**** BEGIN LOGGING AT Tue Mar 24 02:59:57 2009 Mar 24 04:27:55 Bah, rpm-native is broken for me Mar 24 04:28:08 do_configure_prepend needs an export EXTRA_OECONF... I think/hope Mar 24 04:30:25 And wtf, we do do_configure_prepend()... do_configure() { ... full function ... } Mar 24 05:30:29 i am trying to build angstrom for oe, i get lots of these DISPC error messages: omap-dss DISPC error: dispc irq error status 0040 <3>omap-dss DISPC error: dispc irq error status 00e2 Mar 24 05:30:29 and the display goes blank Mar 24 09:33:02 VM: Killing process vsftpd Mar 24 09:33:04 swap_free: Bad swap file entry 40024ae3 Mar 24 09:33:23 any idea why vsftpd refuses to start ? Mar 24 09:40:17 why is vsftp the only ftp server in the repo ?? or did i miss one? Mar 24 09:47:22 03Koen Kooi 07org.openembedded.dev * r9682e05eb1 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: angstrom preferred versions: remove libosip version that no-one can explain Mar 24 09:58:43 good morning Mar 24 10:06:50 good morning Mar 24 10:20:41 http://www.linuxfordevices.com/articles/AT4375744671.html Mar 24 10:21:05 I guess we could answer with an article about OE ;) Mar 24 10:21:53 IIRC they use OE/Poky for their stuff. Mar 24 10:23:52 broonie: This more looks like Debian... Mar 24 10:24:58 We have some Simtec hardware here, but the CATS were sold before OE was started. Mar 24 10:26:24 They do both, bigger more general purpose systems have Debian on them. Mar 24 10:26:33 (Daniel and Vince are both Debian developers) Mar 24 10:33:18 * florian should take a liik at their current hardware Mar 24 10:37:12 morning Mar 24 10:37:17 hi hrw Mar 24 10:37:47 Simtec uses Poky Mar 24 10:39:00 hi Mar 24 10:39:09 and Daniel Silverstone works as Simtec<>Poky guy Mar 24 10:39:36 BBD20EUROU board is supported with Poky Mar 24 10:39:37 hrw: cool, they should have started the series with oe :) Mar 24 10:39:51 This looks interesting... http://www.simtec.co.uk/products/BBD20EUROU/intro.html Mar 24 10:39:57 heh Mar 24 10:40:11 florian: OE is not stable enough to use in production? Mar 24 10:40:26 I want to install a php module for my webserver , which runs with angstrom and an at91sam9260.. have anyone done this before? only found a php module for an wrong µC Mar 24 10:41:07 hrw: not really, but at some point you have to make a branch of course. Mar 24 10:41:45 thorsten1: no php packages in angstrom? Mar 24 10:44:08 oh there is an php package :-) Mar 24 10:54:28 03Koen Kooi 07org.openembedded.dev * red9094c688 10openembedded.git/recipes/gtk-webcore/ (midori.inc midori_git.bb): Mar 24 10:54:28 midori: disable hildon interface Mar 24 10:54:28 * if you want a hildon midori, create a midori-hildon recipe Mar 24 11:06:33 2.6.29 released - time to update some boards Mar 24 11:09:56 03woglinde 07org.openembedded.dev * r9a591708b4 10openembedded.git/conf/machine/wrap.conf: wrap: fix tuning Mar 24 11:09:58 03woglinde 07org.openembedded.dev * rb4b4c8664e 10openembedded.git/recipes/uclibc/uclibc-0.9.30.1/uClibc.distro: uclibc-0.9.30.1: enable UCLIBC_HAS_GETPT Mar 24 11:11:32 I'm having a weird problem were task do_rootfs is hanging forever. I only have two packages: the one that builds the rootfs (http://pastebin.com/m3acc3aaa) and the only package (http://pastebin.com/m3585fcab) that is in the previous packages IMAGE_INSTALL list. Mar 24 11:12:40 if I run bitbake -DDD everything looks normal and the last line before the hang is: "DEBUG: mkdirhier(/home/oe/openembedded)" and there are no error messages. Mar 24 11:17:40 thorsten1: I'm using cherokee + php on a at91sam9260 with the angstrom distro. Mar 24 11:17:59 thorsten1: cherokee being a very nice embedded web server. Mar 24 11:18:28 