**** BEGIN LOGGING AT Fri Mar 12 02:59:57 2010 Mar 12 06:03:30 03Martin Jansa  07org.openembedded.dev * r8539282f8c 10openembedded.git/conf/distro/include/sane-srcrevs-fso.inc: Mar 12 06:03:30 fsogsmd: bump SRCREV for fixed build with enabled modem-qualcomm-palm Mar 12 06:03:30 Signed-off-by: Martin Jansa Mar 12 06:03:59 hi guys .. I have a Bus 1 Device 2: ID 1871:1000 usb webcam Mar 12 06:04:16 how do I identify the driver required for it Mar 12 06:04:29 webcam is attached to a gumstix usbhost Mar 12 06:38:47 whatnick: lspci Mar 12 06:42:55 khem: reading the backlog (but I see ant__ is gone) bumped into the udev151 too, made a quick patch in the sanity branch recipe that adds CFLAGS += -Du8=__u8 Mar 12 06:43:11 that basically got around the error) Mar 12 06:47:49 khem: lsusb .. Mar 12 06:48:10 lspci does not exist Mar 12 06:48:20 lsusb lists the hub and the webcan Mar 12 06:48:22 *webcam Mar 12 06:48:37 i guess I can use this -> http://www.gumstix.net/wiki/index.php?title=Webcam_drivers_on_Open_Embedded_Part_1_-_pwc_driver_webcams Mar 12 06:56:52 03Khem Raj  07org.openembedded.dev * rc7f004f2d0 10openembedded.git/recipes/linux-libc-headers/ (2 files in 2 dirs): Mar 12 06:56:52 linux-libc-headers_2.6.31.bb: Use __u8 instead of u8 Mar 12 06:56:52 Signed-off-by: Khem Raj Mar 12 06:58:48 whatnick: yes Mar 12 06:58:57 I meant lsusb Mar 12 06:59:33 eFfeM: I fixed it and pushed just now try it out Mar 12 06:59:57 khem .. which driver family will the 1871:1000 fall in ? Mar 12 07:00:00 uvc ? Mar 12 07:00:16 googling the numbers makes it look like this is the case Mar 12 07:01:20 khem, not behind the system i worked on yesterday evening and here I seem to have a merge issue Mar 12 07:01:27 will try later Mar 12 07:01:36 ok Mar 12 07:01:55 but after typing saw the other commits on headers Mar 12 07:06:44 khem: http://linux-uvc.berlios.de/ <- couple of 1871 devices are listed Mar 12 07:07:59 whatnick: looks ok Mar 12 07:08:03 try it Mar 12 07:09:26 just adding the lsusb numbers to a config file will help attach the uvc driver to the cam ? Mar 12 07:09:41 i just want to capture a few frames to a microsd Mar 12 07:10:00 and need to compile the uvc driver as well ... hmm Mar 12 07:10:09 * whatnick will dig around .. Mar 12 07:28:08 03Sebastian Spaeth  07org.openembedded.dev * rcdc03c758b 10openembedded.git/recipes/abiword/abiword-2.5.inc: Mar 12 07:28:08 abiword-2.5.inc: Fix double quotes typo in HOMEPAGE Mar 12 07:28:08 Signed-off-by: Sebastian Spaeth Mar 12 07:42:03 gm Mar 12 07:42:22 hi Mar 12 07:42:33 hi effem, up early? Mar 12 07:42:39 you're early, aren't you ? Mar 12 07:42:48 yes :-) Mar 12 07:43:10 me early? I normally start around 8, or 7.15 if the mrs has to start working at 7 Mar 12 07:43:14 early bird :-) Mar 12 07:43:36 early bird and late nite hacking do not match well :-/ Mar 12 07:43:51 at least not for me, so I prefer the latter Mar 12 07:44:29 true Mar 12 07:50:55 morning lrg Mar 12 07:51:15 morning likewise Mar 12 07:52:41 khem: I'll add the same patch for 2.6.32, it apply cleanly and is needed as in 2.6.31 Mar 12 07:54:53 03Martin Jansa  07org.openembedded.dev * r7bd302c37f 10openembedded.git/recipes/linux-libc-headers/linux-libc-headers_2.6.32.bb: Mar 12 07:54:53 linux-libc-headers_2.6.32: Use __u8 instead of u8, the same as c7f004f2d028ef23de333345d246358b8cf1dd94 Mar 12 07:54:53 Signed-off-by: Martin Jansa Mar 12 07:55:04 03Martin Jansa  07org.openembedded.dev * rad8685c374 10openembedded.git/recipes/tasks/task-shr-minimal.bb: Mar 12 07:55:04 task-shr-minimal: add pisi, because opimd doesn't support SIM contacts itself Mar 12 07:55:04 Signed-off-by: Martin Jansa Mar 12 07:56:18 JaMa: khem is the same patch not needed by other linux-libc-headers recipes ? Mar 12 07:57:48 * JaMa haven't checked, but udev_151 is imho still D_P -1, so who use it, probably uses also new enough headers Mar 12 07:58:12 and it's also not compatible with much older kernels IIRC Mar 12 07:58:29 Ohh, abiword has recipes going back to 2.4 and the current 2.8.1 includes abiword-2.5.inc Mar 12 07:58:37 Who maintains that? Mar 12 