**** BEGIN LOGGING AT Fri Mar 23 02:59:56 2007 Mar 23 04:56:48 Having problems compiling opie-recorder... i get cannot convert QValueListIterator to const char* for argument 1 to int remove(const char*) Mar 23 04:56:50 | make: *** [xml/qxml.o] Error 1 Mar 23 04:56:52 | make: Leaving directory `/oe/tmp/work/uicmoc-native-2.3.10-r0/qt-2.3.10/src' Mar 23 04:56:53 | FATAL: oe_runmake failed Mar 23 04:56:54 NOTE: Task failed: /oe/tmp/work/uicmoc-native-2.3.10-r0/temp/log.do_compile.20903 Mar 23 04:56:56 NOTE: package uicmoc-native-2.3.10-r0: task do_compile: failed Mar 23 04:56:58 Any suggestions? Mar 23 04:57:03 Oh, I'm using the oz345fam083 branch Mar 23 05:01:16 Glombool: any way you could use the .dev branch? Mar 23 05:03:10 That works, but the version needs newer packages than what I have on my Z Mar 23 05:03:37 I suppose I could build a feed with all the packages compiled... Mar 23 05:06:45 hvontres|home would it work if I copied the openzaurus-3.4.5.conf from the oz345fam083 branch and used that as the target distro? Mar 23 05:09:51 Glombool: not sure. What are you trying to do? Mar 23 05:11:06 I'm hacking the source of opie-recorder so it will work properly on my collie. I am running oz3.5.4. The dev branch has a conf for oz-3.5.5... don't know if that will work properly. Mar 23 05:11:56 Last time I compiled with dev branch I used openzaurus-unstable which may have been the problem. Mar 23 05:17:14 It's kind of annoying because I actually NEED to be able to record audio :( Mar 23 06:21:32 03rwhitby 07org.oe.dev * r14b470b8... 10/ (1 packages/yeaphone/yeaphone_0.1.bb): yeaphone: Added rrecommends for required kernel modules Mar 23 06:21:41 03mreimer 07org.oe.dev * r41b8e1c2... 10/ (3 files in 2 dirs): python-gstreamer: Rename to py-gst. Mar 23 06:21:51 03rwhitby 07org.oe.dev * r012918bd... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages: Added gspcav1, w3cam and webcam-server Mar 23 06:21:59 03rwhitby 07org.oe.dev * r6dc8c4bd... 10/ (4 files in 3 dirs): puppy: Added patch to handle usb header location change in recent kernels Mar 23 06:22:06 03rwhitby 07org.oe.dev * r64bce8eb... 10/ (1 packages/yeaphone/yeaphone_0.1.bb): yeaphone: Added linephonec rrecommends Mar 23 06:22:11 03rwhitby 07org.oe.dev * r11b2a592... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages: Demoted mysql and packages that depend on it. Promoted puppy and ftpd-topfield. Mar 23 06:22:16 wb CIA-11 Mar 23 06:50:13 03rwhitby 07org.oe.dev * rd774ef44... 10/ (3 files in 3 dirs): ixp4xx-kernel: Updated to latest SVN defconfig, which enables a whole lot more netfilter modules. Mar 23 07:25:27 good morning all Mar 23 07:33:50 koen: morning :) Mar 23 07:34:07 morning all !!! Mar 23 07:34:23 ping pb_ Mar 23 07:48:46 morning all Mar 23 08:08:57 dang Mar 23 08:12:33 good morning everyone Mar 23 08:13:14 Hi Laibsc1 Mar 23 08:40:29 ~botmail for zecke: your bugzilla script seems to create a lot of duplicates, see #304, #305 ... #310 in moko bugzilla Mar 23 08:48:43 rwhitby: Does the RDEPENDS to the kernel packages work ? I am using it also in a custom package and does not seem to have any effect when you try to install it Mar 23 08:49:09 03koen 07org.oe.dev * r127d2540... 10/ (1 conf/distro/angstrom-2007.1.conf): Angstrom: ship avahi daemon by default (http://www.hrw.one.pl/2007/03/23/avahi-looks-interesting/) and some small updates Mar 23 08:50:23 koen: ping? Could you look at http://paste.uni.cc/13896 ? Mar 23 08:52:38 NAiL: I suspect insane.bbclass doesn't know about uclibcgnueabi Mar 23 08:52:47 that, or bitbake has a bug Mar 23 08:55:06 koen: great, one less dependency in local.conf Mar 23 08:59:33 koen: appears to be working if I remove insane Mar 23 09:00:24 probably none has build uclibc using insane Mar 23 09:02:27 is there any port planned or going on for samsung i320 smartphone ? Mar 23 09:04:58 morning Mar 23 09:05:52 hey XorA Mar 23 09:06:11 Ifaistos: I'm still waiting for the .29 release before I'm going to look at uclibc Mar 23 09:06:52 XorA : hi Mar 23 09:08:57 .28 works pretty stable, haven't met any major issue Mar 23 09:09:35 Ifaistos: .28 doesn't support EABI on arm :( Mar 23 09:09:59 ok, thats an issue Mar 23 09:09:59 koen: is there a fix in svn? Mar 23 09:10:10 yes, svn works great Mar 23 09:10:17 * NAiL notes that svn from a couple of days back atleast has EABI as an option Mar 23 09:10:27 some just needs to do 'make dist' in svn Mar 23 09:10:51 heh Mar 23 09:10:58 I get cross compile badness Mar 23 09:11:20 and then internal compiler error :P Mar 23 09:13:36 heh Mar 23 09:13:45 I would've expect uclibc to be crosscompile sage Mar 23 09:13:50 safe* Mar 23 09:17:05 yeah, I would too, strangely enough Mar 23 09:19:32 well, RMK found out that the memcopy code in uclibc for arm is severely broken, so I guess we're in for more surprises Mar 23 09:19:38 btw, compiling uclibc with linux-libc-headers that doesn't have UTS_RELEASE in version.h fails (UTS_RELEASE was moved to utsversion.h quite some time ago). Is it an OK hack to just echo "#define UTS_RELEASE ${PV}" >> version.h Mar 23 09:19:46 ? Mar 23 09:19:56 (in linux-libc-headers, of course) Mar 23 09:21:42 UTS_RELEASE was moved out of version.h to stop unneccessary recompiles of large parts of the kernel when it was changed (eg. by incrementing git count) Mar 23 09:22:11 but since the headers are static, this should be OK right? :-) Mar 23 09:22:15 NAiL: I guess so Mar 23 09:28:56 morning Mar 23 09:29:10 hey hrw Mar 23 09:29:51 good morning hrw Mar 23 09:32:32 ah. finally fixed this darn bug: http://bugs.openembedded.org/show_bug.cgi?id=1593 :) Mar 23 09:32:41 !oebug 1593 Mar 23 09:32:42 * * Bug 1593, Status: NEW, Created: 2006-11-16 05:47 Mar 23 09:32:43 * * dev(AT)settopsolutions.com: GDB 6.4 fails - CFLAGS has changed since the previous run? Mar 23 09:32:44 * * http://bugs.openembedded.org/show_bug.cgi?id=1593 Mar 23 09:33:10 thats the one Mar 23 09:34:36 hey florian Mar 23 09:36:49 hi all Mar 23 09:39:07 koen: thx for avahi Mar 23 09:40:30 + usb0 IPv4 Remote Terminal on fic-gta01 SSH Remote Terminal local Mar 23 09:40:30 + eth0 IPv4 Powerbook SSH Remote Terminal local Mar 23 09:40:31 + eth0 IPv4 Remote Terminal on h2200 SSH Remote Terminal local Mar 23 09:40:31 + eth0 IPv4 Remote Terminal on hx4700 SSH Remote Terminal local Mar 23 09:40:31 :D Mar 23 09:41:02 3 toys on desk are powered off Mar 23 09:45:29 hrw: any plan to apply your internal-lib patch soon? Mar 23 09:48:20 yes Mar 23 09:49:30 cool Mar 23 10:08:29 first some other updates Mar 23 10:09:56 03hrw 07org.oe.dev * racb6fba6... 10/ (1 packages/linux/linux-omap1-2.6.18-omap1/defconfig): linux-omap 2.6.18: refreshed defconfig to get it build on RHEL4 (no changes) Mar 23 10:10:44 ~lart jffs2 for not working with nfs Mar 23 10:10:45 * ibot takes a large goose feather pillow and swings it wildly in jffs2's direction, hitting jffs2 and sending jffs2 flying into the closet for not working with nfs Mar 23 10:23:36 Crofton is sleeping... will have to catch him during day and give one image to test on omap5912osk Mar 23 10:26:39 I wonder if he had some succes with getting boards from TI for OE Mar 23 10:28:27 koen: during ELC greet two people from me - Tim Bird and Mar 23 10:28:29 Matthew Locke Mar 23 10:29:00 if this will be no problem Mar 23 10:29:02 will do Mar 23 10:29:08 thx Mar 23 10:29:19 Matthew is doing a presentation on creating your own distro Mar 23 10:29:39 "cp angstrom.conf myleetdistro.conf" Mar 23 10:29:57 or "cat money > hrw" :) Mar 23 10:29:57 03pH5 07org.oe.dev * rfe04f2a0... 