| - mjw (QUIT: Ping timeout: 246 seconds) (~mjw@gnu.wildebeest.org) | 00:42 | |
| - cobra (QUIT: Ping timeout: 244 seconds) (~cobra@user/Cobra) | 01:00 | |
| + cobra (~cobra@user/Cobra) | 01:05 | |
| - nsc (QUIT: Ping timeout: 252 seconds) (~nicolas@149-98-142-46.pool.kielnet.net) | 01:25 | |
| + nsc (~nicolas@228-97-142-46.pool.kielnet.net) | 01:27 | |
| - spew (QUIT: Quit: spew) (~spew@201.141.99.170) | 01:33 | |
| + jacobk (~quassel@2603:8080:b200:7b02::b71) | 01:35 | |
| henesy | minute: moonlight running on imx8 pocket reform | 02:01 |
|---|---|---|
| henesy | havent tried on big reform | 02:01 |
| - Ar|stote|is (QUIT: Ping timeout: 244 seconds) (~linx@149.210.20.157) | 02:14 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 02:14 | |
| + Ar|stote|is (~linx@149.210.21.253) | 02:18 | |
| - Ar|stote|is (QUIT: Ping timeout: 252 seconds) (~linx@149.210.21.253) | 03:14 | |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@2603:8080:b200:7b02::b71) | 03:15 | |
| + Ar|stote|is (~linx@149.210.22.228) | 03:18 | |
| - paperManu (QUIT: Ping timeout: 265 seconds) (~paperManu@198.16.214.40) | 03:23 | |
| - Ar|stote|is (QUIT: Ping timeout: 276 seconds) (~linx@149.210.22.228) | 03:27 | |
| + Ar|stote|is (~linx@149.210.21.62) | 03:31 | |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.21.62) | 03:36 | |
| + Ar|stote|is (~linx@149.210.21.62) | 03:39 | |
| + cobra_ (~cobra@user/Cobra) | 03:45 | |
| - cobra (QUIT: Read error: Connection reset by peer) (~cobra@user/Cobra) | 03:45 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 03:57 | |
| - sts-q (QUIT: Ping timeout: 260 seconds) (~sts-q@89.58.73.112) | 04:32 | |
| + sts-q (~sts-q@89.58.73.83) | 04:34 | |
| - Ar|stote|is (QUIT: Ping timeout: 252 seconds) (~linx@149.210.21.62) | 04:42 | |
| + Ar|stote|is (~linx@149.210.21.172) | 04:47 | |
| - CandidCadaver (QUIT: Remote host closed the connection) (~aloo_shu@85.51.16.143) | 05:05 | |
| - Ar|stote|is (QUIT: Ping timeout: 252 seconds) (~linx@149.210.21.172) | 06:21 | |
| + Ar|stote|is (~linx@149.210.21.152) | 06:26 | |
| - Ar|stote|is (QUIT: Ping timeout: 245 seconds) (~linx@149.210.21.152) | 07:06 | |
| + Ar|stote|is (~linx@149.210.22.1) | 07:11 | |
| - edenist (QUIT: Ping timeout: 245 seconds) (edenist@rdns.robotprofessor.net) | 07:28 | |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 07:52 | |
| + edenist (~edenist@rdns.robotprofessor.net) | 07:54 | |
| + gustav28 (~gustav@c-78-82-52-34.bbcust.telenor.se) | 08:34 | |
| Twodisbetter | ch: if you have the time, I could use some help getting WiFi working again on bookworm. | 09:46 |
| Twodisbetter | https://community.mnt.re/t/getting-wifi-working-on-bookworm/2673 | 09:46 |
| Twodisbetter | No worries if you don't. | 09:46 |
| - amospalla (QUIT: Quit: WeeChat 4.4.2) (~jordi@user/amospalla) | 09:49 | |
| - Ar|stote|is (QUIT: Ping timeout: 276 seconds) (~linx@149.210.22.1) | 09:52 | |
| + Ar|stote|is (~linx@149.210.21.99) | 09:57 | |
| + amospalla (~jordi@user/amospalla) | 09:59 | |
| henesy | is bookworm latest? i thought trixie was | 09:59 |
| josch | henesy: Twodisbetter is trying out using packages from reform.d.n on the pocket with imx8mplus | 10:02 |
| - ehmry (PART: https://quassel-irc.org - Chat comfortably. Anywhere.) (~quassel@217.155.30.169) | 10:02 | |
| henesy | reform.d.n? | 10:04 |
| henesy | ive been out of the loop | 10:04 |
| henesy | is that a usenet | 10:04 |
| josch | henesy: reform.debian.net is the place where i publish the contents of the MNT debian repo built for bookworm | 10:05 |
| josch | it's a website | 10:05 |
| henesy | ah okay yes | 10:05 |
| henesy | is bookworm what unstable is/becoming/from? | 10:06 |
| + mjw (~mjw@gnu.wildebeest.org) | 10:07 | |
| josch | bookworm is the name of the current debian stable release | 10:07 |
| henesy | gotcha | 10:07 |
| josch | what is currently in unstable migrates to testing which will then become the new debian stable release called trixie | 10:07 |
| henesy | oh so this is for running stable instead of unstable | 10:07 |
| henesy | that would be dope | 10:07 |
| henesy | gotcha | 10:07 |
| josch | yes, i'm doing that on my a311d reform | 10:07 |
| josch | and will soon be doing that on rk3588 as well | 10:08 |
| henesy | dope | 10:08 |
| henesy | will rk be coming to crowd supply or do i have to order direct from mnt? | 10:08 |
| henesy | i would be nice to toss that in the big reform | 10:08 |
| henesy | and run stable | 10:09 |
| josch | henesy: https://mntre.com/reform-irc-logs/2024-10-10.log.html#t15:52:03 | 10:10 |
