| - jacobk (QUIT: Read error: Connection reset by peer) (~quassel@47-186-110-207.dlls.tx.frontiernet.net) | 00:13 | |
| - mjw (QUIT: Ping timeout: 252 seconds) (~mjw@gnu.wildebeest.org) | 01:14 | |
| - Ar|stote|is (QUIT: Ping timeout: 268 seconds) (~linx@149.210.0.13) | 01:28 | |
| + Ar|stote|is (~linx@149.210.0.13) | 01:32 | |
| - Ar|stote|is (QUIT: Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) (~linx@149.210.0.13) | 02:21 | |
| + Ar|stote|is (~linx@149.210.0.13) | 02:21 | |
| - buckket (QUIT: Quit: buckket) (~buckket@vps.buckket.org) | 02:32 | |
| + buckket (~buckket@vps.buckket.org) | 02:32 | |
| - nsc (QUIT: Ping timeout: 240 seconds) (~nicolas@i5C74DDCB.versanet.de) | 03:28 | |
| + nsc (~nicolas@i5C74DF17.versanet.de) | 03:30 | |
| noam | minute: an unbricking board? | 04:38 |
|---|---|---|
| - ptrc (QUIT: Remote host closed the connection) (~ptrc@ptrc.gay) | 06:32 | |
| + ptrc (~ptrc@ptrc.gay) | 06:33 | |
| + chomwitt (~chomwitt@2a02:587:7a21:2700:1ac0:4dff:fedb:a3f1) | 07:41 | |
| - chomwitt (QUIT: Ping timeout: 260 seconds) (~chomwitt@2a02:587:7a21:2700:1ac0:4dff:fedb:a3f1) | 08:29 | |
| + chomwitt (~chomwitt@2a02:587:7a21:2700:1ac0:4dff:fedb:a3f1) | 09:20 | |
| - nybble (QUIT: Quit: WeeChat 4.2.2) (~sprang@user/csprng) | 09:22 | |
| - ex--parrot (QUIT: Quit: _b) (~fincham@user/ex-parrot) | 09:24 | |
| + ex-parrot (~fincham@user/ex-parrot) | 09:24 | |
| - ggoes (QUIT: Ping timeout: 255 seconds) (~gregf@fsf/staff/ggoes) | 10:18 | |
| - erle (QUIT: Ping timeout: 260 seconds) (~erle@user/erle) | 10:19 | |
| - GNUmoon2 (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 10:22 | |
| - chomwitt (QUIT: Ping timeout: 260 seconds) (~chomwitt@2a02:587:7a21:2700:1ac0:4dff:fedb:a3f1) | 10:22 | |
| + GNUmoon2 (~GNUmoon@gateway/tor-sasl/gnumoon) | 10:23 | |
| + ggoes (~gregf@fsf/staff/ggoes) | 10:25 | |
| + mjw (~mjw@gnu.wildebeest.org) | 10:29 | |
| + erle (~erle@2a02:8109:da01:6400::aa65) | 10:32 | |
| - erle (QUIT: Changing host) (~erle@2a02:8109:da01:6400::aa65) | 10:32 | |
| + erle (~erle@user/erle) | 10:32 | |
| - GNUmoon2 (QUIT: Remote host closed the connection) (~GNUmoon@gateway/tor-sasl/gnumoon) | 11:04 | |
| + GNUmoon2 (~GNUmoon@gateway/tor-sasl/gnumoon) | 11:05 | |
| - mjw (QUIT: Ping timeout: 264 seconds) (~mjw@gnu.wildebeest.org) | 12:12 | |
| + artfwo (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 12:32 | |
| artfwo | hi! i have just received pocket reform and got it to start, but after switching to a different power supply it no longer charges the battery (the battery just keeps discharging). switching back to previous power supply doesn't help. is there anything i can do to solve the problem? :) | 12:38 |
| artfwo | ah, actually restarting with a power supply plugged in fixes it, but only until i plug it out and in again | 12:39 |
| * Guest5847 -> mjw | 12:44 | |
| josch | the first pocket reform arrived!! \o/ | 12:54 |
| josch | minute: first problem popped up :) | 12:55 |
| minute | woah woah | 13:09 |
| minute | artfwo: hello hello! you're the first person who has received the pocket reform i think | 13:09 |
| artfwo | i don't know, maybe :) | 13:09 |
| minute | artfwo: which power supplies are you using? | 13:09 |
| minute | artfwo: as a quick fix, try to toggle the standby power switch, it will reset the system controller and everything | 13:10 |
| minute | artfwo: the one you needed to toggle to turn on the device in the first place | 13:10 |
| minute | artfwo: maybe the second power supply confused the state machine of the system controller and its stuck | 13:10 |
| minute | artfwo: which power supplies are you using? | 13:10 |
| artfwo | right, well, the "poweroff" command already helps ("reboot" doesn't). i'm using GaN 60W power supply and 15W supply from my phone | 13:13 |
