**** BEGIN LOGGING AT Thu Mar 02 03:00:02 2017 Mar 02 07:17:38 where can i find the fetch log for a package? Mar 02 07:17:47 good morning Mar 02 07:36:26 Heya, i was wondering, isnt the final do_rootfs executed as root? Mar 02 07:53:21 LetoThe2nd: Hey, maybe you know, is the do_rootfs command executed as root? Mar 02 07:54:43 maka_: certainly it is not. Mar 02 07:55:05 maka_: nothing in a bitbake build is executed with root privileges. Mar 02 07:55:44 maka_: (think: how could you build as a standard user otherwise? and, has bitbake ever asked you for your root password) Mar 02 07:56:44 aha, that explains a lot Mar 02 07:57:29 I thought it did, since it also executes commands after the build is done. What i thought, its logged in as root then so it can edit files and all Mar 02 07:59:06 nope, we use pseudo repectively fakeroot to emulate such things where needed. Mar 02 07:59:40 LetoThe2nd: How exactly does that work anyway. i'm using ROOTFS_POSTPROCESS_COMMAND, isnt that executed within the yocto image, so as root? Mar 02 07:59:54 maka_: just from pure sane thinking. why would a process that literally does nothing than move around files in your home directory need root privileges? Mar 02 08:00:09 maka_: its executes as-root for the image, but not for your box. Mar 02 08:01:31 Is there any way to get it to at as root? Cause i want to create a file in /etc/apt/sources.list.d/ but it keeps throwing permission denied Mar 02 08:01:47 to act* Mar 02 08:02:06 maka_: then my bets are that this warning is more than correct and that your script does not execute them on the image, but on your real box. Mar 02 08:02:32 aah, alright Mar 02 08:03:04 Then i'll have to find out how to run it one the image Mar 02 08:03:07 on* Mar 02 08:14:51 LetoThe2nd: I think i found the problem. I have to add ${IMAGE_ROOTFS} infront of the path where i want to create a file Mar 02 08:16:02 maka_: :) Mar 02 08:42:29 LetoThe2nd: any chance you know how to access apt-get within the image ? Mar 02 08:43:27 maka_: no, besides trying to run it like any other command too. it has to be installed in the image, of course. Mar 02 08:43:50 I have it installed Mar 02 08:44:19 well then. Mar 02 08:44:24 When i run it as normal command it executes it on my pc Mar 02 08:44:46 then you probably just missed some path, pseudo, chroot, or whatever again Mar 02 08:45:01 i tried ${IMAGE_ROOTFS}/usr/bin/apt-get update but that didnt work either sadly Mar 02 08:46:11 sorry, but i never ever use such things. i try to avoid any postprocessing thats not properly packetized, and i absolutely never ever do runtime/in-target package management Mar 02 08:46:34 Alright then, thanks anyways ^^ Mar 02 08:47:54 maka_: Do you have problems installing deb files? Mar 02 08:48:14 Bas____: nah man, all going perfectly fine Mar 02 08:57:06 maka_:I created a template recipe for installing deb files, would you like me to share it? Mar 02 11:11:06 extsdk-toolchain fails do_package with QA Issue: meta-extsdk-toolchain: Files/directories were installed but not shipped in any package: Mar 02 11:11:06 /locked-sigs Mar 02 11:11:06 /locked-sigs/locked-sigs-extsdk-toolchain.inc Mar 02 11:11:11 that seems new? Mar 02 11:14:15 Hi guys, I'm following the "Yocto Project Quick Start" guide to create a minimal core image Mar 02 11:15:20 the MACHINE variable is set to: qemux86 Mar 02 11:15:46 is QEMU already included in the Yocto repo? Mar 02 11:16:11 I see the "runqemu" script in the "scripts" folder Mar 02 11:18:27 Grynium: you need to have qemu/kvm on your host machine Mar 02 11:20:57 many thanks for your answer Mar 02 11:21:10 I'm using a virtual machine with Ubuntu 12.04 Mar 02 11:22:06 I think can be enough to "apt-get install qemu" Mar 02 11:23:09 joshuagl: so i sort of want the AB to set SANITY_TESTED_DISTROS="" in auto.conf Mar 02 11:23:16 joshuagl: any objections? :) Mar 02 11:24:17 rburton: I think that probably makes sense Mar 02 11:33:08 rburton: is that a request for me to make the change? :-) Mar 02 11:33:26 joshuagl: please ;) Mar 02 11:33:42 as