10/ (3 files in 2 dirs): linux-handhelds-2.6: add 2.6.20-hh3, update cvs HHV to -hh3. Mar 23 10:30:24 koen: celinux-test distro is rather copy of openzaurus/openomap then angstrom Mar 23 10:30:51 that reminds me... Mar 23 10:31:07 I need to talk to crofton about merging openomap into angstrom Mar 23 10:32:35 koen: iirc crofton moved to angstrom Mar 23 10:32:49 * hrw rebuild python-pygtk2 again Mar 23 10:32:53 hi pvanhoof Mar 23 10:33:12 hey hrw Mar 23 10:41:26 hi, all! Mar 23 10:41:31 !oebug 2033 Mar 23 10:41:32 * * Bug 2033, Status: NEW, Created: 2007-03-23 03:46 Mar 23 10:41:33 * * slapinid(AT)gmail.com: No proper PREFERRED_PROVIDER for gtk-pixbuf-*, build failed with bitbake 1.8 Mar 23 10:41:34 * * http://bugs.openembedded.org/show_bug.cgi?id=2033 Mar 23 10:41:48 any ideas? Mar 23 10:44:56 03koen 07org.oe.dev * re3b3e820... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: prefer dbus-glib 0.72 Mar 23 10:48:15 03koen 07org.oe.dev * r5071c48a... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: some more preferred providers Mar 23 10:54:09 slapin: a temporary workaround is removing the /home/slapin/OE/org.openembedded.dev/packages/gtk+/gtk+-directfb_2.10.9.bb Mar 23 10:54:28 you will have to revert it later again when you want to update your tree however. Mar 23 10:55:05 There was another solution involving editing some bitbake files, but I don't remember that. Mar 23 10:58:17 03hrw 07org.oe.dev * rd92df12b... 10/ (5 files in 3 dirs): Mar 23 10:58:17 python-pygtk2: make it build on 64bit host for 32bit machine (from Poky) Mar 23 10:58:17 Informations why it is needed sent to mailing list. Mar 23 11:00:56 mickey|zzZZzz: read my mail about it and answer please Mar 23 11:10:49 03hrw 07org.oe.dev * r263476da... 10/ (5 files in 3 dirs): python-gst: updated to 0.10.7, added fix to get it built on 64bit (from Poky) Mar 23 11:11:03 03hrw 07org.oe.dev * r27da4e71... 10/ (5 files in 3 dirs): python-pygobject: updated to 2.12.3, added fix to get it built on 64bit (from Poky) Mar 23 11:19:00 03koen 07org.oe.dev * rbaa52bc7... 10/ (3 files in 3 dirs): openmoko-session: start openmoko-mainmenu plugin by default Mar 23 11:31:57 ~seen CosmicPenguin Mar 23 11:32:01 doh! Mar 23 11:32:11 how do I do that? Mar 23 11:32:24 cosmicpenguin is currently on #edev (20h 8m 9s). Has said a total of 9 messages. Is idling for 13h 2m 1s, last said: 'you'll have to use the same libc that you are using for your base system'. Mar 23 11:37:53 slapin, koen: Isn't that an effect of the commit from JustinP to fix bug 2029? Mar 23 11:38:29 Laibsch: what is? Mar 23 11:39:30 bug 2033 is an effect of the work for bug 2029? Mar 23 11:40:20 !oebug 2033 Mar 23 11:40:21 * * Bug 2033, Status: NEW, Created: 2007-03-23 03:46 Mar 23 11:40:22 * * slapinid(AT)gmail.com: No proper PREFERRED_PROVIDER for gtk-pixbuf-*, build failed with bitbake 1.8 Mar 23 11:40:23 * * http://bugs.openembedded.org/show_bug.cgi?id=2033 Mar 23 11:40:30 !oebug 2029 Mar 23 11:40:30 Laibsch: no, its a side-effect of adding gtk-directfb Mar 23 11:40:31 * * Bug 2029, Status: RESOLVED, Created: 2007-03-21 13:05 Mar 23 11:40:32 * * papercrane(AT)reversefold.com: libgnomeui-2.16.1 requires gdk-pixbuf-2.0 >= 2.9.0 which is not built Mar 23 11:40:33 * * http://bugs.openembedded.org/show_bug.cgi?id=2029 Mar 23 11:40:54 for the FUDers we no longer need X :-) Mar 23 11:44:47 hrw, I am up Mar 23 11:44:59 Laibsch: http://bugzilla.gnome.org/show_bug.cgi?id=417446 Mar 23 11:50:17 Crofton: can you test kernel/image on omap5912osk? Mar 23 11:50:27 in a few hours Mar 23 11:50:30 thx Mar 23 11:50:38 I will put it online then Mar 23 11:50:42 ok Mar 23 11:50:58 where does it come from? Mar 23 11:51:18 I can check booting, there will be some barfing because fs image won't match Mar 23 11:51:27 celinuxtest build Mar 23 11:51:38 ok Mar 23 11:51:47 do you have kernel and image? Mar 23 11:51:51 yes Mar 23 11:51:55 ok Mar 23 11:52:09 Why not se Angstrom? Mar 23 11:52:16 http://www.openembedded.org/~hrw/omap/ Mar 23 11:52:30 Crofton: its for celf - they are conservative a bit. Mar 23 11:52:40 heh Mar 23 11:52:45 koen: will you do Ångström propaganda during ELC? Mar 23 11:52:53 The customer is always right :) Mar 23 11:53:09 hrw: probably Mar 23 11:53:25 s/conservative/obsolete/ Mar 23 11:53:45 (I suspect MV doesn't support EABI on 2.4 and 2.6.10) Mar 23 11:54:00 koen: oabi with gcc 3.4.4 works nice Mar 23 11:54:14 http://www.hrw.one.pl/2007/03/23/its-time-to-buy-laptop/ Mar 23 11:54:23 but you can use any fpu or iwmmx Mar 23 11:54:50 koen: if you build only for armv5 then softfloat is all you need Mar 23 11:55:05 Crofton: copied files. Mar 23 11:55:26 ok Mar 23 11:55:34 Iam dling to work machine now Mar 23 11:56:50 ok, got them. Will flash when I mahe it across the street Mar 23 11:57:52 morning all Mar 23 11:59:24 hi Dirk Mar 23 11:59:30 hi Marcin Mar 23 12:03:05 anyone saw perl-native failing on makedepend? Mar 23 12:09:28 not me Mar 23 12:10:17 it fails on feisty :} Mar 23 12:10:42 zecke: you can't configure debian too? Mar 23 12:10:55 hrw: right, and woglinde seldomly helps Mar 23 12:12:36 any ideas? http://rafb.net/p/DE0Ew283.html Mar 23 12:12:36 I'm trying to add my own config files to the new alsa-state package... I can get it to pick up my file by putting it in alsa-state/DISTRO/asound.state Mar 23 12:12:57 but I want machine specific files for my distro... so was trying alsa-state/DISTRO/asound.state_MACHINE Mar 23 12:13:07 alsa-state/MACHINE/ ? Mar 23 12:14:26 zecke: but I want it by DISTRO & MACHINE Mar 23 12:15:47 and alsa-state/MACHINE/ didn't work Mar 23 12:16:36 tkp: why do you want it by distro? Mar 23 12:16:50 tkp: the audio quality depends on the hardware Mar 23 12:17:10 zecke: It's not about audio quality... it's about setting initial sound levels Mar 23 12:17:49 tkp: see, if you say state/MACHINE didn't work we don't need to talk Mar 23 12:17:52 alsa-state/ will work Mar 23 12:18:07 we are going to move from DISTRO specific machine features to MACHINE features Mar 23 12:18:09 mickey|busy: did you read my python mail on oeml? Mar 23 12:18:17 so please don't introduce more of those Mar 23 12:18:30 mickey|busy: that's what I though... alsa-state/DISTRO works... alsa-state/MACHINE does not Mar 23 12:18:31 hrw: briefly, i'll reply when i have a chance, thans Mar 23 12:18:38 thx Mar 23 12:19:19 mickey|busy: why move away from ISTRO specific machine features? Surely they are appropriate sometimes Mar 23 12:20:02 "sometimes", yes. Mar 23 12:20:15 "seldomly" is more accurate Mar 23 12:20:26 yeah Mar 23 12:20:32 * koen is guilty of abusing distro overrides as well Mar 23 12:20:44 most of the times, it's annoying having other distro's using the same machine replicating your efforts Mar 23 12:20:55 mickey|busy: and in this case, you do not consider it appropriate? I mean, I may want my alsa levels set differently to someone elses distro - and I do. Mar 23 12:21:27 feel free to do it Mar 23 12:21:33 well... saying that... at the moment, my conf is the only one in there! Mar 23 12:21:38 but using the alsa-state package seems not correct Mar 23 12:21:51 state's contain usage cases Mar 23 12:21:55 primaril Mar 23 12:21:56 y Mar 23 12:22:32 mickey|busy: surely that's exactly what alsa-state is for. It is an initscript that sets sound levels on boot Mar 23 12:23:01 ok. Mar 23 12:23:03 just go ahead Mar 23 12:23:08 i don't have time for that now Mar 23 