Wizzup | maybe I can remove the gconf entries as well | 00:00 |
---|---|---|
Wizzup | going to sleep first :) | 00:00 |
stan | on one device tonight, when usb is plugged-in and battery is full, i get a reboot loop shortly after kexec exits. if i pull the usb power, it boots fine | 01:19 |
sicelo | I can hotswap sim on d4 with android | 06:46 |
lel | parazyd created a repository: https://github.com/maemo-leste/libicd-provider-tor | 10:29 |
Wizzup | freemangordon: do you think it makes sense to allow specifying custom network_modules lists on a per-IAP basis | 11:28 |
Wizzup | currently this is the code in icd_iap_connect: | 11:31 |
Wizzup | modules = (GSList *)g_hash_table_lookup(icd_ctx->type_to_module, | 11:31 |
Wizzup | iap->connection.network_type); | 11:31 |
Wizzup | iap->network_modules = modules; | 11:31 |
Wizzup | we could perhaps allow customising it or something | 11:31 |
freemangordon | Wizzup: how is that helpful? do you think about srv modules perhaps? | 12:10 |
Wizzup | no,sec | 12:11 |
Wizzup | (need 10 mins or so) | 12:13 |
meldrian | So, it 'seems' that besides the sim slot the sd-slot @ my n900 is broken too. Any chance installing maemo leste via usb? | 12:30 |
Wizzup | meldrian: hm... | 12:30 |
Wizzup | yes, but it's non trivial | 12:31 |
Wizzup | you can load u-boot via 0xFFFF | 12:31 |
Wizzup | and then perhaps load some image from emmc that you then use for flashing emmc | 12:31 |
Wizzup | freemangordon: so I am not sure if srv modules can be chained, and some modules we might want to run before setting the IP on the interface | 12:32 |
Wizzup | freemangordon: srv_provider is always after IP is set | 12:32 |
uvos__ | meldrian: also clean the contacts maybe, from memory the n900 sdcard slot is a bit wierd in that its not self cleaning, beacuse the contacts are simply pressed down instead of sliding across eatch other | 12:33 |
freemangordon | Wizzup: do you have anything particular in mind? | 12:47 |
freemangordon | IOW: what is the usecase? | 12:47 |
Wizzup | freemangordon: one example is a tails like tor setup where you don't want any packages beyond say icmp and dhcp to leave the device, and everything else goes through tor | 12:47 |
Wizzup | s/packages/packets/ | 12:48 |
Wizzup | that requires setting up iptables rules before ip is set to up | 12:48 |
Wizzup | we could just always load the module and make it mostly a no-op, of course | 12:48 |
freemangordon | and what is the issue if all IAPs share common list of modules? | 12:49 |
freemangordon | yeah, as you said | 12:49 |
Wizzup | no issue, we can just do that | 12:50 |
freemangordon | it makes no difference if it will be loaded once or many times, actually memory wise it should be cheaper to load every module for every iap - no additional lists. | 12:50 |
freemangordon | I would say - keep it as it is unless we see we cannot implement something without this functionality | 12:51 |
freemangordon | I guess it will require UI as well | 12:51 |
Wizzup | yes, but the same is true for providers | 12:52 |
Wizzup | I am still trying to get a grasp of the some of the provider specific func, but we're getting there | 12:52 |
freemangordon | Wizzup: agree, I guess we shall discuss how is it supposed to work UI wise | 12:53 |
Wizzup | I did some exploration on the existing parts in connui, but there is more to do | 12:57 |
Wizzup | https://raw.githubusercontent.com/maemo-leste/libicd-provider-dummy/master/README.rst | 12:57 |
freemangordon | right, there is a place for improvement | 12:59 |
freemangordon | and not everything is clear | 12:59 |
freemangordon | It is still not clear to me how do you select SP for a particular IAP | 13:00 |
Wizzup | freemangordon: it's listed as a separate network in the scan | 13:03 |
Wizzup | my current dummy service provider 'matches' with all wlan and dummy aps | 13:03 |
