~chomwitt@athedsl-13986.home.otenet.gr) | 13:39 | |
| chartreuse | The solution is clearly more mnt reforms in the world | 13:40 |
| + chomwitt (~chomwitt@2a02:587:7a1b:1000:1ac0:4dff:fedb:a3f1) | 13:41 | |
| minute | grimmware: awesome | 13:42 |
| - chomwitt (QUIT: Ping timeout: 252 seconds) (~chomwitt@2a02:587:7a1b:1000:1ac0:4dff:fedb:a3f1) | 13:54 | |
| + chomwitt (~chomwitt@2a02:587:7a0e:5a00:1ac0:4dff:fedb:a3f1) | 13:56 | |
| - mtm (QUIT: Ping timeout: 252 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:04 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:06 | |
| - chomwitt (QUIT: Ping timeout: 252 seconds) (~chomwitt@2a02:587:7a0e:5a00:1ac0:4dff:fedb:a3f1) | 14:06 | |
| + chomwitt (~chomwitt@2a02:587:7a04:4400:1ac0:4dff:fedb:a3f1) | 14:08 | |
| - chomwitt (QUIT: Ping timeout: 276 seconds) (~chomwitt@2a02:587:7a04:4400:1ac0:4dff:fedb:a3f1) | 14:20 | |
| + chomwitt (~chomwitt@2a02:587:7a24:2300:1ac0:4dff:fedb:a3f1) | 14:21 | |
| - chomwitt (QUIT: Ping timeout: 252 seconds) (~chomwitt@2a02:587:7a24:2300:1ac0:4dff:fedb:a3f1) | 14:25 | |
| + chomwitt (~chomwitt@ppp-94-67-201-184.home.otenet.gr) | 14:30 | |
| - chomwitt (QUIT: Ping timeout: 248 seconds) (~chomwitt@ppp-94-67-201-184.home.otenet.gr) | 14:39 | |
| chartreuse | Got some fancy 2.5mm Zirconium oxide bearings in, now the question is do I remove the ones in my existing trackball cup, or do I print new one | 14:51 |
| + chomwitt (~chomwitt@ppp-94-67-201-184.home.otenet.gr) | 14:53 | |
| josch | chartreuse: print a new one so that you have a backup | 14:53 |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-150-69.tukw.qwest.net) | 14:53 | |
| chartreuse | I've got a few backups in the form of the original non-bearing one as well as some of my intermediate prints with different bearing angles, as well as more 2.5mm steel bearings still. But yeah probably should do it | 14:54 |
| chartreuse | Just don't have any more of the nice lake blue filament I used last time, have a roll of white loaded at the moment | 14:54 |
| - Jonas___ (QUIT: Ping timeout: 252 seconds) (~Jonas@82-65-231-86.subs.proxad.net) | 14:55 | |
| + colinsane (~colinunin@97-113-150-69.tukw.qwest.net) | 14:55 | |
| josch | choosing the color is as problematic as choosing variable names in code | 14:55 |
| josch | i'd also want to give my trackball buttons a new color but what to choose... | 14:56 |
| minute | https://source.mnt.re/reform/pocket-reform/-/merge_requests/2 | 15:01 |
| minute | thinking how to roll out a sysctl update on pocket. i guess a bash script at first... | 15:01 |
| minute | very good that we have picotool in apt now, thanks again for that josch | 15:02 |
| kop316 | Hello! so I have noticed that wifi performance in the pocket reform hasn't been too good, i basically have to be in the same room as my AP. is this known/is there a way i can help debug? | 15:07 |
| chartreuse | Some people on the forum were mentioning thermal issues with the wifi chipset possibly | 15:09 |
| kop316 | hmm ok. yeah it seems to be ok then after a while it doesn't work well | 15:11 |
| minute | yay, CI worked on first try https://source.mnt.re/reform/pocket-reform/-/jobs/4889/artifacts/browse/pocket-reform-sysctl-fw/build/ | 15:12 |
| minute | kop316: yep i just wanted to say that people reported that putting thermal pad on the wifi chip helps | 15:13 |
| chartreuse | I don't have a pocket so don't know the construction, but could see about either adding a small heatsink or thermal pad or such on it | 15:13 |
| minute | 2mm thick thermal pad is the easiest solution as almost the whole back of the device is a heatsink | 15:13 |
| kop316 | minute: thanks! I just saw the forum post, so I guess I will be getting a thermal pad | 15:14 |
| minute | it's a PCB which has a large exposed copper area inside | 15:14 |
| minute | i switched to bpi cm4 (a311d processor) in my own pocket and it doesn't seem to have this thermal issue | 15:15 |
| minute | (but the wifi has other issues, it has less bandwidth) | 15:15 |
| mtm | hmm, I assume I'll need to remove the thermal pads once I install the rk3588 in my Pocket (or maybe just get/make a new cover) | 15:15 |
| + bkeys (~Thunderbi@45.134.140.153) | 15:23 | |
| bkeys | josch: How do I build the dts files in the Reform kernel into dtb files? | 15:23 |
| chartreuse | The way I did it was start the build of the linux package in reform-debian-pacakges, then cancel it (since it takes forever on a reform), the navigate into the linux folder, copy the system config in to .config then run make dtbs. Josch might have a better way | 15:26 |