you likely saw, ab exploded when it found fedora24 Mar 02 11:33:56 *ahem* Mar 02 11:40:05 RP/rburton: I'm surprised that http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=28430385ad085be87738c42a997d9e100f0b3baa is required Mar 02 11:40:24 we set SDK_INCLUDE_TOOLCHAIN = "1" in the auto.conf when we build a minimal eSDK Mar 02 11:41:32 RP: hi, do you happen to know why this bug was assigned to me? https://bugzilla.yoctoproject.org/show_bug.cgi?id=10801 I suspect it's just because I'm default assignee for scripts. In this case it should probably be reassigned. Mar 02 11:41:33 Bug 10801: normal, Medium, 2.3, eduard.bartosh, NEW , devtool: upgrade: report diff for changed files referred to by LIC_FILES_CHKSUM Mar 02 11:41:47 or is the problem that the minimal SDK is trying to install that but it isn't built? Mar 02 11:44:28 the tests install it and it isn't present. Mar 02 11:44:52 i'll double check the contents of the sdk though Mar 02 12:00:42 joshuagl: I guess my local minimal version was setup slightly differently Mar 02 12:01:05 joshuagl: i.e. I went even more minimal Mar 02 12:02:49 ed2: This was likely Tim trying to sort out his bugs. I've moved it to Paul for now Mar 02 12:04:32 RP: thanks Mar 02 12:16:40 ed2: can you check that ross/mut-wic has everything you sent? Mar 02 12:17:21 rburton: sure, just a moment Mar 02 12:18:12 heers Mar 02 12:19:55 rburton: which branch is it based on? ross/mut? master-next? Mar 02 12:21:31 ed2: its against master Mar 02 12:24:53 rburton: only one commit is missing: 85bda523 wic: raise WicError instead of calling logger.error Mar 02 12:24:58 whoops Mar 02 12:25:01 glad i asked ;) Mar 02 12:25:44 rburton: thank you for asking :) Mar 02 12:25:52 ed2: oh does ed/wic/wip have everything in? Mar 02 12:25:58 just remembered there was that branch Mar 02 12:26:07 i can just pull that instead Mar 02 12:27:24 rburton: It has, but I probably need to rebase it on master and run oe-selftest -r wic Mar 02 12:27:38 i can do that now :) Mar 02 12:28:02 rburton: be my guest :) Mar 02 12:30:38 rburton: if you find something broken there please let me know. I'll try to fix it right away. There are 29 commits pending there … it'd be nice to have them merged Mar 02 12:30:51 yeah Mar 02 12:31:25 ed2: https://autobuilder.yoctoproject.org/main/builders/nightly-wic/builds/631 is the first test Mar 02 12:35:41 * ed2 keeps his fingers crossed :) Mar 02 12:36:56 Hi,i am new here Mar 02 12:37:33 you guys use any hardware for contributing to yocto Mar 02 12:42:58 Does anyone have experience trying to use meta-xilinx-tools? Mar 02 12:50:12 please anyone reply Mar 02 12:50:30 thebird: what do you mean? Mar 02 12:54:04 rburton, just want know about the yocto project,since you guys working in this field thats why, do you use any development board for working on yocto project? Mar 02 12:55:26 personally, i use a x86 NUC Mar 02 12:55:35 there's a raspberry pi and a beaglebone in my drawer somewhere Mar 02 12:56:52 rburton, is this necessary? Mar 02 12:57:05 no Mar 02 12:58:02 rburton, what are the newbies task? Mar 02 13:07:03 rburton: tets are all green. so far so good :) Mar 02 13:08:47 nice Mar 02 13:09:16 rburton: the tests are quite strange, I'd say. Why do we have them? why don't we just run oe-selftest? Mar 02 13:10:37 rburton: for example they build efi images for qemu targets. quemu doesn't yet support efi, so even if they break I'd not care much. Mar 02 13:11:58 ed2: feel free to ensure that selftest covers all of the cases that nightly-wic does. and qemu/efi support just landed... Mar 02 13:13:04 rburton: yes it does if we run selftest on genericx86* and qemu86* targets. Mar 02 13:28:32 * kanavin is studying Go, and while the language may be an elegant successor to C (didn't get into that), the build system is a bit yuck Mar 02 13:29:15 the standard tools default to creating a static executable with all lib dependencies baked into it, for starters Mar 02 13:30:13 also, said tools will pull the lib dependencies from the internet, but make no effort to