Wizzup | so you'll see the twice the amount of networks | 13:03 |
Wizzup | of course, we probably don't want it this way | 13:03 |
freemangordon | mhm | 13:03 |
Wizzup | the identify function could check if the iap is known and has the service enabled in gconf, for example | 13:03 |
Wizzup | in any case, progress | 13:08 |
Wizzup | brb | 13:08 |
Danct12 | sorry for being offtopic, but i just compiled kernel 5.11 for my n900, it seems like it does boot as the kb backlight turns on but display doesn't work and the bl is on | 13:09 |
uvos | sicelo: ^^^ | 13:09 |
uvos | i think some issue existed with 5.11 | 13:09 |
uvos | maybe try 5.13 | 13:10 |
uvos | also keeping the kernel working on n900 is sufficantly on topic | 13:10 |
stan | has anyone scripted something to quickly reconnect wlan when connection is lost? | 13:25 |
Wizzup | no, because that's something that if considered a bug would be better to resolve in icd2, e.g. a lower reconnect timer | 13:27 |
Wizzup | that said, python-conic is in the repo, so if you find python easy, it sohuld be trivial to script | 13:28 |
Wizzup | https://github.com/maemo-leste/python-conic | 13:28 |
stan | thanks! | 13:28 |
uvos | stan: btw the buttons below the screen are driven by i driver i wrote and are conifgured in dts | 13:28 |
freemangordon | stan: is auto-connect enabled? | 13:29 |
uvos | by default they are simply f7 f8 f9 and xf86search | 13:29 |
uvos | freemangordon: its somewhat broken, we know allready | 13:29 |
freemangordon | uvos: hmm, I think it should be fixed, no? | 13:29 |
uvos | no | 13:29 |
stan | i lose my ssh session, on one device. other device has no problems with wlan. | 13:29 |
uvos | its goten worse freemangordon | 13:30 |
uvos | for some reason scanns that are triggerd by icd and completed by wpasupplicant | 13:30 |
uvos | never make it to icd/connui | 13:30 |
uvos | Wizzup has details | 13:30 |
freemangordon | ah, Wizzup has to fix the plugin | 13:30 |
Wizzup | freemangordon: uvos: if you go out of range of a network, you will wait for 5 minute timer | 13:30 |
Wizzup | this is unrelated to the plugin | 13:30 |
Wizzup | this is just how it works on fremantle as well | 13:30 |
freemangordon | yes | 13:30 |
uvos | autoconnect also dosent work for me | 13:30 |
Wizzup | it works for me, but this is unrelated I think | 13:31 |
uvos | so even 5min wat dosent save me | 13:31 |
uvos | ok | 13:31 |
freemangordon | uvos: what is your timer setting? | 13:31 |
uvos | *wait | 13:31 |
uvos | freemangordon: just defaults | 13:31 |
uvos | and autoconnect | 13:31 |
freemangordon | autoconnect how often? | 13:31 |
uvos | whatever is the default i dont have the device atm | 13:31 |
freemangordon | stan: could you please check and report your auto-connect settings? | 13:32 |
Wizzup | his request was clearly worded: reconnect after connection was dropped | 13:32 |
freemangordon | from the status menu | 13:32 |
Wizzup | what | 13:33 |
freemangordon | is the answer | 13:33 |
freemangordon | ah, wait | 13:33 |
freemangordon | we talk ssh :) | 13:33 |
Wizzup | this is just about for example being out of reach of the ap | 13:33 |
Wizzup | or the ap restarting | 13:33 |
uvos | oh btw | 13:33 |
uvos | for some reason if a have a ssh session to a leste device running | 13:33 |
uvos | and the connection drops | 13:34 |
uvos | ssh just hangs | 13:34 |
stan | the search interval is 5 minutes freemangordon . can you be more specific what you want | 13:34 |
Wizzup | you can configure keepalive and such in ssh | 13:34 |
uvos | while if the same thing happens with my laptop the session gets dissconected fast | 13:34 |
uvos | what setting is that | 13:34 |
uvos | Wizzup: thanks | 13:34 |
stan | connect automatically: wifi | 13:34 |
freemangordon | stan: sorry, I misunderstood your issue | 13:35 |
stan | mosh can be helpful as alternative to ssh if connection is unreliable | 13:37 |
