**** BEGIN LOGGING AT Thu Mar 18 02:59:56 2010 Mar 18 03:03:35 heh heh Mar 18 03:03:51 the internet is almost magical Mar 18 03:03:57 I wish I had something to find me downloads for OE Mar 18 03:03:58 oh wait Mar 18 03:04:00 someone did that Mar 18 03:07:54 hmm Mar 18 03:08:03 its a do_fetch based on a .git Mar 18 03:08:09 blaaarrrh Mar 18 03:10:00 possible someone has a clone of it, but how to find it :) Mar 18 03:11:35 exactly lol Mar 18 03:27:54 well this is annoying Mar 18 03:27:59 I found another source but its not like a git file Mar 18 04:09:38 Baaaaaah Mar 18 04:09:44 Did we just bump opkg for one of 3, again? Mar 18 04:20:20 seems to happen for every opkg rev Mar 18 04:20:57 it would be nice if there was a recipe for released versions instead Mar 18 04:25:15 Well, that too Mar 18 04:25:27 * Tartarus just re-begged to have one var for the rev #, and use that for the rest Mar 18 04:25:51 You're making sense. Stop it! :P Mar 18 06:35:41 what should I use if want try angstrom image, bitbake angstrom-x-image or angstrom-gnome-image? Mar 18 06:37:14 Openfree`: it is really up to what you want to use it for Mar 18 06:37:59 zecke, hi, I'm not very clear with the concept "distro" and "image" Mar 18 06:38:32 if I choose angstrom-2008.1, I still free to choose images? because so many images here Mar 18 06:38:59 Openfree`: distro is a set of policies. This includes package naming, package configuration and such Mar 18 06:39:25 Openfree`: image is just a set of installed packages that is wrapped in a filesystem/tar.gz Mar 18 06:40:56 so distro is nothing to do with what image I have to use, for example, if I want gtk, qt, or e17 etc.. Mar 18 06:41:26 Openfree`: right, you can build any image you want Mar 18 06:41:54 zecke: how are we doing with security stuff sorry couldnt devote much time Mar 18 06:42:24 khem: i still owe us an email Mar 18 06:42:28 zecke: and in April I will be busy too :( Mar 18 06:42:41 till apr 19 Mar 18 06:42:41 khem: we are down to ~13 recipes with issues Mar 18 06:42:52 ok cool Mar 18 06:42:55 where is the list Mar 18 06:43:29 Hi zecke, I'm available for the next hour to make any final revisions to the grip recipe patch if needed tonight. Mar 18 06:44:03 khem: I will have to post it but I checked in my oe_audit.py script so in theory you could generate it yourself. Mar 18 06:44:13 ah nice Mar 18 06:44:16 zenlinuxPDX: it looked fine. I think I will take it as is and nuke the "stable" recipe. Mar 18 06:44:25 ok, your call Mar 18 06:45:29 zecke: did you commit this script to contrib ? Mar 18 06:45:37 oh under qa Mar 18 06:45:48 khem: and we have to find a way to follow other distros too. E.g. FreeBSD has no glibc, udev, dpkg. Mar 18 06:46:17 khem: yeah, it is a crappy python script right now. I will rewrite it in GNU smalltalk. :) Mar 18 06:46:54 khem: You will need the auditfile.tbz and unpack it, you will need to run bitbake -s > available, and you will need to strip the first lines from the available file Mar 18 06:47:05 khem: and then you can run ./oe_audit.py from there, I will clean this up. Mar 18 06:47:52 zecke: python is better atleast I will be able to read/fix it too :) Mar 18 06:48:38 zecke: I will run it and get the list now Mar 18 06:49:13 khem: yeah, but I want to learn smalltalk. :) Mar 18 06:50:39 eh why smalltalk Mar 18 06:50:56 one reason to stick with python is to use bitbakes vercmp... or maybe I should rewrite bitbake in smalltalk too Mar 18 06:51:25 khem: it is an awesome system. when even your stacktrace is an object... you can do so many nice things Mar 18 06:53:50 hmm learn Go language Mar 18 06:54:15 http://golang.org/ Mar 18 06:55:04 khem: no, it is lacking the "image" concept of smalltalk. Mar 18 06:56:19 khem: self class decompile: #fooMethod -> prints sourcecode, self class edit: #fooMethod, will let you edit the sourcecode, then recompile the selector, and next time the method is executed it will have the new code Mar 18 06:56:38 hmm Mar 18 06:56:48 khem: or Object class addInstVariable: #zecke-rocks, will add zecke-rocks as instance variable to every existing instance of Object Mar 18 06:57:17 zecke: where is auditfile.tbz Mar 18 06:57:53 * khem -> sleep () Mar 18 06:58:01 khem: http://ports.freebsd.org/auditfile.tbz Mar 18 06:59:00 ah it needs to know where bb is Mar 18 06:59:11 khem: ah right, yeah :} Mar 18 07:01:11 cool got the list Mar 18 07:01:19 will parse it tomorrow morning Mar 18 07:01:21 gn Mar 18 07:01:38 sleep well Mar 18 08:08:21 rkirti, ping Mar 18 08:14:04 I'm trying maemo-angstrom OE with recent org.openembedded.dev branch, but hint multi provider gtk+ problem Mar 18 08:22:12 I got some trouble getting the NICs working on some of my machines. I got 5 "identical" (thats what the manufacture says) Atom PCs. lspci says they all got RTL8111/8168B PCI Express Gigabit Ethernet controller (rev 01). To test my machines I moved my CF card from machine to machine and found that the NICs are only working in one of them. The module that is loaded is the r8169. I have checked both lspci and lshw and havn't found any interesting Mar 18 08:22:12 