| minute | artfwo: ah, i didn't expect poweroff to work, well, that's more comfortable | 13:13 |
| minute | artfwo: how do you see it's charging? negative amps number in battery status on oled? | 13:13 |
| minute | artfwo: i suspect the GaN is working but phone is not? | 13:14 |
| artfwo | minute: it happens with GaN too - I see negative numbers with GaN on oled after first boot, then I unplug it and plug it back, but only positive numbers are there (and also in sway status bar it shows as discharging) | 13:18 |
| artfwo | only after "poweroff" the numbers become negative again and the label above says "Off" | 13:19 |
| artfwo | and yes, zero or positive only with the phone supply | 13:21 |
| minute | artfwo: ok, can you give me the exact charger model? | 13:25 |
| minute | another one in the wild, but still in a box https://social.sdf.org/@jirka/112512798063123723 | 13:25 |
| artfwo | i can even bring it to your hq :) | 13:25 |
| artfwo | it happens also on a fresh start from battery, after cycling the standby power switch | 13:26 |
| minute | artfwo: ah that could be interesting! are you who i think you are? :D | 13:26 |
| AbortRetryFail | casually ask the chat if they can read your mind. :) | 13:26 |
| artfwo | maybe :) | 13:26 |
| artfwo | the exact model is framework power adapter https://frame.work/de/en/products/power-adapter?v=FRANCEPH0F | 13:27 |
| minute | AbortRetryFail: that's how i roll | 13:27 |
| minute | artfwo: ah very interesting! maybe you could come over tonight? or otherwise near the end of the week it's possible during the day | 13:28 |
| minute | artfwo: a thing you can do is install "tio" (or it's maybe preinstalled) and do "sudo tio /dev/ttyACM0" | 13:33 |
| minute | artfwo: you should then get log output from the rp2040 and its PD state machine | 13:33 |
| minute | artfwo: so normally it should output lines like "# [pd] state 1, timeout." | 13:39 |
| minute | when connecting a power supply it should output information about all the PDOs (power delivery objects) it got offered from the supply and which one it chose | 13:40 |
| minute | so far we have tested with Ugreen, Anker, Apple and CUI supplies | 13:41 |
| artfwo | minute: it begins with state 3, but after unplugging it only spews out state 1, timeout -- https://pastebin.com/EEBAzaMB | 13:42 |
| minute | artfwo: this is good so far | 13:42 |
| minute | artfwo: what if you plug it in? | 13:42 |
| minute | artfwo: or it just stays in state 1? | 13:43 |
| artfwo | exactly, i plugged it back in a few seconds after unplugging :) | 13:43 |
| minute | ohh ok | 13:43 |
| minute | to be able to try stuff you would need to install picotool. i have a helper script for that here: https://source.mnt.re/reform/reform/-/blob/master/reform2-trackball2-fw/install-picotool.sh?ref_type=heads | 13:47 |
| minute | i.e. for flashing modified firmware to the rp2040 | 13:47 |
| artfwo | can you flash rp2040 programmatically? | 13:47 |
| artfwo | (on rpi pico i have to press a physical button for that) | 13:48 |
| minute | yes, you can, it can be forced to bootloader via software | 13:50 |
| minute | (if it uses the usb serial function i think) | 13:50 |
| minute | i think it stops working when you use tinyusb | 13:51 |
| minute | artfwo: but i don't actually see anything we could try right now in the code hmmmm | 14:01 |
| minute | it is strange that poweroff fixes it | 14:01 |
| minute | artfwo: if you press some keys in tio, do you get a reaction from rp2040? (you shouldn't) | 14:02 |
| minute | artfwo: i mean a reaction in that terminal, like "acm command" | 14:02 |
| artfwo | minute: nope, no reaction to keys | 14:11 |
| minute | artfwo: ok! | 14:17 |