Danct12 | uvos, i tried 5.13 and it has more interesting result | 14:10 |
Danct12 | it stucks at "starting kernel" and the phone shuts off a few seconds later | 14:11 |
Wizzup | What config are you using? | 14:12 |
Wizzup | if watchdog is a module, it likely won't work well a tall | 14:12 |
Wizzup | iirc someone tried this last week as well, might be worth searching the logs | 14:12 |
Wizzup | parazyd: btw there are also service_attributes, but we don't use that atm | 14:17 |
Wizzup | parazyd: also iap_scan_default_sort_func looks at preferred_type and preferred_id for service providers | 14:19 |
Wizzup | parazyd: finally, it looks like a provider can also have different service-ids, I am not sure what for, but they have their own gconf paths | 14:23 |
Wizzup | uvos: when you find a moment, please file a brief bug for the pdf crash that you're seeing | 14:24 |
Danct12 | Wizzup, omap_watchdog and twl4030_watchdog is set to y | 14:25 |
Wizzup | Danct12: ok, then you have another problem I think :) | 14:27 |
Wizzup | did you grep logs for someone who tried it more recently? | 14:27 |
Danct12 | i think i'm the only one who tried it on 5.13 | 14:27 |
Danct12 | probably i'll go back to 5.11 and use config_fb_omap instead | 14:28 |
Danct12 | i have seen an email on the mailing list that their n900 doesnt show anything but i checked and the fix was applied | 14:28 |
Danct12 | no dice | 14:28 |
Wizzup | ok, building kernel is tricky and often (at least iny my case) pebcak, but ymmv, and it's good to stay on top | 14:29 |
Wizzup | still, I think the person who tried this the other week was doing 5.13 | 14:29 |
Wizzup | sicelo might remember | 14:29 |
Danct12 | then let's blame him >:) | 14:29 |
Wizzup | blame? :P | 14:30 |
Wizzup | freemangordon: wonder what this does, looks like that key being set results in ... something being added, but the key value itself doesn't seem to matter? https://github.com/maemo-leste/connui-common/blob/master/connui/iap-scan.c#L1227 | 14:31 |
Wizzup | freemangordon: I'm making notes, so no need to jump in now, jfyi | 14:31 |
Danct12 | blame because it works fine but doesnt work for others | 14:32 |
Wizzup | Danct12: what defconfig did you use again? | 14:38 |
Danct12 | https://paste.ubuntu.com/p/H8DwC2MfMZ/ | 14:38 |
Danct12 | this | 14:38 |
Wizzup | is that a standard defconfig? | 14:39 |
Danct12 | i actually started from the linux-armv7 defconfig from arch linux arm, above is all changes that i done | 14:39 |
Wizzup | how about omap2plus_defconfig ? | 14:39 |
Wizzup | stan: you could try changing this: | 14:53 |
Wizzup | gconftool -R /system/osso/connectivity | grep search search_interval = 300 | 14:53 |
Wizzup | so /system/osso/connectivity/search_interval | 14:53 |
Wizzup | not sure if it will be honoured | 14:54 |
Wizzup | gconftool -t string -s /system/osso/connectivity/srv_provider/DUMMY/custom_ui/dummy-provider/type_icon_name gvim | 15:18 |
Wizzup | this works :p | 15:18 |
Wizzup | https://wizzup.org/provider-icon.png | 15:23 |
Wizzup | https://wizzup.org/provider-icon-2.png | 15:24 |
stan | gconftool -t int -s /system/osso/connectivity 10 | 15:25 |
stan | gconftool -R /system/osso/connectivity |grep search // search_interval = 300 | 15:25 |
Wizzup | you need to set the search_interval key. | 15:25 |
Wizzup | not the connectivity _directory_ | 15:25 |
stan | gconftool -t int -s /system/osso/connectivity/search_interval 10 | 15:26 |
stan | gconftool -R /system/osso/connectivity |grep search | 15:26 |
stan | search_interval = 10 | 15:26 |
stan | search_interval = 300 | 15:26 |
stan | the key has a preceding space perhaps? | 15:27 |
stan | not pursuing this. using a closer AP | 15:38 |
stan | thanks | 15:38 |
Danct12 | Wizzup, screw this i'm going to start again from omap2plus | 15:39 |
