**** BEGIN LOGGING AT Thu Oct 10 02:59:58 2013 Oct 10 07:24:20 good morning Oct 10 07:25:28 hi mckoan Oct 10 08:16:25 hi stefan Oct 10 08:17:04 hi woglinde Oct 10 08:17:09 woglinde: how are things? Oct 10 08:18:26 I am sleepy again like yesterday Oct 10 08:18:34 damn weather Oct 10 08:18:46 cold in good old Germany? Oct 10 08:18:54 and darkish I guess Oct 10 08:19:11 will be good to be back next month :) Oct 10 08:28:17 woglinde: you can take comfort thinking that is bad weather in the sunny Italy too :-( Oct 10 08:29:15 Sun is shining in UK Oct 10 08:29:22 Bad its cold anyway Oct 10 08:29:32 s/bad/But( Oct 10 08:33:50 i think someone forgot to pay the gas bill in the UK Oct 10 08:34:03 or the UK's thermostat needs to be turned up again Oct 10 08:40:51 rburton: at least here temp dropped a lot over the last days Oct 10 08:41:12 I remember sitting int the sun at the Great Windsor park next to the castle on Sunday Oct 10 08:41:27 And this morning I was freezing when waiting for the train Oct 10 08:44:10 stefan_schmidt_w: http://www.wunderground.com/weatherstation/WXDailyHistory.asp?ID=ICAMBRID29&day=10&year=2013&month=10&graphspan=week Oct 10 08:44:33 stefan_schmidt_w: it appears the weather forgot to get warm this morning Oct 10 08:45:46 rburton: yup, just decided to stay at night level. Can't be bothered to warm up a bit :) Oct 10 08:58:15 morning a ll Oct 10 09:03:52 moin Oct 10 09:05:18 gm Oct 10 09:07:42 JaMa: I start with http://www.openembedded.org/wiki/OE-Core_Standalone_Setup Oct 10 09:08:06 sanity_info is created on first run aparently Oct 10 09:08:09 +p Oct 10 09:20:39 hi guys!, I built gstreamer1.0, plugins good, plugins base and plugins bad, well but after installing them on my beaglebone, i cannot see the plugins under gst-inspect-1.0, i added the PACKAGECONFIG_append_pn-gstreamer1.0-plugins-good = "v4l wavpack" but well i think its not getting installed/included in the image. Oct 10 09:21:06 kevin_rz2: what exactly are you adding to IMAGE_INSTALL? Oct 10 09:21:25 i am aware that The Plugins get built seperate. but is there a way to include them? Oct 10 09:21:56 i added gstreamer1.0 \ gstreamer1.0-plugins-base \ gstreamer1.0-plugins-good \ Oct 10 09:24:00 to IMAGE_INSTALL ofc, i added them to the cloud9-image. After installation i can see gstreamer1.0, plugins-good, and base but the actual plugins are missing. i can see the .ipk's in the deploy folder but all of them are not included in the image itself, and i dont wanna install them one by one, is there a way to include all the built plugins in the image while its building? Oct 10 09:27:04 kevin_rz2: I think you need to be installing gstreamer1.0-plugins-base-meta and gstreamer1.0-plugins-good-meta Oct 10 09:28:05 assuming it's similar to gstreamer 0.1x the gstreamer1.0-plugins-good and gstreamer1.0-plugins-base *packages* don't actually contain anything ; their -meta equivalents will pull in all the built plugins Oct 10 09:31:54 so i'l try gstreamer1.0-plugins-good-meta and see if they get included. and i shall report back. THanks Bluelightning Oct 10 09:32:01 np Oct 10 09:34:37 Hi Oct 10 09:35:45 in my build there are two meta-layers each having bbappend files with the same name, how will they execute? Oct 10 09:36:52 pt: they will both be appended to the recipe, in ascending order of the layer priority for each layer Oct 10 10:16:26 JaMa: the make of Gentoo does indeed fail the test. Though, the bug http://savannah.gnu.org/bugs/?30612 is set as resolved long ago.. Oct 10 10:16:36 