**** BEGIN LOGGING AT Thu Feb 04 03:02:58 2021 Feb 04 03:05:42 I clone down from a mirror which resyncs every 24 hours. So it could also be that. At that point I had probably ought to mirror feeds as well. Feb 04 03:05:56 let's try pointing at a not-mirror and see what we get. Feb 04 03:10:48 are you sure it even upgraded Feb 04 03:40:07 Where are the mvebu targets at? What Target System chipset? Feb 04 04:01:03 Grommish: mvebu has three separate platforms Feb 04 04:38:45 >KGB-1< https://tests.reproducible-builds.org/openwrt/openwrt_tegra.html has been updated. (0% images and 98.2% packages reproducible in our current test framework.) Feb 04 05:20:52 magix: Gotcha.. I managed to find it :) Feb 04 08:08:35 https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=7c8c4f1be648aff9f1072ee27a2cc8f6a4a788ef am I missing something or is there no change here? Feb 04 08:08:51 ldir-: working on fix now Feb 04 08:09:22 ack :-) Feb 04 08:15:01 zorun: Hauke: shouldn't we announce preparation of 19.07.7 in upcoming days/weeks? :) Feb 04 08:15:07 mangix: ^ Feb 04 08:27:32 ynezz: https://github.com/openwrt/packages/pull/14647 posted PR. Feb 04 08:28:55 anyone have an opinion on https://github.com/openwrt/packages/pull/14646 ? It's also applicable to base FWIW. Feb 04 08:39:37 ynezz: this hostapd issue is severe? Feb 04 08:40:46 fine for me in any case! Feb 04 09:07:57 zorun: hostapd has a problem in the p2p code, this is not used much in OpenWrt Feb 04 09:08:10 only the specail p2p hostapd version has it activated Feb 04 09:08:43 ynezz: you promised to wait 1 week for the CVE number for the IPv6 problem Feb 04 09:08:50 I would do an release after that Feb 04 09:09:43 +1 Feb 04 09:10:14 someone should send out some email, that we're planning to do that Feb 04 09:10:38 don't know if there are some pending fixes to be backported or such Feb 04 09:14:20 zorun: IIUC it's wpa-supplicant (client) P2P related, so probably "just" issue in wpa-supplicant-p2p Feb 04 09:16:29 This can result in an attacker within Feb 04 09:16:30 radio range of the device running P2P discovery being able to cause Feb 04 09:16:30 unexpected behavior, including termination of the wpa_supplicant process Feb 04 09:16:33 and potentially arbitrary code execution. Feb 04 09:16:43 so RCE Feb 04 09:50:29 ynezz: ok, thanks Feb 04 09:50:56 I can send the email this evening, if somebody does it first, fine Feb 04 09:51:18 I would like to test and integrate the workaround for the mt7621 switch issues, I'll try to do that this week-end Feb 04 09:53:00 https://bugs.openwrt.org/index.php?do=details&task_id=2628 Feb 04 13:12:30 hello, it seems to me like even latest OpenWrt uses lua 5.1... lua 5.3 has been released back in 2015 Feb 04 13:13:06 can anyone shine a light on why this component has not been updated for such a long time? Feb 04 13:13:18 any deep technical reasons for this? Feb 04 13:13:54 think of it like python2 vs python3 Feb 04 13:14:31 you should ask lua5.3 projects why they're not using lua 5.4..... ;) Feb 04 13:15:45 to be honest, I am not deeply familiar with the lua ecosystem. Only just started studying the topic when I got to work with OpenWrt very recently Feb 04 13:16:00 never really needed lua before Feb 04 13:16:19 karlp: so this is an API break? Feb 04 13:16:38 in that case, I'd just ignore the version and look at things more interesting toyou personally. Feb 04 13:16:51 for using it, it's mostly the same. Feb 04 13:17:55 karlp: well, there is a practical matter, really. I'm trying to build OpenWrt components using Yocto with meta-openwrt and was wondering why there is a version rollback in the layer Feb 04 13:18:05 because latest Yocto has lua 5.3 Feb 04 13:18:54 again, unless you're explicltly bothered by it for more concrete reasons, just let it go. Feb 04 13:19:16 so I am still confused whether it is a serious technical necessity or just for the sake of being consistent with OpenWrt userspace Feb 04 13:20:44 karlp: I'd rather not have an additional lua 5.1 recipe in my metadata. that's basically