**** BEGIN LOGGING AT Thu Aug 15 02:59:58 2019 Aug 15 04:04:34 any chances latest mt76 in 19.07 or 18.06? Aug 15 04:07:35 aparcar[m]: instead of dind (docker in docker) you should be able to use podman/buildah Aug 15 04:19:40 champtar: thanks will check it out Aug 15 05:52:42 are opkg updates distributed using a CDN? they've been v slow past 3-4 months Aug 15 06:34:12 aparcar[m]: well, it makes no sense to deactive it now if it's already done :) Aug 15 06:35:36 ynezz: I sent you a pm regarding troubled DNS on your worker. Did you receive it? Aug 15 06:39:04 indeed, need to get used to this new type of failures :) http://docker:2375/v1.40/auth: dial tcp: lookup docker on 8.8.8.8:53: no such host Aug 15 06:40:01 DNS works fine, the request looks strange Aug 15 06:42:41 probably the docker inside the container is trying to instruct host's docker via some API call over HTTP instead of that /var/run/docker.sock UNIX socket? Aug 15 06:46:26 BTW you don't need to run dind anymore if you've access to that host's UNIX socket inside container Aug 15 06:46:32 or do you? Aug 15 06:46:40 ^ aparcar[m] Aug 15 08:27:04 Strykar: not yet Aug 15 09:43:14 anyone using zram? Aug 15 09:43:49 looking for a way to test if it is loaded and availble. Ideally something like test -e /sys/modules/zram/refcnt Aug 15 09:43:55 but need a specific path to test Aug 15 10:22:00 <[mbm]> jow: /sys/module/zram/refcnt works Aug 15 10:35:29 [mbm]: jow: probably won't work if zram is builtin, though - zram registers /sys/class/zram-control (as a directory), so probably best check for that (also rebust against renaming of the module) Aug 15 10:40:11 [mbm]: KanjiMonster: thanks Aug 15 10:40:20 <[mbm]> even builtin stuff should get a /sys/module entry Aug 15 10:40:53 <[mbm]> eg, I've got a /sys/module/8250 but no 8250 in lsmod Aug 15 10:44:19 [mbm]: I see e.g. no ag71xx on my ar71xx device (or any other ar71xx drivers) - maybe only builtin modules with parameters get a directory there Aug 15 10:45:20 <[mbm]> hmm, that's a good point, zram doesn't have a parameters directory Aug 15 10:47:04 it has a parameter though, "num_devices" ... Aug 15 10:52:04 <[mbm]> yeah, dunno why that doesn't show up in /sys/module/zram Aug 15 10:52:40 <[mbm]> additionally I grepped for all the /sys/module stuff that didn't have a corresponding lsmod and found that they all had a parameter directory Aug 15 12:21:01 [mbm]: while you're here - do you happen to know the license of the original OpenWrt logo artwork? Aug 15 12:21:12 (https://openwrt.org/lib/tpl/openwrt/images/logo.png) Aug 15 12:21:49 <[mbm]> well, given that I drew the damn thing it's whatever I want it to be ;) Aug 15 12:22:00 <[mbm]> there's also an svg version somewhere Aug 15 12:22:24 yeah, but I couldn't find it anymore Aug 15 12:23:28 <[mbm]> https://web.archive.org/web/20110928110702/http://downloads.openwrt.org/people/mbm/openwrt.svg Aug 15 12:24:37 so is it okay to claim that the openwrt logo is under a CC0 (public domain) license? Aug 15 12:25:19 <[mbm]> yeah, some sort of cc license for the promotion of openwrt Aug 15 12:27:39 do you recall when you created it? Before 2009 ? Aug 15 12:28:15 <[mbm]> it was sometime around 2006, it went with the site redesign Aug 15 12:28:22 <[mbm]> lemme dig around archive.org Aug 15 12:30:16 <[mbm]> ... which seems to be down ... Aug 15 12:32:46 http version works here Aug 15 12:35:22 <[mbm]> seems to have been some time late june/july 2006 Aug 15 12:37:22 allright, noted. Aug 15 12:37:25 thanks! Aug 15 12:46:38 <[mbm]> hmm, has nobody created an luci for network bonding or am I just not seeing it? I'm playing around with 2.4 and 5ghz wifi clients to the same network and figured I'd combine them Aug 15 12:52:13 nope, this does not exist as far as I recall Aug 15 13:40:51 [mbm]: while you're here, restoring the irc cloaks of the project members would have the side effect of enabling more channel ops to deal with occasional issues Aug 15 13:41:22 gch981213: why do you add eth1 to the dts files in c7c807cb8c3fd65 ? the commit message just describes what you did, not anything about the why of it. Aug 15 13:49:15 karlp: To register mdio only, without corresponding eth block. commit 83d2dbc59924c (incorrectly) split eth/mdio driver. The added eth1 nodes has a compatible string of "syscon", "simple-mfd" which made its subnode (mdio-bus) registered as a separated platform device. Aug 15 13:50:29 karlp: As for why that's needed: phy used by eth0 can only be accessed through mdio on eth1. Aug 15 15:56:41 so before, eth0 referred to phy4, via a node within mdio0, but now I have an eth1 node, (which I don't have, and you have to know the compat strings to understand that) and no mdio node, so it just feels less readable, but if this is all part of the "make ath79 great" and having proper phy/mdio drivers then *shrugs* I guess this is the future :) Aug 15 16:18:23 quick question regarding ar71xx/ath79 and GL.iNET GL-AR150, with ar71xx, I have eth0/eth1 and the interface goes up/down when I plug/unplug ethernet cables. Now with ath79 eth0 is always up. I would say this is a know bug because the led reference the switch instead of the interface Aug 15 16:18:28 https://github.com/openwrt/openwrt/blob/cad1b474d961817dcd33c5cd2bdfc85ff4564798/target/linux/ath79/base-files/etc/board.d/01_leds#L94 Aug 15 16:18:44 is it a limitation of ath79 ? Aug 15 16:40:21 <[mbm]> DonkeyHotei: tbh I have no idea who set up the cloaks in the firat place Aug 15 16:41:33 <[mbm]> (also I'm always here, I just haven't been talkative lately) Aug 15 17:08:20 I've found https://patchwork.ozlabs.org/patch/1037756/ useful and doesn't appear broken to my eyes. Aug 15 17:09:06 wanted ipset counts Aug 15 17:11:55 nbd: I've affected my patch to you (https://patchwork.ozlabs.org/patch/1145995/), if you don't have time to test/commit it I'll find someone else, but because you are the original author of the EAP hack ynezz was waiting for your inputs (ie just ack/nack the idea) Aug 15 18:03:06 champtar: i've nudged him about it on a different channel Aug 15 18:03:24 champtar: i need to call him anyhow, will bring this up and ask him to respond Aug 15 18:04:53 would be nice if he could weigh in on FS# 2428 too Aug 15 18:05:18 DonkeyHotei: link ? Aug 15 18:05:50 nbd originally wrote the patch, and Hauke may have broken it: https://bugs.openwrt.org/index.php?do=details&task_id=2428 Aug 15 18:08:23 DonkeyHotei: sent a note Aug 15 18:08:38 potential consequences for any and all spi-nor devices Aug 15 18:08:46 oki **** ENDING LOGGING AT Fri Aug 16 03:05:29 2019