**** BEGIN LOGGING AT Sun Jun 17 02:59:56 2007 Jun 17 04:02:39 03mickeyl 07org.oe.dev * r33019e0e... 10/ (3 files in 3 dirs): libxosd: refresh use-sane-default-font patch Jun 17 04:48:29 03mickeyl 07org.oe.dev * red459a8f... 10/ (1 packages/e17/expedite_cvs.bb): expedite_cvs.bb: add metadata Jun 17 07:06:43 03koen 07org.oe.dev * raf152d47... 10/ (4 files in 4 dirs): gpe-session-scripts: turn on cursor hiding based on the touchscreen machine feature Jun 17 07:43:01 * * OE Bug 776 has been RESOLVED (FIXED) by Jun 17 07:43:03 * *  libqpe-opie fails to compile with EABI support Jun 17 07:43:05 * * http://bugs.openembedded.org/show_bug.cgi?id=776 Jun 17 07:49:01 * * OE Bug 2502 has been RESOLVED (FIXED) by Jun 17 07:49:03 * *  Angstrom: please make no cursor as default in gpe Jun 17 07:49:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2502 Jun 17 08:01:42 03koen 07org.oe.dev * rc5c4014f... 10/ (4 files in 2 dirs): linux-h1940: set compatible machine Jun 17 08:01:47 03koen 07org.oe.dev * r1a26a920... 10/ (3 files in 2 dirs): LAB-kernel: set compatible machine to h2200 Jun 17 08:01:51 03koen 07org.oe.dev * r2a8af8f1... 10/ (3 files in 2 dirs): a780-kernel, e680-kernel: remove needless indirection Jun 17 08:29:15 03koen 07org.oe.dev * r84163066... 10/ (1 packages/linux/linux_2.6.20.bb): linux 2.6.20: raise the default pref for at91 Jun 17 08:29:20 03koen 07org.oe.dev * re57bfc86... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: lock at91sam9263ek to 2.6.20 for now Jun 17 08:33:04 hi all Jun 17 08:33:37 hey steliosk Jun 17 08:35:45 koen : still trying to figure out a way to handle libusb Jun 17 08:36:08 03koen 07org.oe.dev * r69e09858... 10/ (3 files in 3 dirs): libmatchbox 1.9: apply BGR fixes from poky Jun 17 08:36:21 steliosk: I'd say "fix gcc" Jun 17 08:36:26 koen : i still wonder why libtool things there are gcc shared libs.... Jun 17 08:37:03 koen : i also think that is the problem Jun 17 08:46:20 its regarding the shared libs with avr32 gcc Jun 17 08:46:53 morning all Jun 17 08:48:10 RP hi ! Jun 17 08:59:24 Hello Jun 17 08:59:38 armeb-angstrom-linux-uclibcgnueabi-gcc Jun 17 09:04:29 my uclibc did not compile because it tries to include curses.h , I had to install the dev package on ubuntu is that suspected behaviour or should gcc initial compile a curses? Jun 17 09:27:17 hi all Jun 17 10:00:21 03pH5 07org.oe.dev * r9ffceb89... 10/ (1 conf/distro/angstrom-2007.1.conf): magician: use 2.6.21-hh8 Jun 17 10:00:26 03pH5 07org.oe.dev * r34a87b8d... 10/ (1 conf/machine/magician.conf): magician: add a few modules in case they are not compiled in Jun 17 10:00:30 03pH5 07org.oe.dev * r945d147b... 10/ (1 packages/gnome/gnome-cups-manager_0.31.bb): gnome-cups-manager: package icons, too Jun 17 10:00:39 03pH5 07org.oe.dev * rcc76c091... 10/ (1 packages/linux/linux-handhelds-2.6/magician/defconfig): linux-handhelds: disable HWTIMER in magician defconfig Jun 17 10:00:45 03pH5 07org.oe.dev * ra5289b2d... 10/ (1 packages/python/python-pygobject_2.12.2.bb): python-pygobject-2.12.2: package -dev first Jun 17 10:00:49 03pH5 07org.oe.dev * r636893a3... 10/ (1 packages/python/python-pygtk2_2.10.3.bb): python-pygtk-2.10.3: package -dev first, fix -dbg packaging Jun 17 10:05:26 03koen 07org.oe.dev * r381f062d... 10/ (1 packages/linux/linux-2.6.20/at91sam9263ek/defconfig): 2.6.20: update at91 defconfig Jun 17 10:08:52 Hi florian, what is the status? Jun 17 10:09:53 Laibsch: nothing new yet, i just turned on the workstation again Jun 17 10:10:29 you got the necessary information, though, right? Jun 17 10:11:17 Laibsch: I guess so... Jun 17 10:11:44 so what is that information? Jun 17 10:11:51 Any problems to be expected? Jun 17 10:13:39 Laibsch: Which versions of Drupal we use and how to upgrade these. Jun 17 10:13:53 yes? Jun 17 10:14:05 which versions do you use? Jun 17 10:14:25 Can you just share that information, please? Jun 17 10:15:44 Laibsch: OE and the main website use quite old ones that are not compatible with php5.. the Angstrom one is up to date. Jun 17 10:17:15 how are you going to resolve this? Upgrade all drupals to the latest version, php4 alongside php5? Jun 17 10:17:23 When do you expect you can do this? Jun 17 10:29:41 03koen 07org.oe.dev * rb0f2b6b6... 10/ (6 files in 4 dirs): flumotion: add 0.4.1 from poky Jun 17 10:31:12 Laibsch: It depends... I'll try to update the main one today or tomorrow, but i'm not sure who feels responsible for the oe one. Jun 17 10:31:48 you need a hand? Jun 17 10:32:59 more then one ;) Jun 17 10:33:20 Well, I have two :-) Jun 17 10:39:22 "Asterisk 1.4.5 was released and this is the first release that will Jun 17 10:39:23 properly build under the oe build system" Jun 17 10:39:44 * koen looks at the ipkgs that apparently couldn't have existed Jun 17 10:39:45 heh Jun 17 10:39:59 ~lart morons Jun 17 10:39:59 * ibot puts on a hockey mask and jumps out at morons Jun 17 10:40:40 http://www.angstrom-distribution.org/repo/?action=details&pnm=asterisk Jun 17 10:48:43 koen: "build without nasty hacks"? ;) Jun 17 10:49:30 ~lart ibot Jun 17 10:49:31 * ibot throws a AN/M-8 smoke grenade at sup3rkiddo Jun 17 10:49:37 :S Jun 17 11:03:24 <_diego_> it's possible to compile a package (ie: kernel) with a different version of gcc? Jun 17 11:30:14 03pH5 07org.oe.dev * r786b1075... 