2026-05-12.log

aloo_shuhow many of these 90W do you think need to become heat in order to get felt?00:51
aloo_shuI don't think it's wise to normalize this, apart from the cable, it can be poor contact in the plug and that's going to degrade both sides00:55
aloo_shuactually if it's the plug getting hot, it's likely that there's either poor contact or some semi short due to conductive dirt or mechanical imprecision00:56
aloo_shuheat btw increases the resistance of metal so the poor conductivity kind of persists itself00:58
+ paperManu (~paperManu@204.244.197.237)01:08
+ paperManu_ (~paperManu@204.244.197.237)02:19
- aelius (QUIT: Ping timeout: 244 seconds) (~aelius@user/aelius)02:22
+ aelius (~aelius@user/aelius)02:35
- mjw (QUIT: Ping timeout: 256 seconds) (~mjw@gnu.wildebeest.org)02:36
- jn (QUIT: Ping timeout: 244 seconds) (~quassel@user/jn/x-3390946)02:58
+ jn (~quassel@2a0a:a54a:ed40:0:20d:b9ff:fe49:15fc)02:59
- jn (QUIT: Changing host) (~quassel@2a0a:a54a:ed40:0:20d:b9ff:fe49:15fc)02:59
+ jn (~quassel@user/jn/x-3390946)02:59
- paperManu (QUIT: Ping timeout: 244 seconds) (~paperManu@204.244.197.237)03:55
- paperManu_ (QUIT: Ping timeout: 264 seconds) (~paperManu@204.244.197.237)03:55
+ paperManu (~paperManu@204.244.197.237)03:57
- paperManu (QUIT: Ping timeout: 246 seconds) (~paperManu@204.244.197.237)04:01
+ paperManu (~paperManu@79.127.134.45)04:04
- kop316 (QUIT: Remote host closed the connection) (m-6f6zq6@static.138.159.90.157.clients.your-server.de)04:07
+ kop316 (m-6f6zq6@static.138.159.90.157.clients.your-server.de)04:08
- paperManu (QUIT: Read error: Connection reset by peer) (~paperManu@79.127.134.45)04:34
+ spew (~spew@user/spew)05:38
- spew (QUIT: Quit: nyaa~) (~spew@user/spew)06:46
- wickedshell (QUIT: Ping timeout: 245 seconds) (~wickedshe@2601:8c0:c7c:3572:4d7a:9cb8:f693:f993)06:57
* lukateras -> luka07:02
- luka (QUIT: ) (sid744851@user/lukateras)07:47
+ luka (sid744851@user/lukateras)07:47
* luka -> lukateras07:51
* lukateras -> luka07:51
- RandyK (QUIT: Remote host closed the connection) (~RandyK@user/randyk)08:03
+ RandyK (~RandyK@user/randyk)08:04
+ wickedshell (~wickedshe@2601:8c0:c7c:3572:3756:9c40:b769:1c1e)09:04
- wickedshell (QUIT: Ping timeout: 246 seconds) (~wickedshe@2601:8c0:c7c:3572:3756:9c40:b769:1c1e)09:40
+ mjw (~mjw@gnu.wildebeest.org)11:14
+ wickedshell (~wickedshe@2601:8c0:c7c:3572:679c:85a1:8f2e:cdd6)11:47
+ rodion_goritskov (~rodion_go@mail.goritskov.com)11:53
- mjw (QUIT: Remote host closed the connection) (~mjw@gnu.wildebeest.org)12:13
+ mjw (~mjw@gnu.wildebeest.org)12:14
+ paperManu (~paperManu@204.244.197.237)12:33
- andreas-e (QUIT: Ping timeout: 268 seconds) (~Andreas@2a02-8434-b6a3-e901-facc-8e87-8e54-890d.rev.sfr.net)12:35
+ andreas-e (~Andreas@2a02-8434-b6a3-e901-facc-8e87-8e54-890d.rev.sfr.net)12:35
- mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org)13:21
* Guest8124 -> mjw13:25
- Ar|stote|is (QUIT: Quit: No Ping reply in 180 seconds.) (~linx@149.210.0.44)13:56
+ Ar|stote|is (~linx@149.210.0.44)14:01
+ paperManu_ (~paperManu@204.244.197.237)14:19
- aloo_shu (QUIT: Read error: Connection reset by peer) (~aloo_shu@85.51.18.199)14:57
+ aloo_shu (~aloo_shu@85.51.18.199)14:57
- paperManu (QUIT: Ping timeout: 246 seconds) (~paperManu@204.244.197.237)14:59
- paperManu_ (QUIT: Ping timeout: 268 seconds) (~paperManu@204.244.197.237)14:59
+ paperManu (~paperManu@modemcable141.205-200-24.mc.videotron.ca)15:12
Svpjosch: i FINALLY managed to run through the batteries, so far it's sucking 1.4 amps and getting barely warm, almost to the point where im unsure if im the one warming it up lol15:27
Svpoh should mention: i'm using a lenovo 100W usb-c brick15:33
+ spew (~spew@user/spew)15:38
joschSvp: that sounds like you have found much better couplings than i did -- can you share a link?15:53
[tj]https://altelectron.org.uk/notice/B6DpYIQ6TwldVRCwcq15:54
[tj]pocket is such a handy tool15:54