thorsten1: I'm also running the same combination on a PXA270 but I noted the at91sam9260 since that is what you are using. Mar 24 11:23:49 I also looked in the log files and saw nothing that would lead me to the reason of the hang. Is there any thing else I can do or enable to give me better insight into why the do_rootfs task is hanging? Mar 24 11:49:48 Anyone an idea how many contributors we have right now? Mar 24 11:50:51 good question Mar 24 11:51:21 * florian currently registers OE for LinuxTag Mar 24 11:53:07 I will attend Mar 24 11:55:19 florian, review some of the weekly reports from cbrake Mar 24 11:56:12 hmm... who has access to the git server - just counting the keys would be a good estimate Mar 24 12:02:54 282 contributors listed by git shortlog Mar 24 12:03:08 git shortlog|egrep "\):$"|wc Mar 24 12:03:18 florian: the scropts from Greg K. and Jonathan Corbot that they use for the Linux "who committed" statistics are open-source Mar 24 12:03:33 florian: they use the feature of git that can record the author independend of the committer Mar 24 12:08:10 hrw: wow, that's quite a lot... but it lists externals too Mar 24 12:09:25 so we get all one-time contributors in. for linuxtag it would be better to count all the people with commit access in order to get an impression about the size of the developer community Mar 24 12:12:28 159 committers Mar 24 12:12:38 git log --pretty=fuller|grep ^Commit:|sort -u|wc Mar 24 12:12:55 thats for whole repo so also gets bk times Mar 24 12:13:10 ah great! Mar 24 12:13:13 hrw: many thanks Mar 24 12:13:40 now, what percent are "active" Mar 24 12:14:05 they didn't ask for active ;) Mar 24 12:14:27 who might be the biggest employer of OE contributors? OH? Mar 24 12:14:34 122 in past BK time Mar 24 12:16:02 florian: Biggest employer? "miscelleanous" ! :-) Mar 24 12:16:33 heh Mar 24 12:16:41 i need an OE employer ;) Mar 24 12:17:04 schurig: ok, do you haveits address? ;) Mar 24 12:17:22 florian: "git log --pretty=fuller e3d1546d2c1ec9fd00af7780ba41250496153b15..HEAD|grep ^Commit:|sort -u" will give you list of emails Mar 24 12:17:39 florian: It has two addresses. For incoming money, use my address. For invoices, use yours. Mar 24 12:18:35 hehe Mar 24 12:18:36 hrw: but some are twice in the list, you should use "sort -u --ignore-case" Mar 24 12:19:23 hrw: and than some poeple, like Sergey Lapin, are still twice there. Can't they stick to one e-mail address ?!?!? ;-) Mar 24 12:20:26 rschuster: You are a bad guy too ;) Mar 24 12:21:17 ah most of them are working for gmail ;) Mar 24 12:25:34 ok, submitted a draft... nwo food. I'll work on the details later. Mar 24 12:42:05 ~curse kernel Mar 24 12:42:10 May the fleas of a thousand camels infest your most sensitive regions, kernel ! Mar 24 12:42:47 fscking 'trying to set up times as Virtual Wire IRQ' Mar 24 12:42:56 s/times/timer Mar 24 12:43:07 so goodbye qemu/x86 Mar 24 12:43:30 03Koen Kooi 07org.openembedded.dev * r20ad023c27 10openembedded.git/recipes/linux/ (59 files in 12 dirs): linux-omap: add recipe for 2.6.29 and update git version to that as well Mar 24 12:54:04 uf. -no-acpi helped Mar 24 13:12:22 good morning Mar 24 13:12:43 hi mister poullet Mar 24 13:17:53 I've played with my config files and broke it all: http://pastebin.com/d1bcfdcc3 Mar 24 13:18:13 bitbake doesn't start anymore... Mar 24 13:18:58 what change did you make? Mar 24 13:20:23 03Angus Ainslie 07fso/milestone5.5 * rd4b76af992 10openembedded.git/ (5 files in 4 dirs): Add orrery to the feed Mar 24 13:22:39 Crofton|work: I wanted to add some new dirs to the collections setup Mar 24 13:23:26 Crofton|work: and to change beagleboard/beagleboard in configuration/beagleboard