**** BEGIN LOGGING AT Sat Dec 15 02:59:58 2018 Dec 15 05:32:02 jwh: what is ncm? Dec 15 05:35:12 ethernet over usb commonly used by LTE dongles Dec 15 05:35:23 with a weird serial type port exposed for control Dec 15 05:35:50 something causes netifd or the proto script to send empty commands though Dec 15 05:36:00 eventually it might work but most often it needs restarting Dec 15 08:48:23 ynezz: Just checked LSDK and found that ag7240 asserted only mac resets in fast_reset function. I assume both resets works separately but only asserting both of them resets register values. Dec 15 08:49:35 my temporary fix is available here: https://github.com/981213/openwrt/tree/ath79-4.19 (including unfinished 4.19 bump because I'm using this branch recently :) ) Dec 15 11:57:02 build #1258 of x86/64 is complete: Failure [failed targetupload] Build details are at http://phase1.builds.lede-project.org/builders/x86%2F64/builds/1258 blamelist: Hans Dedecker , Kevin Darbyshire-Bryant Dec 15 12:23:03 hi! Dec 15 12:26:02 hello Dec 15 12:32:38 can some one aprove https://github.com/openwrt/openwrt/pull/1590 ? Dec 15 13:07:28 hm Dec 15 13:07:35 Sat Dec 15 13:06:05 2018 daemon.notice netifd: ncm (5561): sending -> AT+CGDCONT=1,"IP","3internet" Dec 15 13:07:38 Sat Dec 15 13:06:06 2018 daemon.notice netifd: ncm (5561): sending -> AT^NDISDUP=1,1,"3internet" Dec 15 13:08:46 don't normally need both, NDISDUP should also define the pdp context Dec 15 13:09:00 wonder if the dongle is bringing up both sessions Dec 15 13:15:34 jwh: imho scm.sh shell script is ugly by design Dec 15 13:15:43 ncm.sh Dec 15 13:15:46 yeah Dec 15 13:16:02 I have a feeling the dongle might be establishing both contexts Dec 15 13:16:08 network are seeing weird behaviour Dec 15 13:16:34 Im thinking to write it on C Dec 15 13:19:52 imho script does not handle 2 cases: Dec 15 13:19:53 1. replyg device Dec 15 13:19:53 2. device lost LTE link Dec 15 13:20:13 yeah Dec 15 13:20:21 it gets really upset Dec 15 13:24:46 I have only 2 huawei for tests ) Dec 15 13:42:43 heh Dec 15 13:42:51 well, I have a traditional one using ppp Dec 15 13:43:07 and a 3372h I converted to 3372s, stick mode Dec 15 13:43:19 lost my other nice LTE one Dec 15 13:49:38 mine is 3272 and 3372, both in ncm/cdce mode only Dec 15 13:49:58 ah cool Dec 15 13:50:09 I think my old one was the 3272 also Dec 15 13:50:23 but I have NCM enabled as well as the old serial ports Dec 15 13:50:50 NCM so it works on windows easily, ppp incase I plug it into a linux box without support Dec 15 13:51:02 or older windows, ofc Dec 15 13:51:31 no modeswitch either, that sucks Dec 15 15:01:00 Wow! kernel 4.19 support is finally added! Dec 15 15:19:51 build #683 of layerscape/armv8_64b is complete: Failure [failed images] Build details are at http://phase1.builds.lede-project.org/builders/layerscape%2Farmv8_64b/builds/683 blamelist: Hans Dedecker , Kevin Darbyshire-Bryant , Hauke Mehrtens Dec 15 17:44:42 build #1259 of x86/64 is complete: Success [build successful] Build details are at http://phase1.builds.lede-project.org/builders/x86%2F64/builds/1259 Dec 15 18:06:14 hmmm, ER-12 is only 250 bucks Dec 15 18:09:26 I know what I want for christmas Dec 15 18:41:44 hm, is there something with a nice webui or what have you that can aggregate commit logs from arbitary git repos and display them in a useful format? Dec 15 18:42:06 want something a bit better than an RSS feed type thing Dec 15 18:44:16 nbd: will this mt76 commit eventually be cherry-picked to 18.06 or is the time for backports over? 4fd788282274e132b974e074a35cf8729c72819e Dec 15 18:46:14 openwrt master is pretty stable for me, running a MT7628N device for 16 days now without problems, including 802.11s mesh Dec 15 18:46:45 maybe i should buy another to be able to run both branches in parallel **** ENDING LOGGING AT Sun Dec 16 02:59:59 2018