**** BEGIN LOGGING AT Tue Apr 21 02:59:58 2015
Apr 21 07:57:01 <mago_> woglinde: about our discussion on including manifests into rootfs yesterday, wouldn't it make sense to have a recipe provide the manifest as a file that gets installed into the rootfs like any other file? cause the manifest is available at any time during the build, it doesn't necessarily need to be put in at the end as a postprocessing step?
Apr 21 07:59:06 <woglinde> mago sure, when it fit your needs
Apr 21 08:15:53 <pompomJuice> meh, I cant figure it out. How can you fill a bitbake variable with a shell command's output?
Apr 21 08:16:36 <pompomJuice> ex. var_FOO= "echo 'hello world'"
Apr 21 08:16:55 <pompomJuice> then var_FOO == 'hello world'
Apr 21 08:19:01 <bluelightning> morning all
Apr 21 08:21:39 <pompomJuice> howdy
Apr 21 08:22:37 <mago_> morning
Apr 21 08:27:36 <bluelightning> hi pompomJuice, mago_
Apr 21 09:00:25 <Jin^eLD> mornin
Apr 21 09:09:41 <Jin^eLD> hmm
Apr 21 09:10:06 <Jin^eLD> I tried a number of things in an attempt to have a dependency on an ext3 image when creating my own image and all of these attempts failed
Apr 21 09:11:36 <Jin^eLD> I tried IMAGE_TYPEDEP and a patch from master - did not help
Apr 21 09:12:24 <Jin^eLD> I tried using my processing function not in IMAGE_CMD but as a python function - that also fails
Apr 21 09:12:54 <Jin^eLD> what is the proper way to really depend on some other image being generated? (i.e. when you neext ext3 which you then process into something else?)
Apr 21 09:14:42 <bluelightning> Jin^eLD: IMAGE_TYPEDEP_live = "ext3"
Apr 21 09:14:52 <Jin^eLD> yeah well that did not work for me for some reason
Apr 21 09:14:56 <Jin^eLD> that's what I started with
Apr 21 09:15:28 <Jin^eLD> although I think this time my problem may be somewhere else and I was mistaken, the workaround that I did via the python function did work after all, sorry :) seems another problem led to my recent build failure
Apr 21 09:28:51 <pompomJuice> I need a bitbake python expert to please help me make this compile -> http://pastebin.com/Gkz2qRBz
Apr 21 09:30:16 <pompomJuice> bitbake says: SyntaxError: invalid syntax (<string>, line 2)
Apr 21 09:30:29 <pompomJuice> kind of a error description is that?
Apr 21 09:30:35 <pompomJuice> it says nothing
Apr 21 09:30:49 <pompomJuice> yea your error exists on planet earth would have been as helpfull
Apr 21 09:38:31 <pompomJuice> meh nm i got it
Apr 21 09:45:36 <kanupatar> hi guys
Apr 21 09:46:01 <kanupatar> may i know, we have upstart or systemd ported on arm based oe linux
Apr 21 09:51:19 <bluelightning> kanupatar: systemd is supported, you just need to enable it
Apr 21 09:51:37 <bluelightning> http://www.yoctoproject.org/docs/current/dev-manual/dev-manual.html#selecting-an-initialization-manager
Apr 21 11:02:08 <kanupatar> bluelightning: thanks...which is better systemd / upstart or sysvinit for fast boot up?
Apr 21 11:22:17 <koen> depends on your use case
Apr 21 11:22:24 <koen> and no, we will not do your homework for you
Apr 21 12:15:48 <Cyric> guys someone has tried to compile a qt5 hello world application on OE?
Apr 21 12:16:09 <Cyric> or is working with QT and OE?
Apr 21 12:28:33 <pompomJuice> if you have a sheel function inside your bitbake recipy that contains varible assignment such as foo=${FOO}, when is ${FOO} evaluated? When the function runs or when bitbake is parsing recipies?
Apr 21 12:29:43 <pompomJuice> because if I add to ${FOO} like so -> d.appendVar( 'FOO', 'foo append' ) ==> foo does not contain 'foo append'
Apr 21 12:30:04 <pompomJuice> what is the point of d.appendVar then?
Apr 21 12:30:18 <Crofton|work> cyric looks at meta-qt5
Apr 21 12:31:16 <Cyric> yes but that layer it doenst have QMAKE5 
Apr 21 12:31:32 <Cyric> has been removed by i do not know who... 