Mar 24 13:23:58 mpoullet|work: oh...I didn't intend to bring you there in the wood :-) Mar 24 13:24:52 I don't reember how, but I come too to these cryptic errors (cooker.py,...) Mar 24 13:25:08 mpoullet|work, look over your changes carefully Mar 24 13:25:23 bitbake is now known for helpfull error messages :( Mar 24 13:25:45 Crofton|work: I'm trying to build a oe demo to convince my colleagues to use oe Mar 24 13:25:50 it's hard Mar 24 13:25:56 mpoullet|work: on fritzs? Mar 24 13:25:56 :) Mar 24 13:26:40 anybody using Freescale i-MX21 based board here ? Mar 24 13:26:56 mckoan: yes, I do Mar 24 13:27:05 mckoan: but I'm still at the hack-the-kernel level Mar 24 13:27:19 schurig: which kernel version? Mar 24 13:27:21 hrw: for some in-house projects first, Mar 24 13:27:22 mckoan: as I don't have a working JTAG for imx21 yet and need to run my kernel via HaRET Mar 24 13:27:39 mckoan: GIT with some arm-related patches Mar 24 13:28:07 schurig: whick board? M9238MX21ADS ? Mar 24 13:28:25 mckoan: no, a PDA-like device for transportation and logistics Mar 24 13:29:01 mckoan: with Barcode-Laserscanner, GPS, GSM-GPRS, Bluetooth, Touchscreen, Keyboard, Mic, Loudspeaker and a hefty price-tag Mar 24 13:29:19 mckoan: but damn robust, you can use it to work nails into wood Mar 24 13:30:26 schurig: do you have a linux running? do you get the prompt? Mar 24 13:30:29 american one schurig ? Mar 24 13:30:32 mckoan: when I was sidetracked with other projects, I was at 2.6.29-rc4, and write some patches the i.MX21 clocks, a keypad driver, touchscreen driver for SPI-slave mode Mar 24 13:30:56 Genesis: american nails? american woods? No, european. Mar 24 13:31:14 mckoan: I have the kernel booting, but I'm not yet on the prompt Mar 24 13:31:59 http://www.amltd.com/ do some "PDA-like device for transportation and logistics" Mar 24 13:32:02 mckoan: without JTAG, I don't dare to modify the bootloader or the NAND-Flash, and I had issues with MMC. Just yesterday MMC-driver-fixes for i.MX21 got posted by Martin Fuzzey in the ARM mailing list Mar 24 13:32:19 Genesis: http://www.mn-solutions.de do it better! Mar 24 13:32:23 hihi Mar 24 13:32:36 schurig: do you want a JTAG debugger? Mar 24 13:33:20 I should start a new 2.6 port for M9238MX21ADS clone Mar 24 13:33:23 mckoan: I already have an Olimex USB-Tiny here (and used that dead-slow device with an AVR32 target), but this device needs a 3.3V ref voltage from the target, and I need to solder this yet Mar 24 13:33:48 schurig: Olimex USB-Tiny is a toy ;-) Mar 24 13:33:51 mckoan: I also have a (very fast!) BDI2000 here, but no software for it to handle Freescale i.MXxx-CPUs normally, only with PXA software Mar 24 13:34:14 schurig: I understand your problem now :-( Mar 24 13:35:06 mckoan: Sascha Hauer from Pengutronix is the Linux i.MXxx maintainer, he's quite nice to work with, but he's not the best maintainer, e.g. you never know when a patch is in his git tree and will go to RMK Mar 24 13:35:54 schurig: I'm trying to contact him in ptxdist irc Mar 24 13:37:19 mckoan: you already know http://git.pengutronix.de/?p=imx/linux-2.6.git ??? Mar 24 13:37:50 heh.. vortex86sx is fun... Mar 24 13:39:47 mckoan: the device that I have in the works now is at http://www.mn-solutions.de/leftmenu/prod/handhelds.html, the last one on the page Mar 24 13:39:53 schurig: I saw some patches for imx1 and imx27 but not imx21 Mar 24 13:40:40 schurig: but there are i.MX21 patches there as well, go to http://git.pengutronix.de/?p=imx/linux-2.6.git;a=shortlog and search for "arm/imx21", there you'll find my patches Mar 24 13:41:01 mckoan: (I spoke to myself :-) but there are i.MX21 patches there as well, go to