**** BEGIN LOGGING AT Sun Jun 10 02:59:57 2007 Jun 10 03:01:01 * * OE Bug 2461 has been created by jin@mediatomb.cc Jun 10 03:01:03 * * ipkg upgrade does nothing, allthough newer packages are available in the feed Jun 10 03:01:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2461 Jun 10 03:01:14 psokolovsky: please take a look if you find the time :> Jun 10 03:01:27 I debugged it as well as I could but I got stuck in the end Jun 10 03:01:35 crap, 5 a.m... Jun 10 03:01:41 night everyone Jun 10 03:01:42 :> Jun 10 04:09:16 morning Jun 10 04:35:47 ~oemirrors Jun 10 04:36:01 i heard oemirrors is extra, extra, read all about it, oemirrors is OpenEmbedded monotone mirrors are ewi546.ewi.utwente.nl, monotone.nslu2-linux.org, opensource.wolfsonmicro.com also see http://www.openembedded.org/wiki/GettingStarted Jun 10 05:49:32 hey mickey|taipei, welcome to my timezone :-) Jun 10 05:49:41 (ish) Jun 10 07:39:57 good morning all Jun 10 08:41:52 koen: http://www.angstrom-distribution.org/building-angstrom could you add a mtn pull to it? Jun 10 08:42:33 zecke: sure Jun 10 08:45:37 hi alkl Jun 10 08:46:05 hi all i meant of course ( am i still sleeping ? ) Jun 10 08:48:59 koen: we had two bitbake bug reports due people following angstrom instructions Jun 10 08:49:14 heh Jun 10 08:56:09 koen: more than we had for the MasterMakefile Jun 10 09:02:13 zecke: any news on the 1.8.4 release? Jun 10 09:03:20 no Jun 10 09:36:47 can i overcome the md5 checksum when a package starts and make him just use the .tar.bz2 inside /sources ? Jun 10 09:39:51 rob_w: use md5sum to create a .md5 file, then OE will use it Jun 10 09:40:24 rwhitby, seems to not happen for kernels ... it gets its checksum by kernel.org i feel Jun 10 09:40:56 Hello Jun 10 09:41:27 rwhitby, i am working these days on a ixp425 implementation.. the manufcaturer points to snapgear project ,,, i i would love to use the their snapgear kernels inside oe Jun 10 09:41:50 maybe i better make a new bb file for it Jun 10 09:41:58 rob_w: rephrase please? Jun 10 09:42:43 zecke|breakfast, is it true that when vanilla kernel is donwloaded it also loads the md5 file from kernel.org ? Jun 10 09:42:48 rob_w: why use snapgear when OE supports ixp425 nicely Jun 10 09:42:55 rwhitby, CORRECT Jun 10 09:42:57 rob_w: no Jun 10 09:43:08 zecke|breakfast, hmm Jun 10 09:43:25 rob_w: let us start the other way around. What is your issue? Jun 10 09:43:39 rob_w: why use the snapgear kernels? mainline linux has great ixp425 support, the missing patch are available from the nslu2-linux project (ethernet driver, mac settings, rtc fixups) Jun 10 09:44:03 zecke|breakfast, i just packed up the snapgear kernel as vaninall 2.6.17.tar.bz2 and put it in sources ,, remade the md5 but then it breaks at NOTE: The MD5Sums did not match. Wanted Jun 10 09:44:45 rob_w: well, you want to touch conf/checksums.ini to disable all md5 sums ;) Jun 10 09:44:52 koen, thats my plan in total but it seems the snapgear kernels include some manufacturer changes .. which they dont comment , so i plan to start using their source intitally .. so all is compile by the same gcc Jun 10 09:44:57 zecke|breakfast: or just rename the snapgear file? Jun 10 09:45:21 koen, then if that boots i wanna work out the manufacturer changes and patch them in standard oe ixp425 area`S Jun 10 09:46:56 the npbe+ n friends compiles well already. but a basic ixp425 kernel doenst boot the unit currently .. i only have the snapgear kenerl which boots but then maybe breaks at init() as it was compiled with a different gcc ( i assume) Jun 10 09:47:15 then my oe rootfs Jun 10 09:47:16 rob_w: is the machine id correct in your bootloader? Jun 10 09:47:26 good question Jun 10 09:47:43 that's the most common reason for mainline kernels not booting on manufacturer boards Jun 10 09:47:44 i read some