| henesy | dope | 10:11 |
| henesy | rip that power switch right after | 10:11 |
| henesy | pushing the moonlight repo build up to blob storage here and ill post it in a bit | 10:12 |
| henesy | the prebuilt debian repo for moonlight-qt didnt work for me | 10:12 |
| henesy | the binary is included ofc | 10:13 |
| henesy | it's just a chunky repo | 10:13 |
| henesy | but i can get a comfy 30fps on wifi on imx8 pocket and sound goes fine | 10:13 |
| henesy | it's also built with the 'slow video' option for systems that lack dedicated gpu which is supposed to help | 10:14 |
| henesy | but it's documented in the moonlight docs | 10:15 |
| josch | henesy: if you have some experience in packaging stuff for debian, you could propose the package on mentors.debian.net so that it gets included in the official repos | 10:15 |
| henesy | i have maintained packages in the past | 10:15 |
| henesy | debian i usually shy away from bc they have high standards ;) | 10:15 |
| henesy | but not a bad idea | 10:15 |
| henesy | i have some slackbuilds under my name tho i bet they are busted these days | 10:16 |
| henesy | but it's slackware so maybe not! | 10:16 |
| henesy | back in the early pi days i rolled a bunch of debs for various pi stuff that wasnt officially supported | 10:17 |
| henesy | but never got any upstreamed just ran my own repo | 10:18 |
| henesy | https://staticssa.blob.core.windows.net/static/moonlight-qt_pocket-imx8.tgz | 10:18 |
| henesy | have a 600MB brick | 10:18 |
| henesy | ive only tested with the sunbeam server, not the nvidia one | 10:19 |
| henesy | oops it looks like i have been contacted about my slackbuilds and missed t hem | 10:23 |
| henesy | i wonder if slackware would boot on the reform -- it has an aarch64, i assume i could just use the debian kernel? | 10:24 |
| ch | yeah that could work | 10:27 |
| henesy | hmmm maybe a project for the weekend | 10:28 |
| josch | there are/were arch linux people who just extract the debian kernel *.deb in an AUR package to make it work | 10:28 |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 11:04 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 11:05 | |
| henesy | neat! | 11:19 |
| gustav28 | pondering whether to buy a reform now or when I have more money, heh. I guess there is not much reason to get the a311d now that the rk3588 is out? ideally, anything I build should work on a311d but it sure is nice to have a fast machine for other slow software though. | 12:38 |
| gustav28 | *soon out | 12:38 |
| + paperManu (~paperManu@198.16.214.40) | 12:46 | |
| josch | gustav28: some people on the forum are selling theirs, maybe that's a better fit for your budget? | 12:48 |
| minute | josch: gustav28: also wanted to suggest that | 12:49 |
| josch | the one on kleinanzeigen.de seems to have been sold, nice! :) | 12:49 |
| - xha (QUIT: Quit: WeeChat 4.4.2) (~xha@user/xha) | 13:03 | |
| gustav28 | my long term plan is to replace my macbook pro, so i would really want the new motherboard + ideally very clean/unused hw. currently thinking if I should go with a311d as a start and then when I'm ready to sell my mbp maybe there are even newer SoMs out. or just wait a bit and order a fully specced out rk3588. | 13:05 |
| gustav28 | also seems like most are sold on the forum? | 13:06 |
| josch | personally, i can recommend a311d. I'm using it in my reform as my daily driver and only computer at home. | 13:08 |
| amospalla | Twodisbetter: what is the output of 'find /usr/lib/firmware -type f | grep -E "bdwlan30|otp30|qwlan30|cfg.dat|qcom_cfg.ini"' on your pocket? | 13:10 |
| gustav28 | .. but you are getting a rk josch? :) | 13:11 |
| Twodisbetter | amospalla: I'll tell you when I can. At work now and have wait for a lull in the workload. | 13:13 |
| * mjw -> Guest8814 | 13:13 | |
| - Guest8814 (QUIT: Killed (zinc.libera.chat (Nickname regained by services))) (~mjw@gnu.wildebeest.org) | 13:13 | |
| * Guest5911 -> mjw | 13:13 | |
| + Guest8814 (~mjw@gnu.wildebeest.org) | 13:14 | |
| amospalla | Twodisbetter: when you can. Reading forums I think I understood something I did not before. I think that by manually installing the deb package in the NEW state it should just work. | 13:15 |
| josch | gustav28: i'm getting an rk3588 but not as my main system but so that i can finally hack on stuff like debian-installer or reform-tools scripts properly without risking to destroy my running system: https://floss.social/@josch/113045524553564309 | 13:15 |
| amospalla | So there is no need at all to manually mv files. | 13:15 |