10/ (1 packages/libxosd/libxosd_svn.bb): Jun 17 11:30:14 libxosd: QA issue: package libxosd-dbg before libxosd-examples, Jun 17 11:30:14 so that ${bindir}/.debug doesn't end up in libxosd-examples. Jun 17 11:30:21 03pH5 07org.oe.dev * r688b2cde... 10/ (1 packages/linux/linux-handhelds-2.6_2.6.21-hh8.bb): linux-handhelds-2.6: add 2.6.21-hh8 Jun 17 11:30:27 03koen 07org.oe.dev * rccf66183... 10/ (1 packages/python/python-gst_0.10.7.bb): python-gst: add missing dependency on gst-plugins-base Jun 17 11:30:33 03koen 07org.oe.dev * rcd15f8b1... 10/ (1 packages/flumotion/flumotion_0.4.1.bb): flumotion: hack packaging Jun 17 11:42:01 * * OE Bug 2508 has been created by darkstarsword(AT)gmail.com Jun 17 11:42:03 * * Minor discrepancy between keymap on Angstrom Spitz and original Sharp Distro Jun 17 11:42:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2508 Jun 17 12:20:21 <_diego_> there is a easy way to compile OE but not the kernel? Jun 17 12:29:06 _diego_: ASSUME_PROVIDED? Jun 17 12:30:04 <_diego_> i've already tried the following in local.conf: Jun 17 12:30:20 <_diego_> += " virtual/kernel " Jun 17 12:30:37 <_diego_> and +=" lart-linux-2.6 " Jun 17 12:31:40 <_diego_> but when i run again bitbake, it compiles again lart-linux-2.6 Jun 17 12:49:56 03koen 07org.oe.dev * r71e45474... 10/ (1 packages/flumotion/flumotion_0.4.1.bb): flumition: add more rdepends Jun 17 12:50:00 03koen 07org.oe.dev * rdee9a233... 10/ (1 packages/zope/zope_3.3.1.bb): zope: install python files in /usr/lib/python2.4 instead of /usr/lib/python Jun 17 12:50:05 03koen 07org.oe.dev * rd047390e... 10/ (1 packages/flumotion/flumotion_0.4.1.bb): flumotion: even more python depends Jun 17 12:50:23 why is python so hard to install!?!?!? Jun 17 12:57:01 * * OE Bug 2509 has been created by  Jun 17 12:57:03 * * flumotion is missing ihooks module Jun 17 12:57:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2509 Jun 17 13:08:07 morning Jun 17 13:08:22 morning Jun 17 13:28:37 _diego_ perhaps assume provided? Jun 17 13:29:36 <_diego_> as this: Jun 17 13:29:42 <_diego_> SSUME_PROVIDED += " virtual/kernel " Jun 17 13:30:00 <_diego_> *ASSUME Jun 17 13:30:44 <_diego_> i've tried, but it's the same Jun 17 14:03:41 hmm looks like localdef recieves SIG 11 on a amd-64 Jun 17 14:04:16 localedef? Jun 17 14:04:24 qemu, you mean? Jun 17 14:05:17 yeah right, qemu: uncaught target signal 11 (Segmentation fault) - exiting during glibc package Jun 17 14:05:59 IT's know. Jun 17 14:06:00 n Jun 17 14:06:08 qemu doesn't work on amd64 :( Jun 17 14:06:12 what can i do ? Jun 17 14:06:24 And as far as I know there's no fix but disabling binary locale generation Jun 17 14:06:35 read local.conf comments to find out how. Jun 17 14:06:39 ah , i can live with that Jun 17 14:07:17 Me too, because I can swap on Z and generate locales on the device, but generally it sucks. Jun 17 14:08:00 ok great , thanks polyonymous Jun 17 14:08:05 np Jun 17 14:09:47 i just recieved my own dedicated server , you know ,, and i will put all my oe builds on it Jun 17 14:10:19 sounds fun Jun 17 14:10:33 rob_w: Nice. I hope you ordered it with plenty of hard disk space Jun 17 14:10:38 You'll need it Jun 17 14:11:25 Laibsch, i know .. it runs 160GB .. so one chain normally takes 6- 9 GB .. so i can run plenty of different chains Jun 17 14:14:11 its nice to have a oe build runnig but see your local cpu at 0 % ;-) Jun 17 14:17:15 hi there. is there anyone that can answer me questions about the licensing of the open embedded project? Jun 17 14:17:21 rob_w, I've been doing that on my old computer sitting next to me, but that's kinda slow (athlon xp 2200+). Now I'm doing that on my computer (amd64 x2 4200+, I think) and it's tolerable :) Jun 17 14:18:45 polyonymous, do you have a need for meta-toolchain to work ? i do , but i dont have the guts to solve it , yet Jun 17 14:19:24 rob_w, I do, but I have tons of other needs, so I didn't really try to do anything about it. Jun 17 14:20:28 ah ok .. do you think it is doable to hand copy all files in place ( accordiing an old working meta-toolchain image) Jun 17 14:20:49 as i have all needed files in the working oe staging dir Jun 17 14:21:00 no idea, I haven't even taken a look at it. Jun 17 14:21:07 but i need to hand out a similar chain to a other company Jun 17 14:21:16 yeah, ok never mind