difference between the machines. So, can anyone suggest next step to get further? Mar 18 08:26:53 morning Mar 18 08:30:03 Anyone using OE on a 64Bit Host for compilation? Which Distro? Mar 18 08:30:37 marekp: Fedora 12, x86_64 Mar 18 08:31:24 Thank you, i tried Cent-OS 5, but gut into trouble Mar 18 08:31:41 what kind of trouble? Mar 18 08:31:58 first it was ccache Mar 18 08:32:22 Morning Mar 18 08:32:29 and when i disabled ccache, the compilation get stuck after 32 packeged in sha package Mar 18 08:32:31 btw I saw various unexplainable problems on different hosts/distros, so far absolutely not sure what those are related to, maybe OE should use more -native stuff to get the absolutely same result everywhere.. Mar 18 08:34:14 jin^eLD: do you have afs available from the repro in fedora 12? Mar 18 08:35:25 let me yum search for it, what is it? Mar 18 08:35:53 afs -> andrew file system Mar 18 08:36:03 is also called openafs Mar 18 08:36:18 hi all. mickey|tv, vala-dbus-binding-tool_0.1.3.bb fails in do_fetch, http://www.freesmartphone.org/sources/vala-dbus-binding-tool-0.1.3.tar.bz2 gives 404 and so do the others URIs Mar 18 08:37:04 marekp: I never felt lke experimenting with filesystems :) sticking to ext3 Mar 18 08:37:11 s/lke/like Mar 18 08:37:25 bt yes, F12 has openafs kernel modules and utilities in the repository Mar 18 08:37:31 s/bt/but/ Mar 18 08:38:09 Jin^eLD: this filesystem is kind of a global filesystem, that is used to securely access files over network Mar 18 08:38:34 anyway, doessn't matter ;) Mar 18 08:38:41 mogning Mar 18 08:39:20 hi hrw Mar 18 08:41:21 hi Mar 18 08:44:52 Openfree`: pong Mar 18 08:46:40 hrw, mogning :-) Mar 18 08:46:44 Openfree`: it wont work with the recent OE tree. Check my webpage about my project. I need to use maemo-gtk (the version that existed when I was working on this project) and that conflicts all Angstrom PREFERRED_PROVIDERs Mar 18 08:47:27 Openfree`: http://www.cse.iitm.ac.in/~rkirti/maemo-oe/ Mar 18 08:50:47 rkirti, I already figured that, So now I'm trying rewind OE version back, But It's really difficult for me to figure which commit is sanity Mar 18 08:51:42 Openfree`: I see. give me a minute. Mar 18 08:51:45 the problem that is maemo-oe is only a overlay, and it need OE Mar 18 08:51:50 rkirti, thanks Mar 18 08:52:40 also I found that few sources located at repository.maemo.org changed Mar 18 08:52:46 but should be big problem. Mar 18 08:53:41 is maemo-gtk from official OE back that time when u do the project? Mar 18 08:53:54 Openfree`: yeah, repo changing in Maemo is really arbitrary and was a big PITA for me, they made goofups like renaming beta sources to alpha or something like that twice Mar 18 08:54:35 rkirti, that would be BIG nightmare.. Mar 18 08:54:46 Openfree`: no, I couldnt get it merged in mainline OE, because I built it on top of Angstrom, and the way it worked when I was the end of GsoC period, it nearly broke Angstrom - because al provider preferences would conflict. Mar 18 08:55:57 rkirti, could you point me which commit version OE is it working with? Mar 18 08:56:01 Openfree`: check my readme page on the github site if you havent - that was around august 12. I am guessing commits somewhere around July 23 to August 6-ish would work fine with you. Try July 3rd week to begin with. Mar 18 08:56:13 * rkirti slaps herself hard for not recording the OE tree version. Mar 18 08:56:35 I switched to some other OE work and updated my tree periodically for that -so I have a more recent version :-( Mar 18 08:56:37 rkirti, ok, thanks Mar 18 08:57:34 Openfree`: as far as the changes to the conf files are concerned - I think I have them systematically somewhere, probably github. If you dont find them, give me a ping. I ll pull them off my local copy Mar 18 08:58:27 Openfree`: I have had some communication with a certain Dennis (Xyun?) regarding maemo-oe. Are you in any way related to the person ? Mar 18 08:58:53 rkirti, that's me Mar 18 08:58:58 ^_^ Mar 18 08:59:11 Openfree`: *_* :-) sorry. Mar 18 09:00:16 ok thanks for your help, It should takes me a few time to try that.. Mar 18 09:01:08 Openfree`: sure. keep me updated. feel free to ping if you need any help. I am a little sporadic on IRC these days,so mail would be faster. Mar 18 09:01:45 rkirti, ^^ that's fine, thanks, again~ Mar 18 09:09:48 rkirti|afk: merge it into maemo5-compat and/or chinook-compat maybe? Mar 18 09:09:58 * hrw -> off now Mar 18 09:10:23 hrw|gone: thats a good suggestion. Will work on integrating it into maemo5-compat. Mar 18 09:11:01 hrw|gone: pointers on who among the OE devels is working on it ? Mar 18 09:11:18 I did some initial stuff Mar 18 09:11:40 woglinde/thebohemian may use it for their java-for-maemo stuff Mar 18 09:11:49 florian may be interested Mar 18 09:12:02 hrw|gone: how about woglinde ? I knew he was working on chinook-compat during my GSoC Mar 18 09:12:10 oh ok Mar 18 09:12:19 didnt notice that line Mar 18 09:12:21 hrw|gone: thanks Mar 18 10:32:25 03Holger Hans Peter Freyther  07org.openembedded.dev * r39a6937fd6 10openembedded.git/recipes/curl/ (3 files in 2 dirs): Mar 