http://sources.gentoo.org/cgi-bin/viewvc.cgi/gentoo-x86/sys-devel/make/files/make-3.82-archives-many-objs.patch?view=log Oct 10 10:22:04 ant_work: do you mean it fails our sanity tests? or some other test? Oct 10 10:34:06 there is one variable in the bbappend file of lower priority, i do not want to use that variable, the one with higher priority is my bbappend file Oct 10 10:35:02 how do i do that? Oct 10 10:42:29 pt: you'd have to set it back to the old value in your bbappend Oct 10 10:55:06 actually i do not want to use that variable, so i commented that variable in my bbappend (i guess its wrong) Oct 10 10:58:27 bluelightning: I did a build from scratch, refreshing also local.conf and removing sanity_info which was there since som etime Oct 10 10:58:55 pt: can you be more specific? what do you mean by you don't want to use it? Oct 10 10:59:17 ant_work: ah, I see Oct 10 10:59:21 bluelightning: somehow Gentoo's make 3.82-r4 now fails the test Oct 10 11:00:12 doh Oct 10 11:02:55 http://pastebin.com/2UumXP77 Oct 10 11:06:56 http://pastebin.com/QiHwZrjX Oct 10 11:07:10 bluelightning: I suspect here we have another isue with the leading space Oct 10 11:08:54 if I see the missing hsep/vsep commands error, what is the problem Oct 10 11:08:58 * Crofton|work forgets Oct 10 11:09:00 see here another reproducer Oct 10 11:09:02 http://pastebin.com/musi72tf Oct 10 11:09:21 util.a: foo.c bar.c util.a( foo.o bar.o) is bad Oct 10 11:09:29 util.a: foo.c bar.c util.a(foo.o bar.o) is good Oct 10 11:09:56 bluelightning: does your buildhost pass the test removing sanity_info? Oct 10 11:11:57 ant_work: actually I just checked, on both of my build machines (Kubuntu desktop and CentOS 6 server) I have make 3.81 installed not 3.82 Oct 10 11:13:18 heh Oct 10 11:14:55 I'm not seeing build issues here, neither JaMa Oct 10 11:15:11 the checks that fray introduced are fairly strict Oct 10 11:15:12 I'd say is a false positive (or th etest is b0rked;) Oct 10 11:15:23 fray: any thoughts (when you wake up) ? Oct 10 11:23:41 + I don't see failed sanity test Oct 10 11:31:45 JaMa: well, that's the most strange thing Oct 10 11:32:07 have you (re)moved sanity_info for testing? Oct 10 11:33:33 bluelightning: if you check in sanity.bbclass you'll find the a.m. link to the testcase. This works here, thus make is patched enough Oct 10 11:33:52 bluelightning: strangely the check in sanity.bbclass introduced the leading space Oct 10 11:34:33 which seems leading to the make: *** No rule to make target Oct 10 11:36:36 now the question is, does other make versions tolerate that empty space? Oct 10 11:37:13 s/does/do/ doh Oct 10 11:38:56 ant_work: I did many builds without tmp-eglibc Oct 10 11:39:11 and without sstate when testing pseudo Oct 10 11:39:41 I'm talking about build/conf/sanity_info, out of tmpdir Oct 10 11:39:54 is rarely wiped out Oct 10 11:41:34 what's "build" directory? Oct 10 11:42:05 ah I see Oct 10 11:42:19 where local.conf and bblayers.conf are Oct 10 11:42:58 somehow I confused this one with tmp-eglibc/saved_tmpdir and tmp-eglibc/abi_version Oct 10 11:43:24 I bet that removing it will show you the problem Oct 10 11:43:36 * JaMa trying now Oct 10 11:44:01 yup it shows that now, sorry for misunderstanding it before Oct 10 11:44:06 ;) Oct 10 11:48:08 can you see the date of that file? Oct 10 11:54:27 not so old Sep 25 22:38 Oct 10 11:55:02 creation? ok, then it did pass the test that day Oct 10 11:55:22 modification Oct 10 11:55:50 seems to be "modified" quite often, in 2nd build I have Oct 6 19:51 sanity_info Oct 10 11:55:55 I'm pretty sure this