| minute | hm weird, i just tested the server-built firmware and now i can't see the rp2040 in lsusb anymore | 15:52 |
| minute | but it otherwise works correctly | 15:52 |
| + andypiper (~andypiper@89.36.117.58) | 16:25 | |
| andypiper | Quick check, what size of SIM card is the slot in the Pocket Reform? from the schematic it looks like microSIM https://www.molex.com/en-us/products/part-detail/788000001 | 16:29 |
| ch | yes, and https://www.crowdsupply.com/mnt/pocket-reform also says Micro SIM card slot | 16:34 |
| andypiper | ah perfect, thanks for confirming, I missed that on the CS page (and the manual does not specify apart from the schematic information AFAICT) | 16:42 |
| - chomwitt (QUIT: Ping timeout: 248 seconds) (~chomwitt@ppp-94-67-201-184.home.otenet.gr) | 17:02 | |
| minute | ok, figured out the usb issue | 17:19 |
| minute | one shall not forget to git submodule --init in pico-sdk | 17:20 |
| andypiper | haha | 17:29 |
| minute | otherwise usb support will just be left out | 17:37 |
| minute | https://community.mnt.re/t/pocket-reform-system-controller-update-looking-for-testers/2277 | 17:37 |
| + mark_ (~mjw@gnu.wildebeest.org) | 17:55 | |
| bkeys | Is there a guide somewhere to get battery status to show up within Linux? I still haven't done that on my Reform | 18:25 |
| - andypiper (QUIT: Quit: My device has gone to sleep. Zzzz…) (~andypiper@89.36.117.58) | 18:26 | |
| josch | minute: hm? how did picotool in apt make your day better today? | 18:26 |
| minute | josch: it was trivial to write this https://source.mnt.re/reform/pocket-reform/-/blob/2ca8b45ebc58aab160c8e86d806b197bc3df9ef8/scripts/update-sysctl-firmware.sh | 18:28 |
| josch | bkeys: you patch the makefile in the way that i have indicated a few days ago -- i can copypaste you details in a few hours | 18:28 |
| bkeys | Okay, maybe then I can include it in my fedora kernel build | 18:29 |
| josch | minute: pico-sdk is available as a debian package as well btw :) | 18:29 |
| josch | minute: maybe then you can avoid git submodule update problems | 18:30 |
| minute | oh hm | 18:44 |
| chartreuse | bkeys: If the .dts file is in an appropriate directory under linux/arch/arm64/boot/dts/ (imx8mq goes in freescale) and added to the Makefile if it wasn't already in it the kernel will build it when compiling | 19:02 |
| * mjw -> Guest9247 | 19:03 | |
| - Guest9247 (QUIT: Killed (osmium.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:03 | |
| * mark_ -> mjw | 19:03 | |
| + Guest9247 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 19:04 | |
| + chomwitt (~chomwitt@2a02:587:7a0d:7000:1ac0:4dff:fedb:a3f1) | 19:06 | |
| - chomwitt (QUIT: Ping timeout: 276 seconds) (~chomwitt@2a02:587:7a0d:7000:1ac0:4dff:fedb:a3f1) | 19:19 | |
| chartreuse | Decide to see what if and which factorio would run under box64. 1.0 and later require newer GL features and fails, even with the overrides, it'll work under software though but slow. I've gotten up to version 0.16.51 working though, a little laggy on the IMX8MQ but would likely be usable on the other CPU options | 19:25 |
| chartreuse | 0.17.79, the next newest one for download on their archives is the same as 1.0, requiring newer features and won't launch (and crashes with the overrides) | 19:25 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 19:25 | |
| chartreuse | Wonder if the RK board would run the newer versions properly, might also have a chance at decent performance | 19:30 |
| + andypiper (~andypiper@45.148.12.75) | 19:42 | |
| josch | chartreuse: factorio is available for arm64 macos -- maybe there is hope for them to compile us some binaries for arm64 linux as well at some point :) | 20:10 |
| - andypiper (QUIT: Ping timeout: 252 seconds) (~andypiper@45.148.12.75) | 20:54 | |
| - kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 20:57 | |
| + kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de) | 20:59 | |
| violet | if i thermal pad the wifi chip, does anyone know what thickness should i be using | 21:08 |
| violet | for the thermal pad | 21:08 |
| + mhoye (~mhoye@li319-32.members.linode.com) | 21:08 | |
| violet | let me try to find the forum post | 21:09 |
| minute | violet: normally the thermal pad is ~2mm thick... not sure by heart how thick/elevated the wifi module is on the imx8mp | 21:09 |
| mhoye | Airport security people get a bit excitable about computers you can see into, I've just learned. | 21:09 |
| mhoye | all ok but whew | 21:09 |
| violet | thats so funny | 21:10 |
| violet | you can see it better! the parts are right there! | 21:10 |
