**** BEGIN LOGGING AT Wed Dec 31 02:59:58 2014 Dec 31 03:37:26 Is there anywhere I can get a list of the packages in openembedded? Dec 31 08:35:19 hello all Dec 31 08:36:40 quick newbie question: I don't the mail I sent through git send-email in http://lists.openembedded.org/pipermail/openembedded-devel/2014-December/date.html. Do I have to do configure something? Dec 31 08:36:56 s/don't the/don't see the/ Dec 31 08:41:18 does the system you did this on have a working MTA that will forward outbound mail? (or did you configure an external MTA) Dec 31 08:42:42 tbr: I think it's fine, I tried to send to my personal mail address (in another domain) and it seemed to worked fine Dec 31 08:43:24 and the sender address git uses is subscribed to that list? Dec 31 08:43:35 (I don't know the list specifics, just general suggestions) Dec 31 08:43:56 tbr: I tried yesterday without subscribing, and I retried this morning after subscribing Dec 31 08:44:08 tbr: oh wait Dec 31 08:44:28 it just appeared in http://lists.openembedded.org/pipermail/openembedded-devel/2014-December/date.html Dec 31 08:44:38 problem solved, thanks for your reply :) Dec 31 08:44:38 :) Dec 31 09:05:16 morning all Dec 31 09:11:46 hey bluelightning Dec 31 09:12:24 * koen wonders what to do with the edison wifi stuff Dec 31 09:12:56 the sources are only in the tarball drop, it's a waste to download that monster for just the wifi Dec 31 09:13:09 but I also want to avoid hosting the code myself Dec 31 09:13:59 that's another issue I'm trying to resolve Dec 31 09:46:07 it's GPL, so licensing is not a problem Dec 31 09:46:32 maybe .tar in the metadata repo Dec 31 09:47:38 bluelightning: for things like https://github.com/koenkooi/meta-edison/commit/28e3f940a33706c7fbd8fd89d91ec2d338568776 do you prefer a .inc or a bbclass? Dec 31 09:48:24 I haven't looked to see if those things are actually needed (or doing something at all) Dec 31 09:50:07 well, at least the deps ought to come in through MACHINE_ESSENTIAL_EXTRA_* / MACHINE_EXTRA_* Dec 31 09:50:44 machine-specific image recipes always feel a bit wrong to me Dec 31 09:51:07 a common inc file would be one way to start though Dec 31 09:58:19 right Dec 31 10:54:03 * bluelightning is finally dealing with php vs. modphp Dec 31 10:54:38 koen: btw, thanks for tidying up the php recipe earlier, this job is a lot easier after that Dec 31 11:01:03 np Dec 31 13:12:42 RP, I've had a report you broke (personally) RM_WORK_EXCLUDE in 6bf06d80c2ce03dfdedac5ad8cf42ef8e36b0ecb Dec 31 13:12:59 We need it to work for images for wic Dec 31 13:26:37 Crofton|work: I don't think RP is around, you might want to file a bug (assigned to him) Dec 31 13:26:54 I need to badger the guy that reported it to do it :) Dec 31 13:27:05 He better not be off doing something dangerous Dec 31 13:27:08 :) Dec 31 13:27:21 not as far as I'm aware :) Dec 31 13:27:37 I had a dream with Dirk, Linus, subsurface, the cia and small linux devices last night Dec 31 13:27:41 I am worried about myself Dec 31 13:28:11 I don't think I can recall having dreams about embedded Linux at all, which is probably a good thing Dec 31 15:19:02 anyone care to comment on the hypothesis that X-compiling is slower than native compile of smae package? Dec 31 15:31:51 Crofton|work: that would depend on a lot of factors, chief among them how fast your "native" platform is Dec 31 15:32:18 we believe gnuradio builds faster native :) Dec 31 15:33:49 what is the native platform? and what kind of time difference are you observing? Dec 31 15:34:35 x86 Dec 31 15:34:37 not sure