**** BEGIN LOGGING AT Sat Jan 11 03:00:53 2020 Jan 11 03:17:02 ugh. These "code cleanups" broke LuCI Jan 11 03:17:05 do_page_fault(): sending SIGSEGV to rpcd for invalid read access from 39383830 Jan 11 05:34:53 build #245 of pistachio/generic is complete: Failure [failed] Build details are at http://buildbot.openwrt.org/master/images/builders/pistachio%2Fgeneric/builds/245 blamelist: Chuanhong Guo Jan 11 09:56:23 Thank you all for this nice 19.07 release Jan 11 10:41:47 Hauke: thank you for waering the release managers hat Jan 11 10:42:03 I promise to be of more help the next time :) Jan 11 10:43:36 I just wanted to get this done ;-) Jan 11 12:36:21 i can't seem to find the changelogs from update to update.. im running 18.06.4 and thinking I should upgrade to 19.07.0, but I want to read the changelog first -- but I must be blind or something cause I can't find it. Anyone mind pointing me in the right direction? Jan 11 12:36:45 openwrt.org Jan 11 12:36:54 ctrl+f and type changelog :) Jan 11 12:37:18 ah, lol Jan 11 12:37:20 thank you :D Jan 11 12:37:49 enjoy Jan 11 12:39:25 lol Jan 11 12:55:17 myeah, all good. updated. Jan 11 12:55:33 err, upgraded*. Jan 11 13:01:46 btw, while reading the frontpage I noticed that we should maybe replace "final" with "initial" Jan 11 13:02:01 "final release of 19.07" sounds as if there will be no more Jan 11 13:02:16 I'd do that if nobody objects Jan 11 13:03:14 good point Jan 11 13:03:23 ack Jan 11 13:03:29 i read that the same way Jan 11 13:03:53 or "first stable release of 19.07" Jan 11 13:21:08 ldir thanks for the cake bump. Jan 11 14:34:47 build #188 of brcm63xx/smp is complete: Failure [failed] Build details are at http://buildbot.openwrt.org/master/images/builders/brcm63xx%2Fsmp/builds/188 blamelist: Chuanhong Guo , Kevin Darbyshire-Bryant Jan 11 16:08:07 * DonkeyHotei just noticed that in menuconfig, under Network, there is both wireless and Wireless Jan 11 16:22:05 hello, in section `Known Issues` from https://openwrt.org/releases/19.07/notes-19.07.0 it's written that: Jan 11 16:22:11 "Sysupgrade from ar71xx to ath79 and vice versa is not officially supported, a full manual reinstall is recommended (...)" Jan 11 16:22:18 do I understand it correctly that I should probably download `*-factory.bin` instead of `-*sysupgrade.bin`? Jan 11 16:22:21 ... and later I can do the upgrade by using LuCI - uploading `*-factory.bin` instead of `-*sysupgrade.bin`? Jan 11 16:23:02 mrl5: please see https://openwrt.org/docs/guide-user/installation/ar71xx.to.ath79 (it's a work-in-progress though) Jan 11 16:23:14 in short: you can try sysupgrade, but there is no guarantee it will work Jan 11 16:23:34 what can happen if it fails? bricked device? Jan 11 16:23:37 zorun: ^ Jan 11 16:23:43 btw thanks for the url Jan 11 16:24:01 softbricking is possible on certain models Jan 11 16:24:10 but only if you force sysupgrade Jan 11 16:26:04 thanks zorun DonkeyHotei - I think that this URL should be more visible (or even linked) in release notes - I did not spot it Jan 11 16:27:39 yes, it will be added to the release notes once the content is validated / more complete Jan 11 16:28:03 if you have feedback for your device, please share! Jan 11 16:29:05 mrl5: worst case you need to reset your openwrt to the defaults, thats why you need to make a backup either way Jan 11 16:29:36 I have tp-link archer-c7-v5 - is there any feedback regarding this device? Jan 11 16:29:57 if not where can I let you know if my upgrade worked? here or ...? Jan 11 16:30:52 yes here Jan 11 16:31:22 cool, lets be in touch then Jan 11 17:13:05 zorun: so I've switched from 19.07.0-ar71xx-generic-archer-c7-v5-squashfs to 19.07.0-ath79-generic-tplink_archer-c7-v5-squashfs-sysupgrade.bin via LuCI Jan 11 17:13:13 (both are OpenWrt 19.07.0, r10860-a3ffeb413b) Jan 11 17:13:22 the switch itself was as usuall but ... Jan 11 17:13:33 0. my 2.4MHz WiFi is not working in LuCI for radio1 I can see "Generic 802.11bg": "Device is not active" - some drivers issue? Jan 11 17:13:50 1. after connecting via ssh I see no /etc/network dir (I was looking for /etc/network/wireless Jan 11 17:14:14 2. in LuCI when clicked to "Network->Wireless" I had some popup, after clicking "OK" or "Confirm" my wifi configuration showed up Jan 11 17:14:26 3. 