| minute | artfwo: then we'll have to look in depth tonight | 14:17 |
| + bkeys (~Thunderbi@45.134.140.153) | 14:59 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 15:17 | |
| + IchikaZou (~IchikaZou@110-30-144-5.adsl.fetnet.net) | 15:41 | |
| - IchikaZou (QUIT: Remote host closed the connection) (~IchikaZou@110-30-144-5.adsl.fetnet.net) | 15:41 | |
| - mark_ (QUIT: Ping timeout: 256 seconds) (~mjw@gnu.wildebeest.org) | 16:22 | |
| - cobra (QUIT: Ping timeout: 240 seconds) (~cobra@user/Cobra) | 16:31 | |
| + chomwitt (~chomwitt@2a02:587:7a21:2700:1ac0:4dff:fedb:a3f1) | 16:32 | |
| - artfwo (QUIT: Quit: Client closed) (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 17:13 | |
| + artfwo (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 17:18 | |
| - artfwo (QUIT: Quit: Client closed) (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 17:25 | |
| + artfwo (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 17:25 | |
| + vagrantc (~vagrant@2600:3c01:e000:21:7:77:0:50) | 17:32 | |
| - artfwo (QUIT: Ping timeout: 250 seconds) (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 17:33 | |
| minute | more pocket https://fosstodon.org/@pr8dan/112513686329927872 | 17:37 |
| minute | also, survival guide https://community.mnt.re/t/mnt-pocket-reform-essentials/2094 | 17:37 |
| josch | boosting *all* the things! | 18:10 |
| josch | the oled screen looks really weird with just two cells XD | 18:11 |
| + jacobk (~quassel@47-186-109-194.dlls.tx.frontiernet.net) | 18:11 | |
| + cobra (~cobra@user/Cobra) | 18:24 | |
| - bkeys (QUIT: Remote host closed the connection) (~Thunderbi@45.134.140.153) | 18:58 | |
| + mark_ (~mjw@gnu.wildebeest.org) | 19:15 | |
| minute | josch: haha | 19:28 |
| + artfwo (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 19:57 | |
| + gustav28 (~gustav@c-2834524e.019-141-67626730.bbcust.telenor.se) | 20:02 | |
| + pocketminute (~reform@p509936c9.dip0.t-ipconnect.de) | 20:35 | |
| - mjw (QUIT: Killed (tantalum.libera.chat (Nickname regained by services))) (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 21:24 | |
| * mark_ -> mjw | 21:24 | |
| + Guest9054 (~mjw@2001:1c06:2488:1400:4fd:39a7:74ac:7bae) | 21:25 | |
| artfwo | on pocket/imx8m plus, how to bring wifi device back, if powered off by accident? i have no idea what happened, but bluetooth and wifi have been switched off after a reboot and while i can restore bluetooth via sway panel, wifi connection doesn't show up anymore | 21:26 |
| artfwo | `nmcli radio` also shows that WIFI-HW is "missing" | 21:26 |
| artfwo | i recall a similar issue has been sporadically happening with imx8m reform, where a reboot would normally fix it | 21:28 |
| minute | artfwo: ok so the wifi is SDIO based, what do you get for `sudo dmesg | grep -i sdio` ? | 21:36 |
| minute | artfwo: also, `sudo rfkill | 21:37 |
| artfwo | minute: `mmc1: new ultra high speed DDR50 SDIO card at address 0001` | 21:37 |
| artfwo | only this, rfkill is not installed unfortunately :) | 21:37 |
| minute | artfwo: ah, rfkill not available even with sudo>? | 21:38 |
| artfwo | ah, ok, it is with sudo | 21:38 |
| artfwo | thanks for the tip | 21:38 |
| minute | artfwo: also, `lsmod | grep qcacld` should show qcacld2 | 21:38 |
| artfwo | ok, so qcacld not loaded, rfkill only sees bluetooth | 21:39 |
| minute | ohh ok... probably kernel update but the wifi driver did not update :/ | 21:39 |
| + reform21161 (~nasendack@p200300ddaf452f0069afe0cc6615afed.dip0.t-ipconnect.de) | 21:39 | |
| artfwo | yeah, the module doesn't seem to exist in /lib/modules | 21:40 |
| minute | artfwo: no, it is in /opt | 21:40 |
| artfwo | the .ko is there | 21:40 |
| minute | yes but probably a different version than uname -v shows? | 21:41 |