out the bootloader <-> id relations Jun 10 09:48:00 what board is it? Jun 10 09:48:08 an adi pronghorn metro Jun 10 09:48:35 where is the machine id in kernel source ? Jun 10 09:49:04 rob_w: openwrt is supporting the proghorn - talk to Kaloz in #openwrt Jun 10 09:49:06 rob_w: http://www.arm.linux.org.uk/developer/machines/ Jun 10 09:49:20 oh cool ,, thanks zecke|breakfast rwhitby Jun 10 09:49:57 rob_w: or look in https://dev.openwrt.org/browser/trunk/target/linux/ixp4xx-2.6/patches Jun 10 09:50:20 there are three pronghorn metro patches in there. Jun 10 09:50:54 i see them thanks Jun 10 09:51:05 rwhitby, how is your compulab going ? Jun 10 09:51:27 what compulab? Jun 10 09:51:50 oh i though you where the one also with a compulab board.. sorry then Jun 10 09:52:09 nope, not me. Jun 10 09:56:51 rob_w: re metro kernel functionality: Jun 10 09:56:53 " rwhitby: yup, except for setting the board's mac address Jun 10 09:56:54 rwhitby: still wondering how to implement that (it's stored in some OTP blocks)" Jun 10 09:57:18 nice Jun 10 10:00:35 rwhitby: thanks for your answer Jun 10 10:00:46 rwhitby: BTW: how is your DebianDeveloper thing going? Jun 10 10:01:07 in the middle of NM, but been *really* slack about it. Jun 10 10:10:03 rwhitby, phuees quite some politic guys in #openwrt ;-) Jun 10 10:10:23 rob_w: I never go near #openwrt Jun 10 10:10:30 try #openwrt-devel Jun 10 10:10:47 hehe Jun 10 10:11:00 the key when approaching a new distro is not to mention any other distro until you know who hates who :-) Jun 10 10:12:02 nslu2-linux people are involved with OE, Debian and OpenWRT, so we're used to trying to dodge politics ;-) Jun 10 10:17:50 rwhitby: hehe Jun 10 10:36:31 hey mickey|fic Jun 10 10:40:09 mickey|fic: have a look at pulse.inc in OE, that one has some more fixes than the one in openmoko svn Jun 10 10:41:47 koen: ok Jun 10 10:42:27 like giving people an upgrade path and working CONFFILES :) Jun 10 10:45:19 mickey|fic: du fehlst uns in #opie.de Jun 10 10:50:40 hmmm Jun 10 10:50:45 1600 revs in 2 months Jun 10 10:50:54 I should update my workstation more often Jun 10 10:51:16 03koen 07org.oe.dev * r13796a3d... 10/ (3 files in 2 dirs): pulseaudio: fix postinsts Jun 10 11:30:52 not all commits make it into commit list? Jun 10 11:32:40 psokolovsky_, ping, we have broken opie build and I'm not yet sure how to work around your last commit. Although if you don't have a solution right away or you aren't here, I think it should be pretty easy to figure it out. Jun 10 11:32:59 polyonymous: ?? Jun 10 11:33:07 palmtop.bbclass: Finally, set qte library variant (MT vs non-MT) at the proper place. Jun 10 11:33:19 Now it fails with -lqt-mt-mt Jun 10 11:33:34 because in qmake's mkspecs we have ${OE_QMAKE_LIBS}-mt Jun 10 11:33:41 (or something to the same effect) Jun 10 11:34:02 polyonymous: well, then we likely should have that in qmake's mkspecs Jun 10 11:34:05 psokolovsky_, that may well mean that you forgot to commit some file? Jun 10 11:34:12 nope Jun 10 11:34:28 to have what? We do have it, but you added it to palmtop.bbclass as well Jun 10 11:34:50 polyonymous: need to remove from qmake's mkspecs, whatever that is Jun 10 11:35:41 psokolovsky_, maybe... the thing is that qmake has something like QMAKE_LIBS and QMAKE_LIBS_MT they would be the same then and non-threaded build won't work then. Jun 10 11:35:59 the problem is that we only support threaded or nontreaded qte, but not both, right? Jun 10 11:36:23 polyonymous: ok, if you know how to fic that, please fix. but individual .bb should not anylonger fish for qte lib name. Jun 10 11:36:25 because qmake specs seems to expect both Jun 10 11:36:57 polyonymous: please give me URI if you'd like me to say sth specific Jun 10 11:37:02 Okay, so you don't know. Neither I do yet, need to give it some thought. Jun 