gordon1so with new kbd firmware i notice that sometimes when the laptop (classic reform) is off for longer period of time - it does not want to turn on with circle+1, i need to do it multiple times or reset the kbd through the menu and do it again15:55
Zabaminute: re. your mastodon posts about SPI - for what it's worth, on the rp2040/rp2350 it is possible to use the PIO instead of the hardware SPI peripheral for better behaviour and not having to deal with the SDK APIs16:00
Svpjosch: https://it.aliexpress.com/item/1005006052488919.html16:20
Svpits lying, it's 2.0, but works well still16:21
- spew (QUIT: Quit: nyaa~) (~spew@user/spew)16:35
aloo_shuno way that 1.4 amps in the usb-c cable equal 90 Watts but ok that was an 'up to'. if the charger driver doesn't negotiate different currents for different situations e.g. what the charge host communicates or what it thinks about battery health & -max currents, then you'd have to lose in excess of 15% in the plug to feel heat16:36
+ mark_ (~mjw@gnu.wildebeest.org)16:49
+ talos3 (~talos@2600:6c5d:f0:1fd0:e267:b8b6:f63c:1696)16:50
- talos (QUIT: Ping timeout: 245 seconds) (~talos@2600:6c5d:f0:1fd0:c5e7:a281:634c:2c2)16:52
* talos3 -> talos16:52
Svpfor what its worth it does seem to do something since the little led shines brighter but i dont have a usb-c snooper to check17:09
+ siviq (~siviq@user/siviq)17:10
- siviq (QUIT: Client Quit) (~siviq@user/siviq)17:13
+ spew (~spew@user/spew)17:20
- mark_ (QUIT: Ping timeout: 244 seconds) (~mjw@gnu.wildebeest.org)17:22
minuteZaba: oh wow @ PIO, i think i heard that before and forgot17:31
minutebuttt i think i got it under control now17:31
minutebig changes are coming17:31
kfxthat sentence stresses me out17:42
kfxstatistically it means mnt is about to be acquired by nvidia17:42
joschaloo_shu: i measured the 90 W using a usb-c power-meter. Here is the IRC log from two years ago when I did the measurements and a video which shows up to 75 W which already exceeds the 60 W of the reform AC adapter: https://mntre.com/reform-irc-logs/2024-01-19.log.html17:44
^aleximagining minute being approached by nvidia and for some reason all we can think of is that scene in the 1998 film Pi where the paranoid hacker is being pursued by the wall street lady17:44
joschSvp: thank you for the link! I think I'll try it out. :)17:49
- mjw (QUIT: Quit: Leaving) (~mjw@2001:1c06:2486:4600:5952:3a9:6e0e:555a)17:49
+ bkeys (~Thunderbi@2600:1005:b183:b915:f059:dc0b:a248:4bef)18:01
aloo_shujosch: hmm not sure I can watch videos now but 15 mystery watts would ondeed generate heat (and represent 20% on top of 60W). I don't know how the 60W AC charger behaves few chargers bluntly switch of at overspec current draws on the secondary side they rather get noisy secondary voltage collapses and perhaps a fuse on the primary side blows before the charger does, but if a usb c power meter I18:10
aloo_shutrust shows 75W, my next step would be checking voltage_now and current_now as seen by the charge controller driver and exposed in sysfs18:10
aloo_shuanother test that would require opening, would be measuring resistance pole by pole, i.e. usb-c breakout board - usb-c cable - reform plug - reform circuit bord 18:10
- _justin_kelly4 (QUIT: Quit: The Lounge - https://thelounge.chat) (~justinkel@user/justin-kelly/x-6011154)18:12
aloo_shuthat's at the same time an educational exercise because you typically discover how probe tip quality, pressure etc., easily add add a few ohms on the meter's reading18:13
+ _justin_kelly4 (~justinkel@user/justin-kelly/x-6011154)18:14
aloo_shuso poor contact[surfaces] - an added 3 Ohms at a 4A charge current - that will do it, and that's the first thing I'd want to rule out 18:15
+ mjw (~mjw@2001:1c06:2486:4600:5952:3a9:6e0e:555a)18:20
joschaloo_shu: maybe you misunderstand. This is not a problem. Just an observation.18:21
- spew (QUIT: Remote host closed the connection) (~spew@user/spew)18:24