12:23:22 sorry Mar 23 12:23:30 we'll clean up afterwards Mar 23 12:23:40 or not :) Mar 23 12:23:44 that wasn't my issue anyway... the issue is that it's not picking up machine specific confgs Mar 23 12:24:00 grrr, my powerbook is emulating a yet engine again Mar 23 12:24:05 mickey|busy, sounds like he has too much work, Mar 23 12:24:23 Crofton: any news from TI? Mar 23 12:24:29 nope Mar 23 12:24:47 I sent rishi an email saying hello, haen't heard from him Mar 23 12:25:13 right, it takes a few days for him to respond Mar 23 12:30:00 * koen is trying to convince the uni to do a project on blackfins and omaps Mar 23 12:30:10 cool Mar 23 12:30:22 blackfin only does uclinux Mar 23 12:31:26 ok. was looking in the wrong place... If bb finds a machine specific file, does it automatically set PACKAGE_ARCH =MACHINE_ARCH? Mar 23 12:31:39 it should Mar 23 12:31:47 ah, cool. nice :) Mar 23 12:33:37 blackfin is atmel? Mar 23 12:34:19 hrw: AD Mar 23 12:34:42 koen: what do you think of it? :) we have a lot of discussions in our company about the use of blackfin Mar 23 12:34:48 I do not like the uclinux idea Mar 23 12:34:54 Jin^eLD: haven't used it yet Mar 23 12:35:02 Jin^eLD: iirc likewise has a few Mar 23 12:35:38 I think the development gets more complicated with uclinux.. but I do not have any real experience with it Mar 23 12:35:44 hi greentux Mar 23 12:36:13 hi hrw... munich airport ;) Mar 23 12:38:46 zecke: next commit if for you Mar 23 12:41:58 03koen 07org.oe.dev * r7c524374... 10/ (1 packages/openmoko-apps/openmoko-rssreader_svn.bb): openmoko-rssreader: add svn version Mar 23 12:47:43 is there a stable version of openembedded ? one without any bugs Mar 23 12:51:06 heh Mar 23 12:51:40 you are either very brave or have your flame-suit on Mar 23 12:51:45 or both Mar 23 12:52:04 seriously though, the answer is: no. stable snapshots are on the agenda for Summer 2007 Mar 23 12:52:38 mickey|busy: any quick thoughts on the OE-ng rfc? Mar 23 12:52:45 koen: sorry, not yet Mar 23 12:52:58 koen, hrw: could you have a look at http://rafb.net/p/rOHeJx35.html? it should run avahi-autoipd on dhcp interfaces if busybox udhpc fails to acquire a lease. Mar 23 12:53:07 empty_mind: You could try poky if you want something that might "just work" Mar 23 12:53:10 ~poky Mar 23 12:53:16 koen: I'm drowning between my HRI project and OpenMoko. in 7 days it'll be over and I have more breath Mar 23 12:53:23 please bear with me until then Mar 23 12:53:23 hmm... poky is http://projects.o-hand.com/poky, or a subset of OpenEmbedded, created by o-hand.com to testdrive some of their technology Mar 23 12:53:40 mickey|busy: ok Mar 23 12:54:00 mickey|busy: I only had time this morning because my lab partner had a hangover :) Mar 23 12:54:06 hehe Mar 23 12:54:10 poor boy Mar 23 12:54:11 koen: on a related note... i'm going to visit pvanhoof in the week from 16-21 April Mar 23 12:54:19 koen: I'd like to take this chance to meet you as well Mar 23 12:54:22 would that work for you? Mar 23 12:54:33 mickey|busy: ELC2007 is then Mar 23 12:54:42 right :( Mar 23 12:54:44 03koen 07org.oe.dev * re1133544... 10/ (1 packages/openmoko-apps/openmoko-rssreader_svn.bb): openmoko-rss-reader: fix description Mar 23 12:54:46 bummer, means we need to adjust that schedule Mar 23 12:54:51 thanks for noticing, hrw Mar 23 12:55:03 pH5: no idea now Mar 23 12:55:04 RF i need to get rid of this bug No such file or directory: '/home/orion/oss/unknown/angstrom/org.openembedded.dev/packages/gtk+/files/./no-xwc.patch' Mar 23 12:55:08 * koen|lab moves to the lab Mar 23 12:55:12 pH5: nice idea Mar 23 12:55:18 empty_mind: two solutions Mar 23 12:55:20 empty_mind: mtn up Mar 23 12:55:24 a) DON'T USE .DEV !!!!!!!!!!!! Mar 23 12:55:28 * hrw started build of all python* to check how much is broken Mar 23 12:55:32 b) remove gtk+-directfb.bb Mar 23 12:55:42 mickey|busy: c) mtn up, bug fixed already Mar 23 12:55:43 aha, second one sounds good Mar 23 12:55:57 koen|away: k. i have added a bitbake bug on that Mar 23 12:56:07 i fear something is wrong wrt. PACKAGES_DYNAMIC and DEFAULT_PREFERENCE Mar 23 12:56:18 there is Mar 23 12:57:08 mickey|busy, and any solution to libx11 compilation error, it says eith xdcmo not found oor xau.pc not found while compiling openmoko Mar 23 12:57:12 mickey|busy : Any chance to get an openmoko dev kit ? Mar 23 12:57:44 Ifaistos: we're out of P0 devices. P1 devices hopefully ship in 2 weeks Mar 23 12:57:50 please get back to me wrt. dev discount then Mar 23 12:57:51 empty_mind: bitbake libxau libxdcmp ;bitbake libx11 Mar 23 12:58:20 * hrw wonder does P1 will get 512M card or 128M one Mar 23 12:58:50 that's where we reduce BOM Mar 23 12:58:51 mickey|busy : thanks Mar 23 12:58:52 *cough* Mar 23 12:59:07 * mickey|busy chuckles given the price difference of a 512M and 128M Mar 23 12:59:16 seriously though, i wanted to have 512M since long Mar 23 12:59:49 mickey|busy: I plan to wait with buying 1G card. if p1 will get 512M then I will not buy larger one Mar 23 12:59:59 ~lart nokia for rs-mmc in 770... Mar 23 13:00:00 * ibot rm -rf's nokia for rs-mmc in 770... Mar 23 13:00:02 hrw: I'm afraid it won't Mar 23 13:00:05 please go and buy one :) Mar 23 13:00:18 i can tell you for sure next week Mar 23 13:00:34 bigger rs-mmc are triple priced compared to microsd/sd/minisd Mar 23 13:01:40 mickey|busy: when I asked Sean about card size he told '512M because 1G are too expensive and 256M only bit cheaper then 512M ones' so I'll wit Mar 23 13:01:44 s/wit/wait Mar 23 13:03:08 Hmm. That bitbake bug has been filed against OE so I didn't see it :/ Mar 23 13:03:58 mickey|busy: Which bug number was the PACKAGES_DYNAMIC and DEFAULT_PREFERENCE problem? Mar 23 13:05:29 let me check Mar 23 13:05:35 I'm almost sure I filed it against BB Mar 23 13:05:53 *sigh* Mar 23 13:05:55 we need a CNAME Mar 23 13:06:04 bugs.openembedded.org Mar 23 13:06:06 vs. bugzilla.openmoko.org Mar 23 13:06:10 i always confuse Mar 23 13:06:25 * Harlekin has joined #opie.de Mar 23 13:06:27 upps Mar 23 13:06:30 cut'n'paste Mar 23 13:06:35 !oebug 2034 Mar 23 13:06:38 * * Bug 2034, Status: NEW, Created: 2007-03-23 05:15 Mar 23 13:06:38 * * mlauer(AT)vanille-media.de: PACKAGES_DYNAMIC vs. DEFAULT_PREFERENCE Mar 23 13:06:40 * * http://bugs.openembedded.org/show_bug.cgi?id=2034 Mar 23 13:07:20 mickey|busy: ok, thanks. For some reason I didn't get that bugzilla mail, probably stuck in a mail queue... Mar 23 13:07:25 np Mar 23 13:07:32 RP: on a slightly related note Mar 23 13:08:02 RP: we should have a phone call or a meeting about soon Mar 23 13:08:27 soon = somewhere in April Mar 23 13:08:56 hrw, no luck, i still get this error http://paste.uni.cc/13920 Mar 23 13:08:57 * zecke has office space to offer for such a meeting :) Mar 23 13:09:23 * mickey|busy wants to visit RP's place since he promised there would be a nice landscape Mar 23 13:09:33 and nice hotel's near by, and nothing to walk... Mar 23 13:09:40 mickey|busy: Did you see my email to the ohm list? Mar 23 13:09:57 RP: no, i'm afraid I'm not subscribed yet :/ Mar 23 13:10:10 ~lart ohm ml Mar 23 13:10:10 * ibot takes a rusty axe and swings it violently, taking ohm ml's head off Mar 23 13:10:25 I'm subscribed, configured to receive mails and no mails from ML Mar 23 13:10:31 * RP wishes he had office space here Mar 23 13:11:09 mickey|busy: We should talk about it, yes. It'll probably have to be arranged with Matthew too Mar 23 13:11:55 RP: sure thing. Mar 23 13:12:46 empty_mind: pester me