**** BEGIN LOGGING AT Mon Dec 03 02:59:59 2012 Dec 03 08:30:35 morning all Dec 03 09:05:49 good morning Dec 03 09:44:03 morning! Dec 03 09:48:15 afournier: gm Dec 03 10:26:35 gm Dec 03 10:35:12 morning all Dec 03 10:38:12 good morning Dec 03 11:08:12 hi all Dec 03 11:09:00 hi florian, pb_ Dec 03 11:20:31 levonmaa: really? What did you got? Dec 03 11:23:17 where is dbus recipe ? Dec 03 11:24:11 oe-core Dec 03 12:07:33 hi bluelightning , hi florian Dec 03 13:31:54 http://lists.linuxtogo.org/pipermail/openembedded-core/2012-November/032251.html anyone? Dec 03 13:41:37 seems like a DISTRO problem to me Dec 03 13:42:55 pb_: for now I have netbase.bbappend which makes package MACHINE specific Dec 03 13:43:08 but would like to know how others handle it Dec 03 13:44:24 I suspect quite a lot of people handle it by not running apache :-} Dec 03 13:44:44 nor mysql Dec 03 13:44:57 right Dec 03 13:45:23 the netbase thing is fine if you do happen to be setting the hostname to $MACHINE, but obviously not everyone does that. Dec 03 13:45:47 writing to /etc/hosts at boot time doesn't work if you have a read-only root filesystem Dec 03 13:45:57 yep Dec 03 13:46:20 use the volatiles stuff Dec 03 13:46:21 :) Dec 03 13:46:34 you could do that if you wanted, but it would make bootup slower Dec 03 13:46:59 using armv8 fastmodel reminds me old armv4 boards Dec 03 15:09:30 i tried to define upstart instead of sysvinit as VIRTUAL-RUNTIME_init_manager in the distro and i get "satisfy_dependencies_for: Cannot satisfy the following dependencies for task-core-boot: sysvinit" when building the image Dec 03 15:09:32 any idea ? Dec 03 15:20:46 otavio: do you have xf86-input-void archive with matching SRC_URI checksums? current archive is different and I cannot compare with old one Dec 03 15:21:00 afournier: I'm going to send patch for that today Dec 03 15:21:18 JaMa|Off: really ? Dec 03 15:21:45 JaMa|Off: is it a big patch ? Dec 03 15:22:27 JaMa|Off: I think I do; let me check Dec 03 15:25:10 JaMa|Off: 93821f21e807260b05431c62437a8b32 ? Dec 03 15:26:58 ericben: are you having any issues ocmpiling Hiawatha recently Dec 03 15:27:16 ericben: for the past couple of days I have been unable to compile Dec 03 15:27:19 ericben: http://ix.io/3yk Dec 03 15:27:30 otavio: that's new one, not matching recipe Dec 03 15:27:37 otavio: will send patch changing checksums today Dec 03 15:27:42 afournier: no, small Dec 03 15:28:08 afournier: https://github.com/shr-distribution/oe-core/commit/6709a248d9a8054f09e77e0466006029ac752638 Dec 03 15:28:57 hey guys - what do I have to do to redirect kernel-outputs to a usb-port during bootup? Dec 03 15:35:26 dFence: console=/dev/ttyUSBx Dec 03 15:36:01 mckoan: does a kernel-module go along with that? Dec 03 15:36:58 dFence: I don't understand your question, sorry Dec 03 15:37:31 mckoan: for the kernel to have a ttyUSB, doesn't it need a corresponding kernel-module, that turns a usb-port into a tty? Dec 03 15:38:19 I think a USB tty generally means an FTDI chip, which has to have the support compiled into the kernel (or a module) yes Dec 03 15:38:48 jackmitchell: so there's no chance of emulating it? Dec 03 15:38:59 dFence: no Dec 03 15:39:06 meeh -_- Dec 03 15:51:30 I guess FTDI is de-facto in usb <-> serial converters Dec 03 15:54:37 well, indeed, there is empirical evidence for that. Dec 03 17:32:47 moin Dec 03 17:35:36 pb_: are you using a read only rootfs with micro at all nowadays? I remember you messing with that sort of thing at some point Dec 03 17:38:12 yeah, all our products have that Dec 03 17:39:29 is the setup for that in a public layer? I'm curious about how people are implementing it with oe/yocto. everybody seems to be rolling their own tweaks at this