18 10:32:25 curl-sdk: Upgrade to 7.19.7 Mar 18 10:32:25 This addresses CVE-2009-0037 Mar 18 10:32:37 03Holger Hans Peter Freyther  07org.openembedded.dev * r0331fa5544 10openembedded.git/ (6 files in 3 dirs): Mar 18 10:32:38 dpkg: Upgrade to 1.14.29 to address a flaw in dpkg-source Mar 18 10:32:38 Addresses CVE-2010-0396. Mar 18 11:08:08 did anyone try runit with OE? or can suggest anything else that would do service supervision? Mar 18 11:08:59 good morning Mar 18 11:09:27 hi florian Mar 18 11:11:14 mtr: wie hast du runit cross compiled? Mar 18 11:11:17 oops wrong window Mar 18 11:11:18 :) Mar 18 11:29:20 morning Mar 18 11:32:49 03Holger Hans Peter Freyther  07org.openembedded.dev * r75d85f77f3 10openembedded.git/recipes/ez-ipupdate/ (ez-ipupdate_3.0.10.bb files/CVE-2004-0980.patch): Mar 18 11:32:49 ez-ipzpdate.bb: Address flaw in syslog handling Mar 18 11:32:49 Address CVE-2004-0980. Mar 18 12:55:12 morning all Mar 18 12:56:01 khem: Interested in being a candidate for the TSC? Mar 18 12:58:31 03Holger Hans Peter Freyther  07org.openembedded.dev * rbd415877ea 10openembedded.git/contrib/qa/oe_audit.py: Mar 18 12:58:32 oe_audit.py: Parse the full output of bitbake -s Mar 18 12:58:32 Instead of being forced to manually edit we can now parse Mar 18 12:58:32 the full output of bitbake -s. Mar 18 12:58:32 03Holger Hans Peter Freyther  07org.openembedded.dev * r1d85542eb6 10openembedded.git/contrib/qa/oe_audit.py: oe_audit.py: Document a bug which was uncovered on chat Mar 18 13:07:24 03Koen Kooi  07org.openembedded.dev * r9355f83bda 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: angstrom: prefer dbus 1.2.22 Mar 18 13:07:28 03Koen Kooi  07org.openembedded.dev * rf6f5cd1ab8 10openembedded.git/recipes/dbus/ (5 files in 2 dirs): dbus: add 1.2.22 Mar 18 13:08:24 Hi All\ Mar 18 13:08:45 How to enable Mouse in FB (terminal ) Mar 18 13:08:50 Am using Angstrom Mar 18 13:42:40 Has anyone got any time to look over some patches in patchwork for me? Mar 18 13:43:02 They're little bugfixes... Mar 18 13:44:37 http://patchwork.openembedded.org/project/openembedded/list/?submitter=253 Mar 18 13:45:14 [oe,v2] commit 099... just needs to be removed, the patch with the correct subject has been applied. Mar 18 14:06:03 03Martin Jansa  07org.openembedded.dev * rd8fd10392b 10openembedded.git/recipes/tasks/task-shr-feed.bb: Mar 18 14:06:03 task-shr-feed: add neolight Mar 18 14:06:03 Signed-off-by: Martin Jansa Mar 18 14:06:03 03Martin Jansa  07org.openembedded.dev * r637b3e4868 10openembedded.git/recipes/xorg-util/ (5 files): Mar 18 14:06:03 makedepend: convert to BBCLASSEXTEND, .inc, move checksums to recipe Mar 18 14:06:04 Signed-off-by: Martin Jansa Mar 18 14:06:05 03Martin Jansa  07org.openembedded.dev * r23799e1ae8 10openembedded.git/conf/distro/include/sane-srcrevs-fso.inc: Mar 18 14:06:05 frameworkd: bump SRCREV for opimd fixes Mar 18 14:06:06 Signed-off-by: Martin Jansa Mar 18 14:06:25 03Martin Jansa  07org.openembedded.dev * r2f48ef9857 10openembedded.git/recipes/ (4 files in 2 dirs): Mar 18 14:06:26 xinit, xinput, libxt: add newer versions Mar 18 14:06:26 Signed-off-by: Martin Jansa Mar 18 14:06:27 03Martin Jansa  07org.openembedded.dev * rc853d5b877 10openembedded.git/recipes/xorg-driver/ (257 files): Mar 18 14:06:27 xorg-drivers: add checksums to recipes Mar 18 14:06:28 Signed-off-by: Martin Jansa Mar 18 14:06:28 03Martin Jansa  07org.openembedded.dev * r4b7cc047cd 10openembedded.git/recipes/xorg-driver/ (3 files): Mar 18 14:06:29 xf86-input-penmount, xf86-video-nv, xf86-video-vmware: newer versions Mar 18 14:06:29 Signed-off-by: Martin Jansa Mar 18 14:06:30 03Martin Jansa  07org.openembedded.dev * rf4ab324b35 10openembedded.git/conf/checksums.ini: Mar 18 14:06:30 checksums.ini: remove entries for xorg-drivers (moved to recipes) Mar 18 14:06:31 Signed-off-by: Martin Jansa Mar 18 14:06:31 03Martin Jansa  07org.openembedded.dev * rea1b207ddf 10openembedded.git/recipes/xorg-util/ (11 files): Mar 18 14:07:20 xorg-util: add checksums to recipe Mar 18 14:07:21 Signed-off-by: Martin Jansa Mar 18 14:07:21 03Martin Jansa  07org.openembedded.dev * r4b21753640 10openembedded.git/recipes/vala/ (vala.inc vala_0.7.10.5.bb): Mar 18 14:07:21 vala: add checksum for 0.7.10.5 Mar 18 14:07:21 Signed-off-by: Martin Jansa Mar 18 14:07:21 03Martin Jansa  07org.openembedded.dev * r4848b50202 10openembedded.git/ (3 files in 3 dirs): Mar 18 14:07:22 om-gta01: fix jffs2 building params, xorg.conf (Thanks to Michael) Mar 18 14:07:22 Signed-off-by: Martin Jansa Mar 18 14:07:23 03Martin Jansa  07org.openembedded.dev * r0ca3942e51 10openembedded.git/recipes/xserver-kdrive-common/ (xserver-kdrive-common/Xserver xserver-kdrive-common_0.1.bb): Mar 18 14:07:23 (15 lines omitted) Mar 18 14:34:39 MWelchUK_work: I will take a look at them sometimes today or tomorrow Mar 18 14:35:17 khem, much appreciated. Mar 18 14:37:14 MWelchUK_work: http://patchwork.openembedded.org/patch/1725/ Mar 18 14:37:21 does the checksum remain same Mar 18 14:38:15 MWelchUK_work: why not migrate to new checksums in same commit? Mar 18 14:38:34 even better Mar 18 14:38:57 