misbehavior is recent Oct 10 11:56:38 have you emerged make 3.82-r4 back in summer? Oct 10 11:58:37 Fri Oct 5 11:16:23 2012 >>> sys-devel/make-3.82-r4 Oct 10 11:58:47 and rebuilt few times since then Oct 10 11:59:59 morning all Oct 10 12:00:05 last rebuilt Wed Sep 11 07:33:14 PDT 2013 Oct 10 12:00:10 gm pb_ Oct 10 12:00:18 hi ant_work Oct 10 12:03:10 who knows about meta-linaro? Oct 10 12:03:13 JaMa: I will try on the Ubuntu I have on another disk Oct 10 12:03:33 pb_: maybe you can enlighten us about Makefile syntax... Oct 10 12:05:05 pb_: http://pastebin.com/2UumXP77 Oct 10 12:07:08 Crofton|work: what's your question about meta-linaro? Oct 10 12:07:35 the default in angstrom seems to fial on libc with hsep/vsep not found erros Oct 10 12:08:01 I think due to F19 having recent tex and those going away Oct 10 12:08:11 I making a number of guesses here :) Oct 10 12:13:26 hopefully I can solve the problem by convincing the person with the issue is to use Angstrom based from 1.4 :) Oct 10 13:40:05 Crofton|work: v2012.12 has fixed the texinfo 5 stuff already, update your setupscripts :) Oct 10 13:41:53 https://github.com/Angstrom-distribution/meta-linaro/commit/e6f8003b039335f7ec75edf9a9b7a84d2d1b911a Oct 10 13:47:32 I talked the guy into using 1.4 instead Oct 10 13:47:48 he now says a fail fetching from infradead Oct 10 13:49:21 the source mirrors are up to date, so all base sources are mirrored Oct 10 13:49:58 not sure if the git2 fetcher ever got fixed wrt snapshots Oct 10 13:53:02 hmm, URL git://git.zytor.com/utils/ranpwd.git just failed for me Oct 10 18:08:29 Is anybody able to fetch stuff from icedtea.classpath.org. I'm trying to bake openjdk-7-jre (meta-java), but it seems that the server is dying with "Proxy Error". Oct 10 18:12:09 mario-goulart hm Oct 10 18:12:37 hm right Oct 10 18:12:48 redirect to icedtea.wildebeest.org Oct 10 18:12:57 I should have updated all recipes Oct 10 18:13:01 my fault Oct 10 18:13:09 thanks for hinting Oct 10 18:13:14 Thanks, woglinde Oct 10 18:13:16 * mario-goulart tries Oct 10 18:13:18 wil do now Oct 10 18:18:33 hm I should introduce a variable for this server Oct 10 18:18:56 woglinde saves the day. Thanks a lot, it works now! Oct 10 18:27:36 mario-goulart git is updated now Oct 10 18:27:50 woglinde: great. Thank you. Oct 10 19:48:10 woglinde: what's the deal with those mercurial web UIs used by icedtea/openjdk? Now icedtea.wildebeest.org seems to be hanging too. Oct 10 20:16:46 can anyone tell me what layer I need to use to get sthttpd? Oct 10 20:19:47 meta-webserver - found it Oct 10 20:25:06 Why isn't sthttpd present in the meta-webserver Oct 10 20:28:46 no one wrote a recipe? Oct 10 20:28:59 use layer.openembedded.org to see if anyone made a reicpe Oct 10 20:30:27 The recipe already exists at https://github.com/openembedded/meta-oe/tree/master/meta-webserver/recipes-httpd/sthttpd hence my confusion. I used it before :-/ Oct 10 20:32:37 maybe not around when branched for yocto1.n Oct 10 20:33:45 ah Oct 10 20:33:55 I wrote a systemd service for it and everything :( Oct 10 20:34:06 copy it from a recent release :) Oct 10 20:35:29 How do I do that in a clean way Crofton|work ? Oct 10 20:35:52 a clean way would be make a custom branch and cherry-pick Oct 10 20:36:11 I doubt anyone has the time to guide me through how to do that? Oct 10 20:36:24 What about an easy way? Oct 10 20:36:25 copy the recipe then :) Oct 10 20:37:32 just literally grab the files and it will parse and work? Oct 10 20:37:53 they shuold I am assuming there is nothing special that