Dec 31 15:34:46 this is something we talk abuot while waiting on builds :) Dec 31 15:36:45 silly question, but you're not comparing the entire compile+install+populate_sysroot+package time to just compiling natively, right? Dec 31 15:39:08 that likely gets lumped in Dec 31 15:39:18 but it still takes ages :) Dec 31 15:39:32 that is somewhat of an apples to oranges comparison then ;) Dec 31 15:39:52 is there a way to get timing info for the tasks? Dec 31 15:39:55 it may be possible to improve Dec 31 15:40:14 yes, for timing info use toaster or dig manually in tmp/buildstats Dec 31 15:40:30 k Dec 31 15:40:52 in case we ever move from the bitching about it phase :) Dec 31 15:52:14 Hmm, anyone run into a leveldb link failure in qtwebkit before? Dec 31 16:27:05 kergoth: what's the link error exactly? Dec 31 16:27:47 leveldb is a new dependency in QtWebKit v5.4.0 for HTML5 Indexed Database feature Dec 31 16:27:50 | LevelDBDatabase.cpp:(.text._ZN7WebCore15LevelDBDatabase12openInMemoryEPKNS_17LevelDBComparatorE+0x34 Dec 31 16:27:51 ): undefined reference to `leveldb::NewMemEnv(leveldb::Env*)' Dec 31 16:27:56 ah Dec 31 16:28:22 kergoth: sorry, I didn't bump into this one yet Dec 31 16:28:34 as I disable indexeddb feature :) Dec 31 16:29:02 if you're not interested in that feature, i'd suggest to disable it as well Dec 31 16:40:33 good idea. now for a qt5 noob, what's the easiest way to do that from the recipe/bbappend? :) Dec 31 16:41:54 totally OT, but You Need A Budget is on sale for $15 at steam for the next 2 hours. quite nice envelope-style budgeting software. Dec 31 16:42:59 Crofton|work: I've been banned from all activities and alcohol atm FWIW :/ Dec 31 16:43:07 timing of said ban sucks Dec 31 16:43:14 * Crofton|work shudders Dec 31 16:43:42 kergoth: the quick & dirty way is to add a patch in your recipe, patching this file : https://qt.gitorious.org/qt/qtwebkit/source/0930236a6e8bef5c2b90f1522a06acf9dee9c15d:Tools/qmake/mkspecs/features/features.pri Dec 31 16:44:04 kergoth: replacing ENABLE_INDEXED_DATABASE=1 to ENABLE_INDEXED_DATABASE=0 Dec 31 16:44:06 ah, right. looks like some of our packageconfigs sed that Dec 31 16:44:07 well, at least everyone knows who is driving people back from the pub Dec 31 16:44:15 thanks for the assistance Dec 31 16:44:24 I've encouraged the reporter to create a bug Dec 31 16:44:34 kergoth: no pb Dec 31 16:45:47 kergoth: the clean way is to tell qmake to remove the feature, but I don't remember exactly the syntax Dec 31 16:46:10 kergoth: and I don't have access to recipes I have at work from home :) Dec 31 16:46:18 np Dec 31 16:48:09 huh, running a devshell confuses zsh's compaudit, due to the current uid being 0 and all. anyone worked around that? getting prompted on every devshell about it gets old Dec 31 16:52:39 kergoth: you could try PSEUDO_UNLOAD=1 Dec 31 16:53:59 its not the command thats the problem, its the shell Dec 31 16:54:22 devshell spawns zsh, zsh loads my dotfiles since HOME Is still the same, but those dotfiles arent owned by the user it thinks its running as :) Dec 31 16:54:30 oh right, I see Dec 31 16:54:59 i'm guessing i can work around it in the zsh config by e.g. disabling its permissions check when id==0, or when the pseudo env vars are set or somesuch Dec 31 16:55:08 * kergoth shrugs Dec 31 17:56:44 <_oldtopman> How do you manually call autogen.sh for a recipe, rather than the autoreconf that normally runs? Dec 31 17:57:22 <_oldtopman> irssi's autogen.sh does a lot more work than it should, and the build process fails when the script is not