- bkeys (QUIT: Read error: Connection reset by peer) (~Thunderbi@2600:1005:b183:b915:f059:dc0b:a248:4bef)18:43
+ bkeys (~Thunderbi@2600:1005:b183:b915:f059:dc0b:a248:4bef)18:44
minutekfx: haha no, big changes in the SC communication :D18:53
minute^alex: lmao @ Pi18:53
minute^alex: i know exactly what you mean18:54
kfxwaking up at night in a cold sweat, having dreamt of an "our incredible journey" post on mastodon18:55
joschminute: the image from the quasar branch now boots successfully on rk3588 pocket reform. But not via EFI u-boot tries efi, fails and then falls back to extlinux.conf: https://mister-muffin.de/p/YAVI.txt19:14
- thejevans (QUIT: Remote host closed the connection) (~m-7r3qil@user/jevans)19:23
+ thejevans (~m-7r3qil@user/jevans)19:25
minutejosch: aw ok interesting19:29
minutejosch: you have a link to the current image for me?19:30
joschminute: i assume for reform next with qcs6880?19:30
joscherrr... 649019:31
joschif the answer is yes, then try this:19:32
joschsudo bmaptool copy https://source.mnt.re/josch/reform-system-image/-/jobs/20638/artifacts/raw/reform-next-system-qcs6490.img.gz /dev/mmcblkXXX19:32
+ mark_ (~mjw@gnu.wildebeest.org)19:49
- bkeys (QUIT: Ping timeout: 246 seconds) (~Thunderbi@2600:1005:b183:b915:f059:dc0b:a248:4bef)19:59
minutejosch: thx20:52
minutei'm currently unifying spi, uart and usb-uart communication (CLIs) in SC20:52
minutewith a new protocol based on s-expressions20:53
minute(but no nesting currently)20:53
joschhow do you plan to do the migration from the old protocol?20:53
minutenot sure yet... maybe it will be somewhat backwards compatible, maybe not at all20:53
josch(in related news i have now also been hit by all kinds of whacky battery readings on pocket reform and classich reform...)20:54
minutebut it's already more robust+much faster than the old variant, even if there's more parsing involved20:54
minutejosch: uh oh... like what? :3 20:54
minuteexactly what i'm working on :D20:54
joschsend over alpha firmware to try is what i'm saying :)20:55
joschdo we have a picture guide to flash classic reform mainboard 3.0? there is a forum user who needs help and i thought next time i flash it i make a more detailed step-by-step guide20:56
minutejosch: haha ok :D it's not complete enough, but very soon21:01
minutejosch: maybe ifixit has one? 21:02
minutejosch: only difference is microusb vs usb-c i think21:02
joschthat and i'm also not sure anymore which flash.sh script to choose and whether to build it or whether there is a download.sh script like for the keyboard etc... i have to check21:02
minutei guess at some point we should polish "MNT Flash & Test" for public consumption21:55
minuteanyway, new SC fw is now largely compatible with the old keyboard fw21:55
minutealso cool that i can trace the output going to the keyboard now21:55
minutelike: # [kbd<] 41 41 00 00 00 00 00 00 mA-0001mV08287  76% P021:55
joschgordon1: sprintek trackstick is alive22:12
joschi built a small trickstick-only qmk firmware and flashed it to an rp204022:12
joschthe thing registers as as: hid-generic 0003:5678:1234.0007: input,hidraw4: USB HID v1.11 Mouse [josch josch-trackstick] on usb-xhci-hcd.3.auto-2.1.1/input122:13
joschmouse movement is all over the place but it's *moving* :D22:13
gordon1good enough, but i thought it supposed to self-calibrate, didn't it?22:15
joschgordon1: in the datasheet https://mister-muffin.de/p/rJQX.pdf on page 5 it talks about pull-up resistors on ps/2 data and clk lines. I read elsewhere that rp2040 doesn't need them, so right now everything is connected directly. I assumed that if resistors were missing it probably wouldn't work at all, no?22:16
joschmaybe -- this is the first time i connect the thing electronically, i'm happy it is doing anything at all because i'm a hardware noob and would have zero chance to debug it if nothing had happened22:17
minutejosch: very nice!22:17
gordon1josch: did you enable pull-ups in your code/qmk config?22:17
joschnot manually, no -- maybe i have to throw them in22:17