| artfwo | but trying to insmod it results in `invalid module format` | 21:41 |
| artfwo | `uname -v` is `6,8,9-1+reform20240518T141249Z (2024-05-18)` | 21:41 |
| minute | artfwo: and the .ko has a different timestamp in the name probably? | 21:42 |
| artfwo | yes, it's Apr 20 | 21:42 |
| minute | ok, i wonder how that happened... did you do `apt upgrade` or upgraded only the kernel module? | 21:42 |
| artfwo | i didn't upgrade anything, but installed a bunch of packages, and restarted the system | 21:43 |
| minute | ah ok, so somehow the package installation must have pulled in a newer kernel package | 21:44 |
| artfwo | the module version is 6.7.12-reform2-1+reform20240503T131044Z | 21:44 |
| minute | ok, it's from the previous kernel version | 21:44 |
| artfwo | right, so can I boot the previous kernel somehow? | 21:44 |
| - reform21161 (QUIT: Remote host closed the connection) (~nasendack@p200300ddaf452f0069afe0cc6615afed.dip0.t-ipconnect.de) | 21:45 | |
| minute | i think you can apt remove linux-image-6.8.9-... | 21:45 |
| minute | it should then rebuild the boot script etc to go back to the older version | 21:46 |
| artfwo | ehh, it tries to remove linux-image-arm64 and linux-image-mnt-reform-arm64 as dependent packages | 21:47 |
| minute | ew | 21:47 |
| artfwo | on other hand, i can reinstall them probably, so the system is upgradeable later? | 21:48 |
| minute | another option would be to use bluetooth or usb tethering (for example to phone) and only `apt install reform-qcacld2` | 21:49 |
| minute | josch: we urgently need some kind of hotfix that ensures that reform-qcacld2 is upgraded whenever linux-image is upgraded. do you have an idea? | 21:49 |
| minute | josch: i didn't expect that other packages would upgrade only the kernel... | 21:50 |
| minute | josch: maybe some post install hook we could put on the kernel package? | 21:55 |
| josch | minute: you cannot upgrade from within an apt hook because apt is running at that point | 21:57 |
| minute | josch: can we Recommends: the package maybe? hm hm | 21:58 |
| josch | minute: you cannot even Depends because that would have the reverse effect on the installation order | 22:00 |
| - pocketminute (QUIT: Ping timeout: 260 seconds) (~reform@p509936c9.dip0.t-ipconnect.de) | 22:01 | |
| minute | josch: any other ideas? | 22:02 |
| minute | josch: does debian have any mechanism for hooks running after apt unlocks the db? | 22:03 |
| josch | minute: honestly i have none -- i was sincere when i told you back then that i didn't know how to implement this. I honestly still don't know. | 22:03 |
| josch | no, it does not | 22:03 |
| minute | ah ok! | 22:03 |
| minute | ok, then we could do a hook that starts a process in the background that waits for the install to be complete | 22:04 |
| josch | yes, that could be done | 22:04 |
| minute | and it checks the platform + kernel + wifi package compat | 22:04 |
| minute | if not compat, it auto-upgrades reform2-qcacld | 22:04 |
| josch | during package installation you are root, so you can theoretically do anything | 22:04 |
| - vagrantc (QUIT: Quit: leaving) (~vagrant@2600:3c01:e000:21:7:77:0:50) | 22:05 | |
| minute | ok, now it's a bit late and i'll get tired so i'll install this tomorrow | 22:05 |
| minute | unfortunately a bunch of people will nuke their wifi now | 22:05 |
| artfwo | reform2-qcacld could be a direct dependency of linux-image perhaps? | 22:05 |
| artfwo | or patched into the image package | 22:06 |
| minute | s/install/implement | 22:06 |
| minute | artfwo: yeah, i was thinking about that... but the image package is quite unwieldly | 22:06 |
| artfwo | ok, huge thanks for the tip anyways! | 22:06 |
| minute | artfwo: no problem, sorry for this major inconvenience | 22:07 |