though Jun 17 14:26:20 polyonymous: you might want to set up an icecc network Jun 17 14:26:23 very easy Jun 17 14:26:26 very nice Jun 17 14:26:34 http://www.openembedded.org/icecc Jun 17 14:26:46 Laibsch, I know and I'm going to :) Jun 17 14:27:07 icecc? distributed compilation? Jun 17 14:27:11 yup Jun 17 14:27:26 can it be cross platform? Jun 17 14:27:34 summatusmentis: yes Jun 17 14:27:35 as in, can a windows machine run it? Jun 17 14:27:44 Well, that I think no Jun 17 14:27:53 oh, ok Jun 17 14:27:55 But take a look at the icecc homepage Jun 17 14:28:00 somewhere on opensuse.org Jun 17 14:28:09 lol, the only other machine on our network at home that has any juice is the desktop :-P Jun 17 14:28:28 summatusmentis: You can boot it to Knoppix when not in use Jun 17 14:28:40 summatusmentis, "desktop" doesn't imply "windows". Jun 17 14:28:55 I have PXE-bootable install with icecc just for that purpose Jun 17 14:29:04 you can run a colinux on the desktop, btw. Jun 17 14:29:07 It is served over NFS Jun 17 14:29:12 does someone here has an idea about the licensing of the open embedded project? I tried to find information about that on the openembedded.org, but I have a few questions about details and it would be cool if someone of you could give me some pointers. Jun 17 14:29:18 (that is on the windows desktop) Jun 17 14:29:28 fewi: Most of it is MIT Jun 17 14:29:30 MIT, I think Jun 17 14:29:33 What is your question? Jun 17 14:29:34 polyonymous: that's true, however, "parents" implies "windows" Jun 17 14:29:56 summatusmentis, that's also configurable. Jun 17 14:30:13 summatusmentis: Most unintrusive (speak preserving family peace) solution is Knoppix or NFS boot Jun 17 14:30:20 summatusmentis, many people think "wife" implies "windows" too, but I had no problems reconfiguring mine. Jun 17 14:30:53 I think the most unintrusive solution is running colinux, because parents can keep surfing while you're compiling. Jun 17 14:30:57 03Laibsch 07org.oe.dev * r64cf723f... 10/ (1 packages/uim/uim.inc packages/uim/uim_1.3.1.bb): uim: fix packaging Jun 17 14:31:04 03Laibsch 07org.oe.dev * re75f9827... 10/ (1 packages/kanjipad/kanjipad.inc): kanjipad: add gtk+ to depends Jun 17 14:31:11 I'm thinking about colinux, what about cygwin? Jun 17 14:31:23 We started a new project involving an iMX27 or 31 and currently I am investigating the possibilities about the basic OS on it Jun 17 14:31:47 summatusmentis, I don't know, but colinux sounds like cleaner solution to me. Less hassle once you get it running Jun 17 14:31:59 summatusmentis: http://en.opensuse.org/Icecream Jun 17 14:32:09 thanks Laibsch Jun 17 14:32:18 * Laibsch agrees with polyonymous Jun 17 14:32:26 I think you want some virtualizer Jun 17 14:32:32 not just the lib like cygwin Jun 17 14:32:46 colinux is virtualized? I didn't know that, ok Jun 17 14:32:50 I am almost certain you will run into big trouble with cygwin Jun 17 14:32:50 and I was asked to clear out the licensing. like: are we forced by the license to open also the application that runs on it etc.pp. Jun 17 14:32:59 summatusmentis: I don't know that either Jun 17 14:33:04 Never used it. Jun 17 14:33:13 But I assume that from polyonymous' comments Jun 17 14:33:15 I'm not sure if virtualizer is a proper term for this implementation but from user perspective yes. Jun 17 14:33:37 right, I've played w/ colinux, I don't remember anything virtualized(specifically) Jun 17 14:33:38 And I have to say, though, that I've never used it either :) Jun 17 14:33:57 ok, after I reformat our desktop, colinux is the first thing on there Jun 17 14:33:59 summatusmentis, the point is that you have a linux system on a windows box. Jun 17 14:34:05 before anti-virus even Jun 17 14:34:14 wait, I thought you don't have to reformat your desktop for it. Jun 17 14:34:16 fewi: From OE I think you are safe Jun 17 14:34:22 polyonymous: no, you don't Jun 17 14:34:28 but ours needs it, desperately Jun 17 14:34:34 Ah ok Jun 17 14:34:38 fewi: But you might want to sent a mail to the dev list just to be sure Jun 17 14:34:42 we're getting programs using asian characters, that I didn't even know existed Jun 17 14:34:56 fewi: If you use OE, we appreciate contributions back Jun 17 14:35:11 fewi: People who just take are frowned upon. Jun 17 14:35:20 fewi: I am not talking financials, not at all Jun 17 14:35:49 fewi: You will have to be sure to comply with the license of the programs that you compile, of course Jun 17 14:36:05 Well, our current stand is that we are happily contributing to projects we are participating from. There are some issues that prevent the actual application being opened Jun 17 14:36:09 If you link your stuff against a GPL source, you need to open your source. Jun 17 14:36:27 fewi: I think IRC is not the right place to discuss this Jun 17 