is version dependent Oct 10 20:40:56 If it works, how can I get it into upstream? Is it as trivial as sending a git diff? Oct 10 20:41:17 you'll need to make the approiate update request Oct 10 20:41:32 which should be cherrypicking from the commit that is in the next rev Oct 10 20:41:45 who is listed as maintainer for meta-webserver? Oct 10 20:43:15 Crofton|work, Paul Eggleton is Oct 10 20:43:29 aka bluelightening Oct 10 20:44:30 indeed, so Mr bluelightning - how can I submit this to you? Oct 10 20:44:54 see if ti worked first :) Oct 10 20:45:11 it did. Build and ready to roll Oct 10 20:45:14 built* Oct 10 20:45:22 exosyst: http://www.openembedded.org/wiki/How_to_submit_a_patch_to_OpenEmbedded Oct 10 20:49:13 fray: ping Oct 10 20:50:10 bluelightning, I'm guessing he realy needs a cherrypick to a branch Oct 10 20:52:24 bluelightning, Crofton|work - I'm gonna file it against the angstrom-staging-yocto1.4 - shall I just call it a backport Oct 10 20:52:25 ? Oct 10 20:52:43 I think so Oct 10 20:57:40 OK - i think that worked :-/ Oct 10 20:59:37 If that's any good - I'd quite like to bring in my sthttpd.service file as well but first things first Oct 10 21:00:06 new work sho0uld go to master, and get cherrypickes in to the branch Oct 10 21:01:02 Crofton|work, is there a guide on how to do that? I'd quite like to get involved as my job uses OE and I'm happy to invest some time learning Oct 10 21:01:47 I do not think we have written the "How we function as a project" guide yet Oct 10 21:01:50 Also - is there a way to do different do_install_append()s for either sysvinit or systemd? Oct 10 21:02:59 I'm looking through existing recipes and figure I must be missing a trick... Oct 10 21:03:38 yopu could submit a patch to mets-webserver for comments Oct 10 21:04:16 exosyst: most of the time that shouldn't be needed; if systemd/sysvinit are not enabled the files installed for one or the other are just deleted automatically at the end of do_install Oct 10 21:05:25 bluelightning, Oh ok, so if I added a line to bring in the service and systemctl enable it that should suffice? Oct 10 21:06:15 I am off to do something foolish Oct 10 21:06:19 on a mountain bike Oct 10 21:06:21 exosyst: I think systemctl enabling happens automatically as well, but check meta/classes/systemd.bbclass for details Oct 10 21:06:26 Crofton|work: have fun Oct 10 21:06:32 thanks Oct 10 21:06:38 Crofton|work, don't die. you're helpful! ;D Oct 10 21:07:01 funny, you sound like my wife Oct 10 21:07:24 bluelightning, Oh ok. well i've sent that patch to the mailing list. If I can fix it up and get it in that's a great start then I can work on getting the service file in too Oct 10 21:07:25 well, the don't die bit Oct 10 21:07:52 yeah - the better half is such a fickle creature Oct 10 21:14:30 good nite Oct 10 21:25:49 bluelightning, I have a working recipe with a systemd service. Should I resubmit to the mailing list as before? Oct 10 22:06:57 I can't get any text rendered using Qt5, even though I have the qtbase-fonts-* packages installed.. Oct 10 22:26:12 trying adding fontconfig-utils.. Oct 10 22:30:54 output of fc-list is empty Oct 10 22:34:22 hmm, fonts are installed in /usr/lib/fonts, and they are not getting picked up by fontconfig. After I created a symlink to /usr/lib/fonts -> /usr/share/fonts, and re-ran fc-cache, now my Qt-apps are displaying text Oct 11 01:45:28 kroon: don't think i've ever seen fonts installed under /usr/lib... Oct 11 01:45:38 is that some new policy thing? **** ENDING LOGGING AT Fri Oct 11 02:59:58 2013