| gustav28 | josch ah, I see! | 13:18 |
| gustav28 | I guess there must be so many upgrading to rk3588 now and letting their old modules collect dust. would be awesome with an option to buy reform without module and a forum thread for selling/buying modules only | 13:19 |
| gsora | great to see debian sponsoring your work josch! | 13:20 |
| josch | thank you -- i hope that i can soon sink a lot more time into reform stuff again but all the normal things seem to be running smoothly and sadly i cannot help much with pocket related stuff... | 13:20 |
| josch | gustav28: i have a couple of people who are using very old desktops right now. For them, the hopefully upcoming desktop-reform with a311d will be a perfect fit and can just re-use existing modules. | 13:22 |
| josch | I also have a totally broken (i literally have to hit it before it turns on and there is no screen, nor keyboard, nor hdd in it...) thinkpad as my media center which i also want to switch out for the desktop-reform with a311d | 13:22 |
| josch | the a311d is completely capable to play all my video collection (1080p max) | 13:23 |
| + xha (~xha@user/xha) | 13:27 | |
| josch | sorry, it's called Reform Desktop not desktop reform: https://mntre.com/reform-irc-logs/2024-09-27.log.html#t16:21:33 | 13:28 |
| gustav28 | yea that's a good way to reuse the modules! | 13:30 |
| gustav28 | Reform Touch :o | 13:30 |
| gustav28 | wait, the a311d cannot use an ssd? | 13:34 |
| josch | it can | 13:37 |
| josch | but on the classic reform you have to remove your wifi card to put an ssd in the slot | 13:37 |
| josch | using an adapter from minipice to m.2 | 13:37 |
| gustav28 | ah I see | 13:41 |
| grimmware | Uh, my UPS tracking for the rcore says it’s being returned at the request of the sender. | 14:04 |
| minute | grimmware: waaa | 14:11 |
| minute | grimmware: oh wait, UK? | 14:11 |
| minute | shiiiit this slipped through the cracks, UPS wanted invoices for 2 UK shipments for some reason, looks like i sent them too late (maybe) | 14:16 |
| minute | no idea why they want extra copies | 14:16 |
| grimmware | Yeah UK | 14:17 |
| grimmware | How much do you want to bet they’re going to charge me for import twice :P | 14:17 |
| minute | sorry for that :/ maybe they can still make it happen, otherwise we have to send them out again when they arrive back here... no idea why they're bouncing them so quickly | 14:17 |
| minute | did they already charge you?? | 14:17 |
| grimmware | Yeah | 14:18 |
| grimmware | Honestly dude I am amazed that you can keep on top of making products let alone all the arcane bullshit required to ship internationally so please don’t sweat it, it’ll get figured out :) | 14:20 |
| - gsora (QUIT: Ping timeout: 252 seconds) (~gsora@user/gsora) | 14:21 | |
| + gsora (~gsora@user/gsora) | 14:29 | |
| - switchy (QUIT: Ping timeout: 272 seconds) (~switchy@mechboards/switchy) | 15:09 | |
| xha | so i just got my rcore in pocket working, and WOAH that is damn fast :D | 15:30 |
| xha | but would be too cool if everything would just work | 15:30 |
| xha | i have some DMA errors in dmesg when trying to use ethernet | 15:30 |
| xha | does anyone have a quick fix for that? | 15:30 |
| - robin (QUIT: Read error: Connection reset by peer) (~robin@user/terpri) | 15:37 | |
| + robin (~robin@user/terpri) | 15:37 | |
| + CandidCadaver (~aloo_shu@85.51.16.211) | 15:41 | |
| minute | xha: hmm afaik they are "normal" until there's a connection | 15:55 |
| minute | xha: sooo it sounds like the package with the missing components arrived already? | 15:57 |
| xha | yes it did :) | 16:00 |
| xha | hooking it up to serial rn so that i can paste the errors | 16:01 |
| minute | xha: so it sounds like ethernet connection does not work at all, or...? | 16:04 |
| xha | yes, the ethernet connection does not work at all | 16:06 |
| minute | xha: do you have another internet connection on the device? because i would like to get the output of some commands that might need to be installed, like ethtool | 16:09 |
| xha | that is the problem, i don't have another internet connection on that device lol | 16:10 |
| xha | gimme a sec to figure out what the problem is | 16:10 |
| minute | xha: do you have a usb ethernet adapter perhaps? | 16:10 |
| minute | or usb wifi... | 16:11 |
| xha | i don't have either of them | 16:11 |
| xha | is 1500000 the right baud rate for rk3588? | 16:11 |
| xha | because i am just getting weird stuff | 16:11 |
| minute | yes that's the right baud | 16:12 |
