| - gustav28 (QUIT: Quit: Quit) (~gustav@c-2336524e.019-141-67626730.bbcust.telenor.se) | 00:04 | |
| - xuochi (QUIT: Quit: leaving) (~xuochi@138.sub-75-236-210.myvzw.com) | 00:19 | |
| - andypiper (QUIT: Quit: bye) (~andypiper@45.148.12.75) | 00:26 | |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 01:23 | |
| josch | aaaaaand the linux 6.10 branch was finally merged into main | 01:24 |
|---|---|---|
| josch | very ugly workaround but at least it works... :/ | 01:24 |
| + reform16124 (~lathie@c-24-17-214-36.hsd1.wa.comcast.net) | 01:39 | |
| - reform16124 (QUIT: Client Quit) (~lathie@c-24-17-214-36.hsd1.wa.comcast.net) | 01:40 | |
| - mtm (QUIT: Ping timeout: 255 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:04 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 02:06 | |
| - bluerise (QUIT: Ping timeout: 252 seconds) (~bluerise@p5b0acdf4.dip0.t-ipconnect.de) | 03:24 | |
| + bluerise (~bluerise@p5b0ac2ab.dip0.t-ipconnect.de) | 03:25 | |
| - cobra (QUIT: Ping timeout: 244 seconds) (~cobra@user/Cobra) | 03:43 | |
| + cobra (~cobra@user/Cobra) | 03:48 | |
| - jn (QUIT: Remote host closed the connection) (~quassel@user/jn/x-3390946) | 04:36 | |
| + jn (~quassel@2001-4dd3-e14e-0-20d-b9ff-fe49-15fc.ipv6dyn.netcologne.de) | 04:37 | |
| - jn (QUIT: Changing host) (~quassel@2001-4dd3-e14e-0-20d-b9ff-fe49-15fc.ipv6dyn.netcologne.de) | 04:37 | |
| + jn (~quassel@user/jn/x-3390946) | 04:37 | |
| - bkeys (QUIT: Quit: bkeys) (~Thunderbi@45.134.140.153) | 04:43 | |
| - aloo_shu (QUIT: Ping timeout: 252 seconds) (~aloo_shu@85.51.17.53) | 05:14 | |
| + aloo_shu (~aloo_shu@85.51.17.53) | 05:23 | |
| - Gooberpatrol66 (QUIT: Quit: Konversation terminated!) (~Gooberpat@user/gooberpatrol66) | 06:10 | |
| voltagex | how difficult is replacing the whole keyboard? I just saw the v3 keyboard with the US-ish layout and it looks like I'd have a much better time with it. | 07:50 |
| voltagex | (I don't know if I selected the wrong layout when I ordered :/) | 07:51 |
| voltagex | happy to send the other keyboard back to be reused or given to someone who needs it, too. | 07:51 |
| + Gooberpatrol66 (~Gooberpat@user/gooberpatrol66) | 08:00 | |
| + mundane-1emantic (~user@158.120.74.51) | 08:14 | |
| Twodisbetter | josch: why was the ugly workaround necessary? What does 6.10 bring us that makes that effort worth it? (That is an innocent question, and not one meant to be condescending.) | 08:26 |
| - mesaoptimizer (QUIT: Quit: mesaoptimizer) (~mesaoptim@user/PapuaHardyNet) | 10:15 | |
| + mesaoptimizer (~mesaoptim@user/PapuaHardyNet) | 10:15 | |
| + amospalla (~jordi@user/amospalla) | 10:17 | |
| amospalla | josch: minute: if this is of your interest, I upgraded my Pocket to 6.10, now kernel startup messages end on its third line "Begin: Loading essential drivers ...", and then the computer ends booting and screen comes black. I can ssh to it but nothing shows on the screen. | 10:22 |
| amospalla | To be clear, kernel loads, shows three lines of text and comes black after a while. | 10:24 |
| amospalla | dmesg -> https://pastebin.com/JZM7gS2k | 10:27 |
| amospalla | oh well, at the fourth reboot, screen works again. :O :)) | 10:30 |
| josch | Twodisbetter: the problem is this: https://salsa.debian.org/kernel-team/linux/-/merge_requests/1152#note_513085 | 10:31 |
| josch | amospalla: oh no!! :( | 10:31 |
| Twodisbetter | 👍👍 | 10:34 |
| Twodisbetter | josch: thanks! | 10:34 |
| amospalla | Now kernel hangs with an "internal error", it may be a coincidence that my hardware got faulty just right now that I also upgraded. | 10:34 |
| vkoskiv | https://www.phoronix.com/news/GhostWrite-Vulnerability-RISC-V | 10:35 |
| minute | voltagex: to replace the keyboard you only have to remove some screws and un/plug some cables | 10:46 |