http://git.pengutronix.de/?p=imx/linux-2.6.git;a=shortlog and search for "arm/imx21", there you'll find my patches Mar 24 13:41:02 ok thx I'll take a look Mar 24 13:41:20 mckoan: the first URL I gave is just the summary, the shortlog goes further back Mar 24 13:41:45 schurig shone Gerate :-) Mar 24 13:49:11 oh..these devices like Dymo-Etikett Mar 24 13:49:18 s/shone/schÃ¶n Mar 24 13:51:00 mckoan: "Schöne Geräte" would be ok :-) Mar 24 13:51:12 s/schÃ¶n/praktisch Mar 24 13:51:37 ö like the middle sound in "Sir", ä like the starting sound of attention Mar 24 13:51:52 how's deutsch for 'ruggedized' ? Mar 24 13:52:04 and for sch "Sch" I have refer you to some scotsman :-) Mar 24 13:52:10 ant_work: robust? Mar 24 13:52:29 i have just built a x11-image for omap3evm. when i boot the kernel i get messages like Unknown HZ value! (82) Assume 100, Buffer I/O error on device mtdblock2 and lots of DISPC errors and my network interface does not get up Mar 24 13:56:34 the unknown Hz is a recent thing iirc (>2.6.26) Mar 24 13:58:06 someone hinted about hacking procps Mar 24 13:59:23 somebody told me 'ignore them' ... Mar 24 14:03:40 ant_work which board you are using? Mar 24 14:03:55 it got something to do with defconfig i think Mar 24 14:04:15 or file system, i download same kernel and compile and run it i dont get these messages Mar 24 14:05:02 I see it on armv5te (pxa255) Mar 24 14:05:03 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=460331 Mar 24 14:05:08 seems solved Mar 24 14:05:23 in OE we have procps_3.2.7.bb Mar 24 14:06:50 sh1: CONFIG_HZ=100" ? Mar 24 14:08:01 ant_work:i dont know, i need to look into that Mar 24 14:08:24 well, it does not help here with linux-rp-2.6.26... Mar 24 14:08:50 btw I have cpufreq enabled Mar 24 14:09:07 ant_work: i dont think cpufreq is the issue here Mar 24 14:09:14 me neither :-) Mar 24 14:09:33 it seems it happens on pxa270 too (without cpufreq) Mar 24 14:09:44 I'm too lazy to test ... Mar 24 14:10:14 you see.. 'ubi major, minor cessat' .. there are other nastier bugs... Mar 24 14:10:32 (e.g. powerman) Mar 24 14:12:31 do we still have kernel-hackers here btw? Mar 24 14:15:14 yes I guess so :) Mar 24 14:15:27 :-) Mar 24 14:16:28 florian...our Zaurus Odyssey is ending abruptely agains 2.6.29 (RMK vs. Pawel Machek) Mar 24 14:16:49 cause: lack of maintainers...(RP) Mar 24 14:18:01 ant_work: sound like i should read the arm-linux-kernel list Mar 24 14:18:17 hi all... I'm having problems bitbaking an image... the error I get is: Mar 24 14:18:17 fakeroot, while creating message channels: No space left on device Mar 24 14:18:17 This may be due to a lack of SYSV IPC support. Mar 24 14:18:17 fakeroot: error while starting the `faked' daemon. Mar 24 14:18:18 kill: usage: kill [-s sigspec | -n signum | -sigspec] pid | jobspec ... or kill -l [sigspec] Mar 24 14:18:33 anyone any idea about the cause for that? Mar 24 14:27:30 florian: http://lkml.org/lkml/2009/1/28/102 Mar 24 14:27:45 Russel King seems fed up Mar 24 14:28:41 heh Mar 24 14:28:44 rmk is like this Mar 24 14:28:50 so is Pavel Mar 24 14:29:31 rofl Mar 24 14:29:33 lalalala Mar 24 14:30:00 ^_^ Mar 24 14:30:08 heh Mar 24 14:30:17 two rather smart and forceful personalities Mar 24 14:31:48 Is it possible to specify PROVIDES as a regular expression, or at least a mask? Mar 24 14:32:27 florian: we could live with 2.6.26 if some minor power mgmt issues would be solved Mar 24 14:32:49 Like, I'd like to create a simplier linux kernel package, with all the modules and the kernel in one package. But lots of packages have kernel-module-foo in RRECOMMENDS.. Mar 24 14:32:59 even, Omegamoon/Cortez (Zubuntu) has a semi-working 2.6.28 Mar 24 14:33:12 Ã¹just no power management... Mar 24 