| minute | sometimes usb uart adapters have some trouble with that speed | 16:12 |
| minute | you can try power cycling the adapter | 16:13 |
| minute | xha: can you try in networkmanager to force the negotiation to manual and speed to 100 mbit or even 10 mbit on the adapter? | 16:13 |
| minute | on the ethernet i mean | 16:13 |
| minute | ah, looks like we ship mii-tool | 16:15 |
| minute | xha: can you try: sudo mii-tool -v end0 (with cable plugged in and without) | 16:17 |
| minute | xha: and can you try: sudo mii-tool -R end0 | 16:17 |
| xha | in all of these cases i am getting EINVAL ("SIOCGMIIPHY on 'end0' failed: Invalid argument") | 16:25 |
| xha | tried -v and -R both with cable pluggin in and without | 16:25 |
| xha | i would like to send you the dmesg spam... but i am cannot get serial working, is there anything special? i am using a openbsd host but it should work the same way, right? | 16:27 |
| xha | hmm, the handbooks says "Wire up a 3.3V USB-to-UART adapter to 3 of the following | 16:30 |
| xha | pins of connector J17" but connector J17 is the "Hard Powerdown Interface"? | 16:30 |
| xha | it should be J16 i think. | 16:31 |
| mjw | cool, looks like the pocket reform is finally shipping from crowdsupply | 16:34 |
| mjw | Unfortunately, the 4G LTE Cellular M.2 Module module you ordered is no longer available and we don’t have an alternate part to offer you. | 16:34 |
| mjw | Do people know an alternative and where to buy? | 16:34 |
| jfred | mjw: Which variant did you order, the US model or the European model? (I suspect in either case you can probably buy the same part from an electronics distributor even if Crowd Supply doesn't have it) | 16:44 |
| mjw | US | 16:45 |
| mjw | Looked at shop.mntre.com but also not there | 16:45 |
| ch | EU didnt have lte modules iirc | 16:45 |
| ch | digikey has EU models, probably they have US models if you're looking in the US shop | 16:46 |
| bluerise | uhh, RK3588 shipping announcement | 16:47 |
| jfred | hm, I was hoping I could see exactly which model was in mine via lsusb but it's a bit vague... I think it's the EM06-A though? | 16:48 |
| mjw | thanks, crowdsupply doesn't give much info, just: Crowd Supply Basics 4G LTE Cellular M.2 Module (certified for North America) | 16:50 |
| Twodisbetter | amospalla:-/usr/lib/firmware/qcacld2/qwlan30.bin | 16:52 |
| Twodisbetter | /usr/lib/firmware/qcacld2/bdwlan30.bin | 16:52 |
| Twodisbetter | /usr/lib/firmware/qcacld2/otp30.bin | 16:52 |
| Twodisbetter | /usr/lib/firmware/wlan/qcacld2/qcom_cfg.ini | 16:52 |
| Twodisbetter | /usr/lib/firmware/wlan/qcacld2/cfg.dat | 16:52 |
| amospalla | Twodisbetter: thank you. | 16:53 |
| jfred | EM06-A seems to be out of stock at lots of places (would like to confirm that's the right one though) but asiarf seems to have it: https://asiarf.com/product/quectel-em06-iot-m2m-optimized-lte-a-cat-6-m-2-module-em06-alar/ | 16:53 |
| amospalla | Twodisbetter: running https://file.amospalla.es/misc/mv_reform_firmware.sh should do it. | 16:56 |
| amospalla | Take note you did this manually somewhere so you can undo if needed anytime. | 16:56 |
| jfred | some similar-looking models on Digikey seem to be marked obsolete though (EM06 series, but with GPS as well), so maybe there's a newer product available that's replaced it and that's why it's out of stock in many places? | 16:57 |
| xha | minute: these are the right connectors though? https://xha.li/paste/out.png | 16:58 |
| xha | obviously i am not running the pocket reform like this with the module out, this is just for the photo | 16:59 |
| xha | i am unsure honesly. it's also not a linux problem and probably also not one with my uart-to-usb connector. it supports up to 3M baud | 17:00 |
| xha | i mean it's not a openbsd problem. i also tried it with a linux host | 17:01 |
| xha | i am also pretty sure that the uart-to-usb connector works because i've used it for other boards previously. with the pocket i am just getting output like "wfwwwvvvvvvfffwffffffffffffffffffffffffffffffffffffffffff3fffffffffff3fffff3ffffff3fffff3fffffffffdfdf1ddDDDDF3DDDDDDDDDF" all the time, and from experience, that is usually happening when i configure the wrong baud rate. | 17:03 |
| hramrach | they use some RM500U here https://www.discomp.cz/turris-omnia-5g-kit_d127623.html with OpenWRT so it's likely possible to use with Linux in general but did not test it | 17:08 |
| hramrach | and EP06-E here https://www.discomp.cz/turris-omnia-lte-kit_d113352.html | 17:09 |
| minute | xha: yes, those are the right ones | 17:11 |