14:36:36 Especially not on Sunday ;-) Jun 17 14:36:51 We are in the process of setting up a foundation (or os I think) Jun 17 14:36:53 I've seen that it is possible to configure "LICENSE" and as far I understood this will prevent me from compiling GPL stuff if "GPL" is not in that list? Jun 17 14:37:12 fewi: NO Jun 17 14:37:13 ok, shall I enter that list: "openembedded-devel@lists.openembedded.org" and discuss it there? Jun 17 14:37:14 no Jun 17 14:37:20 fewi: yes Jun 17 14:37:23 That is best Jun 17 14:37:52 ok, thank you very much (for answering on sunday :) ) Jun 17 14:38:04 The LICENSE field is used to make sure that OE publishes sources when it compiles GPL-covered binaries Jun 17 14:38:13 I don't think it is used for anything else Jun 17 14:38:24 no problem Jun 17 14:38:30 See you on the list Jun 17 14:38:37 basically we would need the kernel plus a filesystem, not very much more. Jun 17 14:39:02 I guess the readers would be interested to learn as much about your project as you can disclose without running into trouble ;-) Jun 17 14:39:46 after all it should be entertaining too Jun 17 14:40:16 yes, for fewi getting roasted for disclosing company internals ;-) Jun 17 14:41:20 And for us to see it :) Jun 17 14:41:27 it is a kind of musical instrument and there is a lot of 'IP' in the DSP so that part can not be opened... it is moving from an old 8bit host architecture to something with beef. currently we are looking for possibilities that allow us not to write a whole OS Jun 17 14:42:09 Oh, I think mickey|fic is going to like that ;-) Jun 17 14:42:14 Sounds fun :) Jun 17 14:42:18 Can you use it with a guitar? Jun 17 14:42:35 yes, I consider that as a fun job that also provides money :) Jun 17 14:42:43 you could also use it with a guitar, yes. Jun 17 14:43:12 Oh, I am sure that mickey|fic would appreciate a complimentar copy of the hardware ;-) Jun 17 14:43:24 He is one of the main guys who got OE rolling Jun 17 14:44:12 but it is not a requirement Jun 17 14:44:16 of course ;-) Jun 17 14:44:26 :) Jun 17 14:44:36 well, currently it does not even boot, so ... Jun 17 14:45:22 it is pretty early stage right now, that is why I can open my eyes a bit more. Jun 17 14:47:53 anyone here is pretty open to Linux etc. but is a bit concerned about the TIVO incident. it is not that we want to use stuff without contributing back but there are some parts that can not be opened. therefore I Jun 17 14:48:26 like to check it before (also to be clear and not greedy) Jun 17 14:53:25 hi, all Jun 17 14:53:44 when building angstrom-x11-image, IMAGE_LINGUAS seems to have no effect - any ideas? Jun 17 14:53:56 more than that, forcing that in image.bbclass sets up only base locale, but no package locales... Jun 17 14:54:22 Is there any way to build localized image? Jun 17 15:00:27 @Laibsch: gonna subscribe, see you there.. Jun 17 15:04:47 fewi: You can also work through gmane.org if that is what you prefer Jun 17 15:55:26 re Jun 17 15:57:25 wb koen|770 Jun 17 15:57:50 polyonymous, i couldn't find 'hooks' or 'ihooks' in the py manifest Jun 17 15:58:03 hey Laibsch Jun 17 15:58:32 good to see Jun 17 15:58:34 you Jun 17 15:58:38 just in time Jun 17 15:58:47 koen, I don't even know what 'py manifest' is :) I'm whatever but a python guru, but the ${WORKDIR}/install/python-misc/ has the file. Jun 17 15:59:01 I got courageous and did a "bitbake -c fetchall world" Jun 17 15:59:04 ah,nice Jun 17 15:59:22 I am stuck now with some maemo stuff Jun 17 15:59:28 fewi : GPL licence had not only to do with OE but with the way your program links to GPL stuff Jun 17 15:59:35 packgs/python/python-2.4.4-manifest Jun 17 15:59:43 the maemo image depends on osso-gnome-vfs2 Jun 17 15:59:47 which is not available Jun 17 15:59:56 a) remove it from meta-maemo Jun 17 16:00:16 b) set default-pref for maemo-image and meta-maemo to -1 to exclude it Jun 17 16:00:22 c) something else? Jun 17 16:00:51 koen: what do you suggest? AFAIK, maemo is not really developped so actively in OE anymore, is it? Jun 17 16:00:52 c) Jun 17 16:00:52 koen, whatever it is, I think the ihooks.* are there and that should serve your purpose ;-) I don't really feel like becoming a python guru, although it seems to be inevitable :) Jun 17 16:01:03 wait for maemo4 Jun 17 16:01:24 Then I think b) is appropriate Jun 17 16:01:35 And reverse b) when maemo4 is there. Jun 17 16:01:44 polyonymous, could yiu add python-musc to flumotion? Jun 17 16:02:14 koen, sure I can, but I've never built and don't even know what's that, so I'm not sure if I'm an appropriate person to do it. Jun 17 16:03:00 flumotion is missing some rdepends Jun 17 16:03:18 I take it python-misc is one :) Jun 17 16:03:33 yes :) Jun 17 16:03:42 Well, I'll add it now, anyway, dunno if it fixes things but shouldn't break them :) Jun 17 16:04:08 and