Yes - it's just a change in location. Mar 18 14:39:08 I moved *a lot* checksums in semi-automated way today.. so I hope we should slowly move all checksums to recipes soon Mar 18 14:39:26 03Michael 'Mickey' Lauer  07org.openembedded.dev * rdbf2045454 10openembedded.git/recipes/images/fso2-demo-image.bb: fso2-demo-image: new image recipe; FSO-compliant image w/ zhone2 as smartphone application Mar 18 14:39:58 TBH, I didn't think that much about it, just updated what was there. Mar 18 14:43:56 mickey, vala-dbus-binding-tool_0.1.3.bb fails in do_fetch, http://www.freesmartphone.org/sources/vala-dbus-binding-tool-0.1.3.tar.bz2 gives 404 and so do the others URIs Mar 18 14:44:10 mickeyl, ^ Mar 18 14:45:02 ao2: you can update it to http://downloads.freesmartphone.org/sources/ Mar 18 14:46:36 JaMa, thanks. I didn't know the URI changed. Will send a fix later to oe-devel Mar 18 14:56:29 khem, Is that patch acceptable or do you want me to move all the fftw scripts to the new style? Mar 18 14:57:42 any particular reason you are using an old fftw? Mar 18 14:58:42 Crofton, I think we may be using the new one now, this patch has been hanging around here a while now. Regardless - the script exists, but fails... Mar 18 14:58:54 ok Mar 18 15:06:07 morning Mar 18 15:07:24 m Mar 18 15:07:32 er gm Mar 18 15:07:39 khem: wrt "base.bbclass: pre-create SRC_URI checksums to include in the recipe", you use 'uname'. Do we have standards wrt naming of these checksums? Mar 18 15:09:08 I see JaMa just used e.g. 'SRC_URI[archive.md5sum]' Mar 18 15:09:33 iirc I added a couple of SRC_URI[tarball.md5sum] Mar 18 15:09:56 I know it's just cosmethic but still... Mar 18 15:13:22 khem: good morning Mar 18 15:13:27 it can be anything basically Mar 18 15:24:30 'bitbake opkg' fails for me on do_fetch. I get an HTTP 404 error for trying to fetch http://www.angstrom-distribution.org/unstable/sources/trunk_opkg.googlecode.com_.svn_526_.tar.gz Mar 18 15:25:15 I don't even see where this URI is listed in the .bb or know where it comes from Mar 18 15:25:37 mikeul: that's normal, first it tries to download from "distro" mirrors Mar 18 15:25:55 mikeul: when fails with 404 it continues with other mirror and then it use the URI from .bb Mar 18 15:25:57 OK, then it tries 3 others, which also fail. Mar 18 15:26:17 mikeul: then it should use svn for checkout from repo directly Mar 18 15:26:19 ah, OK, yes I see the URI from .bb at the end finally. Mar 18 15:26:48 Is that default behavior for any do_fetch? Mar 18 15:27:23 so I shouldn't be bothered by the first failures, and should just examine why svn is failing? Mar 18 15:27:32 yes Mar 18 15:28:21 that .tar.gz will be available as soon as angstrom builder will build newer revision of opkg and store that checkout on that URI Mar 18 15:32:14 thanks, JaMa. I'm guessing I have proxy problems with SVN, then. But at least I don't have to worry about the other URIs. Mar 18 15:40:02 03Koen Kooi  07org.openembedded.dev * rdbbe0cae85 10openembedded.git/contrib/feed-browser/includes/config.inc: feed-browser: add unstable configs for beagle and hawk Mar 18 15:40:06 03Enrico Scholz  07org.openembedded.dev * rb2acb9dfb8 10openembedded.git/recipes/ffmpeg/ffmpeg.inc: (log message trimmed) Mar 18 15:40:06 ffmpeg: moved to native staging Mar 18 15:40:06 Old .inc file listed every single header which was to be staged. This Mar 18 15:40:06 causes problems when upgrading to recent 0.5.1 version or when using Mar 18 15:40:06 older versions because headers might not exist there resp. new ones Mar 18 15:40:06 need to be installed. Mar 18 15:40:07 This patch converts the .inc file to new native staging method and Mar 18 16:18:14 meh Mar 18 16:18:18 quite Mar 18 16:53:18 ls Mar 18 16:53:20 oops Mar 18 17:03:26 gm Mar 18 17:03:53 hi likewise Mar 18 17:03:54 topic out-dated? GIT no longer down? Mar 18 17:04:15 hi Jin^eLD Mar 18 17:04:19 heh Mar 18 17:04:24 git is up again I think Mar 18 17:08:23 I think someone forgot to change the topic. Mar 18 17:08:45 03Antonio Ospite  07org.openembedded.dev * ra9387736cf 10openembedded.git/recipes/vala-dbus-binding-tool/vala-dbus-binding-tool.inc: vala-dbus-binding-tool: fix HOMEPAGE and SRC_URI Mar 18 17:15:31 what do you think people about having bash as /bin/sh? Mar 18 17:16:13 not good rather by automatic, but what about "bash-as-sh" script which will "ln -sf /bin/bash /bin/sh" on install and "ln -sf /bin/busybox /bin/sh" on uninstall? Mar 18 17:16:17 hrw: I am stubborn and use dash Mar 18 17:16:40 hrw: uninstall of what? Mar 18 17:17:01 likewise: "bash-as-sh" package Mar 18 17:17:08 hrw: I rather set bash at last point of host system boot, and return to dash on first step of shutdown Mar 18 17:17:33 if dash is installed yes Mar 18 17:17:40 hrw: you mean, it will overwrite /bin/sh on the host? Mar 18 17:17:46 no - on target Mar 18 17:17:53 ah Mar 18 17:19:03 could be useful, but have to be careful messing with /bin/sh, can't let there be any point in time where its not usable.. and of course, a broken bash binary would give you an unusable system, rather than a user you might not be able to