point Dec 03 17:40:34 I think we pushed all, or nearly all, our changes into oe-core and the public meta-micro. I'll have a quick look and see if anything major is still missing. Dec 03 17:40:54 iirc there were a couple that RP didn't like and that we didn't have time to rework. Dec 03 17:41:54 ah, nice. okay, will do. thanks. Dec 03 17:48:32 doesn't seem to be much. a minor fix for dbus (missing /var/run/dbus when /var/run is volatile) and the gdk-pixbuf loaders thing Dec 03 17:50:28 afaict, all our other local changes are unrelated to read-only-rootfs Dec 03 17:50:37 cool Dec 03 17:52:24 hm Dec 03 17:52:39 I might need readonly too Dec 03 17:53:31 FWIW fixing read-only-root is a priority for the next release Dec 03 17:57:14 how's meta-micro different from yocto-tiny? Dec 03 18:00:35 good question. dunno, I never quite understood why the yocto folks decided to go off and do their own thing there. Dec 03 18:00:43 not invented at intel, I guess. Dec 03 18:03:11 hahah :) Dec 03 18:03:21 they forgot to read the layer index :) Dec 03 18:06:18 it definitely wasn't NIH Dec 03 18:06:40 I don't think Darren was aware of meta-micro when he started work on his project Dec 03 18:06:50 not even sure there was a layer index at that time Dec 03 18:07:01 ah, right, fair enough Dec 03 18:11:51 pb_: what's the size of your image for your products (I don't know which products you do development for, btw) Dec 03 18:12:00 kergoth, you probably don't want to see our rootfs post_process hacks / birthday script for ro root... Dec 03 18:12:04 hehe Dec 03 18:12:12 poky-tiny rootfs is 3.2 mb, excluding the bzImage I guess Dec 03 18:12:13 as soon as I saw the dupe, I did make sure both parties knew what was up Dec 03 18:12:14 eren: pretty big for the current generation, but this one includes webkit and a load of other good stuff. Dec 03 18:12:18 if there was a standard/clean way to do it, that would be awesome... Dec 03 18:12:35 which reminds me, I need to beat the meta-zynq drum again Dec 03 18:13:12 pb_: oh, ok Dec 03 18:13:27 I don't actually seem to have a non-webkit rootfs lying around right now so I can't easily give you a number for what it looks like without that. Dec 03 18:14:22 but, it depends greatly on what you want to put in there. iirc, when we were doing images for arm926 thumb with uclibc, you could get a fairly respectable amount of functionality in 1MB of flash. Dec 03 18:14:32 excluding the kernel, obviously Dec 03 18:15:19 I'm planning to delve into some networking work, for which the boards have usually 16mb rom. Routerstation Pro seems good but I read that it's ended. Some folks at LWN commented that they decided to use Netgear WNDR3700 over Routerstation Pro Dec 03 18:15:22 you make our 256MB flash sound expansive... Dec 03 18:16:33 I will basically need 3-to-4 wlan interface that routes packets for our hamradio network project Dec 03 18:17:59 i just saw last night a weird rasp-pi issue with wifi dongles and USB power Dec 03 18:19:11 nah, rasp-pi is not easy to get. I cannot wait Farnell forever Dec 03 18:20:39 I can only think rasp-pi as a cheap device that runs XBMC to make your LCD/LED TV smart :) Dec 03 18:20:46 pb_, bluelightning, I learned of meta-micro near the end of the prototype for poky-tiny Dec 03 18:21:15 pb_, bluelightning, this was also all new ground for me, so it was a lot of exploring the different ways to accomplish the end goal Dec 03 18:22:11 pb_, bluelightning, eventually it made sense to have something tightly integrated with yocto so we incorporated the work I had done, but was in no way meant to slight or lessen the work of meta-micro Dec 03 18:36:08 otavio: ping Dec 03 18:46:01 levonmaa: pong Dec 03 18:51:24 otavio: there now is an