i only have my 770 with me Jun 17 16:04:14 thanks Jun 17 16:04:35 yeah, that doesn't sound like the most convenient device for editing oe tree :) Jun 17 16:08:01 * * OE Bug 2509 has been RESOLVED (FIXED) by Jun 17 16:08:03 * *  flumotion is missing ihooks module Jun 17 16:08:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2509 Jun 17 16:18:01 * * OE Bug 2510 has been created by  Jun 17 16:18:03 * * tetex fails do_compile Jun 17 16:18:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2510 Jun 17 16:18:14 * koen|770 is happy that noone noticed that mtn server has been moved Jun 17 16:18:29 ~hail ssh -L Jun 17 16:18:31 * ibot bows down to ssh -L and chants, "I'M NOT WORTHY!!" Jun 17 16:18:34 oh that was really smooth migration. Jun 17 16:18:45 heh, smart :) Jun 17 16:19:46 ssh -C -L even Jun 17 16:20:27 does -C help a lot? Jun 17 16:20:58 it seems to help Jun 17 16:21:29 hi all Jun 17 16:21:59 meaning mtn's network layer isn't overly efficient :) Jun 17 16:22:25 cia is now lagging a few minutes, but we'll lok at that next week Jun 17 16:28:02 koen: It still has the same IP over here. Jun 17 16:28:38 Oh, of course Jun 17 16:28:41 Laibsch, the wonders of portforwarding :) Jun 17 16:28:48 Yes, just realized Jun 17 16:34:32 This drupal wiki is a mess Jun 17 16:35:30 Good evening Jun 17 16:54:13 ~lart gprs Jun 17 16:54:13 * ibot hurls dozens of incontinent, insomniac, hungry kittens with tiny little razor-sharp claws and a wide variety of contagious intestinal parasites at gprs Jun 17 16:55:28 poor kittens. Jun 17 16:56:43 hey HopsNBarley Jun 17 16:57:02 hi koen - working on sunday? Jun 17 16:57:30 no, visiting my parents Jun 17 16:57:52 fathersday and all that Jun 17 16:59:06 03polyonymous 07org.oe.dev * r96278dc8... 10/ (1 packages/flumotion/flumotion_0.4.1.bb): flumotion: add python-misc rdepend for ihooks and close bug #2509 Jun 17 16:59:11 03sirfred 07org.oe.dev * rc74d0df4... 10/ (3 files in 3 dirs): xserver-kdrive-imageon : New version 1.2.0 Jun 17 17:00:04 sirfred: cool, is there hx4700 support already? Jun 17 17:00:17 pH5: I'm afraid not. Jun 17 17:01:56 ~lart ati just because Jun 17 17:01:56 * ibot whips out his power stapler and staples ati just because's foot to the floor Jun 17 17:02:08 pH5: I've just modified the current patches to work against 1.2.0 mainstream. To get hx4700 support, I'm afraid we should need some specs of the card. :-( Jun 17 17:05:01 pH5: I'm going to ask people of the tcpmp project. Perhaps they've signed some NDA with ATI, but they have IDCT support for imageon 4200 Jun 17 17:07:47 sirfred: oh, that sounds interesting. I hope you can get something out of that. I fear they just have some kind of API into the binary wince ati driver? Jun 17 17:10:38 pH5: No, they are filling the ringbuffer to perform hardware IDCT + Motion Compensation, as if they knew what they are doing. Jun 17 17:11:43 pH5: This is not really related with w3220 (hx4700) support, but perhaps they could share some specs. Jun 17 17:11:54 xvmc would be nice Jun 17 17:12:02 * sirfred agrees Jun 17 17:12:53 Anyway, it's going to be a hard task. The first thing is to find out how the idct stuff works (in general). Jun 17 17:13:54 Then, probably the microcode and features of the w4200 are more advanced than the ones in the w100/w3220. Perhaps the microcode commands are not compatible at all. Jun 17 17:15:03 But, yes, it would be nice, as the major problem now is the bandwidth we need to send the uncompressed image to the card. This way, we probably could play full-screen vga res videos. Jun 17 17:16:53 ibot should be extended to understand 'just because'. Jun 17 17:23:25 sirfred: ok, so good news but only a little :) Jun 17 17:24:15 pH5: Yes, only a little. :) Jun 17 17:53:21 http://www.itsawonderfulinternet.com/ Jun 17 18:06:42 ibot: botmail for florian: ping Jun 17 18:08:46 CoreDump|home: lol, karate kid III xD Jun 17 18:14:28 great Jun 17 18:18:26 indeed, found it on slashdot of all places Jun 17 18:57:22 re Jun 17 19:04:11 hi koen Jun 17 19:07:39 hey woglinde Jun 17 19:07:52 damn c Jun 17 19:08:05 why there isnt a easy way to get time in nanoseconds Jun 17 19:10:01 woglinde, because ns didn't exist when c was invented ;) Jun 17 19:10:43 koen hm yes that makes sense Jun 17 19:11:04 hm as I see the magic is in librt Jun 17 19:11:20 woglinde: The timeval structure has a tv_nsec field. But I don't know what functions make use of it. Jun 17 19:11:21 2.6.23 should have hrt for most archs Jun 17 19:12:02 Only nanosleep, that I'm aware of Jun 17 19:12:42 woglinde, did you send mickeyl your new mtn key yet? Jun 17 19:22:09 yay Jun 17 19:22:24 flumotion seems to work Jun 17 19:23:31 Trying to track a problem with weird installed=yes in gstreamer .la files, I see that oe_libinstall in base.bbclass only tries to replace installed=yes