08:04:03 spaetzWrk: check out git log Mar 12 08:04:26 or perhaps git blame Mar 12 08:06:03 spaetzWrk: lots of recipes need some serious cleanup Mar 12 08:33:54 mickey|office: can you remove the branch sanity for me that I accidentally created ? (not eFfeM/sanity! ) Mar 12 08:36:50 eFfeM: I had the same problem (now there exist likewise/nios2 and nios2... the latter is up-to-date and I should actually merge that one). Mar 12 08:37:06 03Steffen Sledz  07org.openembedded.dev * rf3d6c69826 10openembedded.git/contrib/angstrom/sort.sh: angstrom: add hipox machine to feed sorter Mar 12 08:43:41 Is anyone else having trouble building cups_1.2.12? Mar 12 08:58:45 eFfeM: what's its name? Mar 12 08:59:34 sanity Mar 12 08:59:52 (NOT eFfeM/sanity; forgot the eFFeM prefix on a push, want to retain the one with prefix) Mar 12 09:00:26 gah, it's snowing again Mar 12 09:00:29 *sigh* Mar 12 09:01:31 gone it is Mar 12 09:01:36 (the branch, not the snow) Mar 12 09:01:47 03Martin Jansa  07org.openembedded.dev * rc20d845b55 10openembedded.git/recipes/linux-libc-headers/linux-libc-headers-2.6.32/scsi-use-__u8.patch: Mar 12 09:01:48 linux-libc-headers-2.6.32: forgot to add patch Mar 12 09:01:48 Signed-off-by: Martin Jansa Mar 12 09:02:50 mickey|office: thanks for removing the branch; wish you could remove the snow also with just one *sigh* Mar 12 09:03:07 hehehe Mar 12 09:03:09 that'd rock Mar 12 09:03:12 git branch -D snow Mar 12 09:03:35 * mickey|office cranks up heating Mar 12 09:03:53 mickey|office: using which build? ;-) Mar 12 09:04:42 I've just added a new heater, it's called 'i5-750'. Mar 12 09:04:46 hahaha Mar 12 09:04:48 right Mar 12 09:04:57 any bitbake will do ;) Mar 12 09:09:31 n00b q: how do I diff a dir in my branch with the main dev head remote thought this would do it Mar 12 09:09:33 git diff recipes/gettext/ org.openembedded.dev:recipes/gettext/ Mar 12 09:10:10 hmm Mar 12 09:10:13 dunno whether that works w/ remote Mar 12 09:10:22 if in doubt, just branch into a local branch and then compare Mar 12 09:12:16 i thought you could remote diff too, but comparing between two branches would also be ok, but no idea how to do that either Mar 12 09:12:58 git diff Mar 12 09:13:02 should work Mar 12 09:13:35 eFfeM: btw you should probably revert "udev 1.5.1: added hack to make it compilable" now that linux-libc-headers are fixed Mar 12 09:14:34 ant_work: yes, this is only in my branch, but was trying to sync and got a non-building gettext Mar 12 09:14:42 :/ Mar 12 09:16:55 ant_work: guess i goofed somewhere merging, but not sure yet where Mar 12 09:17:33 I launched a rebuild from scratch with patched scsi headers and had no errors Mar 12 09:18:08 doing that right now too Mar 12 09:26:22 03Koen Kooi  07org.openembedded.dev * r00cb5a7688 10openembedded.git/recipes/cdparanoia/cdparanoia_svn.bb: cdparanoia svn: convert to new style staging Mar 12 09:51:04 03Koen Kooi  07org.openembedded.dev * rf4b036e09a 10openembedded.git/recipes/angstrom/angstrom-uboot-scripts/ (palo43-dsplink.cmd pico-dsplink.cmd): angstrom-uboot-scripts: add scripts for using pico DLP + DSP and gumstix palo43 + DSP Mar 12 10:06:39 good morning Mar 12 10:08:02 morning flo Mar 12 10:08:15 hi mickey|office Mar 12 10:08:49 florian: i hate to annoy, but if you have a chance, please remember the OEDEM reimbursements Mar 12 10:10:24 Hmm, suddenly having problems building cups 1.2.12, even though it or it's .inc file haven't changed in ages. Mar 12 10:10:43 mickey|office: heh, I checked tha account (and my inbox) half an hour ago. I can't find a mail with your bank account info... Mar 12 10:10:55 florian: ok, give me a minute :) Mar 12 10:11:08 http://pastebin.com/NCFacUw7 Mar 12 10:11:42 mickey|office: I will take care of it tomorrow when I'm back home. Mar 12 10:12:02 sent Mar 12 10:12:04 thanks Mar 12 10:12:20 ah you're still away Mar 12 10:12:25 didn't knwo that Mar 12 10:12:27 take your time Mar 12 10:12:40 * florian slowly returns from Embedded World chaos to "normal" mode. Mar 12 10:13:10 mickey|office: not really... but I won't return home from the office