| amospalla | just for the record, after switching the physical power button it boots again correctly. | 10:48 |
| [tj] | https://lobste.rs/s/rcexlr/mnt_pocket_reform_first_impressions | 10:50 |
| [tj] | all the comments I have seen have been positive, during the crowd funding there were a lot more about the price, but that seems to have died down | 10:51 |
| amospalla | When attaching the display ribbon, is there anything special to do, besides loosen the connector tab, pushing the ribbon into it with the fingers, and close the connector tab? | 11:13 |
| amospalla | I suspect I may have not connected it correctly and that is why my Pocket now randomly fails. I replaced it a few days ago. | 11:14 |
| - blast007 (QUIT: Quit: When the pin is pulled, Mr. Grenade is not our friend) (~blast@user/blast007) | 11:41 | |
| + blast007 (~blast@user/blast007) | 11:45 | |
| - mundane-1emantic (QUIT: Ping timeout: 276 seconds) (~user@158.120.74.51) | 12:01 | |
| amospalla | Sorry for pinging before, but the coincidence made me think it could be related to the kernel upgrade. | 12:48 |
| - iank (QUIT: Quit: ZNC 1.8.2+deb2+deb11u1 - https://znc.in) (~iank@fsf/staff/iank) | 13:25 | |
| + iank (~iank@fsf/staff/iank) | 13:27 | |
| + rwv (1fa29d61cb@2a03:6000:1812:100::6b) | 13:27 | |
| - mjw (QUIT: Ping timeout: 255 seconds) (~mjw@gnu.wildebeest.org) | 13:40 | |
| - mtm (QUIT: Ping timeout: 260 seconds) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:02 | |
| + mtm (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 14:06 | |
| + bkeys (~Thunderbi@45.134.140.153) | 14:39 | |
| + mjw (~mjw@gnu.wildebeest.org) | 14:43 | |
| Twodisbetter | One thing I am noticing with the Pocket is if you have to force off it due to a hanged state or kernel crash, etc. When you boot again you'll be dropped at initramfs prompt and it will not boot further. If you power off here and back on, it will do some file repairs and then boot successfully. Just odd that it doesn't automatically do this the first time. | 15:15 |
| Twodisbetter | This is with an encrypted nvme drive being used | 15:15 |
| + jacobk (~quassel@47-186-105-237.dlls.tx.frontiernet.net) | 16:50 | |
| Twodisbetter | the reason it seems that it is dumping to the initramfs prompt is that it is unable to mount the nvme drive after repairing things. | 16:58 |
| josch | minute: i added the reform-next system image and got: ERROR: Uploading artifacts as "archive" to coordinator... 413 Request Entity Too Large id=5308 responseStatus=413 Request Entity Too Large status=413 token=glcbt-64 | 17:07 |
| josch | minute: https://source.mnt.re/reform/reform-system-image/-/jobs/5308 | 17:07 |
| minute | ok, got DSI (display) working on pocket/rk3588 | 17:13 |
| minute | josch: woops @ too large | 17:13 |
| minute | Twodisbetter: yep, i'm also wondering why it needs 2 boots to repair the disk | 17:14 |
| minute | maybe the first boot just notices there is a problem and registers that for next time | 17:14 |
| sigrid | no hdmi adapter required? | 17:15 |
| minute | nope | 17:16 |
| sigrid | that's great | 17:23 |
| minute | but: rcore has no wifi on board, but i made a port for it with sdio+usb+uart signals. also there's the asiarf card that could be used instead, at least in pocket, if one doesn't need wwan | 17:34 |
| minute | lol https://hardwarerecs.stackexchange.com/a/18456 | 17:37 |
| minute | there's an adapter that allows you to plug an intel wifi card into a m-key slot | 17:37 |
| sigrid | haha nice | 17:39 |
| sigrid | probably great for diskless systems that have to run over wifi, for whatever reason | 17:39 |
| minute | the rk3588 module has 256GB eMMC so one could also use that and don't use an nvme | 17:39 |
