**** BEGIN LOGGING AT Fri Sep 30 02:59:57 2011 Sep 30 03:55:14 NOTE: Package libgles-omap3 skipping QA tests: ['True'] Sep 30 03:55:14 ERROR: QA Issue: non -dev/-dbg/-nativesdk package contains symlink .so Sep 30 03:55:22 wouldn't skipping QA avoid that error? Sep 30 05:49:06 hello Sep 30 06:31:19 good morning Sep 30 07:50:45 JaMa: :) Sep 30 07:51:10 it's without xserver-common change.. but that could be applied later.. Sep 30 07:59:55 JaMa: I've found conflicting examples for x11-common/etc/X11/Xserver Sep 30 08:03:59 examples? Sep 30 08:04:10 http://dos.openmoko.pl/xserver-common Sep 30 08:04:33 (talking about handheld) Sep 30 08:09:20 do_package_qa() {} <- nice solution Sep 30 08:46:17 is git down? Sep 30 08:46:24 ssh: connect to host git.openembedded.org port 22: Connection timed out Sep 30 08:48:21 ka6sox: ^ Sep 30 08:48:46 JaMa: could be, just tried git pull ant it stuck Sep 30 08:50:58 I am having problems with Bitbake 1.12.0 and providers, here is full details: http://pastebin.com/KxdEy6Kb Sep 30 08:51:44 I have included mkfs-ubifs to my image and there are two providers, so I get this message: NOTE: consider defining a PREFERRED_PROVIDER entry to match mkfs-ubifs Sep 30 08:51:59 So, I added PREFERRED_PROVIDER_mkfs-ubifs = "mtd-utils" Sep 30 08:52:08 davidlt: this doesn't work for _runtime_ providers Sep 30 08:52:11 But Bitbake never checks it Sep 30 08:53:01 JaMa: So what would be the right way to get rid of that Bitbake message? Sep 30 08:53:05 yes it's more strict with this warning now Sep 30 08:53:19 davidlt: set right PACKAGES in mtd-utils-tests Sep 30 08:53:32 so it doesn't include mkfs-ubifs Sep 30 08:54:00 It doesn't sounds like a nice hack Sep 30 08:54:11 e.g. vim-syntax has 3 providers Sep 30 08:54:11 or change RPROVIDES if it's caused by RPROVIDES Sep 30 08:54:35 mtd-utils-tests producing mkfs-ubifs sounds like a bug :) Sep 30 08:55:00 so it's not nice hack, but proper fix :) Sep 30 08:56:26 mtd-utils-tests requires mtd-utils, that's why it provides mkfs-ubifs Sep 30 08:56:40 require mtd-utils_1.2.0+git.bb Sep 30 08:57:55 then refactor common parts to mtd-utils.inc and keep mkfs-ubifs only where it should be Sep 30 10:58:48 hi Sep 30 10:58:56 Can anybody tell me why a count is being prepended to the revision returned by sortable_revision method (gets called from get_srcrev) in fetch2/__init__.py in bitbake? Sep 30 10:59:59 Actually it I am facing a problem that multiple u-boot recipes when get parsed this count gets incremented for each of them and if cache gets invalidated and we try to clean u-boot or rebuild after clean then the u-boot dir in work dir gets different names each time due to this count being incremented for each u-boot recipe being parsed Sep 30 11:05:32 imran: to get upgradable path for target packages Sep 30 11:05:56 imran: as different git hash is not always "higher" version Sep 30 11:38:38 Hi, all! Sep 30 11:39:17 cpufrequtils is not fetching. Are there some kernel.org git repositories mirrors? Sep 30 11:44:51 morning Sep 30 11:56:07 I have a distro/image based on the Angrstrom distro. I can see that the syslogd process is running, but there's no logfile under /var/log. What should I do to get this working? Sep 30 12:00:30 JaMa: thanks for replying, the problem is if you e.g. 1) build u-boot 2) invalidate cache by just changing some conf file and undoing the change 3) build u-boot again and it will actually run all the tasks again and create a new u-boot directory in WORKDIR with just this count different in its name than the previous u-boot dir Sep 30 12:01:49 JaMa: it happens when we have multiple u-boot recipes each with different SRCREV value Sep 30 12:19:40 imran: I know about this issue.. you need to get rid of one of them Sep 30 12:52:31 JaMa: you mean the only solution is to avoid having multiple recipes of u-boot? Sep 30 12:54:12 or use different branch in git (different key in persistence cache) Sep 30 13:27:48 JaMa: thanks Sep 30 15:38:04 ok, I made a simple oe-core image for the usrp-e1xx Sep 30 18:31:02 JaMa: ping Sep 30 18:39:51 pong Sep 30 18:42:18 kergoth: Sep 30 18:56:56 kergoth: sorry gtg, but feel free to pm me or send e-mail Sep 30 18:59:04 JaMa: no worries, sorry for the delay, will do Sep 30 19:32:10 I hate scm conversions. Sep 30 19:47:51 kergoth: heh Sep 30 19:48:00 kergoth: what are you trying ? Sep 30 19:48:20 * khem converted a SVN -> git lately Sep 30 19:48:39 guess what SVN co was 7G in size Sep 30 19:48:51 and git clone is 5.4G Sep 30 19:49:00 svn co time was 31min Sep 30 19:49:09 git clone is 6 mins Sep 30 19:49:09 yeah one of those. svn layout changes in the history are causing headaches, plus trying to split up repos by subdirectory, which is a pain when files are copied or moved between them, or subdirs are renamed, or... Sep 30 19:49:14 khem: hah Sep 30 19:49:17 nice Sep 30 19:49:42 I threw away the svn metadata Sep 30 19:49:52 just the commit history Sep 30 19:50:07 svn is lesser crap than cvs Sep 30 19:50:30 I dug up the commiter Ids and emails from ldap Sep 30 19:50:38 it looks pretty sexy now Sep 30 19:50:41 :) Sep 30 19:50:56 conversion ran for 1 week :) Sep 30 19:51:06 nice Sep 30 19:51:26 I can give you the python script I wrote to dump authorsfile Sep 30 19:51:40 it may not map to what you have Sep 30 19:51:43 but its something Sep 30 19:51:53 what's always a pain is that I always feel like I could do more to improve it, but have to stop at some point due to time constraints, and after its out, you can't really screw with it anymore Sep 30 19:51:58 e.g. adding grafts for manual merges Sep 30 19:52:31 yeah Sep 30 19:52:40 there are restrictions Sep 30 19:52:41 dealt with that doing bitbake & oe. course those were even worse due to the old metadata from previous scm conversions in the commit messages Sep 30 19:52:49 <3 filter-branch Sep 30 19:53:17 one of these days I might put hands on a repo which is 39G Sep 30 19:53:23 with all branch history Sep 30 19:53:32 I dont think git is for such huge repos Sep 30 19:55:36 hmmm so I guess now HP should be next in line for oracle's hostile takeovers Sep 30 19:55:45 since its in pretty bad shape Sep 30 20:05:51 does oe have gerrit setup? Sep 30 20:08:24 nope Sep 30 20:08:33 as far as i know, anyway Sep 30 20:08:36 msm: no but I am contemplating Sep 30 20:08:42 k, just reading comments on the list Sep 30 20:08:45 msm: yocto has plans too Sep 30 20:43:15 obi, **** ENDING LOGGING AT Sat Oct 01 02:59:56 2011