login on Mar 18 17:19:04 * kergoth shrugs Mar 18 17:20:04 sure Mar 18 17:20:18 but now my system is not bootable due to total lack of /bin/sh Mar 18 17:20:36 heh, oops Mar 18 17:20:56 * kergoth sighs, caffeine would be nice Mar 18 17:21:28 hrw: I don't get what a bash-as-sh package would do? Don't we have 'alternatives' support for our shell? Mar 18 17:21:38 right Mar 18 17:21:43 * hrw also needs caffeine Mar 18 17:22:00 well, isn't our update-alternatives a shell script? Mar 18 17:22:11 pretty sure we used upd-alt for /bin/sh once and things got .. unhappy Mar 18 17:22:14 egg & chicken? Mar 18 17:22:17 yeah Mar 18 17:22:32 but, on a related note, there's a C update-alternatives implementation in the chkconfig source tree Mar 18 17:22:51 removing busybox from rootfs gives interesting problems ;d Mar 18 17:23:43 anyway time for me Mar 18 17:26:52 03Steffen Sledz  07org.openembedded.dev * rd7274489fb 10openembedded.git/recipes/libunwind/ (libunwind.inc libunwind_0.99.bb): Mar 18 17:26:52 libunwind: new recipe Mar 18 17:26:52 a portable and efficient C programming interface (API) to determine the Mar 18 17:26:52 call-chain of a program Mar 18 17:26:52 Signed-off-by: Steffen Sledz Mar 18 17:34:27 why does task-base have: if "bluetooth" in distro_features and not "bluetooth" in machine_features ? Mar 18 17:35:03 this seems to suggest if I want to disable bluetooth it has to be in DISTRO_FEATURES and MACHINE_FEATURES ? Mar 18 17:35:28 I'm trying to disable bluetooth and wifi from my image that includes task-base-extended Mar 18 17:43:33 oneshel: that logic is a bit weird, but the intent is that ${ADD_BT} gets defined if the distro wants bluetooth and the machine doesn't have it onboard, but does have an expansion bus that you might plug a bluetooth interface into. Mar 18 17:44:00 if you don't want bluetooth at all, you should be able to just remove it from ${DISTRO_FEATURES} and that should be that. Mar 18 17:44:20 pb_: a-ha, gotcha, ok I'll figure out where DISTRO_FEATURES gets set, thanks! Mar 18 17:45:26 if you add bluetooth to MACHINE_FEATURES (which means that your machine has built-in bluetooth) it will disappear from task-base-extended but it will then pop up in task-base itself. Mar 18 17:45:48 pb_: ah tricky, glad I didn't try that and get more discouraged :) Mar 18 17:45:54 heh Mar 18 17:57:35 * kergoth_ gives in and drinks a mt dew.. so much for no caffeine Mar 18 17:59:59 owch Mar 18 18:00:04 I'm trying a slow cut down Mar 18 18:00:15 last week was normal, this week is putting the coffee machine on 4 cups, not 6 Mar 18 18:00:26 (and using a regular coffee cup, not my giant one) Mar 18 18:00:33 next week will be 2 cups i think Mar 18 18:00:45 preparing for the ice coffee season :) Mar 18 18:03:19 heh, nice, gradual is the way to go.. i tried a couple times cold turkey the past few weeks, not the best plan Mar 18 18:03:33 quiters. Mar 18 18:03:34 :) Mar 18 18:03:48 tracking my intake (along with some other habits) on joesgoals.com. handy little site for habit tracking Mar 18 18:03:51 indeed Mar 18 18:18:43 03Martin Jansa  07org.openembedded.dev * r022ec40a3a 10openembedded.git/recipes/gstreamer/ (2 files in 2 dirs): Mar 18 18:18:43 gst-plugins-base_0.10.28: add updated ivorbis-thumb patch from gst-plugins-bad for build fix with thumb enabled Mar 18 18:18:43 Signed-off-by: Martin Jansa Mar 18 18:18:52 03Martin Jansa  07org.openembedded.dev * r101beae925 10openembedded.git/ (2 files in 2 dirs): Mar 18 18:18:52 elmdentica: use SRC_URI from enlightenment repo instead googlecode, bump version Mar 18 18:18:52 Signed-off-by: Martin Jansa Mar 18 18:25:56 03Martin Jansa  07org.openembedded.dev * rc0dd1bf0f5 10openembedded.git/recipes/librsvg/librsvg-native_2.26.0.bb: Mar 18 18:25:56 librsvg-native: PR bump after libexpat soname change Mar 18 18:25:56 Signed-off-by: Martin Jansa Mar 18 18:37:50 03Michael Lippautz  07org.openembedded.dev * r8471bbd175 10openembedded.git/recipes/lighttpd/lighttpd.inc: Mar 18 18:37:51 lighttpd: Convert to new staging. Recipe cleanup. Mar 18 18:37:51 Signed-off-by: Michael Lippautz Mar 18 18:53:57 PR_append doesn't seem to be appending anything to the image filenames. At what point during the bitbake process is this variable parsed? Mar 18 19:30:37 kergoth: I should probably try and moderate my tea consumption, that is gett excessive :/ Mar 18 19:40:41 hey RP_ Mar 18 19:40:57 hi XorA Mar 18 19:41:16 RP_: I have turned off priv message, was getting too distracting while I work Mar 18 19:42:26 XorA: I gathered that, fair enough :) Mar 18 19:43:09 some people are just allergic to asking on channel :-) Mar 18 19:44:05 XorA: :/ Mar 18 19:46:39 03Graeme Gregory  07org.openembedded.dev * rfdbe9e5b69 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: Mar 18 19:46:39 angstrom-2008-preferred-versions.inc : update angstrom to autoconf 2.65 Mar 18 19:46:39 Signed-off-by: Graeme Gregory Mar 18 19:46:39 Acked-by: Koen Kooi Mar 18 19:46:40 03Graeme Gregory  07org.openembedded.dev * r281bcd9468 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git.openembedded.org/openembedded into