today. Mar 12 10:13:12 was it worth it at least? Mar 12 10:13:58 mickey|office: yes it was... Mar 12 10:27:51 morning Mar 12 10:31:01 hi Longfield Mar 12 10:31:04 hi cbrake Mar 12 10:34:39 how could I add a CFLAG only for a certain subarch to a recipe ? for instance all arm subarch but armv7 Mar 12 10:44:13 CFLAGS_append_$ARCH = "somethign" Mar 12 10:47:20 eFfeM, tinderbox shows you had a cups build failure a few days ago - I seem to be suffering from the same problem, that did you do to resolve it? Mar 12 10:48:00 MWelchUK_work__: not yet if i recall correctly Mar 12 10:49:49 eFfeM, Huh looks like you managed to build it. Is it still broken for you then? Mar 12 10:51:32 no idea, this was on my other system, and no idea exactly what I did to fix it Mar 12 10:51:37 thanks florian Mar 12 10:52:04 MWelchUK_work__: think i remember, one sec Mar 12 10:53:07 http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?h=eFfeM/sanity&id=1edf09ed0be4496bffe3dda78c3d794b5797e568 Mar 12 10:53:13 will patch on mainstream too Mar 12 10:53:50 if someone can tell me what the easy way is to merge a specific commit from a branch into the main git Mar 12 10:53:59 anyway ssl needs to be disabled Mar 12 10:54:27 eFfeM: git cherry-pick 1edf09ed0be4496bffe3dda78c3d794b5797e568 Mar 12 10:54:43 XorA: cool, thanks for the info Mar 12 10:55:23 eFfeM, if that resolves it, it affects 1.2.7 and 1.2.12 as well. Mar 12 10:55:48 actually it is an issue with build order if ssl is not build before you build cups you do not have a problem Mar 12 10:55:52 Might be work adding the EXTRA_OECONF to cups.inc? Mar 12 10:56:06 and afaik 1.2.7 and 1.2.12 is legacy and not worth my time Mar 12 10:56:12 ok :-) Mar 12 10:56:35 my proposal would be to delete them, but as that would curse all the wrath of the community on me, i won't do it Mar 12 10:59:16 Oddly the cups.inc file has "--disable-openssl" Mar 12 11:03:45 03Frans Meulenbroeks  07org.openembedded.dev * r6c42256bef 10openembedded.git/recipes/cups/cups.inc: Mar 12 11:03:46 cups: explicitly disable ssl Mar 12 11:03:46 explicitly disable ssl. If not if openssl is build first configure will enable it Mar 12 11:03:46 resulting in a cross compile badness Mar 12 11:03:46 ultimate solution is add a depends on openssl and fix the badness, but Mar 12 11:03:46 time does not permit this at the moment Mar 12 11:03:46 there you go, decided to change in ic Mar 12 11:03:47 Signed-off-by: Frans Meulenbroeks Mar 12 11:03:47 03Frans Meulenbroeks  07org.openembedded.dev * r006d70070e 10openembedded.git/: Merge branch 'org.openembedded.dev' of git.openembedded.org:openembedded into org.openembedded.dev Mar 12 11:03:49 in inc Mar 12 11:04:36 (and sorry about the merge, forgot --rebase) Mar 12 11:06:55 eFfeM, much appreciated. Mar 12 11:07:04 yw Mar 12 11:39:28 morning Mar 12 11:46:16 morning Mar 12 11:48:31 morning Mar 12 11:49:02 I am suprised. few days since util-linux_ng/e2fsprogs changes and no complains Mar 12 11:49:14 hi hrw Mar 12 11:49:16 *g* Mar 12 11:49:39 hrw-n900: Nobody's speaking to you any more. Mar 12 11:50:05 broonie lol Mar 12 11:50:06 hrw-n900: It doesn't build for me :-) Mar 12 11:50:32 hrw-n900: same old dependency problem: e2fsprogs is being built while util-linux-ng is not yet finished. Mar 12 11:50:43 e2fsprogs then fails Mar 12 11:55:36 uf.. desktop Mar 12 11:55:50 likewise: strange.. e2fsprogs depends on u-l-ng Mar 12 11:57:03 hrw: I know Mar 12 12:43:31 how to build EGL native clutter backend Mar 12 13:00:39 do we have a sane clutter recipe in oe? Mar 12 13:00:49 Hi. The package for xinput-calibrator_0.6.0 uses as SRC_URI a git repository. Is that correct? Mar 12 13:01:09 mario-goulart hm Mar 12 13:02:14 mario-goulart: yes Mar 12 13:03:38 mario-goulart: Tias (author) do not publish tarballs but instead give link to github Mar 12 13:03:56 hrw: Now I have a patch I'd like to apply to OE. But I also sent this patch to the author. What if he applies the patch? How OE knows whether it should apply the patch or not? Mar 12 13:04:26 mario-goulart: 0.6.0 recipe has fixed SRCREV Mar 12 13:04:54 hrw: Ah! Thanks for the tip. Mar 12 13:07:22 So whoever wants to change SRCREV has to be very careful no to apply patches that have been added to the main code tree. Mar 12 13:08:20 yes Mar 12 13:08:53 Right. Thanks. Mar 12 13:09:26 khem, ping? Mar 12 13:11:42 03Michael 'Mickey' Lauer  07org.openembedded.dev * r03ffd91b16 10openembedded.git/recipes/freesmartphone/msmcommd_git.bb: msmcommd: add sane srcrev Mar 12 13:11:53 03Michael 'Mickey' Lauer  07org.openembedded.dev * r491a028063 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 12 13:13:18 When building a cpio ramdisk image, I seem to need a symbolic link to sbin/init in the root for it too boot. This doesn't seem to be being created, am I missing something obvious? Mar 12 13:15:28 That is apart from providing init as a bootarg. Mar 12 13:15:46 03Michael 'Mickey' Lauer  07org.openembedded.dev * r578a475a6c 10openembedded.git/recipes/freesmartphone/msmcommd_git.bb: msmcommd: add missing dependencies Mar 12 14:12:05 hi Mar 12 14:13:35 how to build new ipk for clutter 0.8 from the source Mar 12 14:14:05 bitbake clutter Mar 12 14:14:27 I have add one patch in the clutter 0..8 source and want to create the ipk out of this source Mar 12 14:14:48 * XorA prefers an unclettered desktop :-) Mar 12 14:15:06 :) Mar 12 14:15:49 clutter is an awful name Mar 12 14:16:19 bitbake clutter Mar 12 14:16:20 hey Crofton, back at the helm? Mar 12 14:16:30 yeah Mar 12 14:16:31 hi crofton and xora btw. Mar 12 14:16:34 not operating it well though Mar 12 14:16:36 hi Mar 12 14:16:47 Crofton: Ive seen you guys drive, doesnt suprise me :-) Mar 12 14:16:53 heh Mar 12 14:17:15 I saw two cars go off the road in Munich er "yesterday" Mar 12 14:17:21 Thursday AM Mar 12 14:17:27 * XorA swears out loud a lot Mar 12 14:18:15 *g* Mar 12 14:18:18 hi florian Mar 12 14:18:28 * XorA just remebered a task he forgot to do last weekend Mar 12 14:19:22 xora which? Mar 12 14:19:35 start the TSC elections Mar 12 14:20:10 soggy biscuits Mar 12 14:27:20 woglinde, can u pls explain how Mar 12 14:29:27 siji I suggest you read the openembedded manual Mar 12 14:29:34 on the openembedded site Mar 12 14:29:48 if than there are questions gladly come back and ask Mar 12 14:30:16 woglinde, I think u havent read wht i asked Mar 12 14:31:28 -> 15:14 < siji> I have add one patch in the clutter 0..8 source and want to create the ipk out of this source Mar 12 14:31:37 if you have all done right Mar 12 14:31:43 type bitbake clutter Mar 12 14:32:30 but it will take from the clutter repository right Mar 12 14:32:46 am i right? Mar 12 14:33:22 -> 15:29 < woglinde> siji I suggest you read the openembedded manual Mar 12 14:37:13 hi, is OE_CFLAGS valid? Mar 12 14:37:28 *OE_LDFLAGS Mar 12 14:44:01 changing EXTRA_OEMAKE = "CC='${CXX}' LD='${CXX}' OE_CFLAGS='${CXXFLAGS}' INCPATH='${STAGING_INCDIR}' LIBPATH='${STAGING_LIBDIR}'" to: Mar 12 14:44:03 EXTRA_OEMAKE = "CC='${CXX}' LD='${CXX}' OE_CFLAGS='${CXXFLAGS}' OE_LDFLAGS='${LDFLAGS}' INCPATH='${STAGING_INCDIR}' LIBPATH='${STAGING_LIBDIR}'" Mar 12 14:44:11 didn't make the GNU hash go away Mar 12 14:44:39 I'll try LDFLAGS_APPEND Mar 12 14:44:57 *append Mar 12 14:45:47 most makefiles wont understand OE_XXXX as a flag, I think your thinking oe_runmake does something intelligent Mar 12 14:46:17 in fact OE_CFLAGS looks like something I invented, what recipe is it? Mar 12 14:47:21 rofl Mar 12 14:47:31 I hate it when my bad ideas come back to haunt me Mar 12 14:48:11 Crofton|work: in this case I think the Makefile totally shat on CFLAGS so I "worked" around it Mar 12 14:50:41 ok I've found how to do it Mar 12 14:50:45 I'll hope it'll work Mar 12 14:51:02 I just added LDFLAGS after CXXFLAGS Mar 12 15:02:45 didn't work Mar 12 15:03:08 XorA, fbreader Mar 12 15:03:49 Crofton: you in .de? Mar 12 15:04:18 OE_LDFLAGS='${LDFLAGS}' => fail Mar 12 15:04:18 GNUtoo|oeee: luckilly that isnt one of my cazinesses :-) Mar 12 15:04:24 ok Mar 12 15:04:37 OE_CFLAGS='${CXXFLAGS} ${LDFLAGS}' => fail Mar 12 15:04:37 nope, back in .us Mar 12 15:04:43 GNUtoo|oeee: but you would be wise to read the patches, then you