Wizzup | Danct12: check, try that but make sure watchdog and some other things are not a module | 15:39 |
uvos | interesting volume slider you have there | 15:47 |
uvos | Wizzup: any chance of just making it change pulse master volume as a temporary hack | 15:47 |
uvos | Wizzup: ill do a bug later | 15:49 |
uvos | (pdf) | 15:49 |
Wizzup | uvos: 'it' ? | 15:57 |
Wizzup | the volume applet? | 15:58 |
Wizzup | uvos: oh I see | 15:58 |
Wizzup | uvos: yeah it's https://github.com/maemo-leste/maemo-statusmenu-volume | 15:58 |
Wizzup | I could try to do what you suggest | 15:58 |
Wizzup | as you can see it already deals with in call lvolume as well | 15:59 |
Wizzup | just needs the pa setup and modules | 15:59 |
bencoh | neat :) | 16:06 |
bencoh | I didn't think it would be so "complex" though | 16:10 |
Wizzup | this is a clone of the fremantle one | 16:11 |
bencoh | RE-ed? | 16:11 |
Wizzup | for compatibility reasons | 16:12 |
Wizzup | interoperability* | 16:12 |
freemangordon | bencoh: yes, REed | 16:18 |
Wizzup | freemangordon: iap_settings_get_name has double free | 16:19 |
Wizzup | https://github.com/maemo-leste/connui-common/blob/master/connui/iap-settings.c#L348 | 16:19 |
Wizzup | https://github.com/maemo-leste/connui-common/blob/master/connui/iap-settings.c#L355 | 16:19 |
Wizzup | if s fails to be created at least | 16:19 |
Wizzup | parazyd: https://wizzup.org/provider-icon-3.png | 16:22 |
Wizzup | including italic markup :p | 16:22 |
uvos | Wizzup: btw did you ever kick the icd pluginns? | 16:23 |
Wizzup | uvos: no, not yet, freemangordon is back now, so he might be able to help | 16:23 |
Wizzup | freemangordon: this is not a theoretical double free, I hit it | 16:25 |
Wizzup | (of course I was experimenting with services, but still) | 16:25 |
Wizzup | s/services/providers/ | 16:25 |
bencoh | Wizzup: wait, wth is that "vim" icon doing there? | 16:25 |
bencoh | is it a unicode emoji? | 16:25 |
Wizzup | bencoh: no, this is me testing custom service icons | 16:27 |
Wizzup | service providers* | 16:27 |
Wizzup | you can imagine it could be a tor icon | 16:27 |
bencoh | neat | 16:27 |
bencoh | right, tor ... | 16:27 |
Wizzup | I had tor installed but it did not ship with an icon | 16:28 |
bencoh | I have to admit I like where this whole project is going :) | 16:28 |
Wizzup | me too, I just hope we get where we want to be within a reasonable amount of time :-D | 16:28 |
bencoh | wait, do we have a system-wide torrify-like? | 16:28 |
Wizzup | bencoh: it's been worked on, it's not there yet | 16:28 |
bencoh | is it "just" setting a socks proxy for everything that supports it? or something iptables-enforced? | 16:29 |
Wizzup | I think the thought is to have a way to control the tor service as optional useful thing, but also to have a iptables enforced mode, depending on what the user picks | 16:29 |
Wizzup | parazyd has a better sense of it currently | 16:29 |
Wizzup | he's working on it as part of https://dapsi.ngi.eu/hall-of-fame/maemo-leste/ | 16:30 |
Wizzup | https://maemo-leste.github.io/funding-from-ngi-eu.html | 16:30 |
bencoh | <3 | 16:31 |
Wizzup | :) | 16:32 |
freemangordon | Wizzup: yeah, double free | 16:34 |
freemangordon | you want me to fix it? | 16:34 |
Wizzup | if you can, please | 16:35 |
Wizzup | I can also do it, but I need a break from work and hobby soon | 16:35 |
freemangordon | sure. is it urgent? | 16:35 |
Wizzup | not really, just don't forget :p | 16:35 |
freemangordon | ok | 16:35 |
freemangordon | I won;t | 16:35 |
Wizzup | I've worked around it but actually making dgettext succeed | 16:35 |
Wizzup | s/but/by/ | 16:35 |
freemangordon | I hate double-frees just slightly less then memleaks :) | 16:36 |
bencoh | :] | 16:36 |
bencoh | still hate memleaks more though :) | 16:37 |
bencoh | Wizzup: I tried building vncviewer (from maemo extras) and it segfaults when trying to use it :( | 16:39 |
