2026-08-30.log

elbI thought I remembered a Pocket Reform keyboard case that turned a PR keyboard into a portable USB keyboard, but I sure don't see one on mnt.re00:06
elbregular paychecks are starting up again ~now so I'm queuing up all of the 2.0 boards for my PR00:06
+ simba (~simba@45.133.138.178)00:43
- simba (QUIT: Ping timeout: 247 seconds) (~simba@45.133.138.178)00:48
- mjw (QUIT: Ping timeout: 263 seconds) (~mjw@gnu.wildebeest.org)01:00
+ simba (~simba@45.133.138.178)01:33
- simba (QUIT: Ping timeout: 246 seconds) (~simba@45.133.138.178)01:37
- aperezdc (QUIT: Ping timeout: 268 seconds) (~aperezdc@2a03:6000:6e61:633::43)01:48
+ aperezdc (~aperezdc@2a03:6000:6e61:633::43)02:05
- a3f (QUIT: Read error: Connection reset by peer) (~a.fatoum@flummi.grey.stw.pengutronix.de)02:06
+ a3f (~a.fatoum@flummi.grey.stw.pengutronix.de)02:10
+ simba (~simba@45.133.138.178)02:38
- simba (QUIT: Ping timeout: 262 seconds) (~simba@45.133.138.178)02:42
+ wakest__ (~wakest__@2a02:2455:8220:4700:80f0:91d0:cf5d:1cb0)02:48
- xktr (QUIT: Ping timeout: 253 seconds) (~xktr@user/xktr)02:59
+ xktr (~xktr@user/xktr)03:00
- AnimaInvicta (PART: !!unknown attribute: msg!!) (~AnimaInvi@88-169-191-86.subs.proxad.net)03:33
- paperManu (QUIT: Ping timeout: 265 seconds) (~paperManu@173.206.254.242)03:36
+ simba (~simba@45.133.138.178)03:43
- simba (QUIT: Ping timeout: 255 seconds) (~simba@45.133.138.178)03:47
- op_4 (QUIT: Remote host closed the connection) (~tslil@user/op-4/x-9116473)04:05
+ op_4 (~tslil@user/op-4/x-9116473)04:05
+ simba (~simba@45.133.138.178)04:39
- simba (QUIT: Ping timeout: 254 seconds) (~simba@45.133.138.178)04:44
- wakest__ (QUIT: Ping timeout: 268 seconds) (~wakest__@2a02:2455:8220:4700:80f0:91d0:cf5d:1cb0)05:15
- tomenzgg (QUIT: Ping timeout: 241 seconds) (~tomenzgg@user/tomenzgg)05:41
+ simba (~simba@45.133.138.178)05:44
- simba (QUIT: Ping timeout: 247 seconds) (~simba@45.133.138.178)05:49
- talos (QUIT: Quit: The Lounge - https://thelounge.chat) (~talos@35.151.128.31)06:24
+ talos (~talos@2600:6c5d:f0:1fd0:9cf5:ef9:138a:7c32)06:25
+ simba (~simba@45.133.138.178)06:39
- simba (QUIT: Ping timeout: 250 seconds) (~simba@45.133.138.178)06:44
- AshCurry (QUIT: Ping timeout: 252 seconds) (~AshCurry@user/AshCurry)06:46
+ AshCurry (~AshCurry@user/AshCurry)06:47
joschelb: this? https://shop.mntre.com/products/mnt-pocket-reform-standalone-keyboard?taxon_id=1306:49
joschi only see it fully assembled but not as case without pcb06:49
+ simba (~simba@45.133.138.178)07:40
- simba (QUIT: Ping timeout: 254 seconds) (~simba@45.133.138.178)07:45
+ simba (~simba@45.133.138.178)08:25
- simba (QUIT: Ping timeout: 257 seconds) (~simba@45.133.138.178)08:30
+ simba (~simba@45.133.138.178)08:57
- simba (QUIT: Ping timeout: 240 seconds) (~simba@45.133.138.178)09:01
+ simba (~simba@45.133.138.178)09:59
- simba (QUIT: Ping timeout: 252 seconds) (~simba@45.133.138.178)10:03
+ AnimaInvicta (~AnimaInvi@88-169-191-86.subs.proxad.net)11:18
+ simba (~simba@45.133.138.178)12:01
- simba (QUIT: Ping timeout: 261 seconds) (~simba@45.133.138.178)12:05
+ simba (~simba@45.133.138.178)13:02
- simba (QUIT: Ping timeout: 259 seconds) (~simba@45.133.138.178)13:06
- potash6 (QUIT: Quit: The Lounge - https://thelounge.chat) (~potash@user/foghorn)13:14
- frickler (QUIT: Quit: Lost terminal) (~jens@user/frickler)13:27
+ frickler (~jens@user/frickler)13:30
+ potash6 (~potash@user/foghorn)13:52
elbjosch: Yes!  That's what I remembered, but yeah, I'm going to have an extra fully-populated PCB, so I was just looking for the case.  I even see it on the list now, I guess I missed it because I was looking for a photo more like https://shop.mntre.com/products/mnt-reform-main-box?taxon_id=13 :-)13:52
+ simba (~simba@45.133.138.178)14:11
- simba (QUIT: Ping timeout: 240 seconds) (~simba@45.133.138.178)14:15
- cow321 (QUIT: Ping timeout: 268 seconds) (~cow321@user/meow/cow321)14:16
+ mjw (~mjw@gnu.wildebeest.org)14:19
+ paperManu (~paperManu@173.206.254.242)14:28