have any kind of versioning, it's assumed that latest commit will just work Mar 02 13:30:50 RP: I'm getting FileExists errors from extend_recipe_sysroot - is that a known issue? Mar 02 13:31:11 pohly: tends to mean two recipes installing the same files Mar 02 13:31:17 pohly: or cleanup failed Mar 02 13:31:21 Oh, that. Mar 02 13:31:39 I agree the error message sucks Mar 02 13:31:52 (before someone tells me that) Mar 02 13:31:59 I remember that there's a bug open for the overlapping file problem. Mar 02 13:32:06 right Mar 02 13:32:16 However, in my case it is libgcc.a, so the same package. Mar 02 13:33:01 Hmm, not quite. libgcc-initial is involved. Mar 02 13:33:45 RP: FYI about Go ^^^ Mar 02 13:34:03 I'll try to pull some Real Big Project written in Go, and see how they manage those issues Mar 02 13:34:23 kanavin: they build statically Mar 02 13:34:37 rburton: yep, like chrome Mar 02 13:34:45 The Google Way, I guess Mar 02 13:35:07 kanavin: one problem I'm aware of is that by linking that way, -native tools would bypass pseudo Mar 02 13:35:27 RP: can I find out where the existing recipe-sysroot/usr/lib/x86_64-refkit-linux/6.3.0/libgcc.a comes from? Mar 02 13:35:56 pohly: "find" in sysroot-components ? Mar 02 13:36:25 oe-pkgdata-util, surely Mar 02 13:36:43 tmp-glibc/sysroots-components/corei7-64/libgcc/usr/lib/x86_64-refkit-linux/6.3.0/libgcc.a Mar 02 13:36:43 tmp-glibc/sysroots-components/corei7-64/libgcc-initial/usr/lib/x86_64-refkit-linux/6.3.0/libgcc.a Mar 02 13:36:58 RP: yeah, the reason I'm looking into Go is to have a second data point about interfacing with 'language-specific package managers' in a common framework-ish way Mar 02 13:37:07 pohly: -initial is blacklisted by the class Mar 02 13:37:20 pohly: other way is the sstate-control/ manifests Mar 02 13:37:31 Which class? Mar 02 13:37:32 so that maybe we can have more structure instead of ad hoc hacks like npm support Mar 02 13:37:52 pohly: staging.bbclass in extend_recipe_sysroot Mar 02 13:38:05 kanavin: lots of us love the idea Mar 02 13:39:56 that's where the world is heading anyway Mar 02 13:40:00 RP: it fails for me in do_image_wic. Mar 02 13:40:15 it's just too nice to write programs, and especially get started with writing programs, that way Mar 02 13:40:31 compare with setting up an autotools project lol :D Mar 02 13:40:36 I suspect that's why if mytaskname in ["do_sdk_depends", "do_populate_sdk_ext"] and c.endswith("-initial") does not trigger. Mar 02 13:44:17 RP: what's the reason for the "mytaskname in" condition? Should "-initial" dependencies perhaps always be ignored? Mar 02 13:49:06 rburton: can you do an AB of https://git.yoctoproject.org/cgit/cgit.cgi/poky-contrib/log/?h=akanavin/dnf-rpm4 please? Mar 02 13:56:11 Hello, I have a huge git project with submodules. For this Project i have different recipes which are all using "gitsm". The Problem is that gitsm is copying the full project which takes really long (4GB) for every cmake recipe I'm building. Is it possible that I can download the project just one time and use a symlink or something like that ? (without copying ) Mar 02 13:59:24 Hunk: how about creating a tarball? Mar 02 14:05:29 Kanavin: I want to communicate direct with our git server. So this isn't possible or? Mar 02 14:06:06 I'm not sure why it takes so long every time, though - if you're using the same SRC_URI in all recipes, it should clone the repo just once, and then use the local copy Mar 02 14:06:29 or is the problem that initial clone takes too long? Mar 02 14:07:35 No the cloning is fast. But I have for every cmake a own recipes. For every recipe it copy the cloned repo to his workspace Mar 02 14:08:18 can you show me the recipe? pastebin.com please :) Mar 02 14:11:35 one second Mar 02 14:12:56 pohly: sorry, the code I'm thinking of is in setscene_depvalid which does cover generic -initial deps Mar 02 14:13:59 RP: after various back-and-forth I'm no longer able to reproduce it. But it definitely was trying to hard-link libcc.a from libgcc-initial into the sysroot. Mar 02 14:14:32 