| minute | xha: does the adapter handle 1.5mbaud? | 17:12 |
| xha | it says "up to 3M baud", i guess that includes 1.5? | 17:12 |
| xha | i'll try to look it up though. | 17:12 |
| minute | ok yeah sounds like it | 17:12 |
| hramrach | IIRc those modules require to be connected over USB becuase PCI drivers are not available | 17:13 |
| minute | xha: btw thanks about mii-tool output, so then the error is already with the connection to the PHY it seems | 17:14 |
| xha | can you reproduce the issue at your end or is it a me problem? | 17:15 |
| minute | xha: i've def. seen this issue before but i'm still chasing a module here where it shows up | 17:15 |
| ch | hramrach: be careful, the m.2 slots on the pocket mobo are not equivalent; wwan slot has usb connected, the ssd slot pcie | 17:16 |
| xha | hmm, ok. well, i've looked it up, it might actually be the adapter. it's really old (>10years) and there are some forums which say that it's the most faked adapter ever. and my dad (who bought it back then) likes to buy faked stuff. so i guess i'll buy a better one. | 17:17 |
| xha | is there a way to force the baud rate to a lower one? | 17:18 |
| xha | on the rcore side. because i know this adapter works at lower rates. | 17:18 |
| minute | xha: no way that i know of. also, maybe you can temporarily get a usb ethernet adapter? | 17:18 |
| xha | yeah, i can. i'll just steal^Wborrow one from work tomorrow. | 17:19 |
| minute | ok cool | 17:26 |
| hramrach | ch: yes, these should work with USB connection IIRC | 17:44 |
| hramrach | but they are tested for that device only YMMV | 17:45 |
| jfred | hmm, maybe when I get the rcore module for my big reform I'll put the a311d into my pocket reform after all. the extra speed boost along with still having a WWAN card would be nice | 17:46 |
| minute | xha: ok i have news on ethernet | 17:46 |
| minute | xha: i found a module with the issue, _but_ it went away when booting with the cable unplugged, und plugging the cable in after boot. how did you try it? | 17:47 |
| minute | xha: was the cable connected at boot or only later? | 17:47 |
| midfavila | uh oh, wwan on linux | 17:47 |
| josch | it's not that bad... i used to drive an ericsson f3507g in my old thinkpad back in the day when they came with a sim card slot | 17:49 |
| josch | these days, i either use a usb-stick which abstracts everything for me and just gives me a cdc ether interface via usb (so my linux has to do nothing) | 17:49 |
| josch | or i use a quectel ec25-e in a box that runs openwrt | 17:49 |
| jfred | it's been working pretty well for me so far with modemmanager - and as I'm using a very metered data-only SIM it's nice that I can set a hard limit on data transfer before it auto-disconnects | 17:50 |
| minute | josch: quick lazy question, about your wayfire test: is it written in C++ or C? | 17:50 |
| minute | firedecor test i mean | 17:50 |
| josch | minute: c++ | 17:50 |
| minute | ok then nevermind | 17:50 |
| josch | :D | 17:50 |
| josch | what was your idea? | 17:50 |
| minute | the error looked a bit like trying to use C++ stuff from C | 17:50 |
| josch | ah kk | 17:50 |
| minute | i.e. C++ name mangling | 17:51 |
| josch | i think the linker only sees the mangled names | 17:51 |
| josch | and does not de-mangle them when throwing errors | 17:51 |
| josch | so you have to sent them through c++filt | 17:51 |
| minute | ah yeah what ch wrote there makes sense, that that could be an export that is missing | 17:52 |
| minute | btw there is #wayfire and people were pretty helpful there in the past (longer time ago) | 17:52 |
| xha | minute: both cases aren't working for me | 17:53 |
| josch | thank you! yes, i considered just asking via a github issue :) | 17:53 |
| xha | the dmesg errors with DMA are always there, and they are a ton. i'll paste them tomorrow. | 17:53 |
| xha | also, there where two cases where the keyboard would not work. rebooting fixed it. is that a problem, or just related to the uart connection with the adapter? | 17:55 |
| jfred | josch: I was always jealous of the thinkpads with sim card slots but never was in a position to use one until recently (dang carriers and their expensive data plans) | 17:56 |
| minute | xha: ok so you tried multiple reboots and eth doesn't work, alright. i still need to find another unit where that happens, or we can exchange yours already | 18:00 |
| minute | but might need just a little dts tweak or similar | 18:00 |
| minute | xha: and yes, usb hub not working sometimes is a known issue, still looking into that (possibly usb hub needs a reset) | 18:01 |
