**** BEGIN LOGGING AT Tue Aug 12 02:59:58 2014 Aug 12 03:47:03 #rtsd.zigbee Aug 12 03:47:15 #rtsd.zigbee Aug 12 03:47:17 hello Aug 12 08:34:20 morning all Aug 12 08:41:41 hi bl Aug 12 08:49:30 hi woglinde Aug 12 08:49:56 hi Aug 12 13:16:13 any reason why I cant _override SRC_URI (like SRC_URI_x86-64 += "moo") ? Aug 12 13:16:23 (well, at least it looks like I cant here) Aug 12 13:19:05 _append, not += Aug 12 13:19:17 SRC_URI_x86-64_append = " moo" Aug 12 13:19:23 oh ... thanks :) Aug 12 13:20:10 (and I actually stumbled upon that for another recipe already ... silly me) Aug 12 13:21:21 in theory they operate the same Aug 12 13:21:36 but _append is evaluated later and works most of the times Aug 12 13:21:43 just don't forget the leading space :) Aug 12 13:21:48 :) Aug 12 13:22:09 btw should it be _moo_append, or _append_moo, or both work ? Aug 12 13:27:03 koen: you mean _append_x86-64 ;) Aug 12 13:27:32 bencoh: _append_moo Aug 12 13:28:43 thx :) Aug 12 13:40:07 RP, I am having a fetch issue with powertop, lookslike url changed Aug 12 13:42:09 weird Aug 12 13:42:25 made it through today Aug 12 13:42:30 do we have default mirrors? Aug 12 13:51:16 bluelightning: right, I always get confused :) Aug 12 13:51:47 Crofton|work: http://downloads.yoctoproject.org/mirror/sources Aug 12 13:51:54 from poky.conf Aug 12 13:52:18 what is poky? Aug 12 13:52:33 yesterday powertop fialed on a fetch Aug 12 13:52:37 today it worked Aug 12 13:52:45 I think the src url changed Aug 12 13:54:08 lookslike they swithced to https Aug 12 13:54:40 ah. i thought you asked if powertop was in the oe (pre)mirrors.. Aug 12 13:55:59 just just wondering how it fixed itself :) Aug 12 13:56:07 I am doing an oe-core build Aug 12 13:59:12 Crofton|work: I've found why zeroc-ice was failing in some of my builds with missing iostream.h Aug 12 13:59:23 and? Aug 12 13:59:55 related to the c++ header patch on the list? Aug 12 14:00:14 Crofton|work: older db (/usr/include/db4.8/db_cxx.h) doesn't have this define: #define HAVE_CXX_STDHEADERS 1 Aug 12 14:00:24 and then it includes iostream.h instead of just iostream Aug 12 14:00:33 and we don't have iostream.h in sysroot Aug 12 14:00:46 why different db's? Aug 12 14:00:56 it was failing in builds which included meta-webos layer, which has db-4.8.* Aug 12 14:00:58 mind you I need to get on the db6 issue :( Aug 12 14:01:02 ah Aug 12 14:01:06 OK, thanks for update Aug 12 14:01:35 adding --enable-cxx in db-4.8.30.bb could possibly help too, but I'm trying to get db upgraded in our builds instead Aug 12 14:01:49 I should send a patch for the powertop url before I forget Aug 12 14:02:15 Crofton|work: we also have sources.oe.org mirror Aug 12 14:02:23 if you send the archive to Tom King Aug 12 14:02:23 yeah Aug 12 14:02:36 I'd like to get url fixed before someone else has an isse Aug 12 14:03:08 if it does apply to older releases then please send the archive to Tom as well Aug 12 14:03:29 people don't update metadata so often, so it would be nice to get it resolved without any action from them Aug 12 14:03:34 I suspect someone already has the archive since my build worked this morning :) Aug 12 14:03:37 ka6sox, ping Aug 12 14:04:47 proto=https right? Aug 12 14:05:02 or https:// in SRC_URI Aug 12 14:05:20 Crofton|work: for git it's git://foo;protocol=https Aug 12 14:05:29 similar for svn Aug 12 14:05:32 this is for tarball on https server Aug 12 14:05:42 https://tarball Aug 12 14:05:46 k thanks Aug 12 14:05:57 yeah, I am lazy, and I like talking with you guys Aug 12 14:05:59 :) Aug 12 14:10:03 heh Aug 12 14:13:34 is there some known solution of https:// fetch failing recenly with this: Aug 12 14:13:35 ERROR: Fetcher failure: Fetch command failed with exit code 4, output: Aug 12 14:13:35 OpenSSL: error:14077410:SSL routines:SSL23_GET_SERVER_HELLO:sslv3 alert handshake failure Aug 12 14:13:38 Unable to establish SSL connection. Aug 12 14:14:29 JaMa: fetching from which server? Aug 12 14:15:48 github with older wget (and dylan based buidl) Aug 12 14:16:08 http://github.com/downloads/isis-project/WebKit/WebKit_0.54s.zip Aug 12 14:18:03 ah wget-native is ASSUME_PROVIDED, so it's ubuntu-12.10 issue Aug 12 14:19:32 maybe the server is being more stringent about SSL