10 11:37:05 URI to what? Jun 10 11:37:54 polyonymous: to what to look at Jun 10 11:38:22 build error is http://rafb.net/p/I5H9Lz10.html Jun 10 11:38:28 the rest is in the tree. Jun 10 11:38:39 polyonymous: qt-mt-mt or qte-mt-mt? Jun 10 11:38:47 revision 618bda0557cc4588d6e1f66a2fee4041287cae15 is troublesome Jun 10 11:38:54 zecke, qte Jun 10 11:38:59 polyonymous: URI to that flamy "qmake's mkspecs" Jun 10 11:39:02 zecke, as a qmake guru you may have ideas too. Jun 10 11:39:14 psokolovsky_, packages/qmake/files/*.conf, I think. Jun 10 11:39:22 * koen mentions qt4 Jun 10 11:39:37 koen, qt4 builds fine atm, I think. Jun 10 11:39:47 so why bother with qt2? Jun 10 11:40:10 Well, I'm not sure if we want to improve opie, but we sure want to sustain life in it. Jun 10 11:40:16 I mean oe opie. Jun 10 11:41:18 polyonymous: I have not followed too closely but nromally CONFIG+=thread should toggle threading (choosing the right flags, linking against the right libraries) Jun 10 11:41:30 * koen wonders when the first C&D letter arrives for oe opie Jun 10 11:41:39 polyonymous: good suggestion I have is to remove that "-mt" in linux-oe-qmake.conf Jun 10 11:41:45 zecke, that's what I think as well, but I'm sure psokolovsky_ did have reasons for that revision as well. Jun 10 11:42:12 polyonymous: once I finish this mail, I will be all yours Jun 10 11:42:13 psokolovsky_, that would break either threaded or nonthreaded build. Jun 10 11:42:24 polyonymous: how? Jun 10 11:42:40 psokolovsky_, we'd have the same library for both. Jun 10 11:42:49 it will just rely on external means to setup library name Jun 10 11:42:51 psokolovsky_, what exactly have you been fixing with that rev? Jun 10 11:43:25 polyonymous: need to figure out lib name in individual .bb - quite a bunch of them Jun 10 11:43:44 psokolovsky_, external means are supposed to set up qmake's CONFIG, not library. Jun 10 11:43:55 polyonymous: other solutuion may be to change var name, but that would still need to patch .bb's I guess Jun 10 11:43:59 psokolovsky_, can you point to any single .bb? Just so that I get a picture. Jun 10 11:44:18 psokolovsky_, that's what I'm thinking about and that's why I want to see the problem. Jun 10 11:45:08 xqt2/xqt2_20060509.bb Jun 10 11:45:16 ok thanks. Jun 10 11:45:29 zroadmap Jun 10 11:45:40 snes9x-sdl-qpe_1.39.bb Jun 10 11:45:42 etc Jun 10 11:45:47 yup. enough, thanks. Jun 10 11:46:49 brb Jun 10 11:56:15 hvontres|home: did you try the kexec-cmdline stuff? Jun 10 12:04:23 so basically that's for the bb's that don't seem to use qmake for build. Jun 10 12:08:13 psokolovsky_, I think it's easy, your changes to qmake should be reverted, but QT_LIBRARY setting should be left intact. Jun 10 12:10:57 hi, all! Jun 10 12:11:05 hi Jun 10 12:13:30 a package contains additional package to build, and has line FILES_${PN}_foo += '/some/path', if this path contains .debug, then I add FILES_${PN}_foo_dbg += '/some/path/.sebug'. what should I do in addition for debug package to be recognized? Jun 10 12:13:44 s/'/"/g Jun 10 12:14:38 slapin_nb, recognized as in being packaged? Jun 10 12:15:03 slapin_nb, I think you mean PACKAGES += "${PN}_foo_dbg" ? Jun 10 12:15:33 slapin_nb: or do FILES_${PN}_foo += '/some/path/', that would include and dot directories Jun 10 12:16:14 polyonymous, so it should do it? so, no special treatment for -dbg? Jun 10 12:16:47 koen, in this way it qill be treated automatically? I just want to fix some QA errors this way. Jun 10 12:18:02 the extra slash excludes them from the non debug package Jun 10 12:18:13 they should still go into a debug package as well Jun 10 12:18:21 slapin_nb, depends on what do you mean by "recognized" Jun 10 12:18:21 slapin_nb: FILES_${PN}_foo_dbg for FILES_${PN}_foo-dbg? Jun 10 12:19:09 polyonymous: yeah, do as you see fit Jun 10 12:19:34 