14:33:27 who could I bother ? Mar 24 14:33:35 CosmicPenguin, yeah Mar 24 14:33:39 xl0: no, not. What is your use-case for such a feature? Mar 24 14:34:04 xl0: oh, sorry, two lines below you wrote it Mar 24 14:34:34 ~lart power management on arm Mar 24 14:34:34 * ibot makes a balloon animal out of power management on arm Mar 24 14:35:09 schurig: Any ideas, is there a way to make this work without touching all the packages that have kernel modules runtime recommendations? Mar 24 14:35:36 xl0: you can add lots of RPROVIDES += to your kernel package Mar 24 14:38:15 ant_work: so looks like Zaurus support died Mar 24 14:38:51 hrw: it was ill since inception Mar 24 14:39:33 schurig: but during that time it was keeping alive Mar 24 14:39:47 hrw: the good thing is, after all these years I have learned a lot and now I'd know how to treat a new device (Pandore, where are you?) Mar 24 14:40:34 need go go to customer Mar 24 14:40:49 hrw: what puzzles me is 2.6.28 isn't too far Mar 24 14:41:55 hrw: Omegamoon just need a small help...(he has debian behind) Mar 24 14:42:01 http://www.omegamoon.com/blog/ Mar 24 14:43:52 Is it possible to exclude packages with unsatiafied RRECOMMENDS from the built image? Mar 24 14:54:45 Are there any real Zaurus users? Mar 24 14:55:08 xl0: good question Mar 24 14:58:01 I'd say yes, look at the OESF forums Mar 24 14:59:13 yes! x11 on vortex86sx started ;) Mar 24 14:59:20 ~curse sis driver Mar 24 14:59:21 May you be reincarnated as a Windows XP administrator, sis driver ! Mar 24 14:59:31 ~curse vortex86sx for being so crappy cpu Mar 24 14:59:32 May you be reincarnated as a Windows XP administrator, vortex86sx for being so crappy cpu ! Mar 24 15:02:13 hrw: how is we see x86sx in 2009? Mar 24 15:02:26 which advantages? Mar 24 15:03:27 ant_work: only price Mar 24 15:03:56 ant_work: other then price (and good framebuffer with acceleration) there are no advantages of vortex86sx based devices rather. Mar 24 15:05:14 morning Mar 24 15:07:23 hello kergoth Mar 24 15:08:00 hrw: I'm thinking about a self-built router for my office Mar 24 15:08:52 but now some router/firewalls are soo cheap... Mar 24 15:09:34 ant_work: avoid vortex86sx Mar 24 15:30:45 03Theodore A. Roth 07org.openembedded.dev * rd60324b13d 10openembedded.git/docs/usermanual/reference/class_siteinfo.xml: class_siteinfo.xml: Replace packages/ with recipes/. Mar 24 15:30:55 03Theodore A. Roth 07org.openembedded.dev * r952f5765ac 10openembedded.git/docs/usermanual/chapters/ (6 files): doc/usermanual/chapters: Replace references to packages dir with recipes dir. Mar 24 15:37:38 did someone of you saw situation when X11 starts and simple 'DISPLAY=:0 xrandr' makes it die? Mar 24 15:38:33 death by segfault? Mar 24 15:38:45 no Mar 24 15:39:17 (EE) SIS(0): Unable to map IO aperture. Invalid argument (22) Mar 24 15:39:34 ouch Mar 24 15:40:16 I do not think that driver was ever tested on 486 Mar 24 15:40:37 http://cgit.freedesktop.org/xorg/driver/xf86-video-sis/tree/src/sis_driver.c#n7158 Mar 24 15:40:48 03Angus Ainslie 07fso/milestone5.5 * r8eaa208ce0 10openembedded.git/packages/xorg-xserver/xserver-xorg-conf/om-gta02/xorg.conf: Add xorg.conf file for om-gta02 Mar 24 15:41:04 CosmicPenguin: I know - I have that file opened Mar 24 15:42:12 pci_device_map_range could totally not be 486 friendly Mar 24 15:42:16 I could see that Mar 24 15:43:02 though we use it for geode, so perhaps the problem is more subtle Mar 24 15:43:10 mckoan: you do not want x11 on nanopc ;) Mar 24 15:43:19 CosmicPenguin: geode is more 586 Mar 24 15:43:38 CosmicPenguin: you have fpu, cmov and few other acronyms Mar 24 15:43:42 thats true Mar 24 15:43:51 but none of those should matter in this