**** BEGIN LOGGING AT Fri Jun 20 03:00:00 2014
Jun 20 06:24:51 <nerdboy> JaMa: ping...
Jun 20 07:40:34 <mckoan> good morning
Jun 20 08:16:35 <JaMa> nerdboy: pong
Jun 20 09:50:14 <tasslehoff> do I still need X11 for certain things in a wayland/weston image?
Jun 20 09:52:05 <tasslehoff> tried removing X11 from my distro features, but my core-image-weston didn't work well. no mouse cursor, and problems starting Qt5 apps with the wayland/wayland-egl plugins.
Jun 20 16:02:19 <nerdboy> moin
Jun 20 16:06:05 <nerdboy> JaMa: ack
Jun 20 16:06:25 * nerdboy is awake but the italy game just started
Jun 20 16:09:10 <nerdboy> JaMa: wrt to your blacklist patches, what's the policy for fixing B issues?  is -brokensep enough, or ?
Jun 20 16:22:31 <nerdboy> also, libgnomekbd needs a sed cleanup to work with B
Jun 20 16:22:44 <nerdboy> libgnomekbd_2.32 that is
Jun 20 18:17:21 <Crofton> JaMa, do you have any idas on the ice failure?
Jun 20 18:26:37 <Crofton> is there a way I can force a recipe to do fetch again, even if I already have the source?
Jun 20 18:27:59 <kergoth> -c cleanall to clean + remove sources, then bitbake it again
Jun 20 18:28:20 <kergoth> -c fetch -f wouldn't help since thatd force the re-execution of the task, but he task wouldn't do anything if it thinks the sources are intact
Jun 20 18:28:36 <Crofton> thansk
Jun 20 18:28:50 <Crofton> I've got a fail in a git fetch that needs .netrc
Jun 20 18:29:07 <Crofton> trying on my daisy based build
Jun 20 18:29:18 <Crofton> which had worked
Jun 20 18:30:21 <nerdboy> Crofton: which recipe?
Jun 20 18:30:45 <Crofton> local one
Jun 20 18:30:57 <Crofton> that needs to get to a pw protected github repo
Jun 20 18:31:49 * nerdboy just switched to daisy and getting a buttload of PRINC warnings and Package version going backward errors...
Jun 20 18:32:19 <Crofton> something foul is going on
Jun 20 18:32:24 <nerdboy> Crofton: .netrc works "okay" but keys work better
Jun 20 18:32:26 <Crofton> fetch just kind of hangs
Jun 20 18:32:53 <Crofton> on daisy also it seems
Jun 20 18:33:13 <nerdboy> especially with a group of devs and one login that just doesn't work with .netrc...
Jun 20 18:34:51 <nerdboy> i was thinking the the backward version thing was my local config glitch but apparently not
Jun 20 18:34:59 <Crofton> maybe I jsut need to wait longert
Jun 20 18:37:11 <nerdboy> where t=?
Jun 20 18:48:06 <kergoth> i think its best to just use ssh auth and an ssh-agent
Jun 20 18:48:11 <kergoth> but netrc works better than nothing
Jun 20 18:48:33 <kergoth> we should teach bitbake how to ask the ui to prompt in certain rare cases, if not spawned non-interactiveliy
Jun 20 18:56:09 <Crofton> ok, it looks like the key is be patient
Jun 20 19:00:01 <JaMa> Crofton: have you tried building ice in empty sysroot?
Jun 20 19:00:19 <JaMa> Crofton: maybe it was getting right -I paths by some binconfig before
Jun 20 19:01:02 <Crofton> do we have instructions for setting up an empty sysroot
Jun 20 19:01:16 <Crofton> it seems like dbc++ is trying to include iostream.h
Jun 20 19:01:34 <Crofton> maybe some form og gcc improvement
Jun 20 19:01:44 <Crofton> also wondeing if we need to update db
Jun 20 19:02:22 <Crofton> runexe_, said:
Jun 20 19:02:32 <Crofton> Ummm... iostream.*h*??? Should just be #include <iostream> right?
Jun 20 19:02:32 <Crofton> db_cxx.h would be from the db recipe, right - did they change something?
Jun 20 19:02:33 <JaMa> Crofton: just delete tmp-eglibc
Jun 20 19:02:38 <Crofton> ah
Jun 20 19:02:39 <Crofton>  OK
Jun 20 19:02:48 <Crofton> I was thinking host include 
Jun 20 19:03:04 <Crofton> I am getting somethign setup to build off master again
Jun 20 19:03:14 <Crofton> now that I think I have a stable daisy setup for work
Jun 20 19:03:15 <JaMa> shouldn't be host related, because it built at least once on the same host where I'm running world buidls
Jun 20 19:03:20 <Crofton> will try clean build
Jun 20 19:03:22 <Crofton> yeah
Jun 20 19:03:39 <Crofton> when did it start to fail again?