bencoh | (to be fair it asserts, actually) in case anyone is interested: http://pastebin.notk.org/pastebin.php?show=f63545017 | 16:44 |
freemangordon | bencoh: that's why 'slightly less than' ;) | 16:45 |
bencoh | :) | 16:45 |
Wizzup | bencoh: looks like this actually expects a 16bit X | 16:47 |
Wizzup | I'd have to open the code to see why | 16:48 |
Wizzup | (going for a swim now first :-)) | 16:48 |
bencoh | (enjoy!) | 16:48 |
stan | did you make sailfish port for droid4 bencoh? | 16:54 |
stan | sailfish has a nice qt irc client | 16:55 |
bencoh | I only remotely had a quick look at sailfish years ago, so ... nope | 16:56 |
bencoh | it might run on maemo, assuming it's qt5-based (it most probably is) | 16:57 |
bencoh | except for the silicate qml components ... iirc they make heavy use of those | 16:57 |
stan | irssi works well with droid4 keyboard, so ok | 17:09 |
bencoh | :) | 17:21 |
stan | thank you for the onscreen buttons uvos. i was hoping they could be used to send 'Esc' to SDL. Right now that's a thorny problem, getting our mapped keys to sdl1.2 apps (they are really not designed for translated keys) | 17:22 |
bencoh | I wonder how hexchat would perform | 17:22 |
stan | On the other hand, for good playability i am finding most sdl games really need work to be playable anyway, so it's not a problem to change the input from sym->sym to sym->unicode, and then the Esc works | 17:23 |
stan | i am currently looking how to patch the n900 sdl hildon animation actor to work in 32bpp mode | 17:26 |
stan | powered usb hub is quite nice on droid4 also. can feed droid4 with power, hook up storage, keyboard, mouse | 18:02 |
uvos | stan: which hub works for you | 18:39 |
uvos | stan: i have one that dosent work on mapphones for some reason (no android either) | 18:40 |
uvos | works elsewhere | 18:40 |
uvos | the funding thing is nice but im a bit worryed it could distract from the "basic functionality" stuff | 18:43 |
uvos | but no matter | 18:43 |
stan | i don't know uvos, i can get a picture of the only IC on the board | 18:44 |
uvos | sure also lsof -v might help in this case | 18:44 |
stan | https://www.ebay.de/itm/173307495965?hash=item2859ee861d:g:rwMAAOSw1fJa8agr | 18:44 |
uvos | er lsusb -v | 18:45 |
stan | that one by appearance and price | 18:45 |
uvos | ok | 18:45 |
stan | Bus 001 Device 020: ID 05e3:0608 Genesys Logic, Inc. Hub | 18:46 |
stan | i get both power injection and peripherals in the '3' position uvos | 18:48 |
bencoh | uvos: well, telepathy is apparently part of the funding plan, so ... I'd say it's pretty "basic functionality stuff" | 18:57 |
bencoh | as in, assuming we eventually want a working phone/communication handheld | 18:57 |
bencoh | Am I the only one with otg / usbnet issues on droid4? | 18:58 |
uvos | bencoh: no i have otg issues | 18:59 |
bencoh | ah | 18:59 |
uvos | bencoh: so some devices work | 18:59 |
uvos | and others dont | 18:59 |
bencoh | here it shows up as connected on host, and disappears after a second or so | 18:59 |
uvos | that might be power | 18:59 |
uvos | check dmesg | 18:59 |
bencoh | power? | 18:59 |
uvos | d4 can provide like 200mAh i think | 18:59 |
uvos | if the device uses or negotiates more it gets cut off | 19:00 |
uvos | lots of devices just negotatie the full 500mA because they are stupid | 19:00 |
bencoh | oh mybad, I actually meant: when using droid4 as a usbnet device, that was silly of me | 19:00 |
uvos | Note: mA not mAh above | 19:00 |
uvos | oh | 19:00 |
uvos | you said otg | 19:00 |
bencoh | (when connecting it to a computer) | 19:00 |
bencoh | yeah I know, my bad :) | 19:00 |
uvos | usbnet no idea i dont use that | 19:00 |
stan | cheapo fast ethernet works too: asix 1-1.3:1.0 eth0: register 'asix' at usb-musb-hdrc.2.auto-1.3, ASIX AX88772 USB 2.0 Ethernet, 00:01:0a:0a:c4:7 | 19:26 |