called. Dec 31 17:59:02 _oldtopman, try doing your own do_configure() function in the recipe Dec 31 18:01:54 <_oldtopman> kroon: Right now it looks like this, but it still doesn't seem to work. Dec 31 18:01:55 <_oldtopman> http://hastebin.com/ututamufum Dec 31 18:02:11 <_oldtopman> No mention of any of the echo strings in the build logs, either. Dec 31 18:02:42 _oldtopman, what if you move do_configure() after inheriting autotools ? Dec 31 18:04:22 <_oldtopman> kroon: That works way better now. Dec 31 18:05:45 _oldtopman, I think "inherit" can be thought of as "#include" in C, I might be wrong about that though Dec 31 18:06:07 but the autotools.bbclass defines its own do_* functions Dec 31 18:07:08 <_oldtopman> I see. Dec 31 18:07:56 <_oldtopman> Do you know if anyone's built irssi anyway? It seems impossible to me that someone made a working version with that recipe. Dec 31 18:08:54 no idea Dec 31 18:09:29 its in meta-oe ? Dec 31 18:10:52 <_oldtopman> Yeah, meta-oe/recipes-connectivity/irssi Dec 31 18:11:20 <_oldtopman> It calls for a slightly older version that I can't find (since irssi.org is down), but unless they completely overhauled their build system, there's no way that thing ever worked. Dec 31 18:14:37 _oldtopman, recipe looks fairly recent Dec 31 18:15:24 _oldtopman, but I think JaMa's world build testing only covers build-time errors, not run-rime errors Dec 31 18:17:07 <_oldtopman> kroon: No, I'm suffering from all build-time errors. I wonder what I'm missing then. (obviously, it's not in the broken folder, so it should have worked at some point) Dec 31 18:18:03 SO if I add RRECOMMANDS to a package Dec 31 18:18:19 <_oldtopman> I'll look into it more when irssi.org comes back up - bet the old one works alright. Dec 31 18:18:32 I can expect an image recipe to install the packages in that list, only if it can find them Dec 31 18:18:42 not give an error if it can't Dec 31 19:56:28 Crofton|work, I didn't think RRECOMMENDS got installed automatically Dec 31 20:22:44 I'll find out Dec 31 20:23:04 I need a soft DEPENDS for the case where packageconfig changes what packages are output Dec 31 20:32:50 hmm, cant you specify DEPENDS as well as RDEPENDS for a certain packageconfig ? Dec 31 20:34:36 you can, but maybe i'm not grepping that sentence of yours Dec 31 20:53:19 I meant RDEPENDS Dec 31 20:53:34 DEPENDS is covered in the PACKAGECONFIG Dec 31 20:54:54 Here is an odd one: http://pastebin.com/N14XA1GW Dec 31 21:03:55 file isn't there Dec 31 21:08:41 rdepends and depends are both covered by packageconfig, as kroon indicated, it's the 4th entry in the comma separated list Dec 31 21:09:04 ah ok, I need to read up then, might help Dec 31 21:32:09 the recipe spits out a bunch of packages Dec 31 22:06:15 it will add the RDEPEND to the ${PN}? Dec 31 22:09:52 in my experience, bitbake/rpmdeps automatically detects rdepends for elf binaries/shared libraries Dec 31 22:10:59 that is fine Dec 31 22:11:39 I am just dealing with a messy package Dec 31 22:11:52 buildhistory is great for detecting these kinds of changes Dec 31 22:12:28 yeah, I am just trying to make an easier way to install all the output Dec 31 22:12:33 gotta run Dec 31 23:40:45 Here is an odd one: http://pastebin.com/N14XA1GW Dec 31 23:40:58 hmm, a Christmas miracle? Dec 31 23:40:59 http://pastebin.com/N14XA1GW Dec 31 23:41:14 note commit date on fb6b87bf67a2dbe6b50143eb8429c736f61fea2e Dec 31 23:47:44 Crofton|work, odd Jan 01 00:52:35 Crofton|work, fwiw, i get the same kinds of errors building ntp **** ENDING LOGGING AT Thu Jan 01 02:59:59 2015