| xha | yes i tried multiple reboots already | 18:03 |
| midfavila | josch fair enough. it's been a while since i've used a linux box with WAN. was an internal modem and i had to spend hours trying to figure out the right BIOS settings | 18:04 |
| xha | minute: we can exchange devices, i wouldn't mind. but i can also test. whatever is easier for you. | 18:04 |
| midfavila | then i had to write a script to bit-bang the piece of junk into an active state and wehhh | 18:04 |
| josch | oh wow o0 | 18:05 |
| minute | xha: ok thanks. i found another one here that exhibits the issue, lets see if it persists so i can look for a real fix | 18:06 |
| xha | two other issues: (1) sometimes the pocket reform won't turn on (keyboard has the animation that it does, but it doesn't) and (2) keyboard just didn't turn off when shutting down (hyper+enter b says that the reform is on, while it is actually off). probably related to the usb hub thing though? | 18:09 |
| xha | ACTION afk | 18:09 |
| minute | xha: 1. yes that's a known issue on pocket, unrelated to rk3588 2. yes | 18:24 |
| bluerise | minute: what's the boot_merger RKBOOT/RK3588MINIALL.ini necessary for? so that you can change some settings? | 18:25 |
| bluerise | OpenBSD's ports build simply passes these to the u-boot build, no need for RK's tool: | 18:25 |
| bluerise | RK3588_TPL= rk3588_ddr_lp4_2112MHz_lp5_2400MHz_v1.16.bin | 18:25 |
| bluerise | RK3588_BL31= rk3588_bl31_v1.45.elf | 18:25 |
| minute | bluerise: idk, copypasta from collabora... it might not be needed | 18:26 |
| bluerise | ah ok | 18:26 |
| noam | I need to get those steel side covers installed ASAP, there's four bugs visible through the bottom acrylic inside my laptop woo | 18:32 |
| noam | a few of which I'm fairly sure died | 18:32 |
| noam | because they went in series with the cells | 18:32 |
| minute | xha: ok i found one where ethernet didn't work the first 4(!) reboots and then it started working from the 5th reboot | 18:37 |
| minute | noam: debug it! | 18:38 |
| noam | minute: I will, but they'll just go back in :/ | 18:39 |
| minute | eek | 18:39 |
| noam | I need to check when the order is gonna ship out >_< | 18:39 |
| noam | hahaha perfect | 18:40 |
| noam | > Subject: MNT Research Shop Shipment Notification #R598... | 18:40 |
| noam | it shipped out this morning :D | 18:40 |
| noam | Or... huh, no, that didn't include the steel bars? I coulda sworn I ordered them at the same time :/ | 18:41 |
| + spew (~spew@201.141.99.170) | 18:47 | |
| minute | noam: maybe you had a combined order? the notification won't include it possibly | 18:54 |
| minute | rk3588 gmac0 dts node is really complex... it even has AXI parameters | 18:55 |
| noam | minute: looks like I decided not to order it and then forgot | 18:55 |
| noam | >_< | 18:55 |
| minute | noam: which country are you in? crowd supply also has the steel port covers in stock | 18:55 |
| noam | Presumably because - yeah | 18:55 |
| noam | I think I intended to get these from crowd supply since, at the time, the covers were available there but the 3588 wans't | 18:55 |
| noam | and then just... didnt do that | 18:56 |
| noam | Probly cuz I end up looking at pocket reform and talking myself out of it :D | 18:56 |
| minute | bluerise: did you immediately jump into openbsd stuff on rk3588? | 19:10 |
| * mjw -> Guest5145 | 20:00 | |
| - Guest5145 (QUIT: Killed (copper.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 20:00 | |
| * Guest8814 -> mjw | 20:00 | |
| + Guest5145 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 20:00 | |
| minute | ok so the ethernet issue seems phy reset related | 20:02 |
| minute | if i disable the snps,reset... definitions in dts, and unload the drivers, and then toggle the reset manually via gpioset and modprobe the driver again, it works | 20:02 |
| noam | minute: also: thanks for LFP and not li-ion | 20:18 |
| noam | fairly sure one or two of my busted cells is leaking | 20:18 |
| minute | ew | 20:19 |
| noam | I don't wanna know what woulda happened if this were li-ion | 20:19 |
| noam | yeah, looks like the blue wrapper is being eaten away from the inside | 20:19 |
| minute | eek | 20:19 |
| noam | This seems like a great candidate for diassembly! I'm smart | 20:21 |
| - jacobk (QUIT: Ping timeout: 248 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 20:31 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 20:32 | |
| bluerise | minute: once I have it, of course ;) | 20:32 |