later - I will look Mar 23 13:13:07 hrw, sure, may i know what time ? Mar 23 13:13:18 3-4h Mar 23 13:14:06 ok Mar 23 13:15:07 hrw: There have been mails? I haven't had any either, I thought there was just no interest! Mar 23 13:15:08 empty_mind: how old metadata you are using? Mar 23 13:15:35 RP: I'm subscribed since feb. did not get even your and koen mail Mar 23 13:16:16 Ah, I saw them, I see there's been nothing more since though Mar 23 13:16:38 empty_mind: I checked on machine which does not have libxau/libxcmdp on host and libx11 is also built there Mar 23 13:17:27 empty_mind: check tmp/staging/arm-linux/lib/pkgconfig does it has xau.pc and xdcmp.pc Mar 23 13:17:36 hrw, i have monotone 0.32 Mar 23 13:18:01 hrw, base revision: 3f0fa5c1d232d37d47f6436fe74919dd3cd003b1 Mar 23 13:18:18 empty_mind: metadata == OE Mar 23 13:18:25 cd org.openembedded.dev;mtn head Mar 23 13:18:29 hrw, ok, that would be about 20 days old Mar 23 13:18:57 empty_mind: so it is obsolete Mar 23 13:18:58 hrw, nope it got updates Mar 23 13:19:04 empty_mind: mtn head Mar 23 13:19:17 hrw,e11335447cdd49b1da98763b404e04753ebfea59 koen@openembedded.org 2007-03-23T12:51:20 Mar 23 13:19:32 empty_mind: did you start from clean ./tmp Mar 23 13:19:37 hrw, i checked the snapshots too, i think this is the latest one Mar 23 13:19:55 XorA, yes Mar 23 13:20:09 23 14:23 < hrw> empty_mind: check tmp/staging/arm-linux/lib/pkgconfig does it has xau.pc and xdcmp.pc Mar 23 13:21:07 hrw, yes it has Mar 23 13:21:48 hrw, if you look at the error pasted, it says xau.pc is searched in tmp/staging/i686-linux/share/pkgconfig Mar 23 13:22:17 hrw, one min, xau.pc is in tmp/staging/arm-linux/share/pkgconfig Mar 23 13:22:54 empty_mind: it does not check in tmp/staging/i686-linux/share/pkgconfig in log.. Mar 23 13:23:53 hrw, xau.pc is in tmp/staging/arm-linux/share/pkg-config and not in tmp/staging/arm-linux/lib/pkg-config Mar 23 13:24:07 empty_mind: how old is this build? Mar 23 13:24:59 hrw, no idea, i used openmoko mokomakefile to update Mar 23 13:25:07 empty_mind: we dropped usage of tmp/staging/arm-linux/share/pkg-config quite long time ago Mar 23 13:25:36 hrw, how can i check the build date ? Mar 23 13:26:05 I do not know how old stuff does openmoko use but they synced with OE last weeks.. Mar 23 13:27:10 cu Mar 23 13:30:08 hi Mar 23 13:31:49 anyone work with PCI on CM X270 before? Mar 23 13:31:52 hmmm Mar 23 13:31:58 java.lang.OutOfMemoryError Mar 23 13:36:01 is there anyone who have a netbook pro running linux ? Mar 23 13:36:53 * koen|lab points to florian Mar 23 13:37:08 he has the buildtools, so he's the only one allowed to publish binaries Mar 23 13:37:21 i'm little desperated to boot it under a true OS :) Mar 23 13:37:24 koen|away, many thanks Mar 23 13:37:30 right Mar 23 13:37:43 ping pb_ Mar 23 13:37:58 florian, i'm very very interested if you give me some help :) Mar 23 13:38:08 koen|away: so you never managed to create someething from the description i wrote, right? Mar 23 13:38:35 a friend gave me a nb pro some days ago ;) Mar 23 13:39:00 meister: did you send me an email lately? Mar 23 13:39:10 florian, true Mar 23 13:39:26 and to koen too Mar 23 13:39:50 i've googled so much to find free people and free netbook pro ... :) Mar 23 13:40:28 meister: ok... i wrote a part of the answer already :-) Mar 23 13:41:40 florian, well, i'll be happy to read it Mar 23 13:43:50 meister: as soon as i'm finished with my work today Mar 23 13:44:08 florian, thx Mar 23 13:46:27 Laibsch: did you figure it out? Mar 23 13:46:47 Laibsch: my commit messages and the bug linked should explain it Mar 23 13:49:17 Laibsch: ah, saw your subsequent message. I really don't think it was with the DEFAULT_PREFERENCE = "-1" I added to libgnomeui-2.16. If it was then some other conf you were using was resetting it (local.conf?). It wasn't for me without that. Note that there is a PREFERRED_VERSION for gtk+ in the angstrom.conf and that version of gtk+ is by default not prefered. See the bug for more. Mar 23 13:53:33 koen already cleared things up. He confirmed it had nothing to with your changes. direct-fb seems to have uncovered some underlying bugs in bitbake Mar 23 13:53:36 !oebug 2033 Mar 23 13:53:37 * * Bug 2033, Status: RESOLVED, Created: 2007-03-23 03:46 Mar 23 13:53:38 * * slapinid(AT)gmail.com: No proper PREFERRED_PROVIDER for gtk-pixbuf-*, build failed with bitbake 1.8 Mar 23 13:53:39 * * http://bugs.openembedded.org/show_bug.cgi?id=2033 Mar 23 13:53:51 ähm, 2034? Mar 23 13:53:57 mickey reported it Mar 23 13:54:59 JustinP: I assume you were also seeing the effects of that bug. The chain of cause and effect was just the opposite of what I was contemplating. Mar 23 14:09:53 greetings, does anyone know the right method of adding kernel modules to load at boot to an image? Mar 23 14:10:55 grep for module_autoload Mar 23 14:11:10 koen|away: ah ok thanks Mar 23 14:11:31 that specifies which modules will be autoloaded Mar 23 14:11:41 ah perfect Mar 23 14:11:45 *_EXTRA_RRECOMMENDS specify which modules to install by default Mar 23 14:14:07 I need to look up the intended uses of DISTRO_EXTRA_*, I've been using DISTRO_EXTRA_RDEPENDS to specify modules and other packages Mar 23 14:16:58 03nail 07org.oe.dev * ra38828c5... 10/ (1 packages/linux-libc-headers/linux-libc-headers_2.6.18.bb): linux-libc-headers: add UTS_RELEASE to version.h. Fixes uClibc builds with these headers Mar 23 14:18:18 WTF? Mar 23 14:18:31 hmm? Mar 23 14:18:34 if I save the file as fax-fft.m the fft() function doesn't work anymore Mar 23 14:18:37 ah Mar 23 14:18:38 * koen|lab stabs matlab Mar 23 14:18:52 * NAiL thought koen|lab went WTF to my commit Mar 23 14:18:55 heh, no Mar 23 14:19:10 That would've been confused since you kinda-ok'd it already :P Mar 23 14:19:14 indeed :) Mar 23 14:26:29 hehe, I just ran into the directFB issue Mar 23 14:27:13 zecke: with angstrom? Mar 23 14:27:25 yes, but I will update now :) Mar 23 14:32:31 Koen : Apache2 is as 2.2.3 / 2.2.3 is not available anymore this should be 2.2.4 Mar 23 14:32:48 ? Mar 23 14:34:22 Koen: apache2-2.2.3.bb should be apache2-2.2.4 Mar 23 14:34:23 * koen|lab doesn't use apache on his boards, only cherokee Mar 23 14:34:48 gerwinin: well, or grab 2.2.3 and mirror it somewhere Mar 23 14:34:57 gerwinin: or check if apachae ha an old directory Mar 23 14:35:02 zecke: good idea Mar 23 14:35:20 koen|lab: even with latest angstrom gtk+ directfb starts to get built Mar 23 14:35:26 koen|lab: and fails due a missing patch Mar 23 14:35:32 zecke: I am trying to patch 2.2.4. no so that it will built Mar 23 14:36:28 * gerwinin includes his patch and prays Mar 23 14:36:44 :( Mar 23 14:39:48 zecke: really strange, since angstrom.conf defines gtk+ as the preferred provider Mar 23 14:40:04 zecke: or is this a ?