org.openembedded.dev Mar 18 19:46:40 03Graeme Gregory  07org.openembedded.dev * r50c0b6a8d9 10openembedded.git/recipes/libffi/ (libffi/autoconf-2.64.patch libffi_3.0.8.bb): libffi_3.0.8.bb : add fix for autoconf 2.64+ Mar 18 19:46:41 03Graeme Gregory  07org.openembedded.dev * r5f8902fd6e 10openembedded.git/: Merge branch 'xora/autoconf-2.65' into org.openembedded.dev Mar 18 19:46:44 03Graeme Gregory  07org.openembedded.dev * r19c27b7d16 10openembedded.git/recipes/php/ (9 files in 5 dirs): Mar 18 19:46:44 php* : fix for autoconf 2.64+ problems Mar 18 19:46:44 Patch was taked from pld linux. As part of the debug happened to Mar 18 19:46:44 update php-native to 5.3.2 so this is added as well. Mar 18 19:46:51 03Graeme Gregory  07org.openembedded.dev * r3d701ecb17 10openembedded.git/recipes/autoconf/ (11 files in 2 dirs): Mar 18 19:46:51 autoconf_2.65.bb : add newest version of autoconf. Mar 18 19:46:51 .64 and .65 are starting to get required by some configure.in scripts Mar 18 19:46:55 03Graeme Gregory  07org.openembedded.dev * r5f911bdab6 10openembedded.git/recipes/libpcap/ (2 files in 2 dirs): Mar 18 19:46:56 libpcap_1.0.0.bb : apply patch to fix autoconf 2.64+ problems. Mar 18 19:46:56 Patch was extracted from libpcap git and refreshed. Have to copy Mar 18 19:46:56 aclocal to acinclude to not lose the new macros when OE overwrites. Mar 18 19:46:56 03Graeme Gregory  07org.openembedded.dev * r4de0798959 10openembedded.git/recipes/alsa/ (alsa-state.bb alsa-state/omapzoom2/asound.state): alsa-state.bb : add default state file with onboard speakers enabled for zoom2 Mar 18 19:48:28 there, autoconf 2.65 fun :-) Mar 18 19:48:36 yey! Mar 18 19:48:41 thanks XorA Mar 18 19:48:48 XorA: let the fun begin :) Mar 18 19:49:00 XorA: Are you bumping automake too? :) Mar 18 19:49:26 RP_: one step at a time :-) Mar 18 19:49:52 XorA: I speak from experience having switched in poky ;-) Mar 18 19:50:40 my poor machine is groaning under the pressure of poky builds + OE builds Mar 18 19:50:59 XorA: You're running poky builds? :) Mar 18 19:51:14 RP_: yes, purple Mar 18 19:51:26 RP_: btw I meant to say, you guys forgot to upload the manual to the website Mar 18 19:51:28 XorA: cool. Which target out of interest? Mar 18 19:51:33 RP_: zoom2 Mar 18 19:51:44 XorA: I'm about to tag 3.2.1, a bugfixed purple Mar 18 19:51:55 XorA: and you'll notice the manual commits today? ;-) Mar 18 19:52:05 RP_: Ive got a working kernel for zoom2 Mar 18 19:52:21 RP_: one if purple is busted as it has SRCREV=AUTOREV in the .bb Mar 18 19:52:37 XorA: Can you please send me patches ;-) Mar 18 19:53:00 RP_: will do, its just finished its compile Mar 18 19:53:17 XorA: I'll hold the bugfix release ;-) Mar 18 19:54:36 * XorA guesses he better remove the Angstron logo Mar 18 19:54:54 XorA: heh ;-) Mar 18 19:55:07 what can I say, did the recipe in OE first Mar 18 19:57:28 XorA, leave it in :) Mar 18 19:58:06 XorA: These days I live in the belief we all want the same thing so things like that don't worry me too much. I would probably disable it if it was enabled for DISTRO != angstrom but life is too short to get worked up about it Mar 18 19:58:22 Crofton: I would, but then I would have to answer the questions 1million times why does my poky boot Angstrom Mar 18 19:58:29 hhahaha Mar 18 19:58:49 it would create more branding confusion Mar 18 19:59:10 For years the linux-rp recipe had OH and OZ logos and switched depending on the distro. Both logos were in Poky and OE Mar 18 20:00:58 RP_: some other cockups in recipe so respinning anyway Mar 18 20:02:06 03Martin Jansa  07org.openembedded.dev * ra533840b92 10openembedded.git/conf/distro/include/preferred-shr-versions.inc: Mar 18 20:02:06 preferred-shr-versions.inc: update SHR to autoconf 2.65 Mar 18 20:02:06 Signed-off-by: Martin Jansa Mar 18 20:03:30 * XorA notices the automake we use has security issues Mar 18 20:03:46 how can automake have security holes .... Mar 18 20:04:29 automake?! Mar 18 20:04:32 Crofton: making all files 777 for a short period of time Mar 18 20:04:36 ah Mar 18 20:04:41 http://lists.gnu.org/archive/html/automake/2009-12/msg00012.html Mar 18 20:06:02 Does ONLINE_PACKAGE_MANAGEMENT means we use OE without package management system like ipk,deb or rpm Mar 18 20:06:17 or I should say if set to 'none' Mar 18 20:07:41 RP_: I wonder how toolchain components do with so new autotools Mar 18 20:09:54 khem: We don't reautoconf them thankfully Mar 18 20:10:06 khem: I did have to patch libtool in binutils in poky :( Mar 18 20:10:21 khem: They claim to need very specific autoconf/automake versions :/ Mar 18 20:10:23 yeah I thought it should not be a cakewalk :) Mar 18 20:11:42 RP_: plus wud it work with older versions people use Mar 18 20:11:51 like 2.18 and gcc 4.1 Mar 18 20:11:59 and glibc 2.6.1 Mar 18 20:12:39 khem: setting O_P_M=none will give you an image with no support for package management on the target device. no opkg, ipkg, etc. Mar 18 20:14:02 khem: As I said, we don't reautoconf for those things so its not as bad as it could be Mar 18 20:14:53 RP_: some older version of glibc ended up autreconfing