| minute | sigrid: wifi thin client? :D | 17:39 |
| sigrid | yeah, like a terminal or something | 17:40 |
| minute | yeah | 17:40 |
| - mtm (QUIT: Quit: Textual IRC Client: www.textualapp.com) (~textual@c-71-228-84-213.hsd1.fl.comcast.net) | 18:08 | |
| minute | anyone tried this DE yet? with some design fixes (font, margins, borders) it could be interesting https://system76.com/cosmic | 18:18 |
| josch | minute: the last system images had a combined size of close to 10 GB. Maybe the default for gitlab is 10 GB? I started collecting ideas of how to mitigate the increased size of the system images here: https://source.mnt.re/reform/reform-system-image/-/issues/28 | 18:24 |
| Twodisbetter | minute: I wish. I think it is going to be pretty good. Do you think it is going to be power efficient. | 18:25 |
| minute | josch: ok, first we need to bump that somehow | 18:25 |
| minute | maybe we can use some deduplicating filesystem on the machine that stores the artifacts | 18:25 |
| minute | ZFS | 18:27 |
| minute | josch: artifact limit is now bumped to 20480 MB | 18:31 |
| minute | josch: what's the local workaround for the "kernel-wedge" issue? | 18:37 |
| minute | ah, that's just a missing tool | 18:38 |
| minute | apt install kernel-wedge | 18:40 |
| ^alex | hoo boy the pico sdk had a major version update | 19:05 |
| ^alex | also ugh everyone talks about _over_clocking their rp2040 but nobody talks about running it at 12mhz system clock | 19:10 |
| minute | mhm mhm https://codeberg.org/pyoif/rtl88x2cs | 19:13 |
| minute | ^alex: finally i can talk about the rp2350... | 19:13 |
| minute | switching to that will probably decrease our energy usage in sleep mode | 19:14 |
| ^alex | minute, yeah, we hope to clock the 2040s at 12mhz largely and let the USB PLL do the thing | 19:17 |
| ^alex | for the USB clock | 19:17 |
| minute | ^alex: nice! | 19:20 |
| ^alex | this requires completely sandblasting whatever setup the runtime does, and manually whacking the clock sources | 19:24 |
| ^alex | but ought to cut down some on the power consumption indeed | 19:24 |
| Twodisbetter | minute: the anodized black hinges on the pcoket reform was a welcome surprise. It looks great! | 20:00 |
| minute | josch: https://source.mnt.re/reform/reform-debian-packages/-/merge_requests/52 | 20:08 |
| minute | Twodisbetter: cool, happy it was worth it ^^ | 20:08 |
| ^alex | our favorite part of the hinges is that the machine folds flat | 20:09 |
| Twodisbetter | yeah the 180 in there is nice | 20:10 |
| jfred | gives it a bit of a Dynabook feel when fully open | 20:20 |
| josch | minute: the system image matrix is growing!! :D | 20:45 |
| minute | yes :D | 20:45 |
| josch | minute: what u-boot is this using | 20:45 |
| minute | josch: just the rk3588 uboot... i never tested it | 20:46 |
| minute | irl | 20:46 |
| minute | because i'm using vendor uboot | 20:46 |
| minute | (yet) | 20:46 |
| josch | we need the reform-rk3588-uboot repo to emit some blob for system images to be produced for the pocket with rk3588 (if that is something you want to offer as a download) | 20:46 |
| minute | yep, wanna | 20:47 |
| minute | but first, i need a working kernel package :3 | 20:47 |
| minute | ah, build failed | 20:47 |
| josch | well, then good you bumped the maximum size to 20 gigs XD | 20:49 |
| minute | ah > initialization of ‘int (*)(struct platform_device *)’ from incompatible pointer type ‘void (*)(struct platform_device *)’ | 20:49 |
| minute | ye olde | 20:49 |
| bluerise | please use upstream u-boot for rk3588 | 20:56 |
| minute | yeah, for shipping. but not during my testing atm | 20:56 |
| minute | there lurk some surprises, for example there's a weird clock that is not set up correctly in mainline uboot and that renders DSI nonfunctional | 20:57 |