particular situation Mar 24 15:44:02 yep Mar 24 15:44:19 anyway 486 was pre-pci cpus.. Mar 24 15:49:56 hrw: yes I would Mar 24 15:52:53 but 486 linux should be able to cope with PCI - invalid argument from mmap is usually because the offset isn't page aligned Mar 24 15:53:04 poorly constructed PCI header? Mar 24 15:54:44 enough fighting for today Mar 24 15:54:54 this device suxx Mar 24 15:55:24 I miss times when I had ATI onboard in 386sx+fpu Mar 24 15:55:50 im use midori browser, for some reason it is not able to browse the internet. how do i specify the ipaddresses of dns servers Mar 24 15:56:48 # Dynamic resolv.conf(5) file for glibc resolver(3) generated by resolvconf(8) Mar 24 15:56:48 # DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN Mar 24 15:56:48 nameserver 192.168.20.13 Mar 24 15:56:48 nameserver 192.168.1.1 Mar 24 15:56:48 search buglabs.net lan Mar 24 15:56:54 ops Mar 24 15:57:05 fireplay: such stuff in /etc/resolv.conf Mar 24 15:57:25 mckoan: gfx chip is nice anyway Mar 24 16:01:29 hrw: thanks, works, udhcp server will put dns names into /etc/resolv.conf Mar 24 16:20:39 where can I find libintl? Mar 24 16:20:55 there is no such library in my staging dir Mar 24 16:26:11 i use mplayer image and i dont have gui. cant i have a gui for mplayer Mar 24 16:27:35 we do not build gmplayer Mar 24 16:27:40 aloisiojr: hi! Mar 24 16:28:42 which package to use to play videos/songs Mar 24 16:30:20 jeremy_laine: hey! Mar 24 16:31:03 aloisiojr: I'm eager to take a look at your work on qt4 packaging, is there a copy of your .bb floating around somewhere? Mar 24 16:31:45 03Jeremy LainÃ© 07org.openembedded.dev * r08f02468d5 10openembedded.git/recipes/linux/ (6 files in 2 dirs): linux-2.6.27: update patches for boc01 Mar 24 16:31:55 03Jeremy LainÃ© 07org.openembedded.dev * rd14a90c128 10openembedded.git/: Merge branch 'org.openembedded.dev' of ssh://git@git.openembedded.net/openembedded into org.openembedded.dev Mar 24 16:32:21 actually i have just a local copy, but i can put it somewhere Mar 24 16:32:48 jeremy_laine: i didn't finish yet... (i need more free time) Mar 24 16:33:46 I'm curious about this: NOTE: Multiple libraries (libnss_hesiod.so.2, libnss_nis.so.2, libnss_nisplus.so.2) found, but LEAD_SONAME 'libc.so' doesn't match any of them Mar 24 16:34:05 what does that mean? is it bad? Mar 24 16:34:43 aloisiojr: sure, a pastebin or the like is good Mar 24 16:37:58 jeremy_laine: cat classes/qt4.bbclass > http://pastebin.com/d4c215264 Mar 24 16:38:19 mpoullet|work: me again: Mar 24 16:38:21 Aug 07 21:30:09 pb__: building glibc-2.6.1 I got this note: NOTE: Multiple libraries (libnss_nis.so.2, libnss_nisplus.so.2, libnss_hesiod.so.2) found, but LEAD_SONAME 'libc.so' doesn't match any of them Mar 24 16:38:34 why the diff in some recipes between functions named for example 'do_install' and 'do_install_${PN}' ? aren't they the same when declared within a recipe? Mar 24 16:39:02 mpoullet|work: Aug 07 22:25:14 ant_: that's harmless, you can ignore it Mar 24 16:39:31 ant_work: ok, thanks, good to know Mar 24 16:39:39 jeremy_laine: cat classes/qmake-defs.bbclass > http://pastebin.com/dfc88e35 Mar 24 16:40:15 aloisiojr: I am just curious, how come these are .bbclass files? it looks as though it's just to compile Qt, not Qt apps right? Mar 24 16:40:55 aloisiojr: also, it looks as though all Qt plugins are put into the "plugins" package without splitting, right? Mar 24 16:41:35 jeremy_laine: no, -plugins package is a meta-package, there are -plugin-* packages Mar 24 16:41:48 aloisiojr: ok, my bad sorry :) Mar 24 16:41:58 jeremy_laine: it's just qt, actualy i create a bbclass file because qt actually build in a particular way... Mar 