psokolovsky_, yup, done already, building now. Jun 10 12:19:43 anyway, thanks for your time. Jun 10 12:22:24 zecke, in opie-reader I have a combobox which is setFocusProxy-ed to the parent and on Z it doesn't drop down on click. Although on desktop it does. I am whatever but a qt programmer, do you have an idea how to work around that? Jun 10 12:23:38 zecke: should we have then patches from #13913, #14058 and #14059 in OE for webkit? Jun 10 12:23:54 hi all Jun 10 12:24:05 hey steliosk Jun 10 12:24:55 koen: #13913 not. One should not use reuse the FrameView Jun 10 12:25:20 koen: as to #14058,#14059 sure but for #14059 another patch is needed to make it not crash ;) Jun 10 12:25:42 koen: with pkgconfig.bbclass we patch the .pc file. What about replacing -I with -isystem? Jun 10 12:26:09 koen: http://bugs.webkit.org/show_bug.cgi?id=14061 but alp needs to attach the patch Jun 10 12:26:24 sounds like a good idea, byt I don't know the impact on the build Jun 10 12:26:35 koen: WebKit will build ;) Jun 10 12:27:26 koen: #14059 will fix the encoding issues you might see while rendering Jun 10 12:28:19 speaking of replacing: http://rafb.net/p/WAcPNS81.html Jun 10 12:28:52 zecke: what do you think about that? Jun 10 12:29:12 and some similar magic in qt4 .bb, may also make sense to incorporate... Jun 10 12:29:56 koen: why do we need that? Jun 10 12:30:08 koen: and aren't the dirs expanded in .la files anyway? Jun 10 12:31:56 I think you're misinterpreting the patch Jun 10 12:32:18 it replaces /data/build/koen/OE/foo/bar/staging/lib with /usr/lib Jun 10 12:32:56 koen: one question! Jun 10 12:33:08 koen: don't we use autotools_do_install to stage headers and libs as well? Jun 10 12:33:19 I'd rather fix the root cause, but I don't know what it is Jun 10 12:33:29 zecke, that's for install, not staging Jun 10 12:33:32 ahh Jun 10 12:33:35 sorry :) Jun 10 12:33:45 polyonymous: don't worry Jun 10 12:34:09 zecke, it's just that sometimes it makes sense to think first and then speak :) Jun 10 12:34:39 polyonymous: this reminds me of Homer Simpson where Marge is gambling too much Jun 10 12:34:57 "because YOU have a gambling problem" Jun 10 12:35:06 zecke, (un)fortunately, I'm totally simpsons-illiterate :) Jun 10 12:35:40 ~botsnack for koen Jun 10 12:35:40 zecke: thanks Jun 10 12:35:56 ~snack for koen Jun 10 12:36:10 * ibot throws for koen a doggy treat Jun 10 12:37:31 koen: is installed=yes in your .la file? Jun 10 12:39:30 no idea Jun 10 12:49:49 * slapin_nb fixed gnom-menus. adding -dbg as first package fixed a problem. Jun 10 12:50:01 * * OE Bug 2462 has been created by slapinid@gmail.com Jun 10 12:50:03 * * QA error on gnome-menus package (.debug), fix inside. Jun 10 12:50:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2462 Jun 10 13:11:40 03polyonymous 07org.oe.dev * rd9d38708... 10/ (1 classes/qmake-base.bbclass): classes/qmake-base: reverted half a patch in order to fix qmake-based qte[-mt] builds. Jun 10 13:12:44 polyonymous: qt4-x11-free compiles fine now Jun 10 13:12:50 Thank you for the work! Jun 10 13:12:53 Laibsch, great! Jun 10 13:12:58 indeed Jun 10 13:13:31 I think it's almost well-polished now, but I have only built uqtreader on top of it, so it needs more testing. Jun 10 13:14:20 Umm.. and I think it needs rdepend on utf-16 something to work properly, but I forgot what exactly and therefore didn't add it to .bb yet. Jun 10 13:16:01 * * OE Bug 2442 has been RESOLVED (FIXED) by bugs.openembedded.org@rolf.leggewie.biz Jun 10 13:16:03 * *  qt4-x11-free fails do_compile Jun 10 13:16:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2442 Jun 10 13:17:01 * * OE Bug 2463 has been created by slapinid@gmail.com Jun 10 13:17:03 * * gnome-panel build error Jun 10 13:17:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2463 Jun 10 13:17:39 03koen 07org.oe.dev * r99aa7520... 