| artfwo | i'll try to get network running via bluetooth or usb tethering | 22:07 |
| + pocketminute (~reform@p509936c9.dip0.t-ipconnect.de) | 22:08 | |
| minute | maybe we could put a Breaks: on the linux-image package for all older versions of reform-qcacld2 | 22:10 |
| minute | josch: would that work? | 22:10 |
| minute | i wonder if apt would figure out that it could update reform-qcacld2 to solve it | 22:11 |
| josch | you'd want something that is both: let the kernel upgrade reform-qcacld2 while at the same time order the postinst of reform-qcacld2 so that it runs after the kernel is installed | 22:14 |
| josch | since we are talking about hacks, maybe hijacking the kernel .deb would indeed be the more reliable option | 22:14 |
| - gustav28 (QUIT: Quit: Quit) (~gustav@c-2834524e.019-141-67626730.bbcust.telenor.se) | 22:15 | |
| josch | let the final job in reform-debian-packages repack the kernel image .deb and inject the reform-qcacld2 module into it | 22:15 |
| josch | because as you said: the kernel package build itself is quite special and fragile enough | 22:16 |
| minute | josch: ohh, a combined .deb? i didn't know that was possible | 22:16 |
| minute | josch: yeah... | 22:16 |
| josch | well... it's a computer so mostly anything is possible -- it's just a question of how hacky it is :D | 22:16 |
| minute | josch: would... you be able to sketch that up? :D | 22:16 |
| minute | i like it because it sounds way more robust | 22:16 |
| josch | from an engineering perspective, with the reform-qcacld2 package we are already in the land of funny hacks :) | 22:17 |
| minute | hehe indeed | 22:17 |
| minute | i'm at the moment working on kernel patches that work around missing kernel features for dsi link rate negotiation | 22:17 |
| josch | no, sorry, this really doesn't interest me -- i very much dislike the way this is going | 22:18 |
| josch | and i was foreseeing trouble in the past already when you proposed this solution | 22:18 |
| josch | this is just adding more trouble on top | 22:18 |
| minute | josch: oh, ok! | 22:18 |
| josch | it is unfortunate that the hardware needs an out-of-tree driver this large :/ | 22:19 |
| minute | josch: well, unfortunately the ath10k people weren't able to help at all | 22:19 |
| minute | josch: ok, lets go back a few steps. out-of-tree drivers are not per se a totally bad/weird/unusual thing, right? i mean, there is dkms for this, but it assumes that the driver is compiling quickly | 22:21 |
| artfwo | can it be packaged with dkms? | 22:21 |
| artfwo | exactly | 22:21 |
| minute | we had it as dkms at first but it takes forever to build on imx8mp | 22:22 |
| minute | which is also a really bad user experience | 22:22 |
| minute | but dkms is a mechanism which runs automatically for kernel upgrades. but as i understand it cannot deploy binary drivers (?!) | 22:23 |
| josch | it doesn't deploy, it just builds the .ko and stores it locally | 22:23 |
| minute | i guess it could also download a .ko and store that locally? would that be better or worse than the .deb that we have in your book? | 22:24 |
| minute | or download and build from source as a fallback in case there's no internet | 22:24 |
| josch | how would it verify that it downloaded the right thing? | 22:25 |
| minute | hashes stored in the dkms package | 22:25 |
| josch | you could do the downloading in a script in /etc/kernel/postinst.d | 22:27 |
| - pocketminute (QUIT: Remote host closed the connection) (~reform@p509936c9.dip0.t-ipconnect.de) | 22:27 | |
| + pocketminute (~reform@p509936c9.dip0.t-ipconnect.de) | 22:28 | |
| minute | hmm https://unix.stackexchange.com/a/533221 | 22:29 |
| - pocketminute (QUIT: Client Quit) (~reform@p509936c9.dip0.t-ipconnect.de) | 22:29 | |
