**** BEGIN LOGGING AT Wed May 14 03:00:00 2014 May 14 06:19:32 morning May 14 07:20:14 good morning May 14 08:35:00 morning all May 14 08:36:11 hi bluelightning May 14 08:40:50 hi mckoan May 14 09:04:42 I am using midori on a yocto build for a atom E680 May 14 09:04:49 and want to make use of videotags in html5 to play rtsp streams May 14 09:04:56 gst-launch-0.10 playbin2 uri=rtsp://........... works perfectly May 14 09:05:03 the html5 page is not, I see the image, but it is verry shocky May 14 09:05:36 while I thought the midori is also using playbin2... May 14 09:06:07 rburton: this sounds like it might be a question for you ^ May 14 09:17:39 wv: midori might be using playbin in a way that means it can't do hardware accelerated decoding May 14 09:18:15 wv: file a bug with midori, if playbin2 works from a terminal then you've confirmed its a midori bug May 14 09:22:40 rburton: would it be possible that it is already fixed in a more recent midori? standard build by yocto is 0.5.5, but I see that there is already a 0.5.8 May 14 09:23:08 wv: certainly worth looking at the changelog May 14 09:24:14 rburton: and would it be something mirori related? Or webkit related? (I mean, where should I file a bug?) May 14 09:24:50 wv: not sure, don't know enough about webkit's internals. start with midori. May 14 09:46:11 rburton, I just downloaded the midori-0.5.8 all package May 14 09:46:22 but there is no configure/wscript/... in it? May 14 09:46:27 how do you compile this one? May 14 09:47:28 sorry, my bad May 14 09:47:41 0.5.5 is tarred including the 0.5.5 directory May 14 09:47:47 0.5.8 isn't May 14 10:00:11 hi guys why I get this error? package-group-custom is listed in packages multiple times, this leads to packaging errors. May 14 10:00:25 where package-group-custom is my own defined package group May 14 10:04:03 rspock: what lines do you have that touch PACKAGES in your packagegroup recipe? May 14 10:05:46 PACKAGES += "packagegroup-custom" May 14 10:06:12 rspock: if your recipe is called packagegroup-custom, that name is already in PACKAGES May 14 10:06:22 (assuming the recipe does "inherit packagegroup") May 14 10:06:50 actually even without that May 14 10:07:14 so basically, comment out / remove the PACKAGES += "packagegroup-custom" line and it should work May 14 10:07:38 In my recipe I have IMAGE_INSTALL = " \... packagegroup-custom \ .. May 14 10:08:05 the recipe is called mc2-dispencer-qt5 May 14 10:08:28 hang on which recipe are we talking about now - an image recipe? or the same packagegroup recipe? May 14 10:08:47 the image recipe May 14 10:08:56 ok, that should be fine May 14 10:09:01 ok May 14 10:10:05 my packagegroup recipe has more than one packagegroup defined May 14 10:10:17 do I have to split them in multiple files? May 14 10:10:34 no, one file for multiple packagegroups is fine May 14 10:10:48 ok, I get it May 14 10:10:51 each individual packagegroup is a package defined within the recipe May 14 10:10:51 thanks you May 14 10:10:54 np May 14 11:29:18 hi yocto. Is there anyway to non-strip binaries in a package? May 14 11:29:47 vicky: INHIBIT_PACKAGE_STRIP May 14 11:30:14 yes. i found . thanks May 14 11:33:21 I want to add midori-0.5.8 in stead of midori-0.5.5 to my common May 14 11:33:43 for this I used the midori-0.5.5 file and adjusted the checksums in the file May 14 11:34:03 however, the 0.5.5 tar file is extracted in a midori-0.5.5 folder (tarred like that) May 14 11:34:38 and the 0.5.8 tar file is tarred to the current folder (without subfolder), but the do_configure looks for a subfolder (which is not present) May 14 11:34:50 How can I adjust the behavior that it doesn't look for a subfoldeR? May 14 11:36:14 the tarball doesn't expand a directory? May 14 11:36:27 that's just… stupid May 14 11:37:18 nope May 14 11:37:21 if i understand correctly, add ;subdir=midori-0.5.8 to the src_uri May 14 11:37:30 to tell bitbake to create a directory and unpack in there May 14 11:38:00 then moan at the midori team for being a bit daft May 14 11:38:44 between the parentheses? May 14 11:39:24 SRC_URI = "http://www.midori-browser.org/downloads/${BPN}_${PV}_all_.tar.bz2;subdir=${BPN}-${PV} May 14 11:39:30 or whatever the current url is, but like that May 14 11:39:34 "this url, put it here" May 14 11:39:45 ok, thanks May 14 11:41:04 wv: bonus points for sending a patch to oe-core when you've test that the new midori isn't worse than the old one! May 14 11:41:46 rburton: okey, I'll keep that in mind ;) May 14 11:41:58 passed the configure step btw, failed at do_install now May 14 11:58:47 rburton, How can I check if my configure went ok? I don't see any makefile coming in the tmp/work/core2-poky-linux/midori/0.5.8-r0/midori-0.5.8 May 14 11:59:06 and am unable to do a make install afterwards also May 14 11:59:08 so it fails May 14 11:59:08 wv: read temp/log.do_configure May 14 12:00:41 rburton, can you check? http://pastebin.com/cycVhYzv May 14 12:00:56 I don't see any failure, but also don't