**** BEGIN LOGGING AT Tue Nov 06 03:00:00 2018
**** BEGIN LOGGING AT Tue Nov 06 03:25:32 2018
Nov 06 07:06:44 <OutBackDingo> ok, everyone still asleep i see :)
Nov 06 08:18:22 <mckoan> good morning
Nov 06 09:09:32 <yocti> New news from stackoverflow: Self-hosted mender fails to update artifact <https://stackoverflow.com/questions/53168706/self-hosted-mender-fails-to-update-artifact>
Nov 06 10:08:06 <timblechmann_> hi! trying to build an image locally, i'm getting this error message: https://pastebin.com/tmxaXV5P ... any idea where this could be coming from?
Nov 06 10:08:40 <timblechmann_> (smells like a python issue, i'm on kde neon18.04, which is based on ubuntu 18.04)
Nov 06 10:13:36 <RP> timblechmann_: looks like unpaired {} in there. Why, don't know
Nov 06 10:18:37 <timblechmann_> RP: i'm not totally sure: ${@'${DISTRO_VERSION}'.replace('snapshot-20181106','snapshot')} ... to me this looks good
Nov 06 10:19:53 <RP> timblechmann_: sorry, yes. Is DISTRO_VERSION set?
Nov 06 10:20:18 <timblechmann_> yes :/ ...
Nov 06 10:21:58 <RP> timblechmann_: if you know where its coming from I'd probably start by simplifying the expression...
Nov 06 10:22:11 <timblechmann_> but i think i've hacked my way around by replacing the whole expression ... not elegant, but it works ... anyway, thanks a lot!
Nov 06 11:09:57 <yocti> New news from stackoverflow: Partitioning Beaglebone internal mmc for 2 kernels <https://stackoverflow.com/questions/53170461/partitioning-beaglebone-internal-mmc-for-2-kernels>
Nov 06 11:18:48 <OutBackDingo> ok  curious, someone can point me to a roootfs encryption recipe ?
Nov 06 11:19:11 * OutBackDingo too tired need a sanity check :)
Nov 06 13:04:17 <OutBackDingo> zzzZZZzzzzzZZZZzzzzzzz
Nov 06 13:04:19 <OutBackDingo> :)
Nov 06 13:16:50 <eduardas_m> hello, fetching libdevmapper with the current thud branch does not work for me: https://pastebin.com/UiEXZQ5c
Nov 06 13:23:44 <OutBackDingo> eduardas_m: https://src.fedoraproject.org/lookaside/extras/lvm2/LVM2.2.02.180.tgz/sha512/6e0a10ab48be4f0c751447c6ab88b8f9fa1dcd1d703de77966e8507b173df70b1987fa0be252c31dfce9e0ee70e5f439db58c8f534e24144a70f254c091554ee/LVM2.2.02.180.tgz
Nov 06 13:23:50 <OutBackDingo> wget the file
Nov 06 13:23:55 <OutBackDingo> put it in downlooads
Nov 06 13:24:12 <OutBackDingo> and touch LVM2.2.02.180.tgz.done
Nov 06 13:24:37 <OutBackDingo> curious what recipe your using for it ?
