**** BEGIN LOGGING AT Thu May 26 02:59:57 2011 May 26 04:33:51 who would I talk to about write access to oe-core-contrib ? May 26 04:34:15 I'm currently pushing to poky-contrib, but for all the oe-core bits I'd rather work directly with oe May 26 06:41:51 good morning May 26 06:42:05 dvhart: ask to OE mailing list May 26 07:46:23 03Paul Eggleton  07master * rd0cffcf1fd 10openembedded.git/recipes/qte/ (4 files in 2 dirs): May 26 07:46:23 qte: enable 18bpp support May 26 07:46:23 This adds 18bpp framebuffer support for devices such as rokre6, utilising May 26 07:46:23 an old patch from mkezx. (You may need to set the framebuffer to 24bpp May 26 07:46:23 using fbset before starting Opie.) May 26 07:46:23 Signed-off-by: Paul Eggleton May 26 07:46:33 03Paul Eggleton  07master * r6d450abfc1 10openembedded.git/recipes/qte/qte-common_2.3.10.inc: May 26 07:46:33 qte: revert changes from 2a1c9e1ee5eb8067592d26100b563dd2ae7be00d May 26 07:46:33 The recipe qpf-qte already provides the font files in lib/fonts split out May 26 07:46:33 into separate packages, so we wouldn't also install them here. May 26 07:46:34 Signed-off-by: Paul Eggleton May 26 07:46:34 03Paul Eggleton  07master * r05509a1986 10openembedded.git/recipes/manufacturers/manufacturers_20031209.bb: May 26 07:46:34 manufacturers: define a SRCREV to fix package build May 26 07:46:35 Signed-off-by: Paul Eggleton May 26 07:47:02 bluelightning yeah May 26 08:01:41 03Martin Jansa  07master * r608e60f0ef 10openembedded.git/recipes/linux/linux-2.6.39/om-gta02/defconfig: May 26 08:01:41 linux-2.6.39: sync om-gta02 defconfig with defconfig from larsc a bit May 26 08:01:41 Signed-off-by: Martin Jansa May 26 08:36:34 hi stefan schmidt May 26 09:20:02 morning all May 26 09:20:47 he bluelightning, thanks for the font stuff May 26 09:21:52 hi woglinde, no problem May 26 09:22:18 I did a from-scratch build for hx4700, Opie starts fine without qpf fonts May 26 09:22:59 unfortunately for that machine there is now a problem with tslib as well as udev... may have to pin those in the machine conf or something May 26 09:23:07 *sigh* May 26 09:24:14 bluelightning did you use tslib-git? May 26 09:24:34 woglinde: tried that... but the only change was in the error message May 26 09:25:00 instead of "protocol is not one I recognise" it was "tslib has not been compiled for the protocol version in use" May 26 09:25:11 it's because of 2.6.21-hh May 26 09:25:25 not sure when this broke though May 26 09:25:43 tslib-1.0 ought to have worked May 26 11:03:17 03Koen Kooi  07org.openembedded.dev * r3b80e76fc6 10openembedded.git/recipes/systemd/systemd_git.bb: May 26 11:03:18 systemd: bump to a rev where hwclock isn't needed anymore May 26 11:03:18 Signed-off-by: Koen Kooi May 26 11:03:28 03Anand Balagopalakrishnan  07org.openembedded.dev * r436621b821 10openembedded.git/recipes/powervr-drivers/libgles-omap3.inc: May 26 11:03:28 libgles-omap3: Add wsegl.h to the destination include directory May 26 11:03:28 * Copy wsegl.h to the destination include directory. May 26 11:03:28 * This file is required to compile Qt's PowerVR graphics driver plugin. May 26 11:03:29 Signed-off-by: Anand Balagopalakrishnan May 26 11:03:29 Signed-off-by: Koen Kooi May 26 11:03:30 03Koen Kooi  07org.openembedded.dev * r86709c3867 10openembedded.git/recipes/udev/udev_git.bb: May 26 11:03:30 udev git: bump SRCREV for systemd fixes May 26 11:03:30 Signed-off-by: Koen Kooi May 26 11:07:49 03Anand Balagopalakrishnan  07org.openembedded.dev * rf63289cfa7 10openembedded.git/recipes/qt4/ (qt-4.6.3.inc qt-4.7.3.inc): (log message trimmed) May 26 11:07:49 qt4: Fix the include path for SGX PowerVR drivers May 26 11:07:49 * Qt's PowerVR graphics driver plugin requires the PowerVR header files May 26 11:07:49 pvr2d.h, wsegl.h to compile May 26 11:07:49 * The current recipes copy the wsegl.h from Qt's src/3rdparty/powervr to May 26 11:07:50 powervr graphics driver plugin directory - src/plugins/gfxdrivers/powervr. May 26 11:07:50 * If/when the header files change, the incorrect wsegl.h will get picked up. May 26 11:07:58 03Koen Kooi  07org.openembedded.dev * r6cb2c8e50d 10openembedded.git/recipes/qt4/ (qt4-embedded.inc qt4-x11-free.inc): May 26 11:07:58 qt4: bump PR for wsegl header change May 26 11:07:58 Signed-off-by: Koen Kooi May 26 11:53:19 03Koen Kooi  07org.openembedded.dev * r578563341a 10openembedded.git/conf/distro/include/angstrom-2010-preferred-versions.inc: May 26 11:53:20 angstrom: prefer the git version of tslib May 26 11:53:20 Recent kernels don't work too well with the old tslib May 26 11:53:20 Signed-off-by: Koen Kooi May 26 12:17:59 anarsoul: btw I notice h1940 and rx1950 are using old kernels in oe-dev - do you have patches to update these? May 26 12:24:47 bluelightning: afaik rx1950 uses 2.6.35 May 26 12:25:05 bluelightning: no oe patches yet, I'm using self-built kernel May 26 12:25:24 anarsoul: how much of your work has hit mainline? May 26 12:25:40 bluelightning: all, except wifi for rx1950 :) May 26 12:25:49 because acx100-max80211 is out of tree May 26 12:26:09 Hi all, I am trying to check my new recipe works before sending it to the mailinglist for inclusion. Do I need to test on branch "org.openembedded.dev" or "2011.03-maintenance" is enough? May 26 12:26:33 new is master May 26 12:26:48 anarsoul: ah, ok... so in theory h1940 could use 2.6.39 and rx1950 could be 2.6.39 + wifi patch? May 26 12:27:52 yep for h1940, for rx1950 wifi driver should be built via acx100-mac80211 recipe May 26 12:28:46 bluelightning: I tested 2.6.39-rc2 on h1940 and rx1950, not sure about final 2.6.39, will check on weekend May 26 12:28:55 anarsoul: ok cool.. I'll test h1940 with 2.6.39 also May 26 12:29:10 bluelightning: you own one? :) May 26 12:29:32 yes, some months ago once I saw you were working on h1940 I bought one :) May 26 12:30:37 I see. I use Angstrom and have used their script to configure everything with the maintenance branch. I thought I could just replace the openembedded directory with a clone from master but I get an error with bitbake May 26 12:30:40 bluelightning: I also use 0001-s3c2410_ts-add-fake-pressure-events.patch from recipes/linux-2.6.35/rx1950, but it's not necessary with recent tslib May 26 12:31:07 bitbake -c clean linphone - ERROR: Error parsing /Data/work/angstrom/angstrom-setup-scripts/sources/openembedded/recipes/images/xorg-image.bb: Could not inherit file classes/rootfs_${IMAGE_PKGTYPE}.bbclass May 26 12:31:23 What am I doing wrong? May 26 12:31:50 anarsoul: well, as of today angstrom in master uses tslib-git so if you're using angstrom I guess that patch should not be needed May 26 12:32:00 Guillaume_: can you pastebin your whole recipe? May 26 12:32:36 actually hrm... that's a standard recipe May 26 12:32:49 bluelightning: my rootfs for h1940 is a bit old, did not rebuild it for ~0.5 year :) May 26 12:33:16 also there was a report about non-working bluetooth on h1940, but I didn't investigate it due to lack of time May 26 12:33:58 http://hs.cxio.info/index.php/2011-05-26_21:00 May 26 12:34:05 ups, wrong channel May 26 12:36:41 bluelightning: you can find the patch here: http://pastebin.com/vk4nnmYy May 26 12:37:15 bluelightning: I put the whole patch as it consist of several files May 26 12:38:24 Guillaume_: hmm, I'm getting the same error here now :/ May 26 12:40:00 bluelightning: have you an idea where the IMAGE_PKGTYPE variable comes from? May 26 12:42:07 bluelightning: I find out this variable is defined at the top of classes/image.bbclass May 26 12:44:16 Guillaume_: this may be because you have set DISTRO to angstrom-2008.1 which is