5GHz WiFi (Qualcomm Atheros QCA9880 802.11acn) working OK though Jan 11 17:14:55 mrl5: did you NOT keep the configs? Jan 11 17:15:06 aw... Jan 11 17:15:08 nope Jan 11 17:15:14 I left it checked Jan 11 17:15:39 mrl5: wifi device path is often incompatible Jan 11 17:16:52 any1 wanna do some debugging session or I should just fallback to ar71xx? Jan 11 17:18:02 mrl5: so did you keep old configs or not? Jan 11 17:18:14 mrl5: ar71xx is def. not the desired end result Jan 11 17:19:34 PaulFertser: yes I have old config Jan 11 17:20:13 backup-*.tar.gz Jan 11 17:20:14 mrl5: I would suggest removing /etc/config/wireless and doing "wifi config". Jan 11 17:21:06 hmm.. probably it was always /etc/config, I've just forget the right dir name Jan 11 17:23:11 mrl5: heh, somehow I still can't understand if in your ar71xx -> ath79 upgrade the old (from ar71xx) configs were applied to the new system or not. Jan 11 17:25:20 they were Jan 11 17:25:31 but 2.4GHz WiFi stopped working for some reason Jan 11 17:25:54 btw after removing /etc/config/wireless I've managed to set up working 2.4G SSID Jan 11 17:26:06 mrl5: because wireless config from ar71xx is not suitable for ath79. Jan 11 17:30:58 ok, can I somehow now restore my old config from .tar.gz backup or I must manually make the configuration? Jan 11 17:31:24 my old config -> my old configuration of 2.4G and 5G ssid' Jan 11 17:31:59 mrl5: .tar.gz is just an archive, unpack it on your PC, change wireless config to suit new system, repack, try restoring... Jan 11 17:32:26 cool Jan 11 17:32:46 I should have done mv istead of rm for that old /etc/config/wireless :D Jan 11 17:33:32 mrl5: but you have backup so can restore from there again Jan 11 17:34:25 yes ofc :D scp is my good old friend so no problem Jan 11 17:34:59 anyways thanks for help PaulFertser Jan 11 17:35:26 I hope my feedback will be useful for other ppl migrating from ar71xx to ath79 Jan 11 17:35:53 (archer-c7-v5) Jan 11 17:36:04 mrl5: for it to be useful we need to draw some conclusions Jan 11 17:36:25 if you have some questions feel free to ask Jan 11 17:36:26 mrl5: what's your lesson, what's missing on that wiki page? Jan 11 17:36:39 or if you want me to check something in the freshly upgraded router Jan 11 17:38:26 1. I'd do the hyperlink leading to https://openwrt.org/docs/guide-user/installation/ar71xx.to.ath79 for phrase "full manual reinstall" which is in "Known Issues" here https://openwrt.org/releases/19.07/notes-19.07.0 Jan 11 17:39:42 I hope zorun does that, I have no edit permissions for that page. Jan 11 17:40:04 or maybe putting this url in "Target transition from ar71xx to ath79" would be a better place Jan 11 17:41:30 and generally I see that my issues with `/etc/config/wireless` are already covered, well... I did not read that - my fault :D Jan 11 17:41:44 mrl5: thank you for the feedback! Jan 11 17:43:21 no problemo, I can later give you the diff between old and new /etc/config/wireless files Jan 11 18:03:07 Hello everyone, quick question. I want to build an image for a router using the latest OpenWrt release, when I clone the repo which branch/tag should I checkout? I have the choice between "openwrt-19.07" and "v19.07.0-rc1/rc2". Which is the correct one? Jan 11 18:04:33 Rjevski: latest release is "v19.07.0" tag Jan 11 18:04:45 Rjevski: rc means "release candidate" by definition Jan 11 18:06:19 Thanks. Will that tag ever change or be moved (do I need to periodically rebuild the image to get updates, or will those be provided by opkg? The target is officially supported by OpenWrt, I'm just building a custom image to include an extra package that can't be included by default due to licensing issues. Jan 11 18:08:20 ok PaulFertser, changing this: https://pastebin.com/up9ZWfbL fixed the issue with 2.4G WiFI ;) Jan 11 18:09:56 thanks again everyone for support Jan 11 18:12:05 Rjevski: unless something extraordinary