update in the arm-testng branch Dec 03 18:51:48 to me it looks like the work/image folder is ok Dec 03 18:52:06 there are some issues with getting them packaged and staged correctly Dec 03 18:52:42 Hi. Is compiling core-image-minimal with uClibc instead of eglibc known to be broken in openembedded-core master branch in git ? Dec 03 18:54:09 levonmaa: awesome Dec 03 18:54:26 levonmaa: I'd say for you to send it for merging in our master Dec 03 18:54:37 levonmaa: as this ought to be the base of further working Dec 03 18:54:51 levonmaa: I know it is not complete but it is much advanced than current code Dec 03 18:55:03 levonmaa: so it could be merged as is IMO Dec 03 18:55:19 JaMa|Off: what do you think? Dec 03 18:58:46 otavio, levonmaa: looks good, I've couple of comments Dec 03 18:58:58 JaMa|Off: shoot Dec 03 18:59:10 https://github.com/levonmaa/meta-qt5/commit/e43791776b3441ab92dd2f3af425cd544fc4a9ef isn't QT_CONF_PATH generated from some bbclass inherit already? Dec 03 18:59:30 JaMa|Off: the variable is not set in some cases Dec 03 18:59:42 also I would prefer to split it in 2 commits (at least uic build and the rest) Dec 03 19:00:12 "qtbase: builds under arm" is not best description of those changes imho Dec 03 19:00:41 JaMa|Off: true Dec 03 19:01:37 otherwise it's good, will build test it hopefully tonight Dec 03 19:03:32 JaMa|Off: so you want the uic commit in one and the rest of the three commits merged into one? Dec 03 19:04:49 no, just split e43791776b3441ab92dd2f3af425cd544fc4a9ef in two and keep the rest as is Dec 03 19:07:06 JaMa|Off: righty-o, sounds good Dec 03 19:46:29 JaMa: made the pull request Dec 03 19:58:45 seen it, will merge it after build test if otavio isn't faster Dec 03 19:59:22 JaMa: roger that Dec 03 20:33:17 JaMa: I am quite busy looking at U-Boot code and another work duties Dec 03 20:33:23 JaMa: I am very sorry ... Dec 03 20:35:00 otavio: np, I should have a bit more time for this now I guess Dec 03 21:00:52 ping ericben Dec 03 21:37:28 anyone of you heard of framebuffer-console? I can't get any output on the display Dec 03 23:30:29 can I rely on udev to create all nodes or should I create them manually? Dec 03 23:33:45 dFence: off topic. try #edev or #elinux or others Dec 03 23:34:59 ok Dec 04 01:53:07 ka6sox, cgit.openembedded.org was overloaded and nearly unresponsive. I've taken some quick measures and sent you an e-mail describing them. Dec 04 01:53:37 Everyone else cgit.openembedded.org's .tar.gz and .tar.bz2 files are temporarily disabled. Dec 04 01:56:35 halstead, thanks...I'm not sure why its to unresponsive...I may move it to a different host (opal to a different dom0) Dec 04 01:59:39 ka6sox, Do you know if the nginx config has been changed recently? I don't recognize those rewrites. Dec 04 02:00:21 heh just when I was using cgit.openembedded.org to generate big pull request Dec 04 02:01:32 JaMa, Do you need the tarballs for that? Dec 04 02:01:45 no.. but it was very slow Dec 04 02:02:01 halstead, no, it should not have changed at all. Dec 04 02:02:09 JaMa, Is it better now for you? Dec 04 02:02:29 seems so, but I've sent tht pull request few minutes ago Dec 04 02:02:45 well started sending and it's still running :) Dec 04 02:05:09 halstead, with him moving things like that I can't swap hosts now...I'll schedule it tomorrow evening after he goes to bed. Dec 04 02:06:15 ka6sox, Okay. Looking at the logs this seems to be caused by a spider trying to get all of the on-demand tarballs. Dec 04 02:06:47 ka6sox, Once all the crawlers get the new robots.txt file we can probably re-enable full functionality. Dec 04 02:11:00 would you like us to do this our will you handle this? **** ENDING LOGGING AT Tue Dec 04 02:59:59 2012