the reason Feb 04 13:21:13 karlp: if I naively just switch to 5.3, should I expect things to break badly? if so, where? Feb 04 13:22:04 won't you have _only_ 5.1? Feb 04 13:22:16 yes, you should expect things to break and or not even compile... Feb 04 13:22:21 you can work on that if that's your thing. Feb 04 13:24:30 karlp: in a distro, yes, you will only have 5.1, but in terms of metadata, you override the default 5.3 recipe in your layer (in Yocto terms) Feb 04 13:25:00 so metadata really has recipes for two versions Feb 04 13:25:09 oh noes... Feb 04 13:25:29 there's support in opðenwrt for having 5.3 as well as 5.1 Feb 04 13:25:34 if you want to do that too :) Feb 04 13:26:05 karlp: hypothetically, or are the Makefiles already there? Feb 04 13:31:17 openwrt/package/utils/lua/ only has 5.1.5 Feb 04 13:31:47 so I'll assume that was just a hypothetical statement that one can add an additional version Feb 04 13:33:56 look in packages. Feb 04 13:36:44 also, do you know hae packge/utils/lua5.3 ? Feb 04 13:37:08 it's been there since mid2019? Feb 04 13:38:23 karlp: thanks. I must be blind since it's right besides the lua directory Feb 04 13:39:24 but I assume one can not trivially change the version and something like LuCI still depends on 5.1? Feb 04 13:40:06 sure, works out of the box right now, we're just playing games with everyone harhar :) Feb 04 13:41:17 this is back to, unless you really want to make this your project, jus stop looking at it. there's no "oh, I can fix this in an hour or so" magic Feb 04 13:42:20 karlp: is being curious still a valid reason to ask questions in this world? :D Feb 04 13:43:15 it is my experience that I learn more the more questions I ask Feb 04 13:44:33 and no, I really do not expect "oh, I can fix this in an hour or so" magic. Not really common in my experience Feb 04 13:49:12 So people if I make a openwrt Discord server will any one join? Feb 04 13:49:38 so, dig into the 5.1 to 5.3 breakingchanges, and play with it if you want, but yes, there are reasons. Feb 04 13:51:36 xback: you were working on the mikrotik ath79 stuff right ? Feb 04 13:54:38 xback: http://www.xback.be/rb912.tar.xz no longer exists Feb 04 13:56:54 ah, that was for ar71xx and not ath79 Feb 04 14:04:33 currently working on ath79 support for that rb912 Feb 04 14:04:41 most is done. just nand remaning Feb 04 14:05:27 the latch driver is now OF capable and working nicely Feb 04 14:31:06 ok Feb 04 14:31:29 xback: are the patches online ? Feb 04 14:31:42 i just got a rb912uag-5hp Feb 04 15:47:34 Anyone use dnscrypt-proxy2 that wants to test a version bump PR? i don't use it but someone was asking for it. https://github.com/openwrt/packages/pull/14650 Feb 04 16:23:20 >KGB-1< https://tests.reproducible-builds.org/openwrt/openwrt_mediatek.html has been updated. (100.0% images and 98.3% packages reproducible in our current test framework.) Feb 04 17:15:51 any opinions about introducing a miniupnpd-igdv1 variant? see https://github.com/stintel/openwrt-packages/commit/eab8654a9e7c8eaa85873a477da9ad6437ac0173 Feb 04 17:17:46 ehr, yeah some obvious errors :)P Feb 04 17:18:08 ah no, I misread Feb 04 17:47:40 nbd: ping Feb 04 17:48:42 and suddenly wireguard says no :/ Feb 04 17:49:54 stintel: how no? Feb 04 17:50:31 rsalvaterra: pong Feb 04 17:50:47 I've seen WireGuard coming up listening on a completely different port than the one I have configured. It's very rare, but it happens. Feb 04 17:51:08 nbd: testing with GTK rekeying offloading hacked out. Feb 04 17:54:36 rsalvaterra: it just stopped working Feb 04 17:55:00 ifdown; ifup doesn't help a thing Feb 04 17:55:11 even rebooted the device, nothing Feb 04 17:59:11 Namidairo: Beyond a doubt, same compile time in the sysupgrade and in dmesg. Recompiled with the OpenWrt mirror and still no luck. Feb 04 17:59:17 Hi openwrt devs. Am I breaking any of the OpenWrt rools? Feb 04 17:59:20 https://forum.openwrt.org/t/openwrt-fans-get-social/87707 Feb 04 17:59:35 Is it OK that I used the OpenWrt logo? Feb 04 18:01:09 I'm not OK with