| josch | it seems that "dkms mkdeb" would allow to package up a dkms module but that doesn't get you any further than the package we have right now, no? | 22:32 |
| josch | https://github.com/dell/dkms/blob/master/README.md#installation-on-systems-with-no-module-source-andor-compiler also seems an interesting read | 22:34 |
| minute | josch: this is also how i understand it | 22:34 |
| josch | how long does the build actually take on imx8mplus? | 22:35 |
| minute | josch: not sure, maybe 15 minutes? | 22:35 |
| josch | uff... | 22:35 |
| minute | dkms seems like it _should_ be exactly the right solution for this usecase | 22:37 |
| minute | but i don't understand yet how to ship binary modules with it | 22:37 |
| josch | so, for a proper debian dkms package (which builds things) the sequence is /etc/kernel/postinst.d/dkms -> /usr/lib/dkms/dkms_autoinstaller -> dkms autoinstall --kernelver $kernel | 22:39 |
| - chomwitt (QUIT: Ping timeout: 272 seconds) (~chomwitt@2a02:587:7a21:2700:1ac0:4dff:fedb:a3f1) | 22:40 | |
| josch | so maybe the thing you want to do is to add a script into /etc/kernel/postinst.d/qcacld_dkms and let that script do the downloading and then call dkms with the downloaded tarball? | 22:41 |
| minute | josch: yes, i reread your message from before and only now understood that there is /etc/kernel/postinst.d, first i thought this was something in the kernel package | 22:42 |
| josch | minute: you could also try removing reform-qcalcd and installing ezurio-qcacld-2.0-dkms and checking how long it really takes to install and how bad the usability is? or did you try that already at some time in the past? | 23:02 |
| minute | josch: i wouldn't have proposed the binary solution if i hadn't tried that originally | 23:02 |
| josch | okay | 23:03 |
| minute | what i meant to say is, we tried the build-from-source strategy first, right | 23:04 |
| minute | just for fun testing that again now | 23:05 |
| minute | build started at 23:05 (installed the package) | 23:05 |
| minute | build fials | 23:05 |
| minute | fails | 23:05 |
| minute | josch: the dkms package is not updated with the strlcpy fix | 23:06 |
| josch | ah i found it: https://source.mnt.re/reform/qcacld2/-/commit/20496c23b602fbc33ff16b9e9496f06da2d3471b | 23:08 |
| minute | yeah | 23:08 |
| minute | josch: until i understand dkms again/fully, this would be my naive approach https://source.mnt.re/reform/qcacld2/-/merge_requests/2/diffs | 23:12 |
| josch | i think these days, apt waits until it acquires the lock, right? | 23:13 |
| minute | exactly | 23:13 |
| minute | but i don't know if my backgrounding approach works | 23:13 |
| minute | will need to test it irl first | 23:13 |
| josch | it's just stuck for a long time on "Building module(s)..." | 23:37 |
| josch | is usability better if the output would be verbose? | 23:41 |
| josch | in /etc/dkms/framework.conf one can set verbose=1 | 23:41 |
| josch | minute: thank you for testing ezurio-qcacld-2.0-dkms and providing feedback -- a fixed version is uploaded now and will be in mirrors in a few hours | 23:45 |
| josch | completely unrelated, reform-setup-wizard on s390x looks like this: https://mister-muffin.de/p/rY9A.png | 23:51 |
| josch | might just be an endian-problem... | 23:51 |
| artfwo | pocket reform cat https://imgur.com/a/5GyZCIU | 23:54 |
| - artfwo (QUIT: Quit: Client closed) (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 23:54 | |
| + artfwo (~artfwo@2a02:8109:8511:1800:6623:20ce:3860:f0f5) | 23:55 | |
| artfwo | pocket reform cat https://imgur.com/a/5GyZCIU | 23:55 |
| josch | cuuute! <3 | 23:56 |
Generated by irclog2html.py 2.17.2 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!