for some weird reasons Mar 18 20:14:59 khem, pb__: I'm still serously tempted to split the build of libstdc++ from the gcc build... Mar 18 20:15:09 RP_: yes, you should do it. Mar 18 20:15:12 or, well, someone should do it Mar 18 20:15:15 pb__: thanks Mar 18 20:15:16 yes Mar 18 20:15:27 khem: Scary, we don't support that and it will break Mar 18 20:15:34 actually spinning out all runtimes from gcc is a good idea may be Mar 18 20:15:42 khem: right Mar 18 20:15:50 RP_: splitting the gcc runtime packages from the compiler itself is, as they say, a plan with no drawbacks Mar 18 20:16:05 pb__: Apart from the gcc build system ;-) Mar 18 20:16:21 That is always a drawback :/ Mar 18 20:16:31 heh. well, yes. but, as my colleagues tell me, that's just a question of pressing the right keys on the keyboard. Mar 18 20:16:54 RP_: email is incoming Mar 18 20:16:59 pb__: Everything to do with computers is just pressing the right buttons Mar 18 20:17:15 I don't think it would actually be very difficult to do: the target libraries are effectively standalone and you could mostly ignore the toplevel cygnus stuff Mar 18 20:17:24 pb__: should be create a new recipe for it ? or just emit packages from gcc-cross or gcc recipes which ever Mar 18 20:17:43 khem: new recipe Mar 18 20:17:50 yeah libstdc++ use to build standalone Mar 18 20:18:02 until one day they jammed it into gcc Mar 18 20:18:10 pb__: I don't doubt its possible, I just feel ill when I look at those files Mar 18 20:18:25 Poky is breaking in that area so I should just do it Mar 18 20:18:33 XorA: thanks for the autoconf work Mar 18 20:18:34 RP_: we can disable the library generation during gcc build Mar 18 20:18:41 pb__: Did you finish you cross merge into staging? Mar 18 20:18:55 actually if we decide to spin out runtime the build sequence will become easier I would say Mar 18 20:19:01 eFfeM: your job now is to find all the hacks in OE and remove them again :-) Mar 18 20:19:09 khem: That is a hope Mar 18 20:19:42 If someone wanted a job they could find all the automake 1.11 fixes in poky and port them too Mar 18 20:19:48 RP_: Adding --disable- should stop building the give dir Mar 18 20:20:11 XorA: thank you but no. I've had enough heat recently, only commit the things that I really need Mar 18 20:20:37 eFfeM: move to a colder country we appreciate the heat up here :-) Mar 18 20:20:54 that reminds me i still have some u-boot patches for calamari lying around Mar 18 20:21:02 google has datacenters in siberia Mar 18 20:21:12 XorA: NL is cold enough Mar 18 20:21:30 XorA: which address did you send to? Mar 18 20:21:37 RP_: rpsys.net Mar 18 20:21:58 XorA: ah yes, its there now :) Mar 18 20:22:11 * RP_ is too imparient evidently... Mar 18 20:22:11 XorA: hope you understand my feelings Mar 18 20:22:40 pb__: gcc-cross should only produce gcc, cpp , g++, gfortran and friends and everything else should be in gcc-runtime right ? Mar 18 20:23:04 eFfeM: Ive been is FOSS software to develop a certain level of asbestos and teflon coating Mar 18 20:23:30 lol Mar 18 20:23:47 XorA: yeah, i understand, but my coating is not yet up to the koen level :-/ Mar 18 20:24:02 eFfeM: wuss ;) Mar 18 20:24:11 eFfeM: you live right near him, go out for beers, he is a nice guy really Mar 18 20:24:44 XorA: we've met once, i think about 4 years ago. He is still something like 200 km away from me Mar 18 20:25:06 eFfeM: there are more flamers you might not have yet come across Mar 18 20:25:36 actually google maps says it is only 151 km Mar 18 20:25:44 khem, in oe ? Mar 18 20:25:59 in other FOSS communities Mar 18 20:26:08 oh yeah, i know Mar 18 20:26:10 OE is very polite Mar 18 20:26:21 wait until you submit patches to arm-linux or mplayer :-D Mar 18 20:26:51 hm, guess I am not too interested Mar 18 20:27:14 Its weird to meet people in real life Mar 18 20:27:18 did some mtd and also some u-boot patches, those are both nice communities Mar 18 20:27:26 few years back someone banned me on mipslinux IRC channel because my client was reconnecting automatically after a disconnect and my company internet connection was having issues Mar 18 20:27:39 Alan Cox doesn't want to lynch me afterall :) Mar 18 20:27:51 hehe Mar 18 20:28:05 03Koen Kooi  07org.openembedded.dev * r971efeefff 10openembedded.git/recipes/qt4/qt4.inc: qt4: make .pc files work with sysroot support Mar 18 20:28:11 you can send ARM related mplayer patches to me and I'll apply it at least here :) http://repo.or.cz/w/mplayer/glamo.git Mar 18 20:31:34 RP_: the cross -> staging stuff is done, I just need to fix the libgcc and other runtime library packaging bits. Mar 18 20:31:44 or wait for you to fix them :-) Mar 18 20:32:12 found pics from when I met koen, it was in 2006 at slugdag: http://www.nllgg.nl/bijeenkomsten_2006_slugdag#1111 Mar 18 20:32:16 try to spot us :-) Mar 18 20:33:11 right now it is generating libgcc (etc) with the wrong architecture Mar 18 20:33:20 XorA: I assume this applies to poky master as well? Mar 18 20:33:24 black shirt Mar 18 20:33:39 yup Mar 18 20:33:49 koen that is, not me Mar 18 20:33:55 RP_: havent tried Mar 18 