pohly: which should not happen :/ Mar 02 14:15:49 @ Kanavin : http://pastebin.com/THeecaXg Mar 02 14:18:40 hi. i need some help booting a custom yocto distro from eMMC on beaglebone! u-boot starts and the kernel is booting, but then it panics because the root partition can not be found... Mar 02 14:19:05 if i enter this at the u-boot command line it boots correctly: "setenv finduuid part uuid mmc 1:2 uuid" Mar 02 14:19:36 i've tried to put that in uEnv.txt, but it didn't work. Mar 02 14:21:09 any ideas? Mar 02 14:21:21 Hunk: I don't know tbh - I haven't used gitsm fetcher, just the basic git fetcher (which does share clones between recipes) Mar 02 14:23:22 @kanavin : With git fetcher it worked. This is just the Problem with gitsm. Do you have to fetch submodules ? Mar 02 14:23:57 joshuagl: does the new cluster publish sstate and if so, to which url? Mar 02 14:24:20 joshuagl: I suspect a small potential flaw in the recent fixes :/ Mar 02 14:24:28 Hunk: we don't have any recipes in oe-core or meta-oe that need to do that. I think you need to fetch submodules directly, if possible Mar 02 14:24:31 RP: IIRC it doesn't currently publish it Mar 02 14:24:53 joshuagl: so how will eSDK know to update from there. Hmm :/ Mar 02 14:25:35 RP: right, it will use the sstate from the other cluster. Probably need to work with halstead to have both published to different locations Mar 02 14:25:52 ok thank you Mar 02 14:26:18 Hunk: fwiw the main git fetcher clones by reference so it should avoid large copies Mar 02 14:29:16 ok I think the gitsm fetcher is copying Mar 02 14:30:42 still waiting for answer :( Mar 02 14:31:30 thebird: what is the question? Mar 02 14:32:37 kanavin, i asked ,what is the task for a newbie? Mar 02 14:32:50 as i am curious to know Mar 02 14:33:12 thebird: http://www.yoctoproject.org/docs/current/yocto-project-qs/yocto-project-qs.html Mar 02 14:33:51 no need to have special hardware, or any hardware at all Mar 02 14:35:42 kanavin, so use used qemu Mar 02 14:35:58 s/so/so you Mar 02 14:36:18 yes Mar 02 14:37:32 so you just build an image and test it over qemu whether it is working fine or not? Mar 02 14:40:31 yes :) Mar 02 14:41:02 I don't even have any hardware on my desk at all, it would be just desk clutter :) Mar 02 14:43:12 kanavin, is there any programming task also? Mar 02 14:44:16 thebird: yes, plenty Mar 02 14:46:06 @Kanavin : is it possible to use local files for building without fetching? Mar 02 14:46:37 Hunk: yes, absolutely Mar 02 14:46:57 with SRC_URI? Mar 02 14:47:01 kanavin, thanks :) Mar 02 14:47:16 Hunk: http://www.yoctoproject.org/docs/2.2/dev-manual/dev-manual.html#building-software-from-an-external-source Mar 02 14:49:38 rburton, joshuagl: we have a second problem with my testsdk change, its basically broken Mar 02 14:49:49 local.conf gets overwritten Mar 02 14:50:52 noooo Mar 02 14:56:58 rburton: I can see the beauty of Go though, they don't want a build system that is driven by configuration files. Builds happen entirely from what is in the source code, which is kinda awesome for someone coming from the autotools world :) Mar 02 15:02:06 ugh, am I missing something? how can checkuri succeed if fedorahosted has disappeared? e.g. http://recipes.yoctoproject.org/rrs/recipedetail/329/ Mar 02 15:02:21 rburton: ^ Mar 02 15:03:43 rburton, joshuagl: tempted to create auto.conf to avoid this... Mar 02 15:04:34 kanavin: because it redirects to the HTML page Mar 02 15:04:40 which is a valid download Mar 02 15:04:44 -> success Mar 02 15:05:07 doesn't check the sum I guess Mar 02 15:05:30 apparently not Mar 02 15:05:50 kanavin: can i suggest 1) fiiing a bug to make checkuri compare the checksum and 2) grep/fix where possible Mar 02 15:07:03 rburton: there's a bug for making an AB test for -c checkpkg, which will do the same thing and more Mar 02 15:07:17 cool Mar 02 15:07:30 (that is, will check for regressions in checking latest upstream version - from known to unknown) Mar 02 15:07:38 long term we either fix checkuri to check sums, and or drop it Mar 02 15:08:14 