Nov 06 13:25:15 <eduardas_m> OutBackDingo: the one currently on thud branch in upstream poky
Nov 06 13:25:28 <eduardas_m> OutBackDingo: I don't modify anything
Nov 06 13:26:00 <eduardas_m> OutBackDingo: oh wait
Nov 06 13:26:10 <eduardas_m> OutBackDingo: I am actually wrong
Nov 06 13:26:25 <eduardas_m> OutBackDingo: I think this one is from meta-openembedded
Nov 06 13:27:37 <eduardas_m> OutBackDingo: oh, I see the problem now, I need to update my meta-openembedded layer
Nov 06 13:27:45 <eduardas_m> OutBackDingo: https://github.com/openembedded/meta-openembedded/commit/57bcd4df9710034dcdba06344edb5f9f62b88d70
Nov 06 13:27:54 <OutBackDingo> eduardas_m: i smell encrypted filesystem work
Nov 06 13:27:57 <OutBackDingo> :)
Nov 06 13:28:41 <eduardas_m> OutBackDingo: no, I do not know how to do that...sadly
Nov 06 13:28:57 <eduardas_m> OutBackDingo: I believe this is just a udisks2 dependency IIRC
Nov 06 13:29:21 <eduardas_m> udisks2 is convenient for USB key mounting
Nov 06 13:35:49 * OutBackDingo needs a bitbake guru right now
Nov 06 13:36:00 <OutBackDingo> this recipes not cutting it
Nov 06 14:08:34 <angelo_ts> Any idea on this mysterious error ?   Exception: UnicodeDecodeError: ‘utf-8’ codec can’t decode byte 0xb0 in position 1022: invalid start byte
Nov 06 14:23:38 <OutBackDingo> angelo_ts: a bit more information
Nov 06 14:26:48 <angelo_ts> It fails when yocto does do_rootfs task ….. at the end
Nov 06 14:27:11 <rburton> a bit more context from the log would be needed
Nov 06 14:28:22 <angelo_ts> https://pastebin.com/SH7H1NCB
Nov 06 14:29:42 <rburton> at a guess you've a package which is non-ascii
Nov 06 14:30:13 <angelo_ts> mm
Nov 06 14:30:31 <rburton>                 index_cmds.add('%s -r %s -p %s -m %s' %
Nov 06 14:30:31 <rburton>                                   (opkg_index_cmd, pkgs_file, pkgs_file, pkgs_dir))
Nov 06 14:30:32 <rburton>  ^  -- somethign in that is non-ascii
Nov 06 14:30:53 <rburton> non-utf8 that is
Nov 06 14:31:04 <rburton> non-ascii and also not valid utf8
Nov 06 14:31:32 <rburton> if you can reproduce on demand, hack create_index to get the string from subprocess, print it and then decode
Nov 06 14:31:37 <rburton> then you can look at index 1022
Nov 06 14:32:29 <angelo_ts> ok thanks
Nov 06 14:33:57 <OutBackDingo> rburton: ping
Nov 06 14:34:08 <angelo_ts> rburton, you are in openembedded-core/meta/lib/oe/package_manager.py ?
Nov 06 14:34:33 <OutBackDingo> quick question encrypted rootfa recipe ? ideas?
Nov 06 14:34:41 <rburton> OutBackDingo: no
Nov 06 14:34:44 <rburton> angelo_ts: yes
Nov 06 14:35:30 <rburton> afternoon kanavin
Nov 06 14:35:39 <kanavin> hello :)
Nov 06 14:35:54 <angelo_ts> rburton, ok, thanks, any help oin how to decode ?
Nov 06 14:36:10 <rburton> angelo_ts: the decode() call is the decode
Nov 06 14:36:26 <rburton> so just print the byte string before deocoding
Nov 06 14:39:44 <angelo_ts> so you mean the line   result = subprocess.check_output(index_cmd, stderr=subprocess.STDOUT, shell=True).decode("utf-8")
Nov 06 14:40:07 <angelo_ts> can i know what package generates it ?
Nov 06 14:40:22 <angelo_ts> (i am not that master of python)
Nov 06 14:41:00 <angelo_ts> someting line print(subprocess.check_output(index_cmd, stderr=subprocess.STDOUT, shell=True)) is it ok ?
Nov 06 15:56:09 <armpit> YPTM - armin is on
Nov 06 15:59:25 <halstead> YPTM: Michael on the call.
Nov 06 15:59:29 <JPEW_> YPTM: Joshua Watt here
Nov 06 15:59:38 <RP> YPTM: Richard joined
Nov 06 16:00:06 <vmeson> YPTM: Randy MacLeod joined.