| bluerise | I think I'll just use upstream u-boot with the DTS patched in | 20:33 |
| bluerise | anyone already tried submiting the DTS to Linux? | 20:33 |
| bluerise | should be easily upstreamable, no? | 20:34 |
| minute | bluerise: yeah but it's not tested enough in the wild yet | 20:34 |
| minute | for example, i'm currently debugging ethernet phy reset | 20:34 |
| minute | also, mainline linux doesn't support DSI or HDMI on rockchip | 20:35 |
| minute | so it's a bit pointless | 20:35 |
| minute | i mean, HDMI yes but not hdmdi1 (second hdmi) | 20:35 |
| minute | s/rockchip/rk3588 | 20:35 |
| minute | i'm in loose contact with the authors of the respective drivers though, just need a bit more time for them to finish their work i guess | 20:36 |
| minute | hmmmmm now i remember there was some race condition issue in linux about the load order of mdio + dwmac driver and who does the reset | 20:39 |
| bluerise | not pointless ;) | 20:49 |
| - jacobk (QUIT: Ping timeout: 260 seconds) (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 20:58 | |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 20:58 | |
| minute | aha, it looks like the reset pulse needs to be much longer. it could be because i think i have a wrong pullup on the reset line (goes with 4k7 to 1v8, but the rk3588 gpio is 3v3) | 21:00 |
| noam | minute: to confirm: the keyboard version from crowd supply has the same layout as Reform classic? | 21:41 |
| noam | that is, alt in between spaces? | 21:41 |
| minute | noam: no | 21:42 |
| minute | the reform classic also doesn't have that anymore | 21:43 |
| noam | :( picture is out of date? | 21:43 |
| noam | This is why i asked :) | 21:43 |
| minute | noam: can you show me the picture? | 21:43 |
| minute | noam: this is the current layout https://shop.mntre.com/spree/products/403/large/reform2-keyboard3-v2-qwerty-us-title.png?1690902504 | 21:43 |
| noam | I saw, but I thought the crowd supply ones might be remaining stock of the older one :( | 21:43 |
| noam | minute: https://www.crowdsupply.com/img/84d5/9790c1dd-d4b8-4b0b-b56c-d9da600684d5/cs-refkbd-black_png_md-fixed-xl.jpg | 21:44 |
| noam | > MNT Reform Standalone Keyboard V3 (Black Case, Brown Switches, QWERTY-US) | 21:44 |
| minute | wow how the hell did that happen | 21:44 |
| noam | nobody's said anything yet? :P | 21:44 |
| minute | thanks for pointing that out argh | 21:44 |
| minute | no | 21:44 |
| noam | Quite welcome :) | 21:44 |
| noam | There wouldn't happen to be any of that model in stock anywhere, would there? :P | 21:44 |
| noam | Or, anyone in here planning to switch to the new version willing to sell it on? :) | 21:45 |
| josch | noam: are you interested in a keyboard v4 with two alt keys between the spacebars? | 21:45 |
| noam | josch: What's the v3 vs v4 difference? | 21:45 |
| josch | noam: rp2040 and rgb led | 21:46 |
| minute | noam: we would have to dig through the archives. the old one is not very comfy though due to the unusual stagger | 21:46 |
| noam | minute: I like it :P | 21:46 |
| noam | It's what's on the Reform classic I have, I think? | 21:46 |
| minute | ok :D | 21:46 |
| noam | Assuming that's the one with the spacebar alt alt spacebar layout :P | 21:46 |
| minute | yeah it was like that on the reform for a longer while | 21:46 |
| josch | ACTION is still typing on that one | 21:46 |
| noam | I basically just want "that but that I can connect to my PC" :P | 21:46 |
| noam | The Reform's keyboard is the most comfy one I've got, I think | 21:47 |
| noam | It always takes me ~20-30 seconds of typing on it to get back into it, but then it's so much more comfy | 21:47 |
| noam | than anything else :) | 21:47 |
| noam | josch: So, I think the answer is yes :P | 21:47 |
| josch | ACTION misses the double alt keys between the spacebar but i had a look at the schematics in the past and it seems like the large spacebar in v4 is a 2U key, so two alt keys would theoretically fit? Because the board itself seems to still have most of the traces to the missing alt key, no? | 21:48 |
| noam | josch: maybe, but I have enough projects as is, I'm not addding another :p | 21:49 |
| josch | noam: yes, the classic reform was original sold with the space-alt-alt-space keyboard with unusual stagger but the next batch of classic reform will sell with v4 keyboards | 21:50 |
| noam | alas | 21:50 |
| noam | speaking of projects, I need to pull the plan9-rk3399 code and start playing with it today | 21:50 |
| + erle (~erle@user/erle) | 22:01 | |
| - spew (QUIT: Ping timeout: 260 seconds) (~spew@201.141.99.170) | 22:06 | |
