**** BEGIN LOGGING AT Fri Oct 19 03:00:00 2018
Oct 19 06:36:38 <ak77> jofr: did you solve your /opt case?
Oct 19 06:41:40 <ak77> kuzulis: this is known, the ext3 image is not not yet in DEPLOY_DIR_IMAGE
Oct 19 06:42:29 <ak77> kuzulis: it's in the IMGDEPLOYDIR
Oct 19 08:19:25 <JaMa> RP: last round of _remove bitbake changes seems to have unexpected change in behavior, ping me when you're around to discuss (or I'll send it to list after gathering more data0
Oct 19 08:19:50 <banach-space> hello! Does the order in IMAGE_INSTALL matter?
Oct 19 08:20:09 <LetoThe2nd> banach-space: it basically should not
Oct 19 08:20:30 <banach-space> So would you say that alphabetical order is a good practice?
Oct 19 08:21:08 <LetoThe2nd> banach-space: *shrug* i tend to group by functionality. i would say that any understandable order is a good practise.
Oct 19 08:22:07 <banach-space> Ta LetoThe2nd !
Oct 19 08:43:40 <kuzulis> Hi all
Oct 19 08:44:03 <kuzulis> ak77: What do you mean about "this is known, the ext3 image is not not yet in DEPLOY_DIR_IMAGE" ?
Oct 19 08:45:15 <LetoThe2nd> kuzulis: this means that DEPLOY_DIR_IMAGE is a pure output dir that you're not supposed to rely on.
Oct 19 08:46:01 <kuzulis> uglyoldbob: Hi. Many for your help. Sorry for delay..
Oct 19 08:46:24 <kuzulis> uglyoldbob: I tried bbwarn  "${IMAGE_NAME%-*}" and it returns an empty string
Oct 19 08:46:49 <LetoThe2nd> kuzulis: the imagetypes you depend on are provided in IMGDEPLOYDIR
Oct 19 08:49:57 <kuzulis> LetoThe2nd: Ahh.. do you mean, that I need to take the *.ext3 files from there (from IMGDEPLOYDIR)?
Oct 19 08:50:07 <LetoThe2nd> kuzulis: eeeexactly
Oct 19 08:50:14 <kuzulis> :)))
Oct 19 08:50:23 <kuzulis> magic
Oct 19 08:51:17 <LetoThe2nd> kuzulis: actually its just infrastructural enforcement of an already known best practise
Oct 19 08:51:38 <LetoThe2nd> kuzulis: and it removes undeterministic reliance of your builds on the deploy-dir
Oct 19 08:52:03 <LetoThe2nd> (but yeah, i got bitten by that too.)
Oct 19 08:56:46 <kuzulis> LetoThe2nd: Many thanks. Hmm.. but it does not work too.. because I construct that path as $DEPLOY_DIR_IMAGE/${IMAGE_NAME}${IMAGE_NAME_SUFFIX}.ext3
Oct 19 08:57:14 <kuzulis> LetoThe2nd: Where ${IMAGE_NAME} returns a path with the timestamp
Oct 19 08:57:59 <kuzulis> LetoThe2nd: e.g. irp-embedded-qt5-x11-image-apalis-imx6-20181019085311.rootfs.ext3
Oct 19 08:59:04 <kuzulis> but, seems I need in  irp-embedded-qt5-x11-image-apalis-imx6.ext3 which is symlink to thet file
Oct 19 09:01:15 <LetoThe2nd> kuzulis: well, inspect the variables and the respective directory contents. but IMGDEPLOYDIR is certainly the correct starting point.
Oct 19 09:03:07 <kuzulis> LetoThe2nd: seems I need to use the IMAGE_LINK_NAME instead of IMAGE_NAME to generate a full ext.3 path
Oct 19 09:15:18 <kuzulis> LetoThe2nd: Do I need to use an equivalents of IMGDEPLOYDIR but for the U-Boot, Linux Kernel and other stuff by analogy?
Oct 19 09:15:54 <kuzulis> if I want to copy an U-Boot, Linux Kernel's stuff to another location
Oct 19 09:20:27 <LetoThe2nd> kuzulis: no idea, our system is a bit different there.
Oct 19 09:21:01 <kuzulis> ok, many thanks for your time
Oct 19 09:21:37 <LetoThe2nd> np, good luck
Oct 19 09:36:33 <RP> JaMa: hi, around now. Just looking at your email
Oct 19 09:37:03 <kuzulis> LetoThe2nd: I have another question: my yocto does not generates the mbr.bin && boot.vfat files. Do you know where is it?