no longer available in master (use angstrom-2010.x instead) May 26 12:44:28 that's what was doing it here on this machine May 26 12:44:47 unfortunately our current setup just ploughs on ahead even if the DISTRO conf file doesn't exist :/ May 26 12:48:56 bluelightning: DISTRO_DIRNAME is something I can choose arbitrarily? I need to change it and OE_BUILD_TMPDIR to not have conflicts with my old 2008 installation. Right? May 26 12:49:38 Crofton: ping? May 26 12:50:05 pong May 26 12:51:09 Guillaume_: I'm not sure... git grep doesn't give me any results so I don't know what that variable is I'm sorry May 26 12:51:35 zecke, pong May 26 12:54:51 Crofton: with the sffsdr... do you know how the DSP is turned on? May 26 12:55:07 not really May 26 12:55:14 Crofton: or if you can reset the dsp from userspace? May 26 12:55:21 man dsplink :) May 26 12:55:40 there is some trick, but I do not know what it is May 26 12:56:41 bluelightning: thanks, I will take you informed when the (long) compilation finish ;) May 26 12:58:23 zecke you may want to ask koen or jkridner May 26 12:59:53 thanks May 26 13:00:07 not sure how good they are at the DSP though May 26 13:00:19 zecke, also look at the TI 2e2 website May 26 13:01:00 thanks, yeah.. trying to read it now May 26 13:16:31 bluelightning: what do you think about netsurf port for opie? :) May 26 13:16:41 bluelightning: opie really lacks nice browser May 26 13:20:29 anarsoul hm it has no javascript May 26 13:20:52 woglinde: who cares about javascript on 32-64mb devices? :) May 26 13:20:58 *g* May 26 13:38:50 anarsoul: who cares about 32-64mb devices? May 26 13:39:34 hrw: those who care about opie May 26 13:39:52 ok, ugly joke it was May 26 13:43:05 640K for every javascript developer May 26 13:44:04 * mwester-laptop grrrrs at hrw's ugly joke. :D May 26 13:44:28 mwester cares for opie? May 26 13:44:39 * mwester-laptop cares for NSLU2s May 26 13:45:25 Well, when I have time -- been rather busy lately, and haven't been able to spend much time on OE and hobby stuff. May 26 13:46:17 g'day kergoth May 26 13:51:28 mwester-laptop: nslu2 was/is nice device May 26 13:54:52 Yep. Still waiting for a suitable replacement... the sheevaplug and similar devices are technically very good, but don't compare to the availability that the NLSU2 (and some of the other devices) had... May 26 13:55:42 Certain devices have attracted a lot of people to the linux world, such as the hardware upon which Opie runs, and the NSLU2, for example. May 26 14:11:46 anarsoul: yes it's definitely a missing feature... something that's been neglected May 26 14:12:24 I don't know if netsurf would be the best platform to build on though May 26 14:13:55 bluelightning: webkit is no-go on low-mem devices, so netsurf is the only choice May 26 14:14:24 anarsoul: there's no way to configure webkit to be lean enough? May 26 14:15:12 bluelightning: well, for 64mb devices it *may* be possible, but definitely not for 32mb May 26 14:17:20 anarsoul: just came accross a discussion, someone else was asking about this and apparently the webkit devs don't target such platforms and don't want it to be configurable for them May 26 14:17:35 I can sort of see their point, cut out too many pieces and it's not webkit anymore May 26 14:21:53 bluelightning: I'm running netsurf (with sdl backend) on zipit z2 with 32mb of ram, and it works fine here May 26 14:23:37 anarsoul: interesting... presumably it's more capable than konq/e was (when it built...)