Apr 21 12:32:27 <Cyric> so without qmake i cannot use neither qmake nor Cmake to generate makefiles
Apr 21 12:32:45 <Cyric> so i do not know how to compile my application for it
Apr 21 12:34:46 <Cyric> I tried to crosscompile meta-toolchain-qt5 but it returns a lots of error of "unsatisfied recommendation for " 
Apr 21 12:35:14 <Cyric> and i do not know how to overcome that neither
Apr 21 12:35:28 <Crofton|work> hmm, hopefully someof the qt5 guys speak up
Apr 21 12:35:49 <Cyric> i really hope so :)
Apr 21 12:37:23 <Crofton|work> denix, ?
Apr 21 12:37:33 <JaMae> Cyric: there is qmake in meta-qt5 qtbase-native recipe
Apr 21 12:40:39 <woglinde> jama someone wanted it for target or so
Apr 21 12:41:20 <JaMae> there were some changes to make it usable on target as well
Apr 21 12:41:34 <bluelightning> pompomJuice: at parse time, effectively
Apr 21 12:41:47 <pompomJuice> sigh
Apr 21 12:41:59 <pompomJuice> I am still trying to keep passwords out of my CVS
Apr 21 12:42:05 <pompomJuice> but extrausers.bbclass makes this impossible
Apr 21 12:42:08 <bluelightning> pompomJuice: you can still append to it from within an anonymous python function
Apr 21 12:42:10 <pompomJuice> because it does this:
Apr 21 12:42:20 <pompomJuice> user_group_settings="${EXTRA_USERS_PARAMS}"
Apr 21 12:42:37 <pompomJuice> and I am trying to add to EXTRA_USERS_PARAMS dynamically using d.appendVar
Apr 21 12:42:39 <pompomJuice> what a fail
Apr 21 12:42:56 <pompomJuice> currently reading up on anonymous functions
Apr 21 12:42:57 <bluelightning> pompomJuice: where are you doing that currently?
Apr 21 12:43:28 <pompomJuice> basically inserting my own shell function before do_rootfs
Apr 21 12:43:52 <pompomJuice> and extra users runs as a ROOTFS_POSTPROCESS_COMMAND_append = " set_user_group;"
Apr 21 12:44:13 <Cyric> JaMae: so how i can add to the final OE image?
Apr 21 12:44:29 <bluelightning> pompomJuice: I meant where are you doing d.appendVar, because that's python, not shell
Apr 21 12:44:36 <pompomJuice> oh
Apr 21 12:45:05 <pompomJuice> actually it was a python function
Apr 21 12:45:28 <pompomJuice> my original was a shell function, but I added another python function to it so that I can do appendVar
Apr 21 12:46:06 <pompomJuice> my python function is called do_passswords ---> addtask passwords after do_fetch_secrets before do_rootfs
Apr 21 12:46:28 <pompomJuice> do_fetch secrets fetches the passwords and dumps them in WORKDIR
Apr 21 12:47:54 <pompomJuice> I am trying to figure out how anonymous functions work, because that is they only lever that I have not pulled because I cannot understand when those anon functions run
Apr 21 12:53:54 <pompomJuice> I am just going to make a copy of extrausers.bbclass and re-implement it properly... im done fighting this system
Apr 21 13:02:42 <bluelightning> pompomJuice: http://www.yoctoproject.org/docs/current/bitbake-user-manual/bitbake-user-manual.html#anonymous-python-functions
Apr 21 13:03:10 <bluelightning> pompomJuice: ROOTFS_POSTPROCESS_COMMAND is definitely too late to set that variable
Apr 21 13:06:00 <pompomJuice> I am going to modify extrausers.bbclass. I am going to convert it from a shell funtion to a python function and then I am going to use d.getVar
Apr 21 13:06:10 <pompomJuice> that will solve my problems
Apr 21 13:06:33 <pompomJuice> anon funs wont work, they execute to early
Apr 21 13:08:50 <bluelightning> pompomJuice: maybe I am missing something, but surely the easiest way to solve this would be a bbappend that sets the variable? (that you don't check in)
Apr 21 13:09:07 <pompomJuice> how would that work?
Apr 21 13:09:33 <pompomJuice> aah I see
Apr 21 13:09:36 <pompomJuice> let me think
Apr 21 13:10:19 <pompomJuice> would some external process create that file?
