**** BEGIN LOGGING AT Sun Dec 20 02:59:57 2020 Dec 20 07:29:31 >KGB-1< https://tests.reproducible-builds.org/openwrt/openwrt_ar71xx.html has been updated. (99.2% images and 97.4% packages reproducible in our current test framework.) Dec 20 07:56:40 why does ar724x not have a wmac or mtd-cal-data in its device trees? Dec 20 08:11:34 also, how do i tell device tree my ar7240 device has 64MB of RAM? Dec 20 08:11:49 and not just 32MB **** BEGIN LOGGING AT Sun Dec 20 08:38:20 2020 Dec 20 08:54:20 found it Dec 20 08:54:29 the first part anywa Dec 20 08:54:31 y Dec 20 08:55:33 in /etc/hotplug.d/firmware/10-ath9k-eeprom Dec 20 09:58:40 datasheet says 64MB, but boot-loop and the vendor firmware i recovered with says 32MB Dec 20 12:55:18 build #447 of bcm27xx/bcm2708 is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/bcm27xx%2Fbcm2708/builds/447 Dec 20 13:52:20 build #480 of bcm63xx/smp is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/bcm63xx%2Fsmp/builds/480 Dec 20 16:19:29 for WLE900VX - https://github.com/torvalds/linux/commit/2dc016599cfa9672a147528ca26d70c3654a5423 Dec 20 16:36:08 >KGB-0< https://tests.reproducible-builds.org/openwrt/openwrt_ath79.html has been updated. (99.2% images and 97.5% packages reproducible in our current test framework.) Dec 20 17:13:06 I'm curious. After reading https://openwrt.org/meetings/20201210 - this documents enabling 802.11w (in optional mode) by default after branching 20.12 - why not do this before and ship with this? Dec 20 17:14:13 What is the benefit in waiting another year before this hits a stable release branch? What is achieved? Dec 20 17:15:13 The shell script change to offer this by default appears to be simple and a one-liner: set_default ieee80211w $((wpa & 1 ? 0 : (wpa & 2) >> 1)) Dec 20 17:52:27 Any chance that this can be merged to master to fix legacy non-RSN client compatibility with EAPOL versions? https://github.com/openwrt/openwrt/pull/3681 (I'll then submit a merge request containing a change to enable 802.11w in optional mode by default https://github.com/nicklowe/openwrt/commit/354e19946f3b030dc1bf685a4663c95ceb5fc7ab ) Dec 20 18:26:21 nlowe: Apple MacBooks pre-2017 fail to connect as soon as MFP is optional Dec 20 18:26:30 iwlmvm fails on required Dec 20 18:29:52 And to be precise: Altering default behavior this close to a release is not really aiming for stability Dec 20 18:30:14 I have tested with a MBP 2014 and 2015 and they do not fail to connect with 10.14 (Mojave), 10.15 (Catalina) or 11.0 (Big Sur) FWIW Dec 20 18:30:15 Yeah what blocktrron said Dec 20 18:30:17 considering we will have to add workarounds for drivers with broken MFP (Brokencom, marvell) Dec 20 18:30:26 * Tapper nods Dec 20 18:30:51 nlowe: 2016 late MBP - not able to connect to a 802.11w optional PSK2 AP Dec 20 18:30:58 Same for a early 2012 MBA Dec 20 18:30:59 We will have a flud of people on the forums asking why openwrt does not work with there iphones Dec 20 18:31:19 Are you sure that's not for older, obsolete macOS releases?/ Dec 20 18:31:38 Was the one prior to Big Sur for the 2016 Dec 20 18:31:45 For the 2012 it was the last supported Dec 20 18:32:16 I am all ready sick to the back teeth of people asking why they cant have 19.06.5 on the wrt54g v2.2 Dec 20 18:32:35 lol Dec 20 18:33:35 Pleas pleas just let the wrt54g routers die! I beg you! lol They wer grate routers in there day I had one and loved it. Dec 20 18:33:39 you should tell them to put in back in the dumpster where they found it, Tapper Dec 20 18:34:09 Also this needs more logic, as rekeying fails in some drivers when 802.11w / 802.11r are enabled at the same time Dec 20 18:34:09 * Tapper Holds his head in his hands. Dec 20 18:34:16 I have tryed. Dec 20 18:34:56 [415043.174783] Rekeying PTK for STA 2a:xx:xx:xx:xx:47 but driver can't safely do that. Dec 20 18:35:09 ask them how come they are not driving the 20+y old cheapest car they can get back then, and demanding it to have all the latest things in it Dec 20 18:36:52 * urjaman is driving a 30yo car... no wrt54g tho Dec 20 18:38:37 nlowe: i appreciate your push into that direction, but for these change, having these issues documented in the forum by people using the snapshots is a big benefit Dec 20 18:39:16 blocktrron A historical compatibility issue that I've seen with Apple iOS clients (these documented support for 802.11w) with required mode is that the 802.11 spec mandated SHA-256 based AKMs were not implemented: https://wlan1nde.wordpress.com/category/ieee-standards-and-amendments/802-11w/ Dec 20 18:43:42 Tapper: i feel your pain Dec 20 18:45:57 urjaman: you can drive the wrt54g too, with most everytrhign stick like you do with your car I can presume (: Dec 20 19:01:08 when you activate sae or sae-mixed, 802.11w will be optional for WPA2 and required for WPA3 like required in the Wifi alliance standard Dec 20 19:01:28 so the default of 802.11w changes Dec 20 19:29:52 My Apple contact in their Enterprise Wi-Fi division has just said MFP/PMF should be compatible from macOS 10.14 and iOS 13 Dec 20 19:30:35 It's broken in iOS 12 with devices like the iPhone XS Dec 20 20:27:58 Please can somebody review and, if appropriate, merge https://github.com/openwrt/luci/pull/4649 - this seeks to avoid people hiding their SSID and then having roaming break, or disabling WMM and wondering why the performance sucks by adding descriptions Dec 20 20:28:28 At least by being explicit, it's clear what is being signed up for Dec 20 20:28:48 Ideally these options would be removed from LuCI but I think that ship has sailed Dec 20 20:29:11 Descriptions are: Dec 20 20:29:12 Where the ESSID is hidden, clients may fail to roam and airtime efficiency Dec 20 20:29:12 may be significantly reduced Dec 20 20:29:18 Where Wi-Fi Multimedia (WMM) Mode QoS is disabled, clients may be limited Dec 20 20:29:18 to 802.11a/802.11g rates. Dec 20 21:42:44 nlowe: all these controversial PRs days before a release, you did you homework ;) Dec 20 21:43:59 :P Dec 20 21:45:08 I'm wondering what's a good mix of technical and layman description. Dec 20 22:43:06 updated openwrt/upstream, https://sdwalker.github.io/uscan/index.html **** ENDING LOGGING AT Mon Dec 21 02:59:56 2020