grep/fix isn't trivial, I have no idea where all those projects have moved Mar 02 15:08:21 we have about 10 of them Mar 02 15:08:33 for instance, I tried to find where the new home of xmlto is, and couldn't Mar 02 15:10:19 that's a big indicator that they're dead Mar 02 15:10:25 as they had a year to move Mar 02 15:11:01 rburton, kanavin: checkuri just checks success/failure, it doesn't download, hence doesn't check the checksum Mar 02 15:11:12 rburton: this is the new fedorahosted, so some might moved here https://pagure.io/ Mar 02 15:12:13 we might do what we always have, and fetch from debian Mar 02 15:14:53 always worth checking if the reverse depends are still valid Mar 02 15:15:33 rburton: what about the AB run, can you do it? Mar 02 15:16:14 what run? Mar 02 15:16:22 oh a new checkpkg run? Mar 02 15:16:46 no, the dnf run :) Mar 02 15:16:52 asked you 90 minutes ago :) Mar 02 15:17:14 oh, sorry :) Mar 02 15:17:33 master has received a ton of updates, and I rebased, and frankly I don't want to do local testing anymore :) Mar 02 15:18:02 RP, joshuagl: new ab hung again Mar 02 15:18:11 centos7, again Mar 02 15:18:22 i keep on pausing that :) Mar 02 15:19:24 paused, re-fired master, fired dnf for kanavin Mar 02 15:19:45 great :) Mar 02 15:20:43 kanavin: on yocto.io Mar 02 15:20:55 yoctolinux.build Mar 02 15:21:03 yes .build exists Mar 02 15:21:08 register! Mar 02 15:21:21 rburton: are things stuck somehow? https://autobuilder.yocto.io/grid Mar 02 15:21:36 ah, they just got unstuck Mar 02 15:24:14 rburton: cool that master is tested with my branch back to back, so I can verify which errors are not my fault :) Mar 02 15:40:59 kanavin: you own boost, do you want 11104? Mar 02 16:26:46 so *thats* why so many newbies are using 1.6 Mar 02 16:26:57 because the meta-ti page on yoctoproject.org says 1.6 is the supported release Mar 02 16:28:06 rburton: and who's fault is that? Mar 02 16:28:39 denix: it says your BSP hasn't been approved, so maybe yours? or jethros? dunno. didn't even know that page existed :) Mar 02 16:29:29 rburton: it says "compatible" not "works" Mar 02 16:29:29 denix: not implying blame, just curious who owns that page. if its something you can easily fix then maybe you should, if not then i'll prod jethro as maybe he knows Mar 02 16:29:49 rburton: that's just the compliance status, which has been in limbo for quite some time... Mar 02 16:29:57 gotcha Mar 02 16:30:16 lets put this on the "rewrite the yocto website" todo list Mar 02 16:30:30 rburton: "yocto compatible 2.0" process is still being finalized and should be running again soon Mar 02 16:30:32 but i wonder if that is why i've noticed lots of newcomers using 1.6 Mar 02 16:31:04 doubt it... Mar 02 16:34:30 denix: as you're around, are you able to update the proxy wiki page? or would you like me to? Mar 02 16:37:04 joshuagl: I'd prefer you to do it, but if you really can't, I can give it a shot... Mar 02 16:37:54 denix: fair enough Mar 02 16:44:00 rburton: can you do a build of golang compiler use oe-meta-go layer on your machine Mar 02 16:44:12 and see if you still get same error ? Mar 02 16:44:43 https://github.com/mem/oe-meta-go Mar 02 16:44:44 rburton: meta-intel is 1.5 :-P Mar 02 16:45:02 rburton: I want to understand if this is a problem with my restructuring Mar 02 16:45:20 denix: lols Mar 02 16:46:56 khem: trying now Mar 02 16:50:45 rburton: assuming we do get a green master, what next? mut? -next? Mar 02 16:52:35 rburton: bitbake go-cross-x86_64 would be enough I guess Mar 02 16:57:02 zeddii: I have to add this fragment https://github.com/kraj/openembedded-core/blob/kraj/master/meta/recipes-kernel/linux/linux-yocto/drm.cfg to boot linux-yocto based images on virtualbox Mar 02 16:57:47 zeddii: we have talked about it in past too. is this now included by default config in kernel ? Mar 02 17:00:28 khem: yeah that worked Mar 02 17:00:41 rburton: ok thanks Mar 02 17:00:56 RP: i'll prune mut down to something sensible and rebase Mar 02 17:00:56 And the patch from me was v2 Mar 02 17:01:04 that you