= problem? Mar 23 14:42:39 hmm I wonder why this multiple providers messgae is not displayed by default... Mar 23 14:44:35 zecke: http://www.openembedded.org/viewmtn/diff.psp?id1=e3b3e82060f9e58eb354761f3dcd6e946bd4d3eb&id2=5071c48a87d5b7fe1522ead80ed52698f40d90af Mar 23 14:44:43 zecke: that should have fixed it, right? Mar 23 14:45:30 * zecke counts the 'R's Mar 23 14:48:00 heh Mar 23 14:48:16 "Forbidden to eat or drink, your account will be blocked if you do" Mar 23 14:48:18 koen|lab: and the string does not stop with a 'S'... Mar 23 14:48:25 * koen|lab doesn't have an account Mar 23 14:48:28 koen|lab: do both! Mar 23 14:48:45 ;) Mar 23 14:48:52 you will get an account just so they can block you Mar 23 14:48:57 zecke: eat xor drink? Mar 23 14:49:31 zecke: which drink? Mar 23 14:49:33 Hello Mar 23 14:49:37 ehm Mar 23 14:49:39 string Mar 23 14:49:41 hey sirfred Mar 23 14:50:44 BTW: gtk-directfb lacks a deps to directFB?! Mar 23 14:51:03 Thu Jan 1 01:29:53 CET 1970 Mar 23 14:51:06 oh Mar 23 14:51:23 my computer is a time machine Mar 23 14:52:37 I would like to know a given package dependency tree, is that possible with bitbake ? Mar 23 14:52:50 bitbake -g Mar 23 14:52:53 I mean, what is triggering a given package to be build. Mar 23 14:53:03 koen|away: Thanks. Mar 23 14:53:14 or bitbake -nv Mar 23 14:53:30 zecke: do you see anything wrong in that diff? Mar 23 14:54:08 TypeError: findBestProvider() takes at least 3 arguments (2 given). Bad luck with bitbake -g Mar 23 14:54:23 RP: http://rafb.net/p/9pXZfJ21.html we should except that and clean the cache Mar 23 14:54:27 sirfred: yes, see the bug report Mar 23 14:57:48 zecke: agreed Mar 23 14:58:07 RP: Hello. Mar 23 14:58:12 hi sirfred Mar 23 14:58:21 RP: Now that I see you, I had trouble building linux-rp_2.6.20-r7.bb Mar 23 14:58:33 RP: should I file a bug report, or do we remember it? Mar 23 14:58:35 RP: I've submitted a bug report and a patch (related with sharpsl_rc) Mar 23 14:58:50 RP: also should multiple providers be printed to the output by default? Mar 23 14:59:13 RP: Bug 2031 Mar 23 15:01:22 zecke: I will commit a fix now ;-) Mar 23 15:01:46 sirfred: ok, I need to look at that, I didn't add the sharpsl_rc driver, JustinP did... Mar 23 15:01:55 RP: do you want to close a bug report? Mar 23 15:01:59 RP: OK. Just to keep you informed. Mar 23 15:02:12 zecke: You filed it already? :) Mar 23 15:02:19 sirfred: thanks :) Mar 23 15:02:29 RP: you're welcome Mar 23 15:03:07 koen|lab: even without the ? it fails Mar 23 15:03:35 zecke: any idea why? Mar 23 15:03:41 * zecke checks his angstrom version :) Mar 23 15:04:21 DISTRO = "angstrom-2007.1" looks correct to me Mar 23 15:04:49 indeed Mar 23 15:04:59 have we uncovered a bitbake bug? Mar 23 15:05:01 Are you aware of a problem building osb-nrcore ? It seems that gtkclist.h is using GMemChunk without including the proper glib header. Mar 23 15:05:26 03rpurdie * r801 10bitbake/ (ChangeLog lib/bb/cache.py): cache.py: Catch truncated cache errors Mar 23 15:05:34 koen|lab: this would violate rule0 Mar 23 15:06:01 re Mar 23 15:08:20 koen|lab: oh well, I will look into it after work... Mar 23 15:09:50 It doesn't seem to be that simple. gtk.h includes gtkclist.h but it also includes gdk.h that includes gdktypes.h, that includes glib.h that includes gmem.h Mar 23 15:10:43 zecke: thanks Mar 23 15:13:25 for some reason, libmudflap isn't being installed correctly. in that libmudflap.so, libmudflap.so.0 and libmudflap.so.0.0.0 all exist, but none of them are symlinks like they should be Mar 23 15:13:38 this is gcc-4.1.1 Mar 23 15:14:46 OH, I see, GMemChunk is deprecated and osb-nrcore is compiled with -DG_DISABLE_DEPRECATED Mar 23 15:17:25 Koen : is cherokee standardly equiped with php module ? Mar 23 15:17:51 gerwinin: not by default Mar 23 15:17:59 it's using fastcgi Mar 23 15:18:21 re Mar 23 15:18:57 I'm afraid the only way to use gtkclist is not defining G_DISABLE_DEPRECATED Mar 23 15:18:58 hrw: Hello Mar 23 15:19:19 * hrw replaced 6 years old cdrw drive to 6 bottles of beer Mar 23 15:19:59 hrw: What did you get, beers or the cdrw ? Mar 23 15:21:44 * mwester_ thinks the beers are the better end of that bargain Mar 23 15:21:52 * sirfred agree Mar 23 15:22:31 beers Mar 23 15:22:58 Beer is liquid bread it is good for you :) Mar 23 15:23:26 hrw: Good Mar 23 15:23:41 03justinp 07org.oe.dev * rc8eb1c62... 10/ (11 files in 2 dirs): linux-rp-2.6.20: add CONFIG_SHARPSL_RC=n for non-spitz machines Mar 23 15:25:09 JustinP: So, the corgi is not supporting the remote control? Mar 23 15:26:01 sirfred: do you have a corgi remote? Mar 23 15:26:07 XorA: No. Mar 23 15:26:17 XorA: I've seen it only in pictures. :) Mar 23 15:26:28 sirfred: the changes to support corgi are quiet easy, just needs someone who has the remote to test it Mar 23 15:27:07 XorA: I don't expect to have one of those Mar 23 15:27:58 What is selected by default under the presence, for example, of : osb-nrcore_20060212.bb and osb-nrcore_0.5.0.bb ? Mar 23 15:28:02 By bitbake, I mean Mar 23 15:28:10 sirfred: one with highest PV Mar 23 15:28:46 PV = "0.5.0+cvs${FIXEDSRCDATE}" is higher than 05.0 Mar 23 15:28:59 koen|lab: regarding modules autoloading... I'm trying to get snd-pxa2xx-ac97 to load at boot, so I added module_autoload_snd-pxa2xx-ac97 = "snd-pxa2xx-ac97" to my machine's conf file, ran bitbake task-base -crebuild, bitbake bootstrap-image, flashed my board with the new image, but snd-pxa2xx-ac97 is not loaded at boot :( Mar 23 15:29:01 XorA: Thanks. Mar 23 15:29:15 sirfred: although in those files there is some magiv to check for gcc 4.X Mar 23 15:30:06 XorA: I see. Mar 23 15:30:07 xuumbi: sounds like a bug Mar 23 15:30:33 koen|lab: dang, ok thanks Mar 23 15:30:48 xuumbi: rebuild kernel Mar 23 15:31:05 hrw: ok I'll try that Mar 23 15:31:14 module_autoload are set in kernel module packages. so if you change them then you need to rebuild kernel Mar 23 15:31:46 hrw: a-ha, good to know Mar 23 15:44:40 I'm trying to compile opie-recorder (still) and now I get this error during compile: | ERROR: "gcc" looks like gcc 4.x Mar 23 15:44:41 | QEMU is known to have problems when compiled with gcc 4.x Mar 23 15:44:43 | It is recommended that you use gcc 3.x to build QEMU Mar 23 15:44:44 | To use this compiler anyway, configure with --disable-gcc-check Mar 23 15:44:46 | FATAL: oe_runconf failed Mar 23 15:44:55 Can someone tell me what I need to do? Mar 23 15:45:20 you need gcc 3.4 on host Mar 23 15:46:08 I have it Mar 23 15:46:23 I think... Mar 23 15:46:25 let me check Mar 23 15:47:11 nm Mar 23 15:47:16 Installing it now :) Mar 23 15:48:47 boy do I ever feel stupid :p Mar 23 15:49:04 Thanks for helping! Mar 23 15:57:34 ~lart the guy which added gtk-directfb Mar 23 15:57:35 * ibot smacks the guy which added gtk-directfb up side the head with a clue-by-4 Mar 23 15:57:53 hrw: I thought this was fixed? Mar 23 15:58:04 in angstrom Mar 23 15:58:09 OK Mar 23 15:58:17 depending on your bitbake version it seems.... Mar 23 15:58:31 hrw: I recompiled my kernel and created a new image, now my module is loading happily, thanks! Mar 23 15:59:06 xuumbi: btw, what hardware you target? Mar 23 15:59:59 03hrw 07org.oe.dev * rb19a10b0... 