happens the tag is not going to be changed. Updates will be available with opkg, yes. But there might be another minor 19.07.1 release later. Jan 11 18:12:51 So I'm supposed to rebuild every time a minor release is out? Or can I move to 19.07.1 through opkg somehow? Jan 11 18:13:29 Rjevski: you can't upgrade kernel version with opkg for sure Jan 11 18:33:36 Thanks for the info. Building the image now! :) Jan 11 18:50:55 PaulFertser thanks i got the partition info in the dts file worked out Jan 11 19:48:25 from latest snapshot when i do make menuconfig i get https://paste.ubuntu.com/p/4DgRZyybXR/ Jan 11 19:50:49 Slimey: i assume you're seeing the curses interface though? because i've been seeing those recursive dependency errors for months. Jan 11 19:51:02 yup Jan 11 19:51:18 this is the first time i've encountered this Jan 11 19:51:23 ok Jan 11 22:27:00 Hi OpenWrt is not being talked about in a nice way on HN https://news.ycombinator.com/item?id=22018408 Jan 11 22:27:35 Some people have OpenWrt down as having hostile devs! Jan 11 22:29:12 I have finished improving https://openwrt.org/docs/guide-user/installation/ar71xx.to.ath79 and added it the release notes https://openwrt.org/releases/19.07/notes-19.07.0 Jan 11 22:30:04 thanks PaulFertser and adrianschmutzler for contributing Jan 11 22:31:43 zorun: thank you for all your work! I know how it's not easy to write release notes and other "unimportant" documents. Jan 11 22:33:30 zorun: your improved notes are so much better than the initial version! Jan 11 22:35:23 thanks PaulFertser! Jan 11 22:36:08 Tapper: that's funny, because I would qualify most HackerNews comments as "hostile" :) Jan 11 22:36:40 Thanks all for the improved notes Jan 11 22:38:11 zorun yeah some times. It would be nice for some one to post over there as a dev and say We are sorry for any thing said in the past but in the new ames of the progect we want to be nice to all people and get more people to work on OpenWrt. Jan 11 22:38:25 zorun: I see you're a wiki pro, so asking here: some procd services are tricky, depend on triggers etc, and I know a trick to help diagnostics: ubus call service list '{"name": "dnsmasq", "verbose": true}' . Where should it be put on the wiki so it's reasonably visible? Jan 11 22:38:39 Some people are thinking about crap that was going on from 5 6 years ago Jan 11 22:39:21 We all need to be as nice as we can to get people too join in with us Jan 11 22:39:49 I would post, but as you can see I cant tipe for toffee! lol Jan 11 22:42:52 The improved notes are just the kind of thing we need to help people to get in too OpenWrt. People feel mutch safer when they have dox to read. Jan 11 22:45:06 Tapper: most of the comments there are ok. Just a few on top discussing the contribution process are negative. And many people there talk about pre-LEDE days. Jan 11 22:45:47 Tapper: also, some people there are apparently uncomfortable with contributing to _any_ big free software project, so no wonder OpenWrt is not an exception. Jan 11 22:50:33 PaulFertser I get what you are saying, but it would be nice for some to post somthing nice over there for people who mite get a bad inpression of us Jan 11 22:54:01 Tapper: well, I'd prefer having statistics like: N new devices were contributed by M people, X of them not being part of the team. Median/max/min days from the first patch version to merging: N/N/N. Jan 11 22:56:14 Tapper: my personal experience is that my patch adding mir3gv2 was nicely and promptly reviewed but then was sitting for months and then just merged. Jan 11 23:00:09 Tapper: hm, now checking the history I see it wasn't exactly like that. Jan 11 23:02:47 Tapper: the most uncomfortable was waiting for review comments between the 1st of Sep and 21 of Oct when I sent a "gentle reminder". So it was less than two month Jan 11 23:04:27 Tapper: but "uncomfortable" is subjective. Statistics are just data, one can evaluate them according to any particular requirements. Jan 11 23:10:56 The problem is that we do not have enought manpower to review all patches in time Jan 