will see why OE_LDFLAGS fails Mar 12 15:05:01 ? Mar 12 15:05:10 I'll look if there are patches Mar 12 15:05:15 GNUtoo|oeee: read the code Mar 12 15:05:18 ah ok Mar 12 15:05:21 Makefile.patch Mar 12 15:05:27 GNUtoo|oeee: the buildsys patch Mar 12 15:07:39 ah it's only present on older versions Mar 12 15:08:12 ahh Mar 12 15:08:15 it was not ported Mar 12 15:28:57 ok passed the GNU hash Mar 12 15:45:47 hi, I got "gnome-game" package error while building image, so I removed that from RDEPENDS_${PN}. But now do_rootfs has failed Mar 12 15:46:27 erro log is "Cannot satisfy the following dependencies ... gnome-game" Mar 12 15:46:42 how to resolve this? Mar 12 15:50:19 hello Mar 12 15:54:00 gnome-game? not gnome-games? Mar 12 15:55:50 ok Mar 12 15:56:00 gnome-games Mar 12 15:57:56 hrw : have u faced this problem, like do_rootfs failed by dependencied issue Mar 12 16:03:22 bala_holmes, look for other occurences of gnome-games Mar 12 16:03:29 and look in the tasks Mar 12 17:15:27 How many developers with write access do we have right now? Mar 12 17:16:10 no idea Mar 12 17:46:27 ok, have a nice weekend everyone Mar 12 17:47:17 See you, hrw. Have a nice weekend. Mar 12 18:21:04 re Mar 12 18:22:38 likewise: howdy Mar 12 18:22:45 likewise: Crofton is in .us Mar 12 18:23:26 somehow external-toolchain is being picked up for me and its causing mess Mar 12 18:23:50 I wonder if someone added some extra PROVIDE to it Mar 12 18:25:00 * Crofton was in .de and .au for the past week or so Mar 12 18:25:16 * florian votes for .at ;) Mar 12 18:25:17 Crofton: oh cool Mar 12 18:25:28 Crofton: some conference ? Mar 12 18:25:39 YEAH Mar 12 18:25:41 er Mar 12 18:25:46 WSR2010 Mar 12 18:25:53 software radio Mar 12 18:25:56 and vacation Mar 12 18:26:02 went skiing in .au Mar 12 18:26:30 nice Mar 12 18:26:46 there is enough snow in Tahoe this year too :) Mar 12 18:28:37 I wonder how can I find which one dependency is pulling in external toolchain Mar 12 18:31:08 florian: are you in .at ? Mar 12 18:31:35 khem: no in .de Mar 12 18:31:55 oh yes I remember you were near muenster ? Mar 12 18:32:08 khem: the family of my father is from there Mar 12 18:32:30 khem bitbake -DDDD Mar 12 18:32:45 khem: comparably yes... its not really far from here. sigen is half the way between cologne and frankfurt Mar 12 18:34:04 cool Mar 12 18:34:11 woglinde: will try it Mar 12 18:34:30 bitbake master is hanging forever on my ubuntu/9.10 x86 box Mar 12 18:34:39 it was working fine a month ago Mar 12 18:57:58 seems fine here Mar 12 18:59:19 its same local.conf for me and same machine with all updates Mar 12 18:59:37 I will try to bisect bitbake if there is some commit thats doing it for me Mar 12 19:02:17 hm Mar 12 19:03:07 kergoth: external-toolchain-csl is being pulled in for console-image how to find out best who is inviting it into the game Mar 12 19:03:26 i'd say bitbake -g Mar 12 19:03:33 khem isnt bitbake -DDD saying more? Mar 12 19:03:42 I removed all PACKAGES and PROVIDES from the recipe still I see it being pulled in Mar 12 19:03:52 kergoth for external toolchain bitbake -g sometimes dont help Mar 12 19:04:27 -g shows all the dependencies, i fail to see how external toolchain is special Mar 12 19:04:27 khem which machine? Mar 12 19:04:29 its just another recipe Mar 12 19:04:32 woglinde: ok I am parsing thru the -D logs Mar 12 19:04:41 qemuarm/minimal-uclibc Mar 12 19:04:55 *g* Mar 12 19:05:06 hm Mar 12 19:07:19 DEBUG: providers for glibc-gconv-iso8859-1 are: ['external-toolchain-csl', 'external-toolchain-generic', 'external-toolchain'] Mar 12 19:07:31 yes Mar 12 19:07:37 I hit this too Mar 12 19:07:41 some time ago Mar 12 19:07:42 so I think I need to teach poor uclibc about that Mar 12 19:08:02 abiword is dragging it in Mar 12 19:08:10 if I remeber correctly Mar 12 19:08:17 or some other gnome stuff Mar 12 19:08:20 could be Mar 12 19:09:56 ugh Mar 12 19:11:45 yes abiword Mar 12 19:12:51 PACKAGES_DYNAMIC += "glibc-gconv-*" Mar 12 19:12:52 PACKAGES_DYNAMIC += "glibc-locale-*" Mar 12 19:13:12 thats the one in external-toolchain-csl.bb Mar 12 19:15:14 woglinde: I dont have abiword being pulled into my console-image Mar 12 19:15:21 so some other package it is Mar 12 19:16:42 hm Mar 12 19:17:43 Added runtime recommendation glibc-gconv-iso8859-1 for /home/kraj/work/oe/openembedded/recipes/gtk+/gtk+-native_2.12.11.bb Mar 12 19:17:51 interesting Mar 12 19:18:01 jupp Mar 12 19:18:03 saw this too Mar 12 19:18:05 why would one add target recommendation Mar 12 19:18:14 bye grepping the recipes Mar 12 19:18:15 to a native package tsch tsch Mar 12 19:18:27 hm BBCLASSEXTEND? Mar 12 19:18:39 or false require Mar 12 19:19:48 RRECOMMENDS_${PN}_linux-gnueabi = " ${NEATSTUFF} glibc-gconv-iso8859-1" Mar 12 19:20:27 RECOMMENDS_${PN}_linux = " ${NEATSTUFF} glibc-gconv-iso8859-1 " Mar 12 19:20:50 why then bitbake adds it for linux-uclibceabi ? Mar 12 19:21:29 could you try it without? Mar 12 19:21:30 this is from recipes/gtk+/gtk+.inc Mar 12 19:21:34 yes Mar 12 19:21:51 native shouldnt have RRECOMMENDS Mar 12 19:22:26 overwritten :) Mar 12 19:22:46 if iherited from Mar 12 19:22:59 RRECOMMENDS_${PN}_linux = " ${NEATSTUFF} glibc-gconv-iso8859-1 " Mar 12 19:23:06 kicks in here for a native package Mar 12 19:23:20 because we are building for x86_64-linux Mar 12 19:23:35 yes Mar 12 19:23:37 as I said Mar 12 19:23:50 the typical DEPENDS-like error for native Mar 12 19:24:55 khem: ahah you noticed that bug as well, I wasnt beleived when I said it existed Mar 12 19:27:02 XorA: yeah and same problem on DEPENDS Mar 12 19:27:14 which I sort of fixed and have a local patch here Mar 12 19:29:47 khem will you comment bernhards uclibc patch series I will not find the time Mar 12 19:30:19 and good nite for today Mar 12 19:30:29 yes Mar 12 19:32:12 khem, another question (that i forgot to put in the introductory message) is why DEVEL_PREFIX != /usr (in staging)? Mar 12 19:32:25 gtk+_2.18.6.bb:BBCLASSEXTEND = "native" Mar 12 19:32:42 still it picks up gtk+-native_2.12.11.bb Mar 12 19:32:44 for me Mar 12 19:33:15 gtk+-native apparently takes precedent over BBCLASSEXTEND Mar 12 19:33:22 according to RP Mar 12 19:33:51 hmmm so we should delete the native packages at the same time then right or can be add PREFERENCES Mar 12 19:34:09 Im not sure Mar 12 19:34:44 Does anyone happen to have a tarball of the source after the recent smpeg bump? I have an image that drags it in and I notice since the recipe was bumped the upstream SVN has been dead so it can't get the source. Mar 12 19:36:56 khem, yes. As long as you have a *native*.bb lying around, BBCLASSEXTEND'ed natives are ignored (see ML where i was a bit unhappy but folks thought that this would be nice since it would trigger updates to old recipes to make the new classextended work) Mar 12 19:38:07 XorA: so why did we leave the native recipes inside gtk+ Mar 12 19:38:39 khem: I didnt know about that quirck until recently and ive been busy Mar 12 19:39:02 XorA: would it be ok to delete the native recipes ? Mar 12 19:39:10 because I dont want to fix them Mar 12 19:40:06 khem, see [BBCLASSEXTEND BUG?] Not taking virtual:native provider into account? http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-February/017527.html Mar 12 19:46:27 khem: I think it would be ok, shouldnt be at controversial as delteting all versions anyway Mar 12 19:46:47 khem: I think angstrom is the only distro makes use of it anyway Mar 12 20:01:27 ok Mar 12 20:01:34 well I will fix the natives for now Mar 12 20:01:41 and then RFc the deletion later Mar 12 20:54:22 florian: good evening Mar 12 20:54:41 oh Mar 12 20:54:46 * florian is back :) Mar 12 20:54:52 hey pb__ Mar 12 20:55:15 heh Mar 12 20:55:17 welcome back :-) Mar 12 21:04:38 :-) Mar 12 21:09:18 Hello all, may I know whick package provide ldd? Mar 12 21:22:06 hi, is the wiki guide for java up to date? Mar 12 21:22:30 my target is to be able to run eclipse on eee701 Mar 12 21:33:16 JT_: if its glibc based system then its glibc Mar 12 21:37:43 XorA, pb__ I sent a patch for gtk+-native to ml it would be nice if you could scan over it Mar 12 21:41:17 In git 1.7.0, the default has changed to --no-chain-reply-to Mar 12 21:41:17 