with installed=no in the staged .la file when a counterpart .lai file exists. Jun 17 19:23:57 What does this mean? Jun 17 19:24:03 hmm Jun 17 19:24:50 the gst .la files point to workdir, so that needs to get fixed as well Jun 17 19:25:21 koen|770: Sure, I only would like to understand the magic behind this. Jun 17 19:30:24 koen|770: It seems that the super-sed line in the oe_libinstall is also able to fix that Jun 17 19:30:46 sed -e 's/^installed=yes$/installed=no/' -e '/^dependency_libs=/s,${WORKDIR}[[:alnum:]/\._+-]*/\([[:alnum:]\._+-]*\),${STAGING_LIBDIR}/\1,g' $dotlai >$destpath/$libname.la Jun 17 19:31:03 Unfortunately, that is only reached if $dotlai exists Jun 17 19:31:11 * mwester-laptop attempts to read that, and his brain coredumps... Jun 17 19:31:28 mwester-laptop: :) Jun 17 19:31:57 i suspect we can do without the .lai Jun 17 19:32:02 mwester: I read it as: mess around with dependency_libs and installed sections, in an obscure way. Jun 17 19:32:14 koen|770: Do you know what the .lai is, anyway? Jun 17 19:32:51 I've googling for a while, and only got that perhaps it stands for 'inconditional' Jun 17 19:32:57 arcane libtool magic Jun 17 19:33:26 * sirfred prays: deliver us from the libtool evils Jun 17 19:34:48 koen no Jun 17 19:34:55 koen havent yet the time Jun 17 19:34:58 maybee next week Jun 17 19:35:06 when my fpga course is over Jun 17 19:36:19 koen|770: I had the hope that perhaps they wanted to say 'if [ -f "$lafile"' instead of '[ -f "$dotlai"' Jun 17 19:36:39 koen|770: Do you find that reasonable? Jun 17 19:37:09 yes, but i'm not really a libtool literate Jun 17 19:38:42 koen|770: Mental sanity is important Jun 17 19:39:33 koen|770: Perhaps asking in the maillist could help Jun 17 19:44:24 koen|770: Anyway, I've checked the staged libgstreamer-0.10.la file and it doesn't contain any reference to the workdir, only to the staging dir. Jun 17 19:44:55 hmm Jun 17 19:45:21 I think the QA alert is caused by the installed=yes, only. Jun 17 19:45:50 gst-plugins-base didn't build because it looked in workdir for gstreamer .la files Jun 17 19:46:52 koen|770: I'm not aware of that problem. gst-plugins-base built fine for me. Anyway, I will test again. Jun 17 19:47:30 i noticed because i use rm_work Jun 17 19:47:38 koen|770: Ah, ok. Jun 17 19:49:19 I feel like walking on mud, dealing with these dark libtool problems Jun 17 19:51:36 i know the feeling Jun 17 19:55:02 sirfred : heh even bitbake knows libtool is evil, yesterday when doing a build libtool got task id 666 Jun 17 19:55:28 steliosk: I remember you said that. Scaring. ;-) Jun 17 19:59:14 03polyonymous 07org.oe.dev * r11b0e95a... 10/ (6 files in 5 dirs): Jun 17 19:59:14 packages/simh: added a The Computer History Simulation Project package. Jun 17 19:59:14 it's nice to have the power of pdp-11 in your pocket, isn't it? Jun 17 19:59:34 heh Jun 17 20:00:24 ;-) Jun 17 20:07:12 ljp, may I borrow you for a while? :b Jun 17 20:14:27 hi mr_nice Jun 17 20:14:59 hm Jun 17 20:15:00 oh Jun 17 20:15:03 reconnect Jun 17 20:18:38 polyonymous: Did you verify successul compilation for uicmoc-native when upgrading to 4.30? Jun 17 20:18:45 It fails for me Jun 17 20:18:48 4.3.0 Jun 17 20:19:12 Laibsch, if I touched it, then I verified it. Jun 17 20:19:30 Well, maybe only collie is affected. Jun 17 20:19:33 I will try Jun 17 20:20:02 I'll try now Jun 17 20:20:32 Yes, it's been already built Jun 17 20:20:45 So feel free to report it in all the glory Jun 17 20:21:11 I'll try with you first ;-) Jun 17 20:21:48 report to me also counts :) Jun 17 20:22:18 does http://paste.stgraber.org/1643 look like anything you'd recognize? Jun 17 20:24:17 hmm Jun 17 20:25:13 wish I could reproduce it.. Jun 17 20:25:42 oh wait Jun 17 20:25:54 I've just teste uicmoc-native, not uicmoc4-native. Jun 17 20:26:04 Lemme see if I have the same problem Jun 17 20:26:08 OK Jun 17 20:26:30 * chouimat is playing with his indentation randomizer ... Jun 17 20:27:27 It's almost clear what's wrong with your build, let's hope it will be the case for my build as well Jun 17 20:29:01 I am sure we will find out Jun 17 20:32:20 NOTE: build 200706172224: completed Jun 17 20:32:28 can you paste your log.do_configure ? Jun 17 20:54:21 Laibsch, can you paste your log.do_configure ? (I take it you got disconnected). My build succeeded. Jun 17 20:56:16 Good night. Jun 17 20:58:36 polyonymous: OK Jun 17 21:03:58 polyonymous: It is about 250K in size Jun 17 21:04:06 I will bzip it and send it to you Jun 17 21:08:33 re Jun 17 21:09:41 hi florian Jun 17 21:12:04 polyonymous: Can you please accept the file? Jun 17 21:12:46 polyonymous: or just get http://oz.leggewie.org/wip/log.do_configure.22728.bz2 Jun 17 21:13:11 Laibsch, sorry, I was away. Jun 17 21:18:53 Laibsch, so far I have no idea what went wrong... Jun 17 