joschelb: but that is the classic reform body14:32
elbjosch: oh, yes, I agree -- I meant more in the sense that I expected an aluminum block with a hole in it :-)15:18
+ simba (~simba@45.133.138.178)15:19
- simba (QUIT: Ping timeout: 267 seconds) (~simba@45.133.138.178)15:24
+ mrdaught (~mrdaught@bras-base-maplon2310w-grc-11-174-89-155-2.dsl.bell.ca)15:33
- mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@bras-base-maplon2310w-grc-11-174-89-155-2.dsl.bell.ca)16:18
+ simba (~simba@45.133.138.178)16:21
- simba (QUIT: Ping timeout: 248 seconds) (~simba@45.133.138.178)16:26
- op_4 (QUIT: Ping timeout: 251 seconds) (~tslil@user/op-4/x-9116473)17:02
+ op_4 (~tslil@user/op-4/x-9116473)17:10
+ mrdaught (~mrdaught@174.89.155.2)17:18
- mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@174.89.155.2)17:21
stephanominute: I'm working with josch to move patches into his linux repo on gitlab. As I update patches, I'm adding my sign off if the changes I made are substantial (not just API changes). Some of the patches, like 0001-v5-add-multi-display-panel-driver.patch, don't have your sign-off, and I'm changing them. Should I add your signoff then mine, since they'll be living in the git tree, not just in a17:39
stephanopatch file?17:39
- Kooda (QUIT: Quit: WeeChat 4.4.4) (~kooda@natsu.upyum.com)17:56
+ Kooda (~kooda@natsu.upyum.com)17:58
+ simba (~simba@45.133.138.178)18:23
- simba (QUIT: Ping timeout: 264 seconds) (~simba@45.133.138.178)18:27
+ tomenzgg (~tomenzgg@24.148.84.215)18:32
- tomenzgg (QUIT: Changing host) (~tomenzgg@24.148.84.215)18:32
+ tomenzgg (~tomenzgg@user/tomenzgg)18:32
- dodo (QUIT: Remote host closed the connection) (~dodo@user/dodo)18:50
- RandyK (QUIT: Ping timeout: 240 seconds) (~RandyK@user/randyk)18:51
+ dodo (~dodo@user/dodo)18:51
+ RandyK (~RandyK@user/randyk)18:52
+ mrdaught (~mrdaught@bras-base-maplon2310w-grc-11-174-89-155-2.dsl.bell.ca)19:21
- pasta161 (QUIT: Ping timeout: 266 seconds) (~pasta161@user/xha)19:24
+ pasta161 (~pasta161@user/xha)19:24
- mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@bras-base-maplon2310w-grc-11-174-89-155-2.dsl.bell.ca)19:25
minutestephano: yes, that sounds good, thank you for being so thorough!19:47
+ cow321 (~cow321@user/meow/cow321)19:50
- streifen (QUIT: Read error: Connection reset by peer) (~streifen@84.215.253.62)19:53
+ streifen (~streifen@84.215.253.62)19:54
- minute (QUIT: Ping timeout: 242 seconds) (~mntirc@softboy.mntmn.com)20:00
+ minute (~mntirc@softboy.mntmn.com)20:01
stephanominute: okay, will do. josch: running this on my imx8m+ and rk3588s now: https://source.mnt.re/josch/linux/-/tree/mnt-v7.2.120:06
stephanojosch: patch 7000 is missing as I dropped it, just FYI. Collabora has a DP controller driver with alt mode upstream. Andy Yan's dw_dp-rockchip.c series. Sebastian Reichel posted a devicetree binding rework set back in April. minute: was the long-term plan to move to this?20:11
stephanoOh, sorry, Andy's at rockchip. Sebastian's at Collabora. 20:16
+ simba (~simba@91.126.182.14)20:26
- simba (QUIT: Ping timeout: 241 seconds) (~simba@91.126.182.14)20:30
- potash6 (QUIT: Ping timeout: 254 seconds) (~potash@user/foghorn)20:32
+ simba (~simba@91.126.182.14)20:40
- simba (QUIT: Ping timeout: 264 seconds) (~simba@91.126.182.14)20:44
joschstephano: i'm trying it out right now with the reform-debian-packages machinery -- did you push the v7.2.1 tag? The script cannot find it.20:48
joschDebian experimental ships 7.2.2 right now but probably fine -- i shout if not :)20:49
joschthe scripts gracefully will pick the mnt-v7.2.1 branch as a fallback for 7.2.220:49
bremnerugh. the fact that patch 7000 is a real thing ;)20:51
- aloo_shu (QUIT: Read error: Connection reset by peer) (~aloo_shu@85.51.17.82)20:53
+ aloo_shu (~aloo_shu@85.51.17.82)20:54
joschbremner: it's not "patch number 7000" luckily -- the scripts do a shallow clone limited to the top 100 commits, so the MNT patches are fewer than that or the script breaks :)20:56