Nov 06 16:00:15 <aehs29> can you guys please share the link
Nov 06 16:00:20 <sjolley1> YPTM Stephen has joined
Nov 06 16:00:27 <vmeson> https://www.yoctoproject.org/public-virtual-meetings/
Nov 06 16:00:50 <RP> vmeson: you just beat me :)
Nov 06 16:01:39 <aehs29> vmeson: RP  thanks
Nov 06 16:01:47 <RP> Meeting notes https://docs.google.com/document/d/15jB6nUJU2wrtnu6w07L9RZpNlj6AoxSTPEX5aELts1g/edit
Nov 06 16:01:47 <aehs29> YPTM Alejandro joined
Nov 06 16:02:09 * RP nudges rburton
Nov 06 16:02:13 <vmeson> topics/crazy ideas for YPTM: ???
Nov 06 16:02:26 <vmeson> 1. status of release
Nov 06 16:02:50 <vmeson> 2. crazy idea: poky should be bisectable
Nov 06 16:03:03 <RP> vmeson: it isn't?
Nov 06 16:03:45 <vmeson> I tried a couple of weeks ago, about 1/3 of the commits didn't build. something about v86d, iirc. I gave up eventually.
Nov 06 16:03:57 <vmeson> I should try again perhaps.
Nov 06 16:04:11 <npongratz> YPTM: Nick is on the call
Nov 06 16:06:07 <denix> YPTM: Denys is dialing in
Nov 06 16:17:18 * vmeson elimintates the dreadded "going forward" expression for sjolley1's google doc.
Nov 06 16:30:39 <RP> vmeson: better than backwards :)
Nov 06 16:31:53 * zeddii sees he missed the call
Nov 06 16:32:16 <JPEW_> zeddii: It is still going
Nov 06 16:32:45 <RP> zeddii: yes
Nov 06 16:35:27 * zeddii figured it was over. A half hour! that's long for how they normally go.
Nov 06 16:35:37 <zeddii> you'll note I'm stalling, and not dialing
Nov 06 16:37:34 <kergoth> huh, i missed it too, oops
Nov 06 16:38:43 <sjolley1> YPTM is still going, discussing YP 2.7 ideas.
Nov 06 16:39:12 <armpit> wow a jvm bitbake
Nov 06 16:39:19 <kergoth> RP: interesting little bug someone else at mentor found. we've had a long standing issue where the varref code will add deps for shell parameter expansion, i.e. ${foo#bar}.. if there's expansions inside such refs, it causes the checksum to shift when it shouldn't. i.e. ${foo#${TOPDIR}/} -> now this function's checksum will change every time TOPDIR changes. might want to eventually allow shell/python codeparser to also hand back additional variable
Nov 06 16:39:19 <kergoth>  exclusions
Nov 06 16:39:29 <kergoth> going to open a bug in bugzilla for it, just thought it was an interesting case i hadn't considered
Nov 06 16:40:09 <kergoth> to clarify, even if TOPDIR is in vardepsexclude, foo#${TOPDIR}/ isn't :)
Nov 06 16:40:11 <kergoth> sjolley1: ah, thanks
Nov 06 16:41:06 <npongratz> YPTM: I don't want to start a drawn-out debate, and I know I'm swimming against the tide, I just wanted to lodge my vote against systemd as default.
Nov 06 16:41:17 <moto-timo> armpit: did you file a bug for eclipse rocko issues? I have a fix, but would like to reference said bug if it exists
Nov 06 16:45:42 <armpit> yes
Nov 06 16:46:14 <armpit> 12977
Nov 06 16:48:33 * vmeson adds memres checking to the November work list/sprint.