| minute | mhoye: haha yeah i've been through that too | 21:11 |
| violet | ah yes in the forum thread they say 1.8mm so i will try 2mm since that seems easy to buy | 21:12 |
| mhoye | hey, do we have a favorite CAD tool in here? | 21:16 |
| minute | FreeCAD, OpenSCAD here | 21:16 |
| minute | and of course KiCAD ;) | 21:16 |
| mhoye | good, thank you. I'm going to get some bits of metal plate cut, and i want to do the work on this machine. | 21:17 |
| minute | nice | 21:17 |
| + chomwitt (~chomwitt@2a02:587:7a0d:7000:1ac0:4dff:fedb:a3f1) | 21:21 | |
| violet | also during long compiles i definitely start to thermal throttle quite a bit on the pocket, which i suppose makes sense. it seems like the 2GHz overclock really works quite well as a boost-clock during a normal using-a-computer workload though | 21:25 |
| violet | and i am very glad that it throttles instead of burning itself up | 21:26 |
| violet | in my ambient temperatures (25-26C or so) it seems to level off mostly 1.6GHz with some dips to 1.2GHz as necessary | 21:27 |
| mhoye | Last week I learned that navy warships have "battle breaker panels" - big switches that lock circuit breakers closed during combat, because electrical systems maybe setting themselves on fire is a better outcome than being dead in the water. | 21:27 |
| mhoye | I'm glad it's a problem I don't have. | 21:28 |
| violet | wow | 21:28 |
| - chomwitt (QUIT: Ping timeout: 276 seconds) (~chomwitt@2a02:587:7a0d:7000:1ac0:4dff:fedb:a3f1) | 21:28 | |
| violet | also probably i can improve this by putting the pocket in front of a fan while it's compiling | 21:30 |
| minute | :0 | 21:31 |
| violet | if that isnt enough on its own i figure i could remove the side panels to let the fan's air flow through the interior too | 21:31 |
| violet | so it can reach more of the exposed copper | 21:31 |
| minute | indeed | 21:32 |
| mhoye | hmmm. Both OpenSCAD and FreeCad do _not_ like being started in Sway. "failing to create EGLContext, error 3009" | 21:50 |
| mhoye | Anyone seen this? | 21:52 |
| mhoye | wonder if I shouldn't be ignoring that unable to run mnt hardware setup error on boot. | 21:54 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 22:21 | |
| Twodisbetter | minute: my pocket reform was just delivered but I am not there for 2 more weeks. Something to look forward to though. Time to order the RK | 22:22 |
| minute | Twodisbetter: yay! | 22:26 |
| minute | mhoye: ah yes. try starting them from a terminal where you do this beforehand: export QT_QPA_PLATFORM=xcb | 22:27 |
| chartreuse | mhoye: I'm able to start FreeCAD and OpenSCAD just fine still | 22:39 |
| + andypiper (~andypiper@45.148.12.75) | 22:52 | |
| + chomwitt (~chomwitt@2a02:587:7a0d:7000:1ac0:4dff:fedb:a3f1) | 23:06 | |
| bkeys | I will state here publicly I will get the Reform kernel working with Fedora on the Reform if it's the last thing I do on this earth | 23:24 |
| josch | o0 | 23:25 |
| bkeys | I have just been at this for days, and I have made great progress on the effort; I'm almost to the end of it, I'm just dedicated to solving this | 23:26 |
| josch | hehe, i remember that feeling :) | 23:26 |
| bkeys | I have the patches applied, the dts in place to be built, the config is being applied; if it would just wrap it all up in a .rpm file it would theoretically work | 23:31 |
| bkeys | I tried a local build on my server and it got some weird seg fault, but the error explicitly said this is an OS/hardware issue and not a software issue. I was trying to build a Fedora image on rhel9 so maybe that was the issue. I'm gonna see if the copr builders using fedora chroots can solve the issue | 23:32 |
| bkeys | The CentOS Stream SIG is working on minimal raw images at my request so it's very possible that combined with my work the reform could run enterprise linux | 23:33 |
| bkeys | Come to think of it as well having -stream on the Reform would be good news for both standalone and rack reform as well | 23:39 |
| - chomwitt (QUIT: Ping timeout: 248 seconds) (~chomwitt@2a02:587:7a0d:7000:1ac0:4dff:fedb:a3f1) | 23:47 | |
| andypiper | Interested in testing a fedora build, if I can do so non-destructively (I assume I'd be able to boot from SD card to do so at some point, although with uboot I don't know how that all fits together) | 23:51 |
| andypiper | thank you for making the effort! | 23:51 |
| * andypiper -> andypiper[afk] | 23:55 | |
| - andypiper[afk] (QUIT: Quit: My device has gone to sleep. Zzzz…) (~andypiper@45.148.12.75) | 23:55 | |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!