Wizzup | uvos: yeah, it should not distract too much from basic stuff, I try to watch for that | 19:34 |
stan | did fxtec send you a pro1 Wizzup ? | 19:47 |
stan | i wonder if them getting linux on that snapdragon means leste could run on some other, prhaps cheaper qualcomm phones | 19:48 |
stan | on the othr hand perhaps we should take 'qual' as a warning | 19:49 |
parazyd | https://github.com/SoMainline/linux/branches | 19:52 |
parazyd | This is for mainline on the fxtec | 19:52 |
parazyd | afaik | 19:52 |
uvos | also https://wiki.postmarketos.org/wiki/Qualcomm_Snapdragon_835_(MSM8998) | 19:52 |
uvos | its early stages | 19:52 |
uvos | but qualcomm dosent mean bad mainline support | 19:53 |
Wizzup | stan: yes I have one | 19:53 |
uvos | MSM8916 has excellent support | 19:53 |
uvos | on par with omap4 really | 19:53 |
parazyd | oh cool I have a dumpling too | 19:53 |
uvos | better even because foss graphics | 19:53 |
parazyd | Didn't know it's also MSM8998 | 19:54 |
uvos | https://wiki.postmarketos.org/wiki/Qualcomm_Snapdragon_410/412_(MSM8916) | 19:54 |
Wizzup | parazyd: dumpling? | 19:56 |
uvos | OnePlus 5 | 19:56 |
uvos | either that or parazyd is hungry | 19:56 |
parazyd | lol | 19:56 |
parazyd | Yeah, the 5T | 19:56 |
parazyd | About to have dinner though :D | 19:57 |
uvos | guten apetit | 20:07 |
parazyd | Thanks :) | 20:09 |
Wizzup | parazyd: chevapi? | 20:13 |
parazyd | No such thing in Cataluña :p | 20:13 |
parazyd | (Wouldn't object though) | 20:14 |
Wizzup | ah right | 20:14 |
stan | Wizzup: does fxtec have opengl or gl-es for pro1 under linux? | 20:34 |
stan | i think the best homebrew portable right now would be build with khadas vim3 | 20:35 |
stan | 5-20v DC-in, open source drivers, decent cortex a72/73 performance | 20:35 |
Wizzup | last time I checked mainline did not have display yet | 20:35 |
stan | ty | 20:36 |
Wizzup | might have changed | 20:36 |
stan | would be fun to someday have OS on emmc and swap cards with just data | 21:50 |
stan | i understand the sense in booting os from sdcard now though | 21:51 |
stan | it's great to backup 'known good working state' with | 21:51 |
bencoh | stan: all of those sound good, as long as you're fine with mixed modem/cpu | 22:10 |
bencoh | otherwise, you're stuck with n900/droid4/pyra | 22:10 |
bencoh | (and pinephone/librem, mybad) | 22:10 |
stan | what is mybad? | 22:12 |
bencoh | ("my bad" for forgetting about pinephone and librem) | 22:12 |
stan | it would be a fun phone name :) | 22:13 |
bencoh | :] | 22:13 |
stan | i can get a green light hooking up usb power to the board with attached screen, but without bottom of case | 22:14 |
stan | it won't start though | 22:14 |
stan | tightvncserver runs fine on droid4 | 23:13 |
uvos | stan: please do not veer far off topic; mapphones can boot leste from emmc just copy leste to some emmc partition and change boot.cfg, cpcap wont release the cpu from resett unless it has a battery, mapphones can boot with just the bare mainboard and a battery but takes a while because mbm waits a while for the display to work. | 23:13 |
stan | thanks! | 23:14 |
uvos | and then times out and boots anyhow if there is no display | 23:14 |
stan | boot.cfg is in /boot/boot on sdcard. does kexec look for boot.cfg on emmc as well ? | 23:15 |
uvos | kexecboot looks for boot.cfg on every partiton it can mount | 23:15 |
uvos | so yes | 23:15 |
stan | that's great | 23:15 |
uvos | please read the documentation | 23:16 |
stan | if booted from emmc, can sd cards be unmounted/removed/remounted while system is running? | 23:16 |
uvos | like on https://github.com/tmlind/droid4-kexecboot instead of asking so manny questions | 23:16 |
stan | ok | 23:16 |
uvos | yes ofc there is nothing special about the sdcard if its not your root | 23:16 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!