| + spew (~spew@185.153.177.94) | 22:07 | |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-78-82-52-34.bbcust.telenor.se) | 22:15 | |
| minute | ok i have a stable workaround for wifi on rk3588, but it's a bit weird why the driver can't do the reset correctly | 22:17 |
| minute | what i do now is remove the reset from the driver in dts, and add init_rk3588() to reform-hw-setup which unloads the driver, toggles the reset line, modprobes the driver again... that works | 22:17 |
| ch | uhuh | 22:18 |
| minute | the best solution would probably be to move this reset to u-boot | 22:19 |
| bluerise | eww | 22:24 |
| minute | bluerise: isn't that how it's commonly done ^^ | 22:28 |
| bluerise | yeah, the modprobe combination is eww :D | 22:29 |
| bluerise | but u-boot tries to have fewer board-specific code | 22:29 |
| minute | bluerise: yeah but we already do something similar for a311d | 22:29 |
| bluerise | but if it can't be like a reset-duration thingy in the DTS... | 22:29 |
| bluerise | guess then u-boot it is | 22:29 |
| minute | not 100% sure if that also has an stmmac inside | 22:29 |
| minute | bluerise: yeah i'm not sure why it doesn't work in the dts. i can see the reset on the scope but maybe it's at the wrong moment | 22:30 |
| bluerise | :( | 22:30 |
| minute | now i just need a workaround until this can be resolved properly, so people can use their stuff... | 22:30 |
| bluerise | time to change the resistor, muhaha | 22:30 |
| bluerise | yeah yeah | 22:30 |
| minute | not sure if the resistor is really the issue... the reset looks pretty clean | 22:31 |
| minute | maybe it needs 2 reset pulses ;) | 22:31 |
| + switchy (~switchy@mechboards/switchy) | 22:35 | |
| noam | This is off topic, but: if a voltage regulator says "absolute limit: 19v supply", i probably shouldn't hook up a nominally 20V supply to it, right? :P | 22:40 |
| noam | figure people who would probvably know :P | 22:41 |
| minute | noam: correct | 22:44 |
| noam | Drats :/ | 22:44 |
| noam | I wonder... hmmmmm. I think I have an AVR board with a power supply rated up to 21V, and a 5V output; and the voltage regulator claims 4V-18V as operating range, that's probably a terrible idea though, I doubt its current output is sufficient >_< | 22:45 |
| noam | I think I have a buck converter here somewhere | 22:46 |
| noam | Oh! Nominal votlage is actually 19. That's... still cutting it way too close, though. | 22:48 |
| josch | minute: to save time in the reform-debian-packages pipeline, could you also at some point have a look again at the reform-handbook copyright situation? | 22:50 |
| minute | josch: ah yeah, writing this down now. do i basically just need to add one LICENSE that covers everything and that's it? | 22:54 |
| josch | that would be one solution, yes | 22:54 |
| minute | ok | 22:54 |
| josch | the problem is that gpl and cern ohl are incompatible | 22:54 |
| josch | so a work that combines content covered under each of these becomes unredistributable | 22:55 |
| josch | since you own the copyright to all of it, you are free to relicense as you wish | 22:55 |
| minute | josch: btw you've probably seen this, which needs !92 in reform-tools: https://source.mnt.re/reform/reform-debian-packages/-/merge_requests/67 | 22:55 |
| minute | josch: ok, noted for tomorrow | 22:55 |
| josch | minute: yes, would you like me to do a quick reform-tools release with just that change? | 22:55 |
| minute | josch: that would be great yeah | 22:56 |
| noam | Ooh, hacked something together with some spare wires and a USB-C adaptor, this might work (or might start a fire, 50/50) | 22:57 |
| minute | i ran the !67 reform-debian-packages pipeline with the variable pointing to that reform-tools branch also | 22:57 |
| noam | Cool, works :) Janky af though, I should solder and tape off the junctions so this doesn't short >_< | 22:59 |
| noam | minute: thanks for the confirmation that I was about to do something much dumber! :) | 22:59 |
| josch | minute: reform-tools MR 92 is merged into main. I just had to run gbp dch --ignore-branch && dch -r && git tag 1.56 | 23:03 |
| josch | oh it seems 0010-clk-rockchip-fix-error-for-unknown-clocks.patch is already part of 6.11.4 | 23:05 |
| josch | minute: I pushed commits into main that should fix things. Feel free to rebase !67 after the pipeline passed in main. Thanks! | 23:19 |
| minute | josch: awesome, thanks! | 23:49 |
| minute | developments around the clock... | 23:50 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