24 16:42:33 qt applictions will use qmake.bbclass, that depends on the qt build to use its tools (qmake/moc/uic) Mar 24 16:43:27 jeremy_laine: as you can notice, i'm staging the native tools and the specs Mar 24 16:44:44 jeremy_laine: the new proposed qmake2.bbclass is not created yet, but it's very simple to do Mar 24 16:46:44 jeremy_laine: a new qt4.inc file was created to list the patches (only 4 til now) and to choose the configure flags Mar 24 16:52:36 aloisiojr: I like the idea of having the tools staged, it avoids having to update 3 different packages whenever we update Qt Mar 24 16:52:58 its interesting that http://docs.openembedded.org/usermanual/usermanual.html#update-alternatives_class refers to bash/busybox being an example of using update-alternatives class yet the bash recipes do not use it - anyone know why? Mar 24 16:54:51 jeremy_laine: exacltly.... further i think we should provide a recipe to build target binaries to qt-tools, to be used in sdks, etc Mar 24 16:55:21 jeremy_laine: sis you look the file, do you have any comments? improvements? Mar 24 16:55:32 s/sis/did/ Mar 24 16:55:50 o fsck... Mar 24 16:56:06 1280x1024 x11/vesa works on vortex86sx. Mar 24 16:56:22 but lack of speed kills Mar 24 16:56:31 zaurus at vga is much faster Mar 24 16:58:29 has support been added to OE for UBIFS yet? Mar 24 16:58:49 iirc koen is using it already Mar 24 16:59:18 heh.. even vortex86sx at vga is slower then zaurus Mar 24 16:59:34 florian: created by hand Mar 24 16:59:41 ah nasty Mar 24 16:59:44 I also use ubifs on device but also created by hand Mar 24 17:00:34 aloisiojr: are you working on qt 4.5 recipe? Mar 24 17:01:44 mckoan|away: yep... actually i'm restructuring the qt build, and i used qt 4.5 as the first recipe Mar 24 17:02:11 we are working on the same :-| Mar 24 17:03:19 aloisiojr: would be possible to join our efforts starting from tomorrow? Mar 24 17:03:19 mckoan|away: ow... so, let's merge :) Mar 24 17:03:47 mckoan|away: of course! Mar 24 17:04:07 will you be here tomorrow late morning? Mar 24 17:05:22 mckoan|away: which time zone are you? :D Mar 24 17:05:54 ITA CET Mar 24 17:06:06 GMT +1 Mar 24 17:06:21 bbl Mar 24 17:07:05 i'm on GMT -3, so it's early morning here Mar 24 17:07:11 yep, can be Mar 24 17:08:20 aloisiojr: where are you there? Mar 24 17:08:34 recife - brazil Mar 24 17:08:47 obrigado, see you tomorrow :-) Mar 24 17:09:20 mckoan|away: arrivederte Mar 24 17:10:03 * aloisiojr thanks google translator \o/ Mar 24 17:10:33 BTW is arrivederci ;-) Mar 24 17:10:51 auhauahauh Mar 24 17:11:01 google sun of *** Mar 24 17:11:54 it usually helps, but you need to know the background of the language you want to use Mar 24 17:12:16 background = main grammatical rules Mar 24 17:12:34 BBL Mar 24 17:12:45 as you can see, i don't Mar 24 17:19:04 see you guys Mar 24 17:19:07 hrw, that is wht I was going to do if it was not supported yet. everything I have read is ubifs is quicker than JFFS2 and just as stable, is that what you have observerd? Mar 24 17:19:21 svolpe: better it is Mar 24 17:19:59 hrw|gone: good, I think I will switch over to it for several projects :-) Mar 24 17:20:26 hrw|gone: JFFS2 is killing me for mount times even with summery support enabled. Mar 24 17:24:25 03Koen Kooi 07org.openembedded.dev * r1e56480635 10openembedded.git/recipes/ (4 files in 3 dirs): libsoup: fix packaging to keep upgrade paths sane, bump PR on affected packages Mar 24 17:24:27 03Koen Kooi 07org.openembedded.dev * rff2318d29d 10openembedded.git/recipes/dsplink/ (ti-codec-engine_2.21.bb ti-dmai_1.20.bb): dmai, cmem: bump PR for kernel change Mar 24 17:24:27 03Koen Kooi 07org.openembedded.dev * r362512f7d3 