Apr 21 13:11:06 <pompomJuice> that sounds like a plan
Apr 21 13:11:19 <pompomJuice> my modifications to extrausers is not working
Apr 21 13:13:30 <pompomJuice> in fact 
Apr 21 13:13:39 <pompomJuice> that might definitely be the best way to do it
Apr 21 13:13:49 <pompomJuice> thanks
Apr 21 13:15:11 <bluelightning> if I can save the world from at least one bbclass fork that is a win ;)
Apr 21 13:18:11 <pompomJuice> :P
Apr 21 14:24:40 <pompomJuice> meh I just cant get this extrausers to work
Apr 21 14:24:51 <pompomJuice> it seems that my backtick cat command was working all along
Apr 21 14:25:34 <pompomJuice> for some reason whatever password you provide does not authenticate...
Apr 21 14:27:51 <OeTester> hello, i get this error: libid3tag/libid3tag_0.15.1b.bb, do_configure) failed with exit code 1
Apr 21 14:28:06 <OeTester> can sombody help
Apr 21 14:28:10 <pompomJuice> inspect the log files
Apr 21 14:28:13 <pompomJuice> what does it say?
Apr 21 14:29:15 <OeTester> autoreconf: Entering directory `.' autoreconf: configure.ac: not using Gettext autoreconf: running: aclocal -I /tmp/arago-tmp/sysroots/armv7a-none-linux-gnueabi/usr/share/aclocal-1.11 -I /tmp/arago-tmp/sysroots/armv7a-none-linux-gnueabi/usr/share/aclocal -I /tmp/arago-tmp/sysroots/armv7a-none-linux-gnueabi/usr/share/aclocal-1.11 -I /tmp/arago-tmp/sysroots/armv7a-none-linux-gnueabi/usr/share/aclocal --force  autom4te: need GNU m4 1.4 or
Apr 21 14:30:22 <pompomJuice> where did you get the recipy?
Apr 21 14:33:37 <OeTester> is not clear for me, some person checks this in in our build system
Apr 21 14:33:59 <OeTester> we move from an older ubuntu system to debian stable
Apr 21 14:34:13 <OeTester> and the build failed
Apr 21 14:34:41 <pompomJuice> aah
Apr 21 14:34:45 <pompomJuice> that makes sense
Apr 21 14:35:00 <pompomJuice> your autotools are out of date then
Apr 21 14:35:04 <pompomJuice> as the error suggests
Apr 21 14:35:27 <pompomJuice> since you are on debian stable (read ancient caveman linux)... it wont work
Apr 21 14:36:08 <OeTester> which version i need ?
Apr 21 14:37:31 <pompomJuice> well as the message suggests... you need v 1.4 or later
Apr 21 14:37:43 <OeTester> our debian use autoconf 2.69
Apr 21 14:38:37 <pompomJuice> automake?
Apr 21 14:38:48 <OeTester> do you mean automake ?
Apr 21 14:39:19 <OeTester> ok, version 1.11 is installed
Apr 21 14:39:30 <pompomJuice> m4 --version
Apr 21 14:39:33 <pompomJuice> what does it say?
Apr 21 14:40:15 <OeTester> m4 (GNU M4) 1.4.16
Apr 21 14:40:54 <OeTester> automake (GNU automake) 1.11.6
Apr 21 14:41:15 <OeTester> autoconf (GNU Autoconf) 2.69
Apr 21 14:41:21 <pompomJuice> make those versions go higher
Apr 21 14:41:28 <pompomJuice> but
Apr 21 14:41:40 <pompomJuice> that might not be your problem
Apr 21 14:42:01 <pompomJuice> see bitbake prepares something called sysroot
Apr 21 14:42:40 <pompomJuice> so that instead of using older versions of tools located on the box (debian in your case), it downloads and compiles these tools instead and places them in sysroot
Apr 21 14:43:40 <OeTester> ok, i'll investigate in this
Apr 21 14:44:32 <OeTester> thank you bye
Apr 21 14:52:50 <Crofton|work> hmmI just had a build fail for binutils-crosssdk-x86_64 for dizzy that was fixed with cleansstate and restarting
Apr 21 14:59:03 <bluelightning> Crofton|work: are there a bunch of errors about not being able to run sed in the log?
Apr 21 15:14:23 <Crofton|work> http://pastebin.com/KwfhiGFz
Apr 21 15:15:11 <bluelightning> Crofton|work: is this do_compile?
Apr 21 15:15:22 <Crofton|work> looks like it
Apr 21 15:15:36 <bluelightning> Crofton|work: do you have the do_configure log?
Apr 21 15:15:48 <Crofton|work> "maybe"
Apr 21 15:16:03 <Crofton|work> remind me in a bit to llok
Apr 21 15:16:26 <Crofton|work> lunch time and I need to let this build finish
**** ENDING LOGGING AT Wed Apr 22 02:59:58 2015