? May 26 14:38:06 Tartarus, I am sending a poll request for the release branch. Let me know if it is OK. May 26 14:43:29 Crofton: thanks for your email May 26 14:46:39 did it help? May 26 14:51:00 03Koen Kooi  07org.openembedded.dev * r24513855be 10openembedded.git/recipes/linux/ (77 files in 10 dirs): May 26 14:51:00 linux-omap 2.6.39: sync with meta-texasinstruments May 26 14:51:00 Signed-off-by: Koen Kooi May 26 15:19:08 Crofton, k, thanks May 26 15:19:43 hi Tartarus May 26 15:19:47 hi mickeyl May 26 15:19:58 hi woglinde May 26 17:09:11 anyone using OE with a 68332 CPU? I didn't find anything via grep in conf May 26 17:12:07 I don't know of anybody doing any 68k stuff with oe at all, though it probably wouldn't be very hard to get it going. May 26 17:15:06 does the 68332 run full linux, or is it a uclinux-only target? May 26 17:19:46 not clear May 26 17:20:00 does the datasheet say that it has an mmu? May 26 17:20:06 I think most of the deeply embedded 68ks don't May 26 17:21:01 don't believe it does - is that the current focus of uclinux kernel, to support mmu-less? I was under the impression that all got merged into mainline a while back May 26 17:21:20 I think its mmu-less May 26 17:21:33 I've only ever used uclinux May 26 17:22:34 I think it is true that the mainline Torvalds kernel supports nommu nowadays, but you still need a uclinux-style userland world in that case May 26 17:23:02 yeah, freescale website seems to be saying that 683xx is all nommu May 26 17:23:26 only 68030, 68040, 68060 are mmu equipped May 26 17:24:43 * pb_ heads to the beer festival May 26 17:24:45 later all May 26 17:28:26 it must be Beer:30 someplace. May 26 17:57:49 quick question... matchbox-keyboard_svn lists ${datadir}/applications as a FILES_${PN}, but for my machine, MACHINE=bug20, I don't want that as an icon, so I want to remove it from the FILES_${PN} list, or in some other way make sure it's not installed when MACHINE=bug20 May 26 17:57:53 any suggestions? May 26 17:58:04 I thought maybe FILES_${PN}_bug20 or somesuch way May 26 17:58:56 or do I need to create a patch that removes matchbox-keyboard.desktop, stick it in files/bug20/mypatch.patch and list it in SRCURI ? May 26 18:00:38 not sure the correct oe way of solving this May 26 18:14:04 jconnolly hm I would try FILES_ first May 26 18:14:07 and than patch May 26 18:15:25 then I'd need to IMAGE_INSTALL matchbox-keyboard-bug20, is that right? May 26 18:16:52 jconnly hm a desktop file is more than an icon May 26 18:18:40 jconnolly hm okay I would do an do_install_prepend_bug20() May 26 18:18:52 which just removes the file May 26 18:19:01 that seems like a good approach, woglinde May 26 18:19:02 or wait append May 26 18:19:08 append is better May 26 18:19:09 indeed May 26 18:19:19 so you dont needed to fiddle the Makefiles May 26 18:19:30 yes, that's smart. Thanks May 26 18:19:50 remove it from ${D} May 26 18:20:07 okay May 26 18:20:10 ask again May 26 18:20:14 when it dont works May 26 18:24:09 hehe. so do_install_append_bug20(){ rm ${D}/${sysconfdir}/applications/inputmethods/${PN}.desktop} May 26 18:24:16 something like that May 26 18:25:24 right May 26 18:41:00 03Martin Jansa  07master * r275c62bfeb 10openembedded.git/recipes/linux/ (4 files in 3 dirs): May 26 18:41:00 linux_2.6.39: revert MMC changes from openmoko.patch in shr.patch May 26 18:41:00 Signed-off-by: Martin Jansa May 26 19:59:50 03Bernhard Guillon  07master * r06addb6a8d 10openembedded.git/recipes/opkg-utils/ (opkg-utils/arfile_header_split.patch opkg-utils_svn.bb): (log message trimmed) May 26 19:59:51 opkg-utils: port arfile.py patch to opkg-utils, bump PR May 26 19:59:51 We need this patch for opkg as well May 26 19:59:51 original commit b3da649ee25374f284d03ebce5843b6326559f73 May 26 19:59:51 original Author: Scott Anderson May 26 19:59:51 original commit message: May 26 19:59:52 ipkg-utils: Make arfile.py handle six digit UIDs May 26 20:06:28 xora ping? May 26 23:49:50 esbenh: ping May 27 00:07:42 * kergoth ponders May 27 00:09:06 anyone here have access to delete personal oe git branches? **** ENDING LOGGING AT Fri May 27 02:59:58 2011