Set sendemail.chainreplyto configuration variable to true if Mar 12 21:41:19 you want to keep --chain-reply-to as your default. Mar 12 21:41:22 interesting Mar 12 21:52:06 khem, i wasn't aware that i should merge such stuff. Mar 12 21:53:00 khem, i ment to be polite and separate out the logical steps. That PREFERRED_UCLIBC_VERSION hack (following binutils and gcc here) is distinct and can go in as is Mar 12 21:55:29 khem, the git mv uclibc.inc uclibc-old.inc was just so make obvious (and bisectable) that i didn't change anything of relevance Mar 12 21:56:40 ok Mar 12 21:58:27 khem, if you prefer merged batches then that's certainly fine with me -- i'll do that next time, perhaps. The coreutils virtualization is another such bunch but is not so big in the end, patch-wise Mar 12 21:59:41 maybe I should fix the fact that newly pluged sdcard are not detected first Mar 12 22:00:20 khem, and i say perhaps since for the most part big patches make review harder imho. But i'll of course do whatever is established behavior #here Mar 12 22:02:19 its ok Mar 12 22:02:42 GNUtoo|oeee: why are they not detected ? Mar 12 22:03:08 don't know...kernel doesn't detect them if you don't put them at boot Mar 12 22:03:15 the driver is usb_storage Mar 12 22:03:21 the device is eee701 Mar 12 22:09:20 how can we negotiate stuff like ipv4/ipv6/socket/no-network plus (most) combinations thereof? Mar 12 22:09:46 khem, thoughts? Mar 12 22:09:55 patch looks ok Mar 12 22:10:03 I will read thru them when I reach home Mar 12 22:10:45 2/3 seems uclibc and uclibc-initial are only needed Mar 12 22:11:06 khem, the patch would break anybody horribly since it turns e.g. networking completely off since noone requests ip4, ipv6 nor socket support Mar 12 22:12:00 plus I dont know if we should move it to uclibc file Mar 12 22:12:10 all versions are encoded in sane-toolchain.inc Mar 12 22:12:40 khem, [keep in mind that he's using official buildroot. Chances that it's bluntly misconfigurated is pretty high.] Mar 12 22:12:52 blindvt: hmmm Mar 12 22:12:54 khem, i somehow have to see what libc has to support Mar 12 22:13:28 khem, the uclibc-arch class would be one nice centryl spot to decide on them Mar 12 22:13:28 blindvt: you should add the features to distro's if you are singling them out Mar 12 22:13:49 blindvt: I would say DISTRO_FEATURES is right place Mar 12 22:14:06 and the bbclass should provide mechanisms to stick it into uclibc .condig Mar 12 22:14:10 .config Mar 12 22:14:40 khem, like i did for '"ipv4"' in there, you mean? Mar 12 22:14:52 blindvt: patch 2/3 is not needed I would say Mar 12 22:14:59 yes Mar 12 22:15:10 so ipv4 can be a distro feature Mar 12 22:15:32 and when enabled then uclibc mechanism should enable it in .config Mar 12 22:16:07 may be some python to get the DISTRO_FEATURES and search for the required features and take action accordinly Mar 12 22:16:56 * khem jogging time ttyl Mar 12 22:18:31 khem, 2/3 saves me from specifying the whole stanza for uClibc: Mar 12 22:18:36 how can i set PREFERRED_VERSION_gcc* = "4.4.3" at once? Or do i have to explicitly spell out gcc gcc-cross gcc-cross-initial gcc-cross-intermediate gcc-cross-sdk ? Mar 12 22:18:59 03Khem Raj  07org.openembedded.dev * rad0976afee 10openembedded.git/recipes/gtk+/ (3 files): Mar 12 22:19:00 gtk+-native: Override RRECOMMENDS for native recipe Mar 12 22:19:00 * Remove old style native recipes. Mar 12 22:19:00 Signed-off-by: Khem Raj Mar 12 22:19:00 Acked-by: Graeme Gregory Mar 12 22:19:31 khem, and it was pointed out to me that there is PREFERRED_GCC_VERSION and PREFERRED_BINUTILS (no common pattern, yes) Mar 12 22:19:33 Yeah, but uclibc lacks -cross/-cross-sdk copies Mar 12 22:19:36 blindvt: ok you need only uclibc and uclibc-initla Mar 12 22:19:42 and put then in sane-toolchain.inc Mar 12 22:19:52 and also change the one for eglibc and glibc Mar 12 22:20:04 to keep it legible Mar 12 22:20:15 yes there is Mar 12 22:20:22 khem, if i put git into sane-toolchain then everybody will default to that, no? Mar 12 22:20:37 nooooooo Mar 12 22:20:43 you put whatever is