10/ (1 conf/distro/include/openmoko.inc): openmoko: we use GTK not GTK-DirectFB Mar 23 16:00:47 hrw: worse, gtk-directfb is missing a patch?! Mar 23 16:02:43 hrw: CM-x270 from compulab Mar 23 16:09:09 hrw, http://rafb.net/p/0iBy1b67.html Mar 23 16:09:59 Crofton__: flash problem? Mar 23 16:10:05 http://rafb.net/p/iNq6Ko97.html Mar 23 16:10:09 it is booting Mar 23 16:10:11 still Mar 23 16:10:36 Setting up IP spoofing protection: rp_filter. Mar 23 16:10:37 note the sh: rw: unknown ,,, Mar 23 16:10:48 Crofton__: why are you filterint out RP? Mar 23 16:11:01 I think that hgappens on my angstrom build Mar 23 16:11:04 RP? Mar 23 16:11:15 this is hrw's image :) Mar 23 16:11:41 also mount /proc is failing Mar 23 16:12:07 I suspect it mounts it twice Mar 23 16:12:09 I can log in Mar 23 16:12:49 http://rafb.net/p/aGM0xd71.html Mar 23 16:13:08 hrw, let me know if you need me to look at anything specific Mar 23 16:14:40 Crofton__: it boots and allow to login so it works. All I wanted to know. THX Mar 23 16:14:43 How does bitbake calculate dependencies from the RDEPENDS information? Mar 23 16:15:22 ok :) Mar 23 16:15:57 florian: with magic Mar 23 16:16:07 hum Mar 23 16:16:14 hrw, I can login vi ssh also Mar 23 16:16:40 If i build task-whatever i get the task-whatever ipk but not the stuff that ipk would depend on. Mar 23 16:16:46 Crofton__: thanks a lot Mar 23 16:17:07 florian: with bitbake 1.8? Mar 23 16:17:25 koen|away: both with 1.6.6 and 1.8 Mar 23 16:18:10 s/koen|away/koen|lab/ Mar 23 16:20:03 koen|lab, the rp_filter stuff happens by magic. I am not even sure what it is :) Mar 23 16:20:38 Crofton__: spoof filter :) Mar 23 16:21:01 Crofton__: it just means that you DEPENDS are fed to RP and he vets them for correctness Mar 23 16:22:37 now you guys are making my head hurt Mar 23 16:23:33 florian: Its right not to build the dependent packages since nothing in the task file says to do that. I'm guessing you have debian.bbclass turned off? Mar 23 16:25:08 RP: no idea, i tried to build task-sdk for the gerneric distro for the generic qemuarm Mar 23 16:25:22 florian: 1.6.6 and 1.8.0 work differently, 1.6.6 had BUILD_ALL_DEPS which was a bad idea, 1.8 has the rdeptask flag for a given task. See the top of debian.bbclass for an exmaple of each Mar 23 16:25:45 florian: You probably want meta-sdk, not task-sdk as the meaning is different Mar 23 16:26:40 florian: There might be a rdeptask flag missing from meta-sdk too at a quick glance Mar 23 16:28:10 RP: no, i definitely want only the task-sdk Mar 23 16:28:41 cbrake: ping Mar 23 16:30:01 florian: A task package is defined to only build a package which has dependencies of other tasks. It does not build those other packages. That is the defition of a task package Mar 23 16:30:19 florian: meta* packages are intended to build a target and all its dependencies Mar 23 16:30:44 what tool could I use to image an ext2/3 filesystem onto an IDE flash module? Mar 23 16:30:56 RP: but the meta-package would try to build the sdk compiler i do not want and which is simply broken Mar 23 16:31:02 lol http://fun.drno.de/pics/new-firefox-logo.jpg Mar 23 16:31:13 florian: So define a new meta package Mar 23 16:31:32 I'm currently using the bootimg.bbclass to burn two partitions - 1 dos with syslinux as the boot partition, and the other a standard ext2 partition Mar 23 16:31:33 florian: but be clear about the difference between a task and a meta package Mar 23 16:31:53 chouimat: old... Mar 23 16:32:00 It was ok when imaging onto a 128M drive, but for a 512M one it's way too slow Mar 23 16:32:00 hrm... i'm quite sure that worked before, some documents use thi as well. Mar 23 16:32:05 hrw: first time for me :) Mar 23 16:32:29 florian: If you inherit debian.bbclass it will work as a side effect of debian.bbclass Mar 23 16:32:45 python-cairo python-egenix-mx-base python-fuse python-gammu python-m2crypto python-mysqldb python-native python-psyco python-pylirc python-pyode python-pyqt4 python-sqlobject python-sword Mar 23 16:32:54 those ones does not build on 64bit.. Mar 23 16:33:12 xuumbi: hello Mar 23 16:33:55 xuumbi: off to lunch, be back in an hour ... Mar 23 16:35:06 cbrake_lunch: ah nuts, just missed you Mar 23 16:35:51 RP: ok, i'll try--- Mar 23 16:37:20 I wish i had an idea which patch caused gcc-cross-sdk to fail. Mar 23 16:43:28 Koen: compiled with cherokee worked great Mar 23 16:53:37 chouimat: http://developer.mozilla.org/wiki-images/en/e/e0/Moz_ffx_openStandards_1024x768.jpg Mar 23 16:54:27 hrw: :) Mar 23 16:58:23 hrw: very nice =) Mar 23 17:00:49 RP: This didn't really help - i created a special meta-sdk that only rdepends on some task-sdk-wahtever things and inherits debian.bbclass. Mar 23 17:00:58 RP: same effect like before. Mar 23 17:03:13 who made fetch quiet? Mar 23 17:03:40 hrw: New logging structure in trunk - it goes to a log file, the same as shell output Mar 23 17:03:59 florian: I don't know then :-( Mar 23 17:04:05 RP: no way to get it back now? Mar 23 17:04:52 hrw: Not at present. We need to work out how best to handle it but sending to stdout isn't an option with the UI Mar 23 17:04:58 ok Mar 23 17:05:34 ~curse belnet.sf.net Mar 23 17:05:36 May the fleas of a thousand camels infest your most sensitive regions, belnet.sf.net ! Mar 23 17:08:35 *gnaa* Mar 23 17:08:55 How do i get rid of the new bitbake?! Mar 23 17:09:59 * hrw will probably go to 1.8 Mar 23 17:10:08 hrw, Mar 23 17:10:52 ERROR: Task 0 (/home/hrw/devel/oe/org.openembedded/packages/python/python-m2crypto_0.13.1.bb, do_rebuild) failed Mar 23 17:11:05 why rebuild fails? Mar 23 17:12:55 hrw: BBINCLUDELOGS doesn't work anymore, that was in the changelog Mar 23 17:13:00 (in trunk) Mar 23 17:14:25 ok Mar 23 17:15:35 Its the start of the UI breakage and will get worse before it gets better... Mar 23 17:16:03 Downgrading seems to be not that easy... ./setup.py doesn't seem to overwrite everything :-( Mar 23 17:19:12 * mwester_ makes a not to *NOT* sync up for a while, until bb works again... :) Mar 23 17:19:21 s/not/note/ Mar 23 17:23:07 sirfred|afk: I do want to support the corgi and akita but I'm not sure what the correct way to support them is. The current driver is known not to even compile for non-spitz. Mar 23 17:24:22 sirfred|afk: even if that patch allowed it to compile, the CE-RH1 wouldn't work with the driver because the values would be different. Do you have a CE-RH1? Mar 23 17:26:59 cbrake: hope you had a good lunch :) Mar 23 17:27:09 xuumbi: I sure did :-) Mar 23 17:27:29 cbrake: have you had any experience with audio on the cm-x270? Mar 23 17:27:46 xuumbi: yes, I am currently using it -- playback only ATM. Mar 23 17:28:05 cbrake: ah cool, have you noticed any noise with the 2.6.16 kernel? Mar 23 17:28:34 xuumbi: yes, I have noticed some noise, but thought that was due to the base board design. Mar 23 17:29:08 cbrake: I've been noticing a LOT of noise Mar 23 17:29:24 cbrake: how have you set levels? I'm using alsamixer Mar 23 17:29:56 cbrake: I notice that if I mute "headphones" the noise goes away, but the playback volume (even with master volume at 100%) drops to an almost inaudible level Mar 23 17:31:41 cbrake: if I use compulab's 2.6.12.2 kernel and file system image audio is crystal clear Mar 23 17:32:15 cbrake: have you encountered any problem with the CM x270 with PCI cards? Mar 23 17:35:50 xuumbi: hmm, very strange Mar 23 17:35:55 xuumbi: let me try a few things here ... Mar 23 17:36:16 TinY_Apple: I am currently not using PCI with the cm-x270. What problems are you seeing? Mar 23 17:36:35 cbrake: oh that's great, much appreciated Mar 23 17:37:00 RP: ok, with 1.6.6 it works so its a bug in 1.8 Mar 23 17:43:17 xuumbi: I notice the volume drops as well if I mute headphones Mar 23 17:43:41 cbrake: ok that's good to know thanks for testing that Mar 23 17:43:53 xuumbi: I'm not hearing any noise right now Mar 23 17:44:15 hi all Mar 23 17:45:20 xuumbi: I'm currently updating the cm-x270 to the 2.6.20, so maybe that will fix it. Due to other distractions -- will probably be another week before I'm finished. Mar 23 17:45:23 cbrake: ok interesting, I'm going to experiment with different boards Mar 23 17:45:48 cbrake: oh cool, are you bringing over all of compulab's 2.6.16 patches? Mar 23 17:46:00 xuumbi: no, just a base set to start with Mar 23 17:46:22 xuumbi: will probably drop PCI, etc for the first pass. Maybe someone else will pick that up. Mar 23 17:46:33 cbrake: ok good luck :) Mar 23 17:46:55 