bremnerphew!20:57
joschthe Debian import scripts i mean, the Debian kernel packaging needs patch files which are generated using git-format-patch from the git branch (at least soon will be)20:57
joschi'm now looking into netbsd as probably the only kernel with a clear policy against slop afaik... hrm...20:57
+ murphnj (~murphnj@user/murphnj)21:06
- stephano (QUIT: Ping timeout: 253 seconds) (~stephano@71.238.14.13)21:06
bremneryeah, I'm curious how that works out.21:07
+ enwu9 (~enwu@user/enwu)21:12
+ Zaba_ (80b9b4b35e@2a03:6000:1812:100::116)21:12
+ jedavies_ (~jedavies@plan9.info)21:12
+ qbit_ (~qbit@user/qbit)21:12
+ cararemixed_ (sid724089@id-724089.helmsley.irccloud.com)21:13
+ rick____ (~rick@m.n0id.space)21:15
+ vkoskiv_ (~vkoskiv@109-204-186-7.cust.valoonet.fi)21:16
+ stephano (~stephano@71.238.14.13)21:17
stephanojosch: ops, forgot to push that tag. should be all set now.21:17
- jedavies (QUIT: *.net *.split) (~jedavies@plan9.info)21:18
- vkoskiv (QUIT: *.net *.split) (~vkoskiv@109.204.186.7)21:18
- rick_ (QUIT: *.net *.split) (~rick@m.n0id.space)21:18
- qbit (QUIT: *.net *.split) (~qbit@user/qbit)21:18
- cararemixed (QUIT: *.net *.split) (sid724089@id-724089.helmsley.irccloud.com)21:18
- exfalsoquodlibet (QUIT: *.net *.split) (a7085e0f71@user/exfalsoquodlibet)21:18
- Zaba (QUIT: *.net *.split) (80b9b4b35e@2a03:6000:1812:100::116)21:18
- enwu (QUIT: *.net *.split) (~enwu@user/enwu)21:18
* jedavies_ -> jedavies21:18
* Zaba_ -> Zaba21:18
* qbit_ -> qbit21:18
* enwu9 -> enwu21:18
* cararemixed_ -> cararemixed21:18
+ exfalsoquodlibet (a7085e0f71@user/exfalsoquodlibet)21:21
orvajosch: Same. Won't probably fulfil my desktop needs for now (too much gaming, too new hardware), but looking to try it out in my VPS/homeserver (which incidentally are Debians)21:28
orvaWas supposed to try it out this weekend but had some ear troubles and managed to roll my ankle, so sleep has been lacking21:29
joschstephano: perfect, scripts find kernel 7.2.2 and pick your mnt-7.2.1 branch but now I get:21:31
joschdpkg-source: error: expected ^@@ at line 10 of diff '/builds/josch/reform-debian-packages/linux/linux.orig.Z8nH9Y/debian/patches/reform/0007-sn65dsi86-use-hs-clock-of-samsung-dsim-host-directly.patch'21:32
joschwill investigate in a few hours21:32
joschorva: ouch, quick recovery to you!21:32
+ paperManu_ (~paperManu@173.206.254.242)21:42
minutestephano: it's not that simple @ alt-mode. alt mode in our case is external to the rockchip21:44
minutestephano: my patches are mostly about using the eDP part of the dual mode hdmi/edp controller for dp alt mode21:47
minutestephano: the usb-c/dp combo phy is a different thing that is native to the rk3588, and not explored yet by me21:48
+ simba (~simba@91.126.182.14)21:54
- simba (QUIT: Ping timeout: 242 seconds) (~simba@91.126.182.14)21:59
+ cli (~m-vsauiy@user/cli)22:03
- paperManu_ (QUIT: Ping timeout: 249 seconds) (~paperManu@173.206.254.242)22:05
- spew (QUIT: Quit: nyaa~) (~spew@user/spew)22:08
joschstephano: could you please rewrite the commit message of this commit: https://source.mnt.re/josch/linux/-/commit/8d10bc4b4e268b7815027b377377367c7df81d75 it messes up some patch format parsers22:20
josch(and is probably not intended as part of the commit message in the first place)22:21
+ simba (~simba@91.126.182.14)22:26
stephanojosch: will do. i think the Subject: was populated with the filename slug instead of the real one-liner. regardless there was also a bug in the script I wrote to merge in these patches. I'll fix this now.22:26
joschstephano: oh no, there are more commit messages like that (dtsi-add-hdptx, sdio-improve-wifi-speed, WORKAROUND-meson-plane-disable-afbc-32x8)22:29
- simba (QUIT: Ping timeout: 260 seconds) (~simba@91.126.182.14)22:30
joschstephano: ah those were probably created by your script? The old 7.1.10 branch didn't have those in its commit messages.22:30
