**** BEGIN LOGGING AT Sun Oct 25 02:59:57 2020 Oct 25 14:20:25 @zeddii: waiting for k3s ;) Oct 25 14:21:59 @zeddi, moto-timo: My understand is that k3s and k8s are quite incompatible. I mean you can not deploy stuff from an k8s cluster to an k3s cluster. This would need some more hacking. Oct 25 14:22:07 My understanding Oct 25 14:57:55 the dependencies and networking of the current k3s need quite a bit more work. I'll be hacking on it this week while listening to ELCe. Oct 25 14:58:01 * zeddii should go buy some guiness Oct 25 22:10:47 I have a recipe that generates a library(*.so and *.a) file only. It seems the rpm name takes the library name as suppose to recipe name. How do I change the behavior Oct 25 22:12:26 Eg recipe-A generate libfoo.so. I was expecting "recipe-A.rpm" however, I am getting libfoo-1.2.0.rpm Oct 25 22:13:55 I run bitbake -e recipe-A.bb | grep libfoo There was nothing. Anything I am missing here. Oct 25 23:11:39 Hadi: debian.bbclass does that by default, as it aligns with debian distributions. it's done that way to let multiple library versions be installed at the same time on device that upgrade Oct 25 23:12:06 Hadi: INHERIT_DISTRO ?= "debian devshell sstate license remove-libtool" as set in conf/distro/defaultsetup.conf is what does it. if you set that yourself in local.conf or your own distro .conf, you can leave debian out Oct 25 23:33:11 Thanks Kergoth Oct 25 23:47:04 Perhaps this is trivial to you. When do I need to enable debian in a distro? I am using rpm packaging in my destro. Should I care about DEBIAN? Oct 25 23:49:01 Hi Kergoth, thanks for you response Oct 25 23:49:16 Perhaps this is trivial to you. When do I need to enable debian in a distro? I am using rpm packaging in my destro. Should I care about DEBIAN? Oct 26 00:22:19 Hadi: the rationale for why that's done is here: https://www.debian.org/doc/debian-policy/ch-sharedlibs.html#run-time-shared-libraries Oct 26 00:22:34 oops, they're gone Oct 26 00:45:16 zeddii: the moment I tried to build cri-o I thought it asked for libselinux... I can revisit it Oct 26 00:46:09 RobertBerger: I don't know about compatibility, but k3s is probably a better fit for on target? I don't have enough experience doing what I am trying to do with it yet. Oct 26 00:52:35 zeddii: I'll go back and review OverC again Oct 26 00:54:07 moto-timo: so you do, or don't want selinux as a cri-o depends ? We have it in there, but as a "conditional" meta-selinux dependency. Oct 26 00:54:56 it's odd though. It somehow snuck into the full depends, but yes, it is supposed to be conditional .. not in the sense of cri-o, but in the sense of meta-virt as a whole Oct 26 00:55:08 I can't buid it at the moment, due to a selinux build issue though Oct 26 00:55:23 /opt/bruce/poky/build/tmp/work/core2-64-poky-linux/libsepol/2.9-r0/recipe-sysroot-native/usr/bin/x86_64-poky-linux/../../libexec/x86_64-poky-linux/gcc/x86_64-poky-linux/10.2.0/ld: ../cil/src/cil_verify.lo:/usr/src/debug/libsepol/2.9-r0/libsepol-2.9/src/../cil/src/cil_mem.h:39: multiple definition of `cil_mem_error_handler'; ../cil/src/cil.lo:/usr/src/debug/libsepol/2.9-r0/libsepol-2.9/src/../cil/src/cil_ Oct 26 00:55:24 mem.h:39: first defined here Oct 26 00:55:24 | collect2: error: ld returned 1 exit status Oct 26 01:15:37 zeddii: personally, I don't currently want SELinux Oct 26 01:16:10 zeddii: but that might change the moment I get a secure system request Oct 26 01:17:22 zeddii: and also the moment I'm back on my centos8 or fedora-32 builders Oct 26 01:17:40 * moto-timo pretends to go to bed and be on Dublin-ish time Oct 26 01:19:06 zeddii: I wonder if we can do the selinux depends in dynamic-layers bbappend? Oct 26 01:19:37 I need to investigate if it really is required, I'll have a look at the cri-o code and see what's using it. Oct 26 01:20:17 my stance was that all of meta-virt couldn't depend on meta-selinux, just for cri-o, and we made some anonymous python for it, but I don't recall if we can packageconfig it out of existence as a dependency. Oct 26 01:20:32 if so, it can be a package config and avoid yanking it in. **** ENDING LOGGING AT Mon Oct 26 02:59:57 2020