an OpenWrt discord or whatever other hipster chat alternative Feb 04 18:01:48 stintel why? Feb 04 18:02:08 because we already have enough "communication platforms" Feb 04 18:02:16 Is it just discord you don't like? Feb 04 18:02:25 IRC, github issues/PRs, gitlab, mailinglist, forum, flyspray Feb 04 18:02:52 This is not for devs all tho they can join if they like. Feb 04 18:03:16 I don't want to get involved with dev stuff! Feb 04 18:03:17 Namidairo: I just needed a spare day to get my head straight. Redisocvered the -V option. Feb 04 18:03:48 it's not just discord, it's all of those hipster resource eating crap chats Feb 04 18:04:06 Lol Feb 04 18:05:04 stintel it's not going to do openwrt any harm tho is it? Feb 04 18:06:50 it's going to scatter communications even further, I think it actually might Feb 04 18:06:53 pkg_hash_fetch_best_installation_candidate: kmod-batman-adv arch=mips_24kc arch_priority=10 version=5.4.94+2021.0. Feb 04 18:07:17 yet DISTRIB_ARCH='mips_24kc' in /etc/openwrt_release Feb 04 18:07:25 I am the admin of the server and it is just for chater and some help with setting things up. Feb 04 18:07:49 If any posts about dev stuff I will point them to the forums or here. Feb 04 18:08:20 If it gets out of hand I will shot it down k Feb 04 18:11:37 Gonna kick it off without per-device rootfs / multiple targets options and see if that helps. That has to be it Feb 04 18:11:39 oh no Feb 04 18:12:38 I meant to control-k. I diff'd the control from the snapshot and found different kernel commit dependency -- any place I can find the commit of the running kernel? Feb 04 18:13:13 s/snapshot/OpenWrt-compiled package from snapshot/ Feb 04 18:14:44 Woah. Feb 04 18:14:52 Other packages work o_O Feb 04 18:15:49 Sigh. Got it. Feb 04 18:16:16 Obviously if you're installing a single package, opkg doesn't do dependency resolution for you. I followed the dependency tree manually and it installs now Feb 04 18:17:15 that error message is probably only so poor because all advice points away from manually copying packages one-by-one and instead copying the whole tree and pointing to it like a remote Feb 04 18:17:25 (or setting up an actual remote) Feb 04 18:24:48 What I really need is a rubber-ducky extension to weechat. Copies the visual context of the current channel but messages go to nobody. It'd be as good as asking for help but with the added bonus of annoying nobody :^) Feb 04 20:37:38 I just spent nearly 6 hours doing DM's on the forum working thru an issue with someone. Discord woud help make that a lot easier, but as Tapper said already, he's pointed out it isn't an Official communications avenue for reaching devs.. although, not like IRC is either most times Feb 04 20:38:43 Anything official contact wise would be going thru the forum or website or corporate contact Feb 04 20:40:16 Grommish: why DMs? something that could not see the light of day (closed source)? Feb 04 20:42:12 Grommish: dnscrypt-proxy2, it seems that nobody wants to try it, yet. You can try to run tested it and also you can edit a PR a little bit. In the commit message, you can remove [] and use just packagename: foo anyway, it is possible that the maintainer, who is there listed will get to it eventually. Feb 04 20:43:12 jow: This has bugged me for a while - stats graph page hangs at 'loading data' if the session has timed out. Most other pages redirect to login page but stats graphs fails to do that. Any ideas? Feb 04 20:43:19 Borromini: because it involved walking the guy thru git steps, rebasing and dealing with tabs vs spaces in the Makefile. Boring stuff that didn't need to clutter everything.. Besides.. debian git doesn't have git restore apparently Feb 04 20:43:55 Pepe: Yeah, I don't use it myself, but someone asked about it and I was bored.. The build seems to work and I changes the default filenames in ./files to reflect the new config TOML it uses Feb 04 20:44:17 Pepe: Will remove the braces Feb 04 20:45:24 and I'll remove it from the commit messages when I squash before repushing the