**** BEGIN LOGGING AT Wed Aug 28 02:59:57 2019 Aug 28 05:00:20 aparcar[m]: i did not set the status Aug 28 05:03:55 morning Aug 28 05:05:12 i've a custom packaged rpcd plugin that exports some object + methods I want to access using https Aug 28 05:05:39 if I want uhttpd to generate certificates, I need /usr/bin/openssl (for details see package/network/services/uhttpd/files/uhttpd.init ) Aug 28 05:06:11 is that good idea (clean design) to make my package select openssl? Aug 28 05:06:49 something like DEPENDS:=+rpcd +openssl-util Aug 28 05:07:34 the thing is, my rpcd plugin doesn't directly depend on openssl, so I was looking for something like uhttpd-ssl (meta?) package Aug 28 05:11:38 blogic: sorry likely my bad then! But would you be willing to check it? Aug 28 05:12:02 Hauke ping Aug 28 05:22:56 what the schedule of https://downloads.openwrt.org/releases/19.07-SNAPSHOT/targets/ramips/mt7621/ builds? still Aug 24 Aug 28 05:32:21 ds_shadof: what do you mean? Aug 28 05:46:03 aparcar[m], i want build with this commit https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=c3327f48dccc548f1a18948e61ce465c968620fd Aug 28 06:31:48 good morning all Aug 28 06:32:43 is somewhere documented how to install snapshot images with kernel.bin and sysupgrade bin? Aug 28 06:54:45 jow: can I ask you to look at https://patchwork.ozlabs.org/patch/1151976/ ? [PATCH 1/2] base-files: use JSON for storing firmware validation info **** BEGIN LOGGING AT Wed Aug 28 07:16:43 2019 Aug 28 07:28:58 ds_shadof: git pull & make ? Aug 28 07:38:38 aparcar[m], hard times for me now to build by myself :) Aug 28 07:39:02 ds_shadof: wait a day it will be in snapshots tomorrow Aug 28 07:39:45 aparcar[m], that's wonderful Aug 28 08:26:50 blogic ping? Aug 28 08:30:11 blogic: ping hi i saw you introduced linkit smart 7688, i played a bit with upstream uboot which contains support for mt7688, problem is that in upstream uboot not yet landed patch for spi flash and changed partitions Aug 28 08:36:01 blogic: and i noticed change of installation process in skw92a commit - https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=e42327aa890e64f8bf5e620c2b2a1fd609ce20fb Aug 28 08:45:40 indy__: change from what to what? Aug 28 08:48:07 russell--: change in first message or in second message? :) Aug 28 08:48:51 the skw92a was adding support, so it changed from nothing ;-) Aug 28 08:53:11 FYI: mt76x8 ethernet driver is now in the net-next with the PHYLINK API, switch isn't. Aug 28 08:55:27 russell--: i see in snapshot ramips images *-initrams-kernel.bin and *-sysupgrade.bin for all supported devices Aug 28 08:56:11 indy__: and what's your question? Aug 28 08:57:55 in commit message are described 3 installation processes so they are also valid for all ramips boards? Aug 28 08:58:14 no, they are valid for that board Aug 28 08:58:38 or were when the commit was prepared Aug 28 09:00:15 indy__: what board do you have? Aug 28 09:00:20 blogic: anyway, would be happy if you could merge the procd patch :)! Aug 28 09:00:30 good night Aug 28 09:04:32 hmm... I'm confused, again. afair 802.11s supports choosing different channels for mesh an data, i.e. stations can negotiate to use a different channel for data to have more airtime in the mesh in total, isn't it? Is that feature implemented in OpenWRT? If so: What about multi radio boxes - can different radios in different bands? i.e. can I have a mesh in 2.4 Ghz with station selecting 5 Ghz channels for data, if available? Aug 28 09:06:12 https://openwrt.org/docs/guide-user/network/wifi/mesh/80211s doesn't say much about this Aug 28 09:08:33 yanosz: what means "for data"? Aug 28 09:09:31 if node A want's to sent data (no mesh mgmt) to node B, both can agree on a dedicated channel for doing so Aug 28 09:10:09 management frames are sent through the one and only mesh channel, though Aug 28 09:10:19 no idea Aug 28 09:10:32 i've never heard of that