joschthey have a rp2040 specific ps/2 driver and my assumption was that maybe it is doing things like that automagically22:18
joschi'm reading other keyboard drivers in the qmk source tree to get inspired22:18
gordon1i do not remember what ps/2 is and if it has any error correction or parity, but absence of pullups may end up in reading garbage22:18
gordon1meaning random bits22:18
minuteuff it's really nice to have zero delay on the SC-SoC interface22:19
joschi'll investigate adding the pull-ups via qmk config next then22:19
joschalso arrived today is the adapter which lets me feed usb to my intel ax210 wifi/bt card -- i'm trying that out now22:20
gordon1josch: you can also check rp2040 datasheet to check what value internal pull-up has and if it is in the required range of 4k-100k 22:20
gordon1(spoiler: it is)22:21
gordon1josch: hmm https://github.com/qmk/qmk_firmware/pull/2593622:28
joschgordon1: oooooh nice find!22:30
gordon1there's also that https://github.com/qmk/qmk_firmware/pull/22532, i really thought that ps2 mouse is a very old thing that is supported by qmk since forever, i wonder what is it about22:30
gordon1yeah, looks like you need first PR if you want to connect it directly https://www.reddit.com/r/ErgoMechKeyboards/comments/1q4gclk/we_dont_need_no_pull_up_resistors_or_reset/22:32
+ RandyK_ (~RandyK@user/randyk)22:34
- RandyK (QUIT: Ping timeout: 276 seconds) (~RandyK@user/randyk)22:34
joschgordon1: i have a few things to try out, thank you! I'm already relieved that i can build qmk just fine (funky build system) and that flashing works and that aparently i didn't fry either of the boards (yet) by connecting something very badly. From here on it should be "just" iterating on the code, flashing, trying it out :)22:35
* RandyK_ -> RandyK22:35
gordon1yeah all this stuff is surprisingly sturdy22:38
joschi recently friend my classic reform motherboard and had to send it to MNT to get it repaired -- i'm scarred :D22:39
gordon1what went wrong?22:42
gordon1i meant rp2040 and all this hobby level stuff, and generally keyboards and input devices - no scary voltages or currents or sensitive chips, if you connect something wrong probably nothing bag will happen22:43
gordon1*bad22:43
gordon1reform mb is a bit more delicate than that22:44
joschgordon1: i 3d printed a small scaffold which allows me to add the port board of the pocket reform with the intel wifi/bt card into classic reform case22:46
joschwhile installing it, i got smoke from one of the diodes and batteries wouldn't charge anymore (but the unit would otherwise run)22:46
joschreplacing the diode didn't fix it (would blow again right away after connecting the batteries)22:46
joschso liazon had to replace the LTC6803: https://floss.social/@josch/11641603120412673122:52
gordon1oh i think i was present at that, at least if that wasn't the second installment of the same story23:00
joschnono, it didn't happen a second time round :)23:01
joschhuh i haven't seen this before:23:01
joschEnter choice: 1:Debian GNU/Linux forky/sid 7.0.4-mnt-reform-arm6423:01
joschRetrieving file: /boot/vmlinuz-7.0.4-mnt-reform-arm6423:01
joschSkipping l0 for failure retrieving kernel23:01
gordon1ok, good23:01
joschoh i know what's wrong23:02
gordon1josch: check df on your boot, i had that when i had only half of a kernel23:02
gordon1iirc23:02
joschthe path is wrong23:03
joschit needs to be /vmlinuz-7.0.4-mnt-reform-arm64 not /boot/vmlinuz-7.0.4-mnt-reform-arm6423:03
gordon1ah, i usually do ln -sf . /boot/boot to catch those errors23:04
joschcreative!23:05
- paperManu (QUIT: Ping timeout: 276 seconds) (~paperManu@modemcable141.205-200-24.mc.videotron.ca)23:07
- luka (QUIT: Quit: Updating details, brb) (sid744851@user/lukateras)23:32
+ paperManu (~paperManu@204.244.197.237)23:32
+ luka (sid744851@user/lukateras)23:32
+ paperManu_ (~paperManu@204.244.197.237)23:36

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