handshaking with the recent OpenSSL debacle? Aug 12 14:19:54 there was a github blog about that iirc Aug 12 14:20:46 ah right Aug 12 14:20:53 JaMa: change the url to https:// Aug 12 14:21:14 that error happens when you try https on port 80 Aug 12 14:21:37 (according to the interwebs) Aug 12 14:23:40 ~morning Aug 12 14:23:40 Good morning! Aug 12 14:31:01 hmm, I have a dev package with a .so that is the real lib Aug 12 14:31:09 this is now an rdepends error Aug 12 14:31:18 how do I fix the packaging Aug 12 14:31:26 the .so isn;t versioned Aug 12 14:33:03 add it to FILES_${PN}, remove it from FILES_${PN}-dev and add 'dev-so' to INSANE_SKIP Aug 12 14:33:05 Crofton|work: you want to include the .so in the 'main' package? Aug 12 14:33:06 (iirc) Aug 12 14:33:11 yeah Aug 12 14:33:16 that would be best Aug 12 14:33:31 i do this in general Aug 12 14:33:31 FILES_SOLIBSDEV = "" Aug 12 14:33:31 FILES_${PN} += "${libdir}/*.so" Aug 12 14:33:37 koen, that sounds about right Aug 12 14:33:40 the answer is to stop using cmake Aug 12 14:33:49 I forget the SOLIBSDEV though :) Aug 12 14:34:02 koen, not an option Aug 12 14:34:32 did I guess the build sys correctly? Aug 12 14:34:44 *g* Aug 12 14:34:54 I need to learn how to fix that cmake-ism Aug 12 14:54:59 koen: https:// fails the same Aug 12 14:55:13 GNU Wget 1.13.4 built on linux-gnu. Aug 12 14:56:49 GNU Wget 1.15 built on linux-gnu. Aug 12 14:57:06 1.15 used in ubuntu 14.04 works fine (with both http and https) Aug 12 15:06:49 friends do not let friends use ASSUME_PROVIDED? Aug 12 15:17:35 so S = "${WORKDIR}" is a bad idea apparently Aug 12 15:17:57 it is Aug 12 15:18:14 I was wondering why packaging linux failed at stripping scripts built for the host Aug 12 15:18:52 Crofton|work, pong Aug 12 15:19:34 is the powertop tarball in the oe mirror? Aug 12 15:20:00 bbl Aug 12 15:20:12 Crofton|work, I'll look Aug 12 15:20:15 but which version? Aug 12 15:20:25 2.6.1 Aug 12 15:20:34 I'll check Aug 12 15:20:36 culs Aug 12 15:20:43 url seems to have changed Aug 12 15:20:48 working on a patch also Aug 12 15:21:22 if you have a URL please pass otherwise I'll put here what I find and you say yes or no before I put it tin the repo Aug 12 15:25:07 Crofton|work, https://01.org/sites/default/files/downloads/powertop/powertop-2.6.1.tar.gz Aug 12 15:26:26 Crofton|work: you could also email the 01.org maintainers, ask them to add a redirect for the old URLs Aug 12 17:30:51 Crofton|work, what do you want me to do with powertop 2.6.1? Aug 12 17:31:20 it needs to be in the mirror Aug 12 17:31:24 that is all Aug 12 17:31:33 I'll send in a patch with the updated url Aug 12 17:32:06 okay Aug 12 17:32:19 I'll get it in when I get back from the mountain Aug 12 19:11:18 now I am motivated to add cgdb to meta-oe Aug 12 21:58:09 hey, what does this message mean? Aug 12 21:58:09 WARNING: QA Issue: locale-base-ar-ma.utf-8 rdepends on eglibc-binary-localedata-ar-ma.utf-8, but it isn't a build dependency? [build-deps] Aug 12 21:58:32 from what I can tell... the eglibc-binary-localedata is provided by the same recipe as the localebase.. Aug 12 21:58:36 (verifying that though) Aug 12 21:59:55 Hmm.. maybe not? Aug 12 22:00:01 (maybe I broke something) Aug 12 22:03:21 fray_, chek history of insane.bbclass Aug 12 22:03:58 it's the test-dependencies check Aug 12 22:04:41 ya, I found that just wasn't sure what is was actually checking.. Aug 12 22:04:45 but I think I see now.. Aug 12 22:04:50 there is a dep, but no provide Aug 12 22:05:17 this one should be in JaMa's world-build logs Aug 12 22:06:19 http://logs.nslu2-linux.org/buildlogs/oe/world/ Aug 12 22:07:21 these QA are being nailed down one by one Aug 12 22:10:00 fray_: insane_qa doesn't see packages provided by PACKAGES_DYNAMIC Aug 12 22:10:30 fray_: so when you explicitly rdepend on something you need to add that something explicitly in PACKAGES variable in the recipe which provides it Aug 12 22:10:50 I'm trying to figure out where that was added.. I suspect it's being done dynamically as well Aug 12 22:10:51 JaMa: heh, I didn't mention that for the klibc-utils, just added a DEPENDS ;) Aug 12 22:12:59 looks like the dep is introduced from libc-package.bbclass **** ENDING LOGGING AT Wed Aug 13 02:59:58 2014