21:20:17 Laibsch, I think you should file a bug attaching compile log and maybe we'll get to fix it... Jun 17 21:20:51 OK Jun 17 21:23:01 * * OE Bug 1981 has been RESOLVED (INVALID) by Jun 17 21:23:03 * *  gtk-doc-native fails sanity test Jun 17 21:23:05 * * http://bugs.openembedded.org/show_bug.cgi?id=1981 Jun 17 21:23:49 koen: you exclude bugs marked as invalid from your weekly bug report, right? Jun 17 21:24:36 that will skew the numbers when comparing with number of opened bugs since a portion of those will eventually become invalid Jun 17 21:29:01 * * OE Bug 2511 has been created by  Jun 17 21:29:03 * * uicmoc4-native fails do_compile after recent update to version 4.3.0 Jun 17 21:29:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2511 Jun 17 21:29:52 polyonymous: do you add the old bb file by default? Jun 17 21:30:07 Laibsch, what do you mean "by default"? Jun 17 21:31:40 I saw that you added the old uicmoc4-native version with your upgrade Jun 17 21:31:53 Laibsch, I did an mtn mv and mtn add Jun 17 21:31:57 I was wondering if you always do that, "by deafult" Jun 17 21:32:14 Well, yes, unless I'm tired and forget to :) Jun 17 21:32:31 I don't think that is generally necessary Jun 17 21:32:40 Let the old stuff die, I'd say Jun 17 21:32:50 Unless there is a good reason not to Jun 17 21:33:13 Well, in this case it's a big tricky thing so I wanted to keep old stuff for reference for a while Jun 17 21:33:28 This case I think it made sense Jun 17 21:33:37 Seeing that the new version has problems Jun 17 21:33:46 That is why I asked about "default" Jun 17 21:33:47 But in general I prefer to have the old stuff around for a while unless it's really trivial. Jun 17 21:34:01 In the default case I think it makes more sense to remove it Jun 17 21:34:27 polyonymous: The old stuff is never further away than "mtn cat $file -r $rev" Jun 17 21:34:29 Well, I never considered any "default" behaviour, normally I make decision at the time of doing it. Jun 17 21:34:35 that is what we have the SCM for Jun 17 21:34:44 Laibsch, that's quite far away ;-) Jun 17 21:34:51 no Jun 17 21:34:59 Let's not clutter things up Jun 17 21:35:04 It means more work Jun 17 21:35:06 I rarely can produce -r $rev out of my head :) Jun 17 21:35:09 MUCH, MUCH more work Jun 17 21:35:21 why is it more work? Jun 17 21:35:27 polyonymous: that is what mtn is also for Jun 17 21:35:33 It can tell you the rev Jun 17 21:36:06 Somebody else will have to come around and wonder "can I remove it", he will ask around, make an entry in removal.txt, send an RFC Jun 17 21:36:10 I know what mtn is for, but you're talking about much more work than what so called "clutter" would give :) Jun 17 21:36:19 all worthless, dumb, unnecessary work Jun 17 21:36:35 Believe, I have cleaned up a lot of the clutter Jun 17 21:36:47 It is a f+cking lot of work Jun 17 21:36:58 And very hard and frustrating, because it is senseles Jun 17 21:36:59 Well, creating clutter wasn't much easier! :) Jun 17 21:37:16 You might remember a couple of days you don't need the old version Jun 17 21:37:23 "real men don't write documentation: if it was hard to write it should be hard to understand" :) Jun 17 21:37:26 But even you will have forgotten in two weeks Jun 17 21:37:32 Nevermind somebody else Jun 17 21:37:45 Well, again, it depends. Jun 17 21:37:52 well, let's take this humorous Jun 17 21:37:53 Trivial bumps deserve no preservation. Jun 17 21:38:00 But I hope my point was received Jun 17 21:38:11 If you bump a widely used dependency it's not exactly the same. Jun 17 21:38:16 polyonymous: Then delete them immediately Jun 17 21:39:01 Yes, trivial bumps are simply mtn mv with no add, correct. But the triviality is not just how trivial are changes in .bb. It's a matter of affected part of the tree as well. Jun 17 21:39:35 Sure Jun 17 21:39:47 I am not saying the old stuff should never be kept Jun 17 21:39:49 But generally, I wouldn't talk about strict rules here - we're not computers, we can make intelligent decisions. And stupid too (computers aren't smart enough to make stupid decisions, are they?) Jun 17 21:40:11 polyonymous: I just please and again ask you to consider the person after you Jun 17 21:40:35 And yes, when I'm talking about no strict rules, I'm not only against "never add", but against "always add" as well. Jun 17 21:40:39 It is easier to add something if something says "oh I need version X" than cleaning the shit up Jun 17 21:40:50 if sombody Jun 17 21:40:56 somebody, eve Jun 17 21:40:59 even Jun 17 21:41:01 ;-) Jun 17 21:41:03 I'm always trying to think a couple of step further. Jun 17 21:41:07 hmm, too late Jun 17 21:41:29 Take this from another angle Jun 17 21:41:52 Time management books are always clear about "try to throw stuff away", it is a burden Jun 17 21:42:13 Here you have it easier