20:34:10 pb__: I knew you were a little enthusiastic with the removal of code from the class ;-) Mar 18 20:34:23 XorA: np, it has to be better than whats there ;-) Mar 18 20:34:37 RP_: heh. well, I still think the old code was wrong, it's just a case of replacing it with the right thing :-} Mar 18 20:35:18 RP_: and, I think the best way to fix it is to generate the target packages (both libgcc and libstdc++ etc) from a separate recipe, rather than trying to have a single hybrid recipe which emits packages for multiple arches. Mar 18 20:37:34 pb__: I agree with splitting it but for different reasons Mar 18 20:40:18 pb__: The fundamental problem is that the current approach can end in a situation where you want to package two different things of different arches Mar 18 20:40:47 Until now we've ways said the "native" binaries didn't need packaging Mar 18 20:47:57 eFfeM: in one pic he seems to be listening to you :) Mar 18 20:48:19 khem historical moment :-) Mar 18 20:48:48 actually was a fun day Mar 18 20:57:26 03Graeme Gregory  07org.openembedded.dev * r1f4479554d 10openembedded.git/recipes/automake/ (automake.inc automake_1.10.3.bb): Mar 18 20:57:27 automake_1.10.3.bb : add latest version of 1.10.X branch. Mar 18 20:57:27 This fixes CVE-2009-4029 Mar 18 20:57:36 03Graeme Gregory  07org.openembedded.dev * r8086a8a5cd 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: angstrom-2008-preferred-versions.inc : move angstrom to the latest 1.10.3 am Mar 18 20:58:23 XorA: any known issues with am-1.11.1? Mar 18 20:58:24 khem, your talk is tonight? Mar 18 20:59:45 XorA: m4 build failed there in do_configure that it needs am-1.11*, so I'm curious why older am-1.10* Mar 18 21:13:05 JaMa: clearing a security bug with a minor update, proper test of 1.11 will take a day or two Mar 18 21:13:12 JaMa: its on the plans Mar 18 22:01:41 03Graeme Gregory  07org.openembedded.dev * r0378972559 10openembedded.git/recipes/linux/ (2 files in 2 dirs): Mar 18 22:01:41 linux-omap-zoomsync_2.6.32.bb : update defconfig to enable power management Mar 18 22:01:41 Also turned off the pm debug messages that were drowning dmesg. Mar 18 22:01:41 Signed-off-by: Graeme Gregory Mar 18 22:09:08 Crofton: yes at 6:30 PDT Mar 18 22:13:09 khem, gl Mar 18 22:13:16 not sure if anyone I know will make it Mar 18 22:16:18 Crofton thx Mar 18 22:16:41 I have 13 confirmed and some 8-9 maybes so its a good sized group. Mar 18 22:21:19 Crofton: no issues. I think depending upon todays talk there may be follow ups Mar 18 22:21:59 if group has enough interest in learning more then we will have more talks related to OE and some deep dives into various parts **** BEGIN LOGGING AT Thu Mar 18 22:39:53 2010 Mar 18 22:58:17 gl khem Mar 18 23:01:43 thanks likewise Mar 18 23:01:55 I am hopeful people are going to like it Mar 18 23:02:30 we'll see the crowds appear in #oe :-) Mar 19 01:06:19 * kergoth thinks maybe project euler exercises weren't the best way to practice Go... math is way too rusty Mar 19 01:57:26 Can I disable the checksums? I can't remember which parameter control this Mar 19 01:58:57 Openfree`: yes you can, but why would you? Mar 19 01:59:21 Openfree`: the checksum check is implemented in base.bbclass, you can find the variable you need to set there (or maybe in the documentation too) Mar 19 02:00:10 I using a overlayer with official OE, and too many missing checksums.. Mar 19 02:00:50 It takes too much time to resume after I add into checksums.ini Mar 19 02:15:58 I asked a related question yesterday Mar 19 02:16:11 about modifying recipes Mar 19 02:16:14 for images Mar 19 02:16:29 is tehre any way to run commands on the root fs before it gets packaged? Mar 19 02:16:47 like running commands to copy some files over to specific places in the rootfs image Mar 19 02:16:54 is there a way to do that from the recipe? Mar 19 02:18:12 in the recipe its easy, you can manipulate the do_rootfs task if you really want to Mar 19 02:18:16 but you can use ROOTFS_POSTPROCESS_COMMAND variable Mar 19 02:18:40 ROOTFS_POSTPROCESS_COMMAND = "rm ${D}${sysconfdir}/inittab; touch somefile" .. or whatever Mar 19 02:18:51 can put that in local.conf, or wherever Mar 19 02:19:03 clearly, postprocess is after, i think there's a pre too Mar 19 02:25:51 zecke, can we check multi chechsums.ini . because when deal with overlay, sources's chksums won't find in OE (oe/conf/checksums.ini) Mar 19 02:26:06 or better split one checksum per package Mar 19 02:28:24 Openfree`: yes, we are migrating there... you can put ;name=WHATEVER add the end of the entry in SRC_URI Mar 19 02:28:37 monolihic checksums is a monster Mar 19 02:28:43 Openfree`: and then SRC_URI[WHATEVER.md5sum] = "sum", SRC_UR[WHATEVER.sha256sum] = "" Mar 19 02:30:45 Openfree`: uh, you can already split conf/checksums.ini per overlay/collection Mar 19 02:30:51 last i looked, anyway Mar 19 02:31:03 and of course the in recipe form is the future Mar 19 02:31:33 zecke, that's good~ Mar 19 02:35:01 damnit keep getting distracted Mar 19 02:35:06 thanks again kergoth Mar 19 02:38:08 np **** ENDING LOGGING AT Fri Mar 19 02:59:56 2010