Oct 19 09:39:44 <LetoThe2nd> kuzulis: nope. those are not generically known files, but something the is probably either board or bootloader specific
Oct 19 09:39:59 <kuzulis> ok, clear
Oct 19 09:41:32 <RP> JaMa: I reproduced with just OE-Core and a bbappend
Oct 19 09:43:59 <RP> JaMa: what is interesting is that code uses OVERRIDES to apply pkgname so it does d.getVar("RDEPENDS")
Oct 19 09:44:08 <RP> JaMa: now I know that I can try and write a simpler test case
Oct 19 11:28:30 <RP> JaMa: patch in -next and on the bitbake list
Oct 19 12:04:48 <JaMa> RP: thanks, cherry-picked to my builds to test it
Oct 19 12:06:13 <JaMa> RP: I'm still a bit surprised that it caused QA error only in these 2 cases and not in all RDEPENDS_${PN}_remove case I have (a lot)
Oct 19 12:12:04 <uglyoldbob> where do i look to troubleshoot my pkg_postinst_blabla functions?
Oct 19 12:26:20 <banach-space> What happens (or should happen) if two layers contain recipies for similar package/git-repo (yet 2 different versions). Is there any overriding taking place is it basically a FIFO?
Oct 19 13:18:01 <kergoth> banach-space: layer priority controls that in the absence of a PREFERRED_VERSION
Oct 19 13:22:08 <rokm> Hello I have one question, why on one VM during extracting SDK I got "exec: /usr/bin/env: argument list too long"
Oct 19 13:22:23 <rokm> on another PC It works OK
Oct 19 13:29:44 <no_such_user> LetoThe2nd: Sorry for the delay replying! Thanks - is there a standardised way of using containers for Yocto? Im quite interested in this as I was already thinking about using a more virtualised environment to standardise my builds (was originally thinking vmware)
Oct 19 13:30:24 <no_such_user> LetoThe2nd: I was also looking into using toaster - can it work with containerised builds?
Oct 19 13:36:08 <kuzulis> Guys, how I can add the 'truncate' command to my script?
Oct 19 13:36:47 <kuzulis> because a script fails with: run.do_image_miaimg.31444: truncate: not found
Oct 19 13:37:39 <kuzulis> I tried to add : do_image_miaimg[depends] = "truncate" or 'truncate-native' but it fails too
Oct 19 13:45:39 <kuzulis> Ahh.. I need in coreutils-native.. sorry
Oct 19 13:50:28 <mrk377> All Question:  sumo and sumo-next are different SHA-1.  Is it best to go with sumo-next due to bug fixes or should I stick with sumo for a release?
Oct 19 13:55:39 <jofr> Is there a difference between "FOOBAR +=" and "FOOBAR_append"?
Oct 19 13:57:08 <jofr> Sorry. it's in the manual.  :P
Oct 19 13:57:09 <BubuIIC> does anyone know what could be the reason why  do_deploy task doesn't deploy anything until I manually clean the sstate cache?
Oct 19 14:03:57 <mrk377> I see cross-compiler changes and using the host gcc instead of the cross-compiler.  That is necessary.  I need sumo-next
Oct 19 14:09:26 <banach-space> Thank you kergoth !
Oct 19 14:47:40 <yocti> New news from stackoverflow: tree of dependencies for python library in yocto build <https://stackoverflow.com/questions/52894506/tree-of-dependencies-for-python-library-in-yocto-build>
Oct 19 15:55:25 <banach-space> If I have a recipe for a package, say foo (which is 100% sure parsed), does this mean that foo is going to be installed? Or do I have to add it to IMAGE_INSTALL (or somewhere else as a dependency)?
Oct 19 15:55:56 <kergoth> no, adding a recipe makes it available to be built, that's all
Oct 19 15:56:12 <kergoth> either bitbake foo or add foo to IMAGE_INSTALL or otherwise add it to a dependency of something you're building
Oct 19 15:56:35 <banach-space> got it, thank you kergoth !
Oct 19 20:09:27 <uglyoldbob> anybody know what might cause a gtk+3 demo program to no properly go fullscreen?
Oct 19 21:03:00 <rburton> uglyoldbob: maximise or full screen? maximise will keep the title bar if you're using matchbox, proper fullscreen should work though
Oct 19 21:04:28 <uglyoldbob> fullscreen. I'm getting black borders around the program. matchbox-terminal doesn't do this
Oct 19 21:04:50 <uglyoldbob> gtk reports window size of 200x200
Oct 19 21:07:38 <rburton> matchbox-terminal is a gtk+ app, so i'd be looking at what your demo app is doing
Oct 19 21:07:46 <rburton> maybe it has size restrictions in the code
Oct 19 21:15:05 <uglyoldbob> https://github.com/uglyoldbob/gumstix_progs
Oct 19 21:33:22 * uglyoldbob will be back monday
**** ENDING LOGGING AT Sat Oct 20 03:00:00 2018