11 23:11:04 Tapper: there's also apparently an issue that some people are trying to contribute to "core" via github and core devs do not like github so it's often taking more time/effort/persuasion to get patches that way. Jan 11 23:12:05 Does it say anything on the page about contributing? Jan 11 23:12:25 Hauke: lack of official maintainer for ath79 is surprising too. Jan 11 23:12:35 On the wiki Jan 11 23:13:27 Like Hauke says man power is a problem for OpenWrt. Jan 11 23:15:10 we need to be a more inclusive community. Well as best we can anyway. Jan 11 23:16:00 Example from github: https://github.com/openwrt/openwrt/pull/2351 . Sent 25 Aug. Author updated to something meaningful on 21st of Sep. Next comment just on 31st Oct. Author updated on 15th Nov, discussed, pushed nice version on 16th Nov. Tried pinging devs on 2nd Dec. Still no review comments. Jan 11 23:17:10 Does this make any sense at all? Jan 11 23:17:30 PaulFertser I am with you mate. I am trying to say that we can change if we want. If I am on here and see any one being shitty to some one I will speak up! Jan 11 23:18:01 Lack of manpower is understandable. But for a project that accepts outside contributions this might seem like really too long for a relatively simple patch. Jan 11 23:18:41 What's even more annoying is that the author has actually listened, improved his work, did everything he was told to do and then "nothing". Jan 11 23:19:42 about to reboot main gateway with 19.07, fingers crossed Jan 11 23:20:51 Tapper: my point is that the timeline from this "github pull request" example is something many contributors would be unhappy about. So no wonder they're writing comments like than on HN and elsewhere. Jan 11 23:22:38 Tapper: but if the patches are accepted without proper scrutiny, the project can easily become shitty in no time. So it would be silly to require the maintainers to be less strict about reviewing. Jan 11 23:25:01 PaulFertser: you mean it helps writing procd init scripts? Jan 11 23:25:05 Hauke: btw, a probably important message regarding release, I guess it needs clarifications: < shibboleth> paging any devs: this is the third time in a row that 18.06 releases have been signed with a different key than the previous one. this ends up meaning that we can't trust image downloads since we find out about a new key only after each release which may as well be a plant. this is not how Jan 11 23:25:11 you do pgp Jan 11 23:25:37 zorun: hm, not really, it helps debugging weird procd startup issues. Like when you think something should trigger on your wan interface but its called differently. Jan 11 23:25:42 (and I'm no wiki pro; tmomas is the real workhorse behind the wiki in its current form, and it's really really improved since 2 years ago) Jan 11 23:28:18 PaulFertser: I have no idea, but probably https://openwrt.org/docs/guide-developer/procd-init-scripts ? Jan 11 23:28:27 there is already a section called "debugging" Jan 11 23:29:24 zorun: ok, thanks! Jan 11 23:30:12 zorun: I really like "client mode wireless" article on the old wiki but now it's nowhere to be found. Is tmomas on IRC? Jan 11 23:34:19 I don't know Jan 11 23:34:42 PaulFertser: thanks for the info about the signing, jow do you know the details what is going on there? Jan 11 23:37:34 zorun: " Thomas Endt ", ok, will write an e-mail then. Jan 11 23:44:41 for what its worth i love you guys and the work that goes into OpenWrt ;) Jan 11 23:45:51 just trying to figure out how to add a new device Jan 11 23:51:28 zorun: IMHO it would be nice to have some rationale explaining to end users why ar71xx is going to be dropped. Jan 11 23:52:01 PaulFertser: exactly what I thought while reading the HN comments :) https://openwrt.org/releases/19.07/notes-19.07.0#target_transition_from_ar71xx_to_ath79 Jan 11 23:53:10 zorun: yeah, looks great now. Jan 12 01:07:27 build #246 of pistachio/generic is complete: Success [build successful] Build details are at http://buildbot.openwrt.org/master/images/builders/pistachio%2Fgeneric/builds/246 **** ENDING LOGGING AT Sun Jan 12 02:59:58 2020