were using Mar 02 17:01:20 yeah Mar 02 17:01:53 OK Mar 02 17:02:41 khem: just checking now :) Mar 02 17:05:57 rburton: ok it adds SSTATE_SCAN_CMD = "true" Mar 02 17:06:08 to go-common.inc Mar 02 17:06:30 whaaaaaat Mar 02 17:08:09 this is the paths compiled into the binaries problem? Mar 02 17:08:37 khem: yeah definitely broken here Mar 02 17:08:50 tig Mar 02 17:09:01 stupid focus follows intent isn't working yet then Mar 02 17:10:21 RP: seems so yes Mar 02 17:10:34 rburton: so did you test v2 or v1 ? Mar 02 17:10:54 v2 Mar 02 17:11:02 ok Mar 02 17:11:23 let me create a deb8 container and try to reproduce Mar 02 17:11:56 ill admit this is deb8 plus a few bits of testing, but everything else works! Mar 02 17:12:24 rburton: I just want to try debian 8 unmodified Mar 02 17:12:46 rburton: we should run this patch trhoguh AB Mar 02 17:12:52 and see if it fails there as well. Mar 02 17:13:01 if not I would blame your host config Mar 02 17:14:05 when the ab is idle i'll add it to an image and give it a spin Mar 02 17:14:11 both builders are working now though Mar 02 17:15:02 rburton: I have a quick fix for the wikilog. I'd appreciate a window to push that at some point :-) Mar 02 17:18:04 joshuagl: i'll fire go tomorrow or something, so grab a window when the current builds end. Mar 02 19:30:40 rburton: sure Mar 02 19:52:06 rburton: I have prepared a debian8/64bit VM here Mar 02 19:52:17 let me see if I can reproduce it here Mar 02 19:57:11 khem: I'm still working on getting llvm-config sorted out and am wondering if you could give me some pointers. Mar 03 00:56:54 RP: debian8 VM I am able to build go-cross-x86_64 fine Mar 03 00:57:11 RP: having trouble reproducing Ross's build env here Mar 03 00:57:39 RP: this is for go addition to Core, can you give this patch a shot ? Mar 03 00:58:22 this is the patch http://git.openembedded.org/openembedded-core-contrib/commit/?h=kraj/go&id=76b3dc7000fa5940fedb3fcebccd9d23b0b46047 Mar 03 01:48:59 Does anyone knows if there is a recipe exists for udunit library in yocto? Mar 03 01:57:51 check layers.openembedded.org Mar 03 02:08:48 Crofton|work: but those are only meta-layers, not individual recipes. Thats why I am asking. Is there an easy way to find it :) Mar 03 02:16:19 Crofton|work: sorry, missed the search recipe part. I will try that. Thanks! Mar 03 02:16:52 that is the quickest way to see what people have done Mar 03 02:17:06 even if you find someones half complete work Mar 03 02:17:42 no mathcing recipes, what is udunit Mar 03 02:18:48 Crofton|work: that was a great help I did not know that you can search recipes :) . But what about personal branch someone might be working. Mar 03 02:19:05 No way to know Mar 03 02:19:19 it does look like it sill be easy to build Mar 03 02:19:57 expat is in core Mar 03 02:21:23 hard part is deciding where to put it :) Mar 03 02:21:44 Crofton|work: ok, might be easy to create one. just want to check if someone is already doing that. Otherwise I try to create one. Mar 03 02:22:04 I've never seen it come up before Mar 03 02:23:03 Crofton|work: ok. good. Then I will try to write a recipe for it. But as you said in which layer ? Mar 03 02:24:49 what does it do? what kind of applications Mar 03 02:25:14 Crofton|work: http://www.unidata.ucar.edu/software/udunits/udunits-2.2.24/doc/udunits/udunits2.html Mar 03 02:25:45 heh Mar 03 02:25:50 this is greek to me :) Mar 03 02:26:16 but I'd suggest meta-openembedded/meta-oe Mar 03 02:26:28 as it isn't clear where else it would go Mar 03 02:26:57 and I am tired :) Mar 03 02:27:23 Crofton|work: haha :) . ok. Mar 03 02:28:18 Crofton|work: ok I will see what I can achieve with this. Thanks for your help. Mar 03 02:30:07 ideally meta-oe should be called meta-kitchensink Mar 03 02:30:44 I have been saying meta-oe should disperse into more specific layers Mar 03 02:31:39 I agree, but in this case I can't see an obvious other layer :) Mar 03 02:35:47 may be recipes should just start appearing in packages themselves like spec files **** ENDING LOGGING AT Fri Mar 03 03:00:00 2017