there Mar 12 22:20:53 and you override it in your local.conf Mar 12 22:20:55 to git Mar 12 22:20:57 or whateevr Mar 12 22:21:24 khem, _twice_ in local.conf Mar 12 22:21:43 khem, plus i have to know that it's foo and foo-initial Mar 12 22:22:47 khem, ok, so i'll add it twice in local.conf, but can we perhaps have a common scheme for gcc and binutils? The naming they currently use is very unintuitive Mar 12 22:22:47 UCLIBC_GITREV = "39548d72297a8c3d8d55e04b8240e4ce09913a82" Mar 12 22:22:47 PREFERRED_VERSION_uclibc_local = "0.9.30+gitr${UCLIBC_GITREV}" Mar 12 22:22:47 PREFERRED_VERSION_uclibc-initial_local = "0.9.30+gitr${UCLIBC_GITREV}" Mar 12 22:23:00 you can put in your local.conf Mar 12 22:23:11 if you think you need only one line Mar 12 22:23:52 you can add PREFERRED_UCLIBC_VERSION Mar 12 22:23:57 and then do Mar 12 22:23:59 i thought and ended up with '' '-initial' but i'll try to rethink that Mar 12 22:24:10 PREFERRED_UCLIBC_VERSION_local = "0.9.30+gitr${UCLIBC_GITREV}" Mar 12 22:24:37 and do this all in sane-toolchain.inc Mar 12 22:24:44 all versions are chosen there Mar 12 22:25:16 some config tweaks go into libc specific sane file Mar 12 22:25:37 same you should do for eglibc and glibc too Mar 12 22:25:41 to be consistent Mar 12 22:26:27 khem, would, perhaps, a commonly useable pattern be something like PREFERRED_VERSION_${toupper(${PN})} = that get's ?= in sane-toolchain and as such is overridable in local.conf ? Mar 12 22:26:49 yes Mar 12 22:27:05 but use explicit names Mar 12 22:27:34 ${toupper(${PN})} might not work and even if they did it will be obscure Mar 12 22:27:49 * khem is really out Mar 12 22:27:50 khem, that way we wouldn't have to spell out that many different 'PREFERRED_GCC_VERSION' vs. 'PREFERRED_BINUTILS' Mar 12 22:28:11 khem, the toupper was not ment literally Mar 12 22:30:39 khem, i, as a user, would not care if it's PREFERRED_FOO_VERSION or PREFERRED_VERSION_FOO or something else, but it should follow one single, common pattern so can use it blindly and not have to look up individual occurances, but anyway Mar 12 23:05:57 khem, i feel pretty dumb to spell out the obvious, but, just to reiterate: an arch constrains (hard/soft constraint, via e.g. float, mmu ,endian, abi, stack growth direction and so forth), a toolchain features/provides (binfmt etc, and highlevel -- or midlevel if you prefer -- stuff like malloc_0, lfs, nls, network, etc); Then, from a TARGET_OS level, a package features/provides stuff, a distro requires stuff Mar 12 23:08:24 khem, ideally i'd like (a tailored) distro telling what it needs and getting stuff setup that it needs (provided that the desired features form a possible combination), see? Mar 12 23:20:40 When I submit a patch to the ML, should I include a Signed-off-by: line for myself, or are sign-offs only intended for people with commit access (who are "signing off" work done by someone else)? Mar 12 23:31:54 You should include one Mar 12 23:32:00 As you're signing off on the work as your own Mar 12 23:32:03 * Tartarus digs up a link Mar 12 23:32:18 http://kerneltrap.org/taxonomy/term/245 Mar 12 23:32:24 That's what the line means Mar 12 23:34:21 Thanks Tartarus. Now that I have git send-email set up properly, my next patch submissions should go more smoothly. :) Mar 12 23:36:56 The zlib stuff was you then, yes? Mar 12 23:37:03 yes Mar 12 23:37:05 Does BBCLASSEXTND="sdk" really work? Mar 12 23:37:17 ie you had zlib-sdk build? Mar 12 23:37:27 I'm actually re-verifying that right now. Mar 12 23:37:47 but I do seem to recall it did create a zlib-sdk package Mar 12 23:38:00 ok Mar 12 23:59:24 yep, zlib-sdk is created just fine Mar 13 00:30:24 03Michael 'Mickey' Lauer  07org.openembedded.dev * r8c6131dc17 10openembedded.git/recipes/dbus/ (dbus-1.3.0/improve-threading-stability.patch dbus_1.3.0.bb): dbus 1.3.0: add patch to improve threading stability Mar 13 01:08:48 not working properly, Audio on SNES9X Mar 13 01:08:52 Sorry Mar 13 01:09:04 not working properly, Audio on snes9x-sdl Mar 13 01:09:11 on omap platform **** ENDING LOGGING AT Sat Mar 13 02:59:57 2010