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 24 17:40:25 So I have a bunch of devs who want to sym link sources to the same place, but it seems that bitbake makes the files writeable by the ownder ONLY. Is this on purpose? Mar 24 17:41:31 BTW I am usig bitbake 1.8.10 Mar 24 17:42:10 waite: what sources you want to symlink ? metadata or the srcs of downloaded packages Mar 24 17:42:28 srcs of downloaded pkgs. Mar 24 17:43:38 Perms for source packages and md5 sums are 0x0644 (read/write owner, read group, read other) Mar 24 17:44:17 So other users fail because the src files are not writable Mar 24 17:44:40 hmm I dont think bitbake does that Mar 24 17:44:59 so who is resp. for do _fetch()? Mar 24 17:45:05 probably check umask in your shell Mar 24 17:45:48 ERROR: '/home/mpoullet/oe/openembedded/recipes/gnome/libsoup-2.4_2.26.0.bb' failed Mar 24 17:46:52 configure doesn't find sqlite http://pastebin.com/d67d2d11b Mar 24 17:47:06 khem, thanks That sounds resonable. Let me dive into that a bit. Mar 24 17:48:54 mpoullet|work: add it to DEPENDS Mar 24 17:50:21 khem: ok, trying this Mar 24 18:05:38 * cbrake digs out an old geode board to do some x86 testing ... Mar 24 18:06:04 cbrake: any opinion on my distutils RFC ? Mar 24 18:10:15 khem: checking Mar 24 18:15:37 khem: seems like a good idea, and I don't seem any obvious problems. I could do some clean builds tonight if that would help for testing. Mar 24 18:17:51 anyone else find it odd that the wget fetcher uses FETCHCOMMAND/RESUMECOMMAND, but not FETCHCMD_wget, and svn uses FETCHCMD_svn, but doesn't use FETCHCOMMAND/UPDATECOMMAND at all.. Mar 24 18:17:54 gogo consistency Mar 24 18:24:02 03Koen Kooi 07org.openembedded.dev * r5a03b8a7b4 10openembedded.git/recipes/dsplink/ti-dmai_1.20.bb: Mar 24 18:24:02 ti-dmai: fetch from svn Mar 24 18:24:02 * gets us some buildfixes for free Mar 24 18:24:02 * no need to download it from my.ti.com, which requires a login Mar 24 18:24:12 03Koen Kooi 07org.openembedded.dev * r6306b3fa6f 10openembedded.git/classes/kernel.bbclass: kernel.bbclass: stage more files Mar 24 18:24:13 03Koen Kooi 07org.openembedded.dev * rdf49f5e7f5 10openembedded.git/recipes/linux/ (linux-omap-2.6.28/beagleboard/defconfig linux-omap_2.6.28.bb): linux-omap 2.6.28: bump PR Mar 24 19:00:09 it seems that supporting another toolchain than the GNU one would require a lot of work...it seems hardcoded in base.bbclass Mar 24 19:01:02 what do you mean? Mar 24 19:01:23 for instance if you want to use another compiler such as ACK or something else... Mar 24 19:01:35 ah there is only one line hardcoded... Mar 24 19:01:36 first of all, anything in base.bbclass can be overridden by another class. Mar 24 19:01:44 ah ok wow Mar 24 19:02:36 so even if there was hardcoding, which i doubt, you could write a class that changes it and add it to INHERIT so it affects all packages. Mar 24 19:03:03 and I bet bitbake.conf can also be overridden by the local.conf Mar 24 19:03:09 wow thanks a lot Mar 24 19:05:32 bitbake.conf includes local.conf, and every other .conf, in a particular order Mar 24 19:05:45 so yes, thats how it was designed, local.conf can override everything thats loaded before it Mar 24 19:06:17 03Jeremy LainÃ© 07org.openembedded.dev * rd0b2a8dd1b 10openembedded.git/recipes/qt4/ (qt-embedded_4.4.3.bb qt4-x11-free_4.4.3.bb qt_staging.inc): qt4-x11-free, qt-embedded: stage mkspecs Mar 24 19:07:53 Gnutoo: the intent was always that machine & distro would be able to override global, and local would be able to override those. of course, thats difficult, since its often local.conf that -defines- machine and distro, thats why those are loaded after it.. Mar 24 19:08:26 ok thanks a lot Mar 24 19:09:47