Nov 06 16:48:42 <moto-timo> armpit: thanks. I'll send patches to eclipse-yocto ML and merge
Nov 06 16:48:52 <armpit> k
Nov 06 16:48:55 <armpit> thanks
Nov 06 16:49:08 <moto-timo> armpit: oxygen/rocko already built, building neon/rocko now and then it's a done deal :)
Nov 06 17:09:32 <moto-timo> armpit: not sure my mic was working. fixes pushed please build rocko again
Nov 06 19:32:26 <armpit> RP I see we are dropping qt3 & qt4 from the build for master. does that mean we are dropping support for them in the 2.7 time frame?
Nov 06 20:12:33 <RP> armpit: I'm not sure they're actually tested now, we dropped the pieces to build from 2.6 iirc
Nov 06 20:12:38 <RP> armpit: maybe before
Nov 06 21:15:40 <derek007> Hi if I delete tmp/deploy/images/qemux86-64/* with rm -rf command, does anyone knows which bitbake command can regenerate it?
Nov 06 21:27:29 <RP> derek007: remove the do_deploy stamps under tmp/stamps/
Nov 06 21:49:24 <derek007> Hi does anyone knows how to force bitbake to rebuild specific receipe, e.g. recipes-devtools
Nov 06 21:50:39 <kergoth> the 'devtools' thing is just a directory on disk, bitbake doesn't group recipes that way. so you can certainly collect the recipes with find or something..
Nov 06 21:56:43 <derek007> Thanks, kergoth, I made some changes, and added a patch to syslinux under recipes-devtools, I just wanted force rebuild it.
Nov 06 21:57:32 <kergoth> there's no need to do that
Nov 06 21:57:50 <kergoth> adding a patch to SRC_URI will change checksums and bitbake knows what to rebuild based on dependencies
Nov 06 21:58:08 <derek007> I did, I have verified my patch already applied on top of source code by bitbake,
Nov 06 21:58:43 <derek007> oh, I didn't change checksums
Nov 06 22:00:03 <kergoth> ?
Nov 06 22:00:31 <kergoth> bitbake tracks what variables are used by what and checksums variables, functions, tasks, and those checksums are included in task stamp filenames
Nov 06 22:00:45 <kergoth> if you change a task, bitbake reruns it and anything that depends on it, recursively
Nov 06 22:00:53 <kergoth> force building anything is unnecessary 90% of the time
Nov 06 22:00:56 <derek007> oh, I don't have to change checksum manually
Nov 06 22:24:43 <rburton> derek007: just bitbake syslinux is all you need
Nov 06 22:37:07 <derek007> thanks rburton
Nov 06 22:37:40 <rburton> but you don't need to do that if you want to build an image with your new syslinux in, as building the image will rebuild syslinux (assuming its part of the image)
Nov 06 22:46:34 <derek007> rburton, it's wired, I made some changes and added a patch for syslinux, file name is mbr.S, and I verified my patch already applied to mbr.S, after bitbake, file name is ./tmp/work/x86_64-linux/syslinux-native/6.03-r0/syslinux-6.03/mbr/mbr.S
Nov 06 22:47:08 <derek007> but I the output file, mbr.bin, it doens't contain my changes
Nov 06 23:24:59 <kergoth> hmm, should probably migrate from pkgutil style namespace packages to native namespace packages at some point, given they're 3.3+. downside is you want all of the namespace packages to use the same method across layers, but layers need to adapt to new releases anyway, and not many layers extend  the namespace packages in oe-core
Nov 06 23:25:10 <kergoth> will create a bug proposing teh change for 2.7
Nov 06 23:27:48 <kergoth> https://packaging.python.org/guides/packaging-namespace-packages/#native-namespace-packages
Nov 07 00:12:01 <armpit> kergoth, do you want me to add that to the 2.7 planning
Nov 07 01:05:42 <kergoth> armpit: if appropriate, sure. it's proposed still, haven't discussed it at all yet
Nov 07 01:08:53 <armpit> done
Nov 07 01:15:04 <kergoth> thanks
**** ENDING LOGGING AT Wed Nov 07 02:59:59 2018