Aug 28 09:10:40 You can have it with ns-3 afair and I guess, its in the 802.11s spec Aug 28 09:11:07 But I'm not sure, wheter it made in to into IEEE 802.11s in the ende Aug 28 09:11:08 end Aug 28 09:13:19 Hmmm ... Common Channel Framework (CCF) Aug 28 09:13:23 http://www.ieee802.org/802_tutorials/06-November/802.11s_Tutorial_r5.pdf Aug 28 09:15:07 looks like its done via a rts/cts mechanism Aug 28 09:16:03 hmm https://wiki.opennet-initiative.de/wiki/802.11s - claims, it didn't make it into the spec Aug 28 09:17:21 no idea why, whether it's implemented nevertheless or if sbdy has done a performance evaluation on it ... Aug 28 09:18:02 it could be implemented in some drafts Aug 28 11:00:02 russell--: i have linkit smart mt7688 Aug 28 11:28:25 russell--: and vocore2 Aug 28 13:10:25 indy__: Currently upstream u-boot is too big. MTK is preparing to submit a SPL supporting LZMA compression. They've submitted 26 patches reworking upstream mt76x8 support today and LZMA SPL will be submitted after this series is applied. You could wait a while for that. Aug 28 13:12:33 indy__: Or you could change the partition table in OpenWrt to give more space to u-boot. Remember to backup calibration data (content of factory partition) before flashing. Aug 28 13:57:49 gch981213: too late :) Aug 28 14:00:21 gch981213: so time to pull u-boot git? Aug 28 14:22:32 Hauke: is there a chance that PMU is in a different address on my device? Aug 28 14:23:11 So - the first problem was caused by a"data abort" - when the kernelunprotects flash write area. I have a small 256kb flash Aug 28 14:23:40 Hauke: but maybe it's not where we expect it to be. Commenting that line to experiment, I see a pmu activation failure Aug 28 16:32:40 good evening / everything guys! :) Aug 28 18:44:34 nbd: will you be backporting 8650201f10 to 19.07 as well? Aug 28 18:44:50 since you just bumped mt76 :) Aug 28 18:53:20 does openwrt respect /etc/gai.conf? How do I prefer IPv4 over IPv6 on a dual stack install? Aug 28 19:07:17 Borromini: done Aug 28 19:20:23 nbd: much obliged, vielen Dank. Aug 28 19:49:39 aparcar[m]: pong Aug 28 19:52:46 Hauke: could you please merge the PR that introduces generic as subtargets? Aug 28 19:59:47 Hauke: meaning https://patchwork.ozlabs.org/patch/1152029/ and https://patchwork.ozlabs.org/patch/1152044/ Aug 28 20:10:53 ynezz: ping? Aug 28 20:25:48 isn't the generic in the directory paths enough as it is? Aug 28 20:25:56 why are we always changing things for your tooling? Aug 28 20:27:08 karlp: there's nothing wrong with consistency is there? Aug 28 20:32:21 karlp: the image path is xyz/generic hower the image prefix is openwrt-xyz-sysupgrade.img instead of openwrt-xyz-generic-sysupgrade.img. Aug 28 20:32:51 didn't you already add more than enough metadata elsewhere to get those names yet? Aug 28 20:33:01 it kkinda just feels like it's never enough, always needs more. Aug 28 20:33:21 karlp: you talk about the json patch? Aug 28 20:37:23 I'm trying to automate and simplify some stuff around OpenWrt. Often I find inconsistencies making (my) tooling harder and so I try to improve thing upstream Aug 28 20:59:02 rmilecki: what do you think about: https://patchwork.ozlabs.org/patch/1152029/ ? Aug 28 21:22:51 Hauke: i don't know Aug 28 21:23:03 jow: can you look at above patch? https://patchwork.ozlabs.org/patch/1152029/ Aug 28 21:28:12 I did so last week with a different targets and the snapshots work fine https://git.openwrt.org/?p=openwrt/openwrt.git;a=commit;h=853e4dd3062df7cb5704b15d6af6730e3194b571 Aug 28 23:00:19 Hauke: rmilecki I mean I can also set the SUBTARGET variable to generic if unset and leave the targets as they are. Aug 28 23:17:16 lynxis: hey, can you let me know whats missing on https://git.openwrt.org/?p=openwrt/staging/lynxis.git;a=commit;h=416cccf398e9589e3de386e05b61b1c46cace20d ? I maybe could help out **** ENDING LOGGING AT Thu Aug 29 02:59:57 2019