xuumbi: I'm having trouble with the SD card driver with new Sandisk 2G cards, plus I would like the to get some of the real-time improvements, etc. Mar 23 17:47:15 xuumbi: it looks like it will be easier to just move everything up than backport the SD card driver. Mar 23 17:47:17 cbrake : i've tried running a Ethernet Card running Realtek 8139 chip, it work fine, but I'm having a lot of problem trying to let my Farsync WAN card to work on CM X270 Mar 23 17:48:36 cbrake: I'm running into all these unexpected problems since we originally developed around compulab's debian image with 2.6.12.2, now I'm trying to make the case for OE and 2.6.16, but things aren't going very smoothly Mar 23 17:49:11 florian: ok, please file a bug so we don't forget Mar 23 17:49:23 xuumbi: nod it takes some work. You running into any other major problems? Mar 23 17:51:02 cbrake: BTW have you tried CM-iVCF? I've problem powering it up on the ATX base Mar 23 17:51:25 cbrake: just getting our software to compile in OE was a big job, our developers use SlickEdit which uses a binary called vsbuild to run everything in the build process, getting the two to talk to each other was difficult Mar 23 17:52:09 TinY_Apple: no, cm-x270 is the only one I've used. I don't have any idas about the Farsync card. Mar 23 17:53:10 cu Mar 23 17:53:24 cbrake: np, i'm now working very hard to try making the Farsync card to run. Wander if there are any PCI expert here Mar 23 17:55:33 What is the proper way of using apt-get in the storage limited cm x270? Mar 23 17:55:52 xuumbi: that sounds painful. OE is not real developer friendly. What I generally do is keep my app source in SVN and build w/ standard autotools. Then I set up a bb recipe the fetches it and builds it when I want an ipk, or need to cross compile. This lets me develop and debug the app on a x86 PC, and then do final dev for the target. Mar 23 17:56:24 xuumbi: * developer friendly -- I mean to application developers who don't want to learn anything about OE. Mar 23 17:56:43 TinY_Apple: if you build with OE, you generally use ipkg instead of apt-get Mar 23 17:57:17 I only know how to get the kernel from OE, not sure how i can get app from OE Mar 23 17:59:07 cbrake: I used only virtual/kernel then find the kernel image not sure wat else i can do with it :P Mar 23 18:00:13 cbrake: that sounds like a good system Mar 23 18:00:23 TinY_Apple: build an entire OS image -- set up for the angstrom distro, and then "bitbake angstrom-bootstrap-image" Mar 23 18:00:39 cbrake: the way I have it now developers will just need the oe toolchain and they can still build with SlickEdit Mar 23 18:01:13 TinY_Apply: I'm using a Dlink DWL-G650 WiFi card on my cm-x270 and it works ok. I did have to throttle the pci bus all the way to keep the LCD readable when the WiFi is turned on. Mar 23 18:01:58 xuumbi: how do you move the toolchain to other dev's PCs? Do you just zip it up and then decompress in a set location? Mar 23 18:02:55 apaulsen: I realised that too, if the PCI is not in full throttle the LCD will jump.. Mar 23 18:03:43 apaulsen: did you compiled the DWL-G650 driver? Mar 23 18:04:37 wow, lots of cm270 users here :) Mar 23 18:05:21 koen: cm-x270 is a really nice part for low volume vertical apps :-) Mar 23 18:06:10 cbrake: yep that's how it is Mar 23 18:06:48 cbrake: very few of them need to build it on their desktops, we have a build server that they can all access to run builds if needed Mar 23 18:07:02 xuumbi: ahh, that helps. Mar 23 18:07:31 I use madwifi, which is present in OE Mar 23 18:10:22 apaulsen: I see, wander if OE have Farsync driver hee.. Mar 23 18:12:15 Ocean: is Farsync supported by the kernel already? Mar 23 18:13:35 Ocean: I just grep'd the 2.6.16 kernel and see drivers for Farsync T series cards Mar 23 18:16:17 * TinYApple slaps TinY_Apple around a bit with a large trout Mar 23 18:17:52 03hrw 07org.oe.dev * r846bde67... 10/ (12 files in 2 dirs): site/*: move mr_cv_target_elf to common as all our targets are ELF systems - this also make libelf buildable Mar 23 18:27:26 03coredump2 07org.oe.dev * rf0ea8dcd... 10/ (5 files in 3 dirs): Mar 23 18:27:26 altboot: - CVS -> SVN Mar 23 18:27:26 - Fix do_install() and do_configure to use instead of fixed path Mar 23 18:27:46 cbrake: ok here's something interesting, I flashed my OE board with the default 2.6.12.2 kernel from compulab, which I've had no noise problems with running compulab's debian image, and I'm getting noise Mar 23 18:28:06 this makes me think the problem is on OE's side Mar 23 18:29:35 bbl Mar 23 18:33:51 xuumbi: interesting. What do you use with the debian image to set audio levels? Perhaps something is different there. Mar 23 18:34:43 xuumbi: also, are you running any tasks that might be starving the audio process? You might have to bump your audio process to RT. Mar 23 18:36:05 cbrake: I'm not sure how we set levels in debian, I'll check, I'll check processes too Mar 23 18:36:46 I'm going to flash a debian image with compulab's 2.6.16 kernel and see what happens... Mar 23 18:39:08 xuumbi: compulab's nand flash driver is notorious for locking the systems for long periods of time. JFFS2 seems to be working better for me. Mar 23 18:39:28 cbrake: I know what you mean, I see all kinds of lockups Mar 23 18:40:05 cbrake: so you're flashing using a tiny linux install in NOR? Mar 23 18:40:15 xuumbi: turning on kernel preemption also helped a lot w/ the compulab NAND driver. Mar 23 18:41:04 xuumbi: yes, we always boot through a small linux rootfs located in NOR that looks for updates on USB. Mar 23 18:41:09 cbrake: ah, good to know Mar 23 18:41:33 xuumbi: but, JFFS2 is really the way to go if you can tolerate the mount times. Mar 23 18:42:17 cbrake: yeah that is a big problem for our product, I don't think management will tolerate the time it takes JFFS2 to mount Mar 23 18:42:38 xuumbi: from everything I've seen, it works better. There has to be some inefficiencies running a standard block fs on top of a translation layer. Mar 23 18:43:53 cbrake: yeah I definitely agree Mar 23 18:45:15 cbrake: tried summary images for jffs2? Mar 23 18:45:57 koen: yes, I'm using that now. It cut it about in 1/2, but w/ 512MB of flash, it is still a long time. Mar 23 18:46:22 koen: I setting up partitions now, so the main partition will just be 64MB. Mar 23 18:46:46 koen: I'm sure RP's JFFS2 patches would help w/ increased read speed as well -- some day ... Mar 23 18:47:24 maybe I should try no compression ... Mar 23 18:48:31 cbrake: did the kernel preemption cause any problems? We're trying to play videos on our compulab board, but the flash driver causes 1/4 second glitches every four seconds. Mar 23 18:49:56 apaulsen: it may be what is causing my SD problems. Mar 23 18:50:15 apaulsen: some early testing indicated that the SD driver in 2.6.16 might not be preempt safe. Mar 23 18:50:23 apaulsen: I've encountered no other problems. Mar 23 18:51:31 cbrake: since I'm not using SD I might have to give that a try. Hopefully my own drivers are safe :) Mar 23 18:51:52 apaulsen: kernel preemption is good stuff :-) Mar 23 18:55:25 cbrake: the problem is that jffs2 needs to recreate the inode tree every boot and keeps it in RAM Mar 23 19:10:05 if my library has a bla-config script, do I have to manually stage it? it's under bin_SCRIPTS in Makefile.am Mar 23 19:10:27 Jin^eLD: inherit binconfig IIRC Mar 23 19:10:40 thx Mar 23 19:12:11 yes, that was it :) thanks a lot Mar 23 19:16:16 there were three TMNT movies? Mar 23 19:16:54 TMNT? Mar 23 19:17:11 hi, newbie question: which MACHINE do I havee to put on local.conf to build an Angstrom image that runs on qemu? Mar 23 19:17:11 ah, turtles Mar 23 19:17:19 qemuarm Mar 23 19:17:26 or qemux86 Mar 23 19:17:28 more specifically qemu-system-arm Mar 23 19:17:50 ok, so I just need to set MACHINE to qemuarm, right? Mar 23 19:17:59 thanks Mar 23 19:21:29 apaulsen: I should ask you as well, are you using audio on the cm-x270? Mar 23 19:22:14 xuumbi: I'm using bluetooth audio, not the onboard ac97 Mar 23 19:22:33 apaulsen: ah ok just wondering, thanks Mar 23 19:31:36 03coredump2 07org.oe.dev * rdf3001c7... 