since you are never really throwing stuff away Jun 17 21:42:17 You can always restore it Jun 17 21:42:40 I'm not reading time management books and I'm not going to either :) Jun 17 21:43:39 Laibsch, if you have a spitz build setup, check if this uicmoc is a collie problem or not whenever you get time. Jun 17 21:43:53 IIRC it failed for both Jun 17 21:44:06 And IIRC 4.2.3 built fine Jun 17 21:44:12 again for both Jun 17 21:44:25 but I will test and report more clearly Jun 17 21:44:26 That would mean I won't be able to reproduce it by building for collie :( Jun 17 21:44:46 I think 4.2.3 never built fine. Well, uicmoc4-native, maybe, but not the whole thing. Jun 17 21:44:48 clean out tmp Jun 17 21:45:01 My suspicion is "undeclared dependency Jun 17 21:45:23 Doesn't look like it, but I can do an overnight build. Jun 17 21:48:07 Laibsch, try adding -no-pch to uicmoc4 configure? Jun 17 21:50:25 strangely enough I have PCH support on and NO -include .pch/release-static/QtCode.... switch Jun 17 21:51:24 Can I add that in the bb or do I need to interrupt bitbake after -c configure? Jun 17 21:51:45 Laibsch, that's on EXTRA_OECONF Jun 17 21:51:48 in Jun 17 21:52:07 and as for missing dep, do you have qmake2 built? Jun 17 21:52:18 qmake2-native that is Jun 17 21:57:32 Ich habe qmake-native Jun 17 21:57:39 aber kein qmake2-native Jun 17 21:57:44 Ist es das? Jun 17 21:57:48 try building qmake2-native? Jun 17 21:57:57 hehe Jun 17 21:58:14 I do have both and maybe das ist es, I don't know. Jun 17 21:58:22 lol Jun 17 21:58:36 Oh man, I was chatting with florian Jun 17 21:58:43 Forgot to push the switch Jun 17 21:58:56 I should be speaking German with you guys, actually, because I'm learning it and I need practice :) Jun 17 21:59:03 4.2.3 built both fine Jun 17 21:59:03 hm qmake-native comes from qt3? Jun 17 21:59:11 for collie and spitz Jun 17 21:59:12 But I'm afraid of miscommunicating things. Jun 17 21:59:20 now, onto qmake2 Jun 17 21:59:21 woglinde, if not qt2 :) Jun 17 21:59:26 and qmake2-native from qt-embeded Jun 17 21:59:30 2 Jun 17 21:59:30 no Jun 17 21:59:36 qmake2 is for qt4 Jun 17 21:59:55 qmake2 signifies version of qmake, not qt. Jun 17 21:59:56 uh strange Jun 17 22:00:05 that suckz Jun 17 22:00:10 that's life Jun 17 22:00:47 so, Laibsch try building qmake2-native, not sure if that's it, but let's see... Jun 17 22:00:50 hm fpga is crazy Jun 17 22:00:59 1 hour for the binary file Jun 17 22:01:19 and if that doesn't help, I'll push my changes in qmake2 bb which are actually for other dependency, don'trememberwhich, but may be of help here. Jun 17 22:01:25 what's fpga? Jun 17 22:01:48 I was looking in ../tmp/angstrom/work/i686-linux/ Jun 17 22:02:00 But there is no qmake2-native* dir even after I built it Jun 17 22:02:15 field programming arrays Jun 17 22:02:16 that would mean you don't have it built :) Jun 17 22:02:22 try bitbake qmake2-native Jun 17 22:02:30 that is what I did Jun 17 22:02:31 and then rebake uicmoc4-native ? Jun 17 22:02:36 ah Jun 17 22:02:38 uhh Jun 17 22:03:08 okay I am going to sleep now Jun 17 22:03:10 have fun Jun 17 22:03:22 Laibsch, neither I do have it :)) Jun 17 22:03:43 weird :) Jun 17 22:04:34 but wait, that may also mean that I do not have qmake2 built now. Jun 17 22:05:03 yes, I don't have qmake2 built, so it's not it. Jun 17 22:06:18 I don't have this -include .pch/release-static/QtCore switch I see in your compile log Jun 17 22:25:58 4.2.3 fails for the spitz Jun 17 22:26:05 bs Jun 17 22:26:12 4.2.3 work for spitz and collie Jun 17 22:26:18 4.3.0 fails for spitz Jun 17 22:26:29 and I am again verifying failure for 4.3.0 for collie now Jun 17 23:12:05 owell, I'll look into it again tomorrow, it's time to sleep now Jun 17 23:13:55 polyonymous: qt4-x11-free fails, too Jun 17 23:14:21 Laibsch, sounds like a total system screwup to me... But we'll look at it tomorrow... Jun 17 23:14:27 sure Jun 17 23:14:35 Just wanted to let you know Jun 17 23:14:43 so that I don't sleep well :)) Jun 17 23:14:47 So you can dream up a solution while zzz'ing Jun 17 23:14:51 heh Jun 17 23:14:54 Exactly;-) Jun 17 23:14:54 Well, goodnight. Jun 17 23:14:58 Goodnight Jun 18 02:29:10 03lenehan 07org.oe.dev * rf8ed821b... 10/ (4 files in 2 dirs): Jun 18 02:29:10 iproute2: Use update-alternatives. Note that this moves the ip command from Jun 18 02:29:10 /sbin to /bin, to match where busybox was putting the ip command. Jun 18 02:59:47 03lenehan 07org.oe.dev * rbdfe6dfd... 10/ (1 packages/iproute2/iproute2_2.6.20.bb): Jun 18 02:59:47 iproute2 2.6.20: Add the 2.6.20 bb file. It seems when I added 2.6.20 I Jun 18 02:59:47 didn't mtn add the .bb file, so only the patches got included. I notice this Jun 18 02:59:47 when the PR change wasn't included as part of the previous commit. **** ENDING LOGGING AT Mon Jun 18 02:59:57 2007