Jun 20 19:03:50 <Crofton> as in did gcc version change?
Jun 20 19:04:45 <JaMa> hard to say, because first it was broken and then some of its dependencies were broken
Jun 20 19:05:10 <JaMa> but I've seen successful build just before merging it
Jun 20 19:05:34 <Crofton> I'll try and poke around
Jun 20 19:05:46 <Crofton> I ahve a stupid fecth that is taking ages to finish first 
Jun 20 19:05:59 <Crofton> I've also pinged the other guys that care
Jun 20 19:06:06 <Crofton> runexe_, and trondeau 
Jun 20 19:06:08 <JaMa> thanks
Jun 20 19:07:02 <JaMa> ugh my buildhistory is too big for git server to handle efficiently
Jun 20 19:07:29 <JaMa> it's running git pack-objects for last 6 hours :/
Jun 20 19:07:58 <mario-goulart> O_O
Jun 20 19:08:06 <Crofton> I assume RP hasa fancy go kart?
Jun 20 19:08:19 <JaMa> being 3GB deep in swap doesn't help
Jun 20 19:08:40 <JaMa> Crofton: I wouldn't say fancy (from the pictures) but fsckingly powerful :)
Jun 20 19:08:44 <Crofton> :)
Jun 20 19:09:30 <Crofton> sounds loud
Jun 20 19:09:37 <Crofton> I am more human/wind power oriented
Jun 20 19:13:14 <nerdboy> the iostream thing depends on version
Jun 20 19:13:34 <nerdboy> anything recent should not be a .h
Jun 20 19:29:12 <Crofton> yeah, I am windering
Jun 20 19:37:00 <Crofton> compiles here. I'll nuke tmp next and try again
Jun 20 20:16:50 <Crofton> JaMa, it builts for an arm machine in a clean sysrrot
Jun 20 20:18:55 <Crofton> trying for qemux86 now
Jun 20 20:29:39 <nerdboy> okay, my daisy sato image also has broken keyboard support
Jun 20 20:30:00 <nerdboy> i'm assuming that's not true for other people?  or is it?
Jun 20 20:30:43 <nerdboy> daisy branches of poky/meta-openembedded/rpi/erlang
Jun 20 20:52:35 <JaMa> nerdboy: honestly I'm not sure how many people tried that combination
Jun 20 20:55:50 <nerdboy> no keyboard on daisy or master
Jun 20 20:56:37 <nerdboy> JaMa: combination being poky/meta-oe?
Jun 20 20:58:14 <nerdboy> was working fine in mid-april (last good build for me) then lost just the cursor keys for a few weeks
Jun 20 21:08:01 <Crofton> JaMa, no luck duping the ice fail
Jun 20 21:09:04 <Crofton> my sysroot has no iostream.h file either
Jun 20 21:10:08 <JaMa> Crofton: I'll try got get more details when I see it again (with tmpdir on tmpfs it's already gone)
Jun 20 21:12:48 <Crofton> thanks
Jun 20 21:14:07 <JaMa> hehe http://errors.yoctoproject.org/Errors/
Jun 20 21:14:56 <JaMa> I don't know if I do too many builds or just get all the errors in my builds :)
Jun 20 21:15:38 * nerdboy wonders why "test" is so much better than "master"
Jun 20 21:16:21 <nerdboy> assuming test is staging/test branch for merge to master...
Jun 20 21:17:07 <JaMa> jansa/test is just old name for jansa/master
Jun 20 21:17:41 <JaMa> I've switched to later, because now I'm running some verifications also from jansa/dylan, jansa/dora, jansa/daisy
Jun 20 21:41:02 <Crofton> hmm
Jun 20 21:41:03 <Crofton> http://pastebin.com/35crpCbw
Jun 20 21:43:02 <Crofton> swig-native is failing
Jun 20 21:45:40 <woglinde> no suprise
Jun 20 21:45:52 <Crofton> why?
Jun 20 21:46:11 <woglinde> because swig is complicated
Jun 20 21:46:31 <Crofton> heh
Jun 20 21:46:32 <Crofton> ok
Jun 20 21:46:41 <Crofton> I was hoping for something more helpful :)
Jun 20 21:46:50 <woglinde> sorry
Jun 20 21:53:42 <Crofton> I am guessing swig has a flaoting depend on libpcre
Jun 20 21:54:24 <Crofton> nope explicit
Jun 20 21:57:29 <nerdboy> newer things typically use cython
Jun 20 21:58:02 <nerdboy> also weird, but probably less wonky that swig...