final Feb 04 20:48:28 I'm sure whoever the maintainer is will get to it and since I can't test it, Itll sit a a draft until it can be tested properly by people who know the package. If nothing else, it'll give them a legup and the guy who wanted it has it now and is able to build it locally if nothing else Feb 04 20:50:15 anyone have any experience with vncserver? Feb 04 20:53:36 Grommish: ok. very kind of you to do that through PM :) Feb 04 20:56:18 Borromini: I'm wondering if OpenWrt can handle vncserver and a console emulator :D and do vnc over ethernet Feb 04 20:59:39 iirc git restore is a new name for an operation you used to do with git checkout... and debian is quite ancient by default Feb 04 21:19:26 Grommish: no idea sorry :( Feb 04 21:20:08 Why would you want to do that when you can just SSH ? Feb 04 21:25:01 EqUaTe: VNC over ethernet would allow me to stream the graphical aspects from the device to a remote point Feb 04 21:25:46 I want to try and turn this box into a Game emulator box using Openwrt Feb 04 21:25:50 not sure why you're saying 'VNC over ethernet' - VNC is an IP-based protocol, so there's that, but you're running a graphical environment on an openwrt box?? why? Feb 04 21:26:06 I don't think openwrt is really the right choice for that tbh Feb 04 21:26:15 it definitely isn't the target market, that's for sure :) Feb 04 21:26:47 Probably not, but if I view owrt as a embedded OS rather than just as router software, I can expand my use cases hehe Feb 04 21:27:12 seems like you're adding a lot of extra work to it for yourself Feb 04 21:27:22 also, VNC really isn't very good for low latency Feb 04 21:27:25 EqUaTe: I want to look into MIPS-based console emulators ala SNES etc Feb 04 21:28:17 and yes, it probabyl going to be a massive cluster anyway I go about it, but. Feb 04 21:29:38 EqUaTe: The latency point you make is one of the things I was going to ask about. A local LAN connection shouldn't be too bad though, no? Feb 04 21:29:55 I've found it usually is Feb 04 21:30:16 Ah, ok.. Thanks :) I"ll have to play around with options then Feb 04 21:30:19 eg: two systems that i have, both cabled, vnc into one from the other, can see the mouse lag (they're both laptops) Feb 04 21:30:27 FreeNX might be more useful Feb 04 21:30:35 or whatever it's evolved into Feb 04 21:32:01 Hmm.. Thanks :) I'll look into it. The device is a dual-core 1ghz mips64 SoC with 1gb ram, 4GB emmc onboard and a SD slot. I figure if I can get console emulation, it would just be a fun toy that's super small Feb 04 21:32:26 rpi clone of some sort? Feb 04 21:32:48 no.. it was a small production run of a device that was meant to be a network security device Feb 04 21:33:58 https://openwrt.org/toh/hwdata/itus_networks/itus_networks_shield Feb 04 21:59:40 https://kb.cert.org/vuls/id/794544 Feb 04 22:04:19 Grommish: is that a new one or is that the one the fuss was about like a week ago? Feb 04 22:09:28 is there a neat way to read out a led brightness status on openwrt or do i just need to alternate between 0/255 matches? Feb 04 22:10:12 what? Feb 04 22:10:31 karlp: my device has a button that does nothing more than turn the led off or on Feb 04 22:11:10 i suppose i just check the value of /sys/class/leds/white\:status/brightness and then write the opposite to it Feb 04 22:12:00 yes...? Feb 04 22:12:10 just wondering if there was a smarter way to do it. Feb 04 22:12:45 what sort of thing would you be hoping for? another file in some other path that had 1/0 isntead of a range? I feel like i'm missing something :) Feb 04 22:13:18 no it's just that this is the first time i'm playing with hotplug and buttons. Feb 04 22:13:29 sorry ;) Feb 04 22:14:03 i read the hardware buttons part on the wiki, and i saw the dts has the kernel key code defined. all pretty neat Feb 04 22:14:08 Borromini: it's the one from last week or so that's had patches issued Feb 04 22:14:20 EqUaTe: thanks, i'm good then :P Feb 04 22:29:51 lovely, OEM behaviour replicated ^_^ **** ENDING LOGGING AT Fri Feb 05 02:59:57 2021