| Twodisbetter | minute: any initial hunches on whether or not the RK3588 will be an easy or difficult thing to get suspend working on? | 20:57 |
| minute | Twodisbetter: probably difficult | 20:58 |
| Twodisbetter | minute: I really wish hibernation was possible on those platforms. I feel like that would be best. Still I really appreciate you and the teams efforts. | 20:59 |
| Twodisbetter | the good news is that when the scrteen is off the pocket is using around 500-700mA | 21:00 |
| Twodisbetter | If we could down clock the cpu when the screen is locked and off, then we could approximate susopend power savings without even using it. | 21:00 |
| josch | bluerise: currently, the collabora fork is being shipped by MNT: https://source.mnt.re/reform/reform-rk3588-uboot/-/blob/main/.gitlab-ci.yml?ref_type=heads#L9 | 21:09 |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 21:09 | |
| josch | instead, we could also clone upstream u-boot and apply collabora patches on top | 21:10 |
| bluerise | figured :) when I get mine I'll use upstream u-boot | 21:11 |
| bluerise | though if their u-boot can initialize the display that would be nice | 21:11 |
| josch | minute: thanks to your bumping of the CI artifact limit, there is now a system image for reform next with rk3588 for you to test: https://source.mnt.re/reform/reform-system-image/-/jobs/5315/artifacts/browse | 21:22 |
| minute | josch: nice!! | 21:22 |
| grimmware | Back from holiday and reunited with my Pocket <3 | 22:10 |
| grimmware | (I intentionally left it behind because I needed to take a break from being on my normal bullshit) | 22:12 |
| ^alex | awoo | 22:17 |
| - colinsane (QUIT: Quit: bye) (~colinunin@97-113-150-69.tukw.qwest.net) | 22:19 | |
| Twodisbetter | so far I have had my pocket for about a week. Made a lot of changes and have gotten up to full speed on touch typing on it | 22:19 |
| + colinsane (~colinunin@97-113-150-69.tukw.qwest.net) | 22:22 | |
| ^alex | we had to completely rearrange the keyboard | 22:23 |
| grimmware | I'm going to do some more work on the keyboard layout, currently I've only made the arrow key cluster into an hjkl-style on the bottom row and moved / to where up was (and made the left space a backspace) | 22:25 |
| grimmware | I'm interested in adding a tap vs hold functionality | 22:25 |
| grimmware | I'd like to make so that I can tap space to get space and hold it to get another layer on the keyboard like I've got on my Corne | 22:31 |
| minute | nice | 22:57 |
| minute | josch: yay, my rk3588+dsi debian packages built went through. tomorrow i'll see if that worked | 22:57 |
| grimmware | Super excited about that. If I have to sacrifice an M.2 slot for WiFi so be it, I can tether | 22:58 |
| josch | awesome! progress \o/ | 23:05 |
| minute | got hdmi audio to work on a311d | 23:27 |
| josch | i... actually had never tried that! | 23:28 |
| ^alex | changes to handle pico-sdk 2.0 are tiny, but you do gotta `rm -rf build/` before i'll build again | 23:28 |
| ^alex | it'll* | 23:28 |
| minute | ^alex: nice | 23:30 |
| josch | ^alex: you are fast! that got released 6 hours ago :D | 23:31 |
| ^alex | it also wants to build a picotool 2.0, at least on macos | 23:33 |
| ^alex | (we do a lot of crossbuilding for ARM microcontrollers on our mac) | 23:34 |
| ^alex | we'll see how this branch handles on the pocket in a minute | 23:34 |
| ^alex | josch, we just closed out a work week of switching the internal tooling to the AWS API v2, we've got SDK migrations on the brain apparently | 23:37 |
| josch | :D | 23:38 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 23:50 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!