find it normal that the usage is printed... May 14 12:03:24 hm, if I compare it with the output of 0.5.5, then it does not look good May 14 12:05:00 presumably theres been some options removed May 14 12:05:21 so compare the options the recipe passes with the options listed there May 14 12:06:47 yups, I saw it also May 14 12:06:52 it's a complete different configure May 14 12:19:39 rburton: ERROR: QA Issue: midori: Files/directories were installed but not shipped May 14 12:19:39 /usr/share/appdata May 14 12:19:39 /usr/share/appdata/midori.appdata.xml May 14 12:20:09 hm, we might need to start doing something with appata May 14 12:20:12 like, deleting it May 14 12:20:24 feel free to either ignore or add those to FILES_${PN} May 14 12:22:00 thanks, I'll check (sorry for the maybe stupid questions, but it's the first time I'm really into adding packages instead of just using them ;) May 14 12:25:10 allright, it compiled :-) May 14 12:25:37 (btw, is there avariable name for /usr/share/appdata? I now added it just with full pathname to FILES_${PN} May 14 12:25:50 it will be ${datadir}/appdata May 14 12:27:11 ok, thanks May 14 14:00:22 Hi. Does anybody know: How far into the build process are changes in RDEPENDS/RRECCOMMENDS considered? If I write a packaga that inherits a class "example.bbclass", and in that class it defines a tast do_something() { bb.appendVar("RDEPENDS", "somedependency") }, will it be considered? May 14 14:01:13 depends if do_something happens before do_package May 14 14:13:32 Denwid: rburton: basically that can't work May 14 14:13:52 oh, i thought if you got the name right, that would May 14 14:14:05 you can't set a variable in one task and have it changed for another task, it's two separate copies of the same datastore May 14 14:14:25 oh, the tasks get a copy of d? May 14 14:14:31 yes May 14 14:14:36 * rburton tries to remember that May 14 14:14:49 there are two ways to programmatically set packaging variables that will work: May 14 14:15:02 (a) set them on an anonymous python function (in which case bitbake will verify they are satisfied) May 14 14:16:08 (b) set them in a populate_packages_prepend() (in which case bitbake won't check them, they only end up in the final package) May 14 14:16:58 (and RDEPENDS/RRECOMMENDS are packaging variables, just in case that wasn't clear) May 14 17:51:20 I've just made an image with Sato demonstration, and the fonts are not rendering in high resolutions. In low resolutions they're ok again. How to enable fonts to work in high resolutions? May 14 20:01:58 Is FILES_${PN}_append a valid statement? If I grep through meta* I find only FILES_${PN}_append_someotherstuff May 14 20:02:37 it's valid, but pointless May 14 20:02:41 just use += May 14 20:03:22 True, now I learned that all the bitbake variables work in recipes, too :) May 14 20:05:01 kergoth: when do I need _append, when +=; do_install_append is nothing unusual May 14 20:05:13 += is valid for variables, not functions May 14 20:13:37 so rule of thumb: _append for functions and += for variables? May 14 20:13:59 only because its impossible to use += with functions due to the syntax May 14 20:15:26 kergoth: is there a reason why the documentation states IMAGE_INSTALL_append? (http://www.yoctoproject.org/docs/current/ref-manual/ref-manual.html) May 14 20:15:44 yes. _append is postponed. it occurs at the end of the parsing process. May 14 20:15:55 if you use += in local.conf, it'll be applied immediately, and a recipe using = will overwrite it May 14 20:16:08 by using _append, it applies at the end of the parse, after the recipe gets parsed, so can affect the recipes May 14 20:17:09 kergoth: so += and _prepend are then the same when it comes to variables in local.conf? May 14 20:17:28 I just explained how they're completeliy *not* the same when used from local.conf. May 14 20:17:32 read what i just wrote again :) May 14 20:18:41 kergoth: thats how I come to it, if I use += then it might be overwritten by the recipe, the _prepend from my understanding is put before the recipes run. so _prepend is equal to += May 14 20:19:01 no, that's completely wrong May 14 20:19:05 opposite of _append (which gets applied after the recipe parsing is done). May 14 20:19:11 again, += is *immediate*. _append is at the end of the parsing process May 14 20:19:24 just as =+ is immediate, and _prepend is at the end of the parsing process May 14 20:19:41 comparing _prepend and += wouldn't make sense, one is a postponed prepend, one is an immediate append May 14 20:19:46 they're completely different operations May 14 20:19:54 append != prepend, += != =+ May 14 20:20:40 afk May 14 21:31:31 busybox udhcpc does not seem to run as a daemon. How do you make sure that you get a new IP address if network is back again or if the network changes? May 14 21:31:41 I only found how/why it was removed: http://patchwork.openembedded.org/patch/33183/ May 15 01:00:31 * ajtag is away: Away **** ENDING LOGGING AT Thu May 15 02:59:59 2014