stephanojosch: yep, that's my script not knowing how to parse these things. I'll force push a fix shortly with the cleanly parsed commits.22:34
stephanojosch: okay, those should be cleaned up now. the script I wrote applied the code fine, but when trying to create commit messages, some of the patches had unique qualities that made my parser choke.22:44
stephano minute: okay, i think i see. the mainboard alt-mode mux is wired to accept the eDP output of the RK3588 HDMI/eDP controller, not the differential DP lanes off the separate USB-C/DP combo PHY.22:57
+ mrdaught (~mrdaught@bras-base-maplon2310w-grc-11-174-89-155-2.dsl.bell.ca)23:25
- mrdaught (QUIT: Read error: Connection reset by peer) (~mrdaught@bras-base-maplon2310w-grc-11-174-89-155-2.dsl.bell.ca)23:27
minutestephano: exactly23:27
+ simba (~simba@91.126.182.14)23:28
minutestephano: my patches are mostly about 1. reimplementing the HPD mechanism with a worker, because the original one with the interrupt just didn't do anything 2. some tuning to allow limiting the number of lanes, and boosting the voltage in certain cases. 23:30
minutestephano: also there was some strangeness where something in drm was resetting the connected state. but this is from when i was still using a "fake" edp-panel, because dp-connector didn't work at all. in the meantime i found a workaround for that, a tiny patch in analogix_rockchip (or similar).23:32
- simba (QUIT: Ping timeout: 267 seconds) (~simba@91.126.182.14)23:33
minutestephano: this whole stuff is very hard to debug because it's still not possible to unload and reload rockchipdrm .kos cleanly whenever the panel-mnt-pocket-reform driver is also loaded. there is some linked list corruption in drm that happens, and one has to reboot23:33
minutethere are also unsolved bugs in vop2 that make things fail when using vp1,vp2 and vp3 together. this was never properly tested by anyone but me i think. because there is no other devicetree with rk3588 that actively uses 3 displays23:35
minutei tracked this down to be actually about the "cluster windows". that's why in my latest patch i move 2 of the vps to "esmart windows" instead, and only hdmi remains in cluster 0.23:36
minutei don't even know where to begin properly describing and reporting all these pretty serious bugs.23:37
stephanominute: Where does that small analogix_dp-rockchip fix for the connected-state-reset bug live? If it fixes the root cause, I think we can drop the powered_down/connector.force stuff in the HPD timer. maybe also revisit the suspend no-op and the AUX error code changes. Were those all working around the same PM issue?23:51
minutestephano: well you don't have mb2.0, right? so you can't really test that atm23:52
stephanominute: lol, well now you're being pragmatic. and yes, still waiting for mb2.0. :)23:52
minutestephano: right. i'm pushing on the gas pedal to get these to people ^^ 23:53
stephanominute: no worries, I'm patient. i just couldn't help poking at this one patch i can't port to 7.2. 23:55
minutestephano: there are a few people who have very new pocket reforms with mb2.0, and there's at least one user in community who was interested in the alt-mode feature. but i don't wanna pester them too much with testing things that might not work. i feel pretty burned from recent events where i broke stuff for people a little bit too much23:55
stephanominute: makes sense. well feel free to throw stuff at me anytime. my oscilloscope and fluke are always ready to go. ;)23:56
stephanoI have a dusty old spectrum analyser too, but if I'm getting that out it means I'm really making a mess of things.23:58
minutestephano: right, that's very cool. i'll put in my todo list now to check asap if the forcing + suspend workarounds are still necessary to keep it working23:58
stephanominute: no rush. you've got plenty to keep you busy.23:59
minuteunfortunately that's true as well. but it's noted down now so i won't forget it23:59

Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!