Jun 20 22:00:58 <XorA> FOSS guys love re-inventing broken wheels :-)
Jun 20 22:02:41 <nerdboy> converting some libimobiledevice stuff from swig to cython (in classic) took some work the first time...
Jun 20 22:07:10 <woglinde> gn
Jun 20 22:09:39 <Crofton> swig-native fail is transient
Jun 20 22:09:42 <Crofton> need to chase that
Jun 20 22:28:37 <nerdboy> okay, this looks completely suspicious => libfakekey0 only exists in the borked build
Jun 20 22:29:51 <nerdboy> ah, it's a depend for matchbox-keyboard
Jun 20 22:30:06 <nerdboy> that's very different...  nm...
Jun 20 23:12:31 <Crofton> JaMa, I do see failure of swig-native
Jun 20 23:28:11 <Crofton> it seems like swig uses pcre-config to find pcre stuff
Jun 20 23:28:23 <Crofton> but, it feels like swig no longer finds pcre
Jun 20 23:28:32 <Crofton> should this surprise me?
Jun 20 23:28:45 <Crofton> or is this related to the binconfig stuff?
Jun 20 23:41:55 <JaMa> yes
Jun 20 23:47:35 <Crofton> bother
Jun 20 23:47:50 <Crofton> it would appear I need to patch swich configure.ac to use pkgconfig?
Jun 20 23:48:49 * Crofton beats RP with a trout
Jun 20 23:49:20 <JaMa> yes for patching configure.ac, no for beating RP
Jun 20 23:49:51 <JaMa> oh well it's Friday night, spank him! :)
Jun 20 23:49:58 <Crofton> rofl
Jun 20 23:49:59 <Crofton> yes
Jun 20 23:50:08 <Crofton> what is the motivation
Jun 20 23:50:16 <Crofton> short answer is fine
Jun 20 23:50:24 <JaMa> for binconfig-disabled?
Jun 20 23:50:29 <JaMa> or spanking?
Jun 20 23:50:31 <Crofton> I supposce these binconfig things are hard to make work cross
Jun 20 23:50:56 <JaMa> yes pkg-config has at least some knowledge when to prefix returned value with sysroot directory
Jun 20 23:51:29 <Crofton> we need to find everything that depends on pcre-config
Jun 20 23:51:50 <Crofton> but it is Friday
Jun 20 23:51:59 <JaMa> not always (it works fine for normal variables like cflags/ldflags, but custom variables are usually returned without prefix and I don't even know if you can define them as "prefix-able"
Jun 20 23:52:05 <Crofton> and I do not feel like fixing swig before I quit working
Jun 20 23:52:35 <JaMa> my builds were finding many failing recipes in last days
Jun 20 23:52:38 <Crofton> well swig is not finding pcre at link time anymore
Jun 20 23:52:47 <JaMa> we need to find people fixing them :)
Jun 20 23:53:01 <Crofton> I need to do swig
Jun 20 23:53:12 <JaMa> good
Jun 20 23:53:36 <JaMa> I've sent long patchset with blacklist for recipes which seem unmaintained because of this
Jun 20 23:53:54 <JaMa> RP sent big patch with many fixes for many recipes
Jun 20 23:54:29 <Crofton> I saw
Jun 20 23:54:32 <Crofton> he missed swig
Jun 20 23:58:18 <Crofton> OK, I've lost interest in working
Jun 20 23:58:27 <Crofton> Monday it is. Sticky note made
Jun 21 00:37:46 <nerdboy> JaMa: is -brokensep an acceptable fix?
Jun 21 00:59:28 <JaMa> nerdboy: yes, for stuff which (seems) to be hard to fix properly
Jun 21 00:59:59 <JaMa> when it's just missing ${S} in some path, then it's bad but still I'll probably accept it
Jun 21 01:00:48 <nerdboy> libgnomekbd actually needs a sed cleanup
Jun 21 01:01:16 <nerdboy> all the other ones i hit at least build with -brokensep
Jun 21 01:01:38 <nerdboy> libgcrypt still has textrels too
Jun 21 01:01:53 <nerdboy> need to dig a little deeper on that one...
Jun 21 01:06:58 * JaMa off to bed
**** ENDING LOGGING AT Sat Jun 21 03:00:00 2014