10/ (1 packages/altboot/altboot_1.1.1+wip-20061123.bb): altboot_1.1.1+wip-20061123.bb: Add fixed snapshot for Zaurus models Mar 23 19:31:45 03coredump2 07org.oe.dev * r2f1eb27f... 10/ (1 packages/altboot/altboot_wip.bb): altboot: Update _wip.bb to 20070323, enabling first Neo1973 aka fic-gta01 support. Mar 23 19:32:43 koen: that puts the config file here: /staging/i686-linux/bin/mipsel-linux Mar 23 19:32:45 is that ok? Mar 23 19:32:51 looks a little weird to me Mar 23 19:33:04 the class should be doing the right thing Mar 23 19:33:22 mhm.. ok then Mar 23 19:35:48 altboot on the NEO? Crazy shit. Awesome. Mar 23 19:35:54 well done, CoreDump|home Mar 23 19:36:30 ;) Mar 23 20:08:04 psokolovsky__: although I'm getting quite interesting values from apm now:) On-line, battery charging (2d+7:08:00) Mar 23 20:10:02 sorry, wrong channel:) Mar 23 20:14:20 cbrake: I found the problem! Mar 23 20:14:48 cbrake: and the problem is me! :) I was using an atx power supply, using the brick that came with the atx board I get almost no noise Mar 23 20:26:13 xuumbi: phew! That still does explain why the headphone being enabled affects volume -- might be time to dump the AC97 regs ... Mar 23 20:40:17 hello all Mar 23 20:40:22 gm Mar 23 20:42:34 hey likewise Mar 23 20:42:59 Crofton__, Jin^eLD: howdy Mar 23 20:44:03 Jin^eLD: I might get a HD H.264 encoder next week, with some work I can then feed HD video into a real-time encoder and capture the TS over IP. Mar 23 20:44:50 ...which I then could test on the 8100? :) Mar 23 20:45:14 H.264 encoder must be really cool btw :> Mar 23 20:45:25 likewise: the one JP wrote? Mar 23 20:46:23 my LAN server/OE compile host (running edgy) is crashing all the time: http://rafb.net/p/XvHDk119.html At first I thought this was https://bugs.launchpad.net/71212 but it seems there is more. Anybody willing to help me find what it is? Mar 23 20:46:25 koen: no, an encoder 20 people worked on for the last 6 years. It has 8 vector DSP chips, 5 vector DSP cores per chip, which do 8 instructions perr cycle Mar 23 20:46:27 I read through all logs recently changed at least twice but there seems to be nothing in it. Mar 23 20:46:41 likewise: ah, a hardware one Mar 23 20:47:02 koen: yup, www.telairity.com Mar 23 20:47:35 Jin^eLD: but I really try to get that Decypher in as well :-) Mar 23 20:48:10 :) Mar 23 20:50:15 re Mar 23 20:51:13 hrw: wb Mar 23 20:51:38 koen, Hopefully rishi doesn't get caught up in this mess Mar 23 20:51:56 http://www.pegasusnews.com/news/2007/mar/23/texas-instruments-lay-233-dallas-workers/ Mar 23 20:52:14 Current OE org.openembedded.dev should be buildable with both bitbake 1.6 and 1.8, right? Mar 23 20:52:33 wow Mar 23 20:52:37 likewise: yes Mar 23 20:52:39 likewise: and 1.9 Mar 23 20:54:44 Crofton: Any idea what particular product ranges are involved? I would expect TI to have multiple D(S)P divisions. Mar 23 20:54:54 not sure Mar 23 20:54:57 koen: tnx Mar 23 20:55:19 the only TI sources I had aren't on IRC much any more Mar 23 20:55:58 likewise: enjoying the new office? Mar 23 20:56:43 konen: yes, spaceous. no, I am searching for stuff on the lab every hour. Mar 23 20:56:53 s/konen/koen :-) Mar 23 20:56:53 heh Mar 23 20:57:13 koen: in other words, it's a *big* mess :-) Mar 23 20:57:47 reminds of the indt office: "we have a new airco, bring a sweater" Mar 23 20:58:10 koen: how's uni again? Mar 23 20:58:15 busy Mar 23 20:58:26 tedious assignments Mar 23 20:58:28 busy on courses? Mar 23 20:58:29 all the globe trotting must cut into study time Mar 23 20:59:05 likewise: signal processing projects, so death by matlab Mar 23 21:01:11 koen: your uni seems to be much into assignments? Mar 23 21:01:22 I remember lots of theory stuff. Mar 23 21:01:24 the ESP course is Mar 23 21:01:32 (not that I remember the theory itself) Mar 23 21:01:42 the linear systems course was all theory Mar 23 21:02:29 the theory is a bit devalued by matlab and maple combined with a fast cpu Mar 23 21:02:38 technology changes, theory gives you the background :) Mar 23 21:02:59 background to change with technology Mar 23 21:03:47 Crofton: the last conference put me in a good position Mar 23 21:03:58 Crofton: I'm taking the test 2 weeks after the rest did Mar 23 21:04:03 heh Mar 23 21:04:04 * koen has copies of the original test :) Mar 23 21:04:12 before the next conference Mar 23 21:04:19 (the jpegs I showed you yesterday) Mar 23 21:04:29 ah Mar 23 21:05:32 anyway Mar 23 21:05:35 'night all Mar 23 21:05:38 interesting, the tests are in English Mar 23 21:14:48 Crofton: Dutch uni's often give courses in English if at least one foreign student is present. At my uni, that was with most of the courses. Mar 23 21:22:52 'night everyone Mar 23 21:24:07 hi likewise! Mar 23 21:24:22 HopsNBarley: hi HopsNBarley! Mar 23 21:25:38 HopsNBarley: that reminds me I would look at the webfolders support for that other os. Mar 23 21:25:45 w2k Mar 23 21:26:11 good trip home? i think my partner is. vista has some issues, but we might be able to work around them. Mar 23 21:29:22 wb crofton Mar 23 21:29:48 HopsNBarley: yes, sure, no jet lag even, I managed to stay awake and crash late. Mar 23 21:57:31 cworth: hi carl Mar 23 21:57:50 hello Mar 23 21:58:17 likewise: hi Mar 23 21:59:12 are there any plans ever for openmoko to support a CDMA phone? I know FIC is planning on announcing more phones in sept. Mar 23 22:00:01 shit, sorry, wrong channel Mar 23 22:00:17 Hello Mar 23 22:00:31 Has anyone been getting compilation errors with konqueror? Mar 23 22:00:48 It used to build for me when I built it about a month ago Mar 23 22:01:17 I tried to build it today with a new tmp folder and updated OE, and it won't build Mar 23 22:01:29 May'st thou be interested in a log? Mar 23 22:04:51 Here it is, can someone check it out please? Mar 23 22:04:53 http://pastebin.ca/407280 Mar 23 22:06:33 B_Lizzard: I would suggest a binary section approach on the revisions since then Mar 23 22:07:03 I did not understand you, could you please elaborate? Mar 23 22:22:01 Someone, please? Mar 23 22:22:16 I searched the bugtracker, nothing pertaining to the problem Mar 23 22:27:42 B_Lizzard: I mean, check out the repository half-way in time between your last good build and your current failing build. Mar 23 22:27:51 B_Lizzard: and see if that one still builds Mar 23 22:27:55 B_Lizzard: etc etc Mar 23 22:27:56 Aha Mar 23 22:28:02 Ok, thanks Mar 23 22:31:21 But if it's a problem with OE, it still needs to be fixed Mar 23 22:34:16 I'm pretty sure it will build with an older revision Mar 23 22:34:30 I haven't changed much on my system Mar 23 22:46:14 cu Mar 23 23:01:44 03mickeyl 07org.oe.dev * r3c6cac2f... 10/ (3 files in 2 dirs): openmoko-apps: fix description in openmoko-messages Mar 24 00:44:47 **** ENDING LOGGING AT Sat Mar 24 02:59:57 2007