**** BEGIN LOGGING AT Sat Apr 06 02:59:58 2013 Apr 06 11:31:49 hello, may be, somebody have worked recipe for qtgstreamer? Apr 06 18:30:27 | Makefile:100: OPTION_EGLIBC_NSSWITCH is disabled, but fixed config file Apr 06 18:30:27 | Makefile:101: *** does not exist: "". Stop. Apr 06 18:30:47 is this a known issue? I tried to compile a systemd image with todays master Apr 06 19:05:32 this is still insanely broken. :( Apr 06 21:12:45 zecke, #oe might be a better palce for that question. But it is the weekend Apr 06 21:22:25 Crofton|work: the manual is wrong.. it asks to set DISTRO_FEATURES = "systemd" Apr 06 21:22:45 Crofton|work: which will cause the various libc-* stuff to not me in DISTRO_FEATURES. :} Apr 06 21:24:08 * Daemon404 stopped trusting the manual and handbook years ago Apr 06 21:24:11 it's always wrong... Apr 06 21:25:04 but this must be a new entry. :) Apr 06 21:39:34 "always wrong" is a bit of an exaggeration Apr 06 21:40:01 but especially in new areas there may be things that need correcting/expanding Apr 06 21:40:23 I have been spending time helping improve the documentation recently Apr 06 21:41:05 I will talk to Scott about fixing that DISTRO_FEATURES = "systemd" thing specifically next week Apr 06 21:50:22 Daemon404: years ago it was in a bad way, it is a lot better these days Apr 06 21:50:51 Daemon404: if inaccuracies are reported they will get fixed too Apr 06 21:51:43 my biggest problem for a long time was that a lot of confg vars simply were not documented Apr 06 21:51:49 not even with source code comments Apr 06 21:51:58 cant to read btbake source code Apr 06 21:52:01 s/cant/had/ Apr 06 21:52:30 just kinda... kept doing that Apr 06 21:53:27 bluelightning: thanks Apr 06 21:54:09 yeah the best way you can help improve the documentation is to report issues in the bugtracker; there's a category for documentation bugs Apr 06 21:55:08 i unfortunately dont use poky for anything but generating toolchains and sysroots for other projects' build coverage anymore Apr 06 21:55:18 so i dont run into many that arent "metatoolchain is broken" Apr 06 22:02:41 Daemon404: fair enough, but please do report any issues you do find and perhaps look again before saying its "always" wrong :) Apr 06 22:02:54 yes Apr 06 22:03:16 i am ashamed to admit, many non-reprted bugs are me being "golly, i dont want yet another bug tracker account" Apr 06 22:03:48 and thats not so reasonable. Apr 06 22:04:19 Daemon404: there are limits to what the project can reasonably do to help... :) Apr 06 22:04:39 indeed Apr 06 22:05:02 actually i know one regression that was introduced lately (few months) ago Apr 06 22:05:08 i should test to see if it still exists Apr 06 22:05:24 (uclibc's static libs no longer worked, and failed linking in libc.a) Apr 06 22:06:54 Daemon404: uclibc isn't a speciality of mine but khem would no doubt be interested to know if that were the case Apr 06 22:07:20 i shall test at some point this weekend i suppose Apr 06 22:08:00 Daemon404: if it were reported, it might get fixed. I can say for sure its less likely otherwise... Apr 06 22:11:05 morning Apr 06 22:13:44 RP_: bluelightning : You don't happen to know the knobs for systemd images/distro features? Apr 06 22:13:50 bluelightning: I think some of the sanity test failures we've seen recently are timeouts from slow booting of qemu with the AB under load :/ Apr 06 22:14:45 RP_: indeed, I saw your patch to increase the timeouts further Apr 06 22:14:51 zecke: I know you probably want sysvinit and systemd in DISTRO_FEATURES Apr 06 22:14:56 RP_: what was the recent dmesg failure ? Apr 06 22:15:09 bluelightning: that I do not know. Another segfault Apr 06 22:15:14 VIRTUAL-RUNTIME_init-manager? Apr 06 22:15:37 zecke: VIRTUAL-RUNTIME_init_manager = "systemd" Apr 06 22:17:02 okay...then it doesn't work. :) Apr 06 22:17:21 zecke: existing build or a build from scratch? Apr 06 22:17:58 ah another typo Apr 06 22:18:23 ../documentation/dev-manual/dev-manual-common-tasks.xml: VIRTUAL-RUNTIME_init-manager = "systemd" Apr 06 22:18:26 ../documentation/dev-manual/dev-manual-common-tasks.xml: VIRTUAL-RUNTIME_init-manager = "systemd" Apr 06 22:18:29 needs to be _ Apr 06 22:18:34 instead of - Apr 06 22:18:35 :} Apr 06 22:19:14 i added sysvinit to the BACKFILL_CONSIDERED.. but will leave it out Apr 06 22:19:19 anyway.. time to sleep now Apr 06 22:19:39 zecke: thanks, will get that fixed Apr 06 22:20:58 thanks Apr 06 22:21:41 RP_: back in the days the Trolltech doc team introduced "snippets".. so the code was still embedded snippet by snippet in the doc but the app was compilable.. I wonder if this could be done for the local.conf file. :} Apr 06 22:22:07 anyway, I am really off now Apr 06 22:22:33 zecke: A local.conf fragment for systemd is certainly a good diea Apr 06 22:22:39 zecke: 'night! Apr 07 01:52:30 Hi, how do i prepend a shell task to do_qa_configure? Apr 07 01:55:20 I'm trying to write a .bb for an autotools package. do_qa_configure dies on a configure problem that I can ignore safely ignore. I want to remove the offending file before running do_qa_configure for this particular package. Apr 07 01:58:30 myopiate: do_qa_configure is a postfunc of do_configure... so you can probably just do whatever you need to do in a do_configure_prepend Apr 07 01:58:45 er, do_configure_append Apr 07 01:59:27 bluelightning: thanks! Apr 07 02:06:31 another issue I'm having is autoreconf is being run on my package... but I don't want it to (as a work around for an upstream issue). Is there a standard way to stop this behaviour? Apr 07 02:18:43 myopiate: you can override do_configure to just run oe_runconf instead Apr 07 02:19:20 myopiate: however we do run autoreconf for a good reason; quite often upstream autotools files can be out-of-date leading to compatibility issues Apr 07 02:19:39 *cough*libtool*cough* Apr 07 02:20:44 yes libtool is special indeed Apr 07 02:25:43 bluelightning: I probably need to look closer at said upstream issue. thanks again! **** ENDING LOGGING AT Sun Apr 07 02:59:58 2013