10/ (1 packages/images/angstrom-minimal-image.bb): angstrom-minimal-image: drop DEPENDS On task-base Jun 10 13:18:00 koen: do you know if the n800 can do SDIO? Jun 10 13:18:10 koen: and when will TI OpenSource the DSP stuff? Jun 10 13:18:29 no idea and 'within 4 months' Jun 10 13:18:38 dspbridge has been gpled already Jun 10 13:21:13 zecke: and afaik nokia is still using dspgateway, which will dissappear soon according to TI Jun 10 13:32:51 zecke: Where does the log for QA_LOG go? Jun 10 13:33:04 It's somewhat ridiculous to have gtk+ built in the process of building opie image :) Jun 10 13:35:12 Laibsch: ${TMPDIR} Jun 10 13:36:26 Well, I looked there and there is nothing in it Jun 10 13:39:47 Laibsch: well, then your QA_LOG var is wrong? Jun 10 13:46:01 * * OE Bug 2464 has been created by slapinid@gmail.com Jun 10 13:46:03 * * scummvm doesn't build for qemux86 Jun 10 13:46:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2464 Jun 10 13:46:55 at ixp4xx-kernel-svnpatch.inc it includes patches from svn after do_patch -- how do i now add my own patches after this do_svn_patch ?? Jun 10 13:47:29 rob_w: why is there svn_patch at all? Jun 10 13:47:44 rob_w: you can directly add svn URLs to SRC_URI for patching Jun 10 13:48:27 History Lesson: http://video.google.com/videoplay?docid=-5046297730700144952 Jun 10 13:50:43 zecke how would i integrate ixp4xx-kernel-svnpatch.inc into fe ixp4xx-kernel_2.6.21.1.bb and added some private patches after this svn work ? Jun 10 13:50:59 zecke: it is set to "yes" Jun 10 13:51:16 Should it be set to ${TMPDIR} Jun 10 13:51:22 "1" Jun 10 13:51:28 aha Jun 10 13:53:13 Laibsch: well, set it to anything Jun 10 13:53:21 924K Angstrom-minimalist-image-uclibc-test-20070610.rootfs.tar.gz Jun 10 13:53:29 Laibsch: and it will write into 'T' at log.qa_package Jun 10 13:53:35 koen: :) Jun 10 13:53:49 * koen wonders if that can easily be made smaller Jun 10 13:53:51 koen: for me someone is regularily deleting intl.h from the staging area Jun 10 13:54:02 koen: sure, remove stuff from uclibc, statically link Jun 10 13:54:50 I still think it's pretty big for init, uclibc, dropbear, tinylogin + scripts Jun 10 13:54:53 * zecke should get an oven that actually works Jun 10 13:55:00 or can i move svn_patch before do_patch ? Jun 10 13:55:06 rob_w: wait a second Jun 10 13:55:10 ok Jun 10 13:55:54 but it's nice to be in the sub-megabyte range Jun 10 13:56:04 koen, tremendous Jun 10 13:56:08 koen: build it as cramfs! Jun 10 13:56:16 koen: just give it a try and see how it will work Jun 10 13:58:08 rob_w: oh, you can't do that easily Jun 10 14:14:33 hmm Jun 10 14:16:13 maybe i can apply my patches at do_patch without interferring the svnpatches Jun 10 14:17:29 rob_w: alternatively fork and remove svnpatch Jun 10 14:17:36 rob_w: and apply them by hand in the SRC_URI Jun 10 14:20:32 coulld i do : addtask svnpatch before do_patch ? Jun 10 14:23:40 hm mthat seems to work Jun 10 14:25:46 rob_w: hehe Jun 10 14:28:42 03koen 07org.oe.dev * r372eff83... 10/ (1 packages/base-files/base-files_3.0.14.bb): basefiles: preserve issue[.net] contents Jun 10 14:29:10 steliosk: now I we can finally see your cool ascii angstrom logo :) Jun 10 14:41:24 03koen 07org.oe.dev * re3af5468... 10/ (1 packages/base-files/base-files_3.0.14.bb): base-files: copy issue files Jun 10 14:42:47 morning Jun 10 14:43:17 good one. Jun 10 14:46:00 * * OE Bug 2465 has been created by jake@cogini.com Jun 10 14:46:02 * * Support /etc/defaults file for gpsd Jun 10 14:46:04 * * http://bugs.openembedded.org/show_bug.cgi?id=2465 Jun 10 14:55:26 koen : heh. It takes time for artists to get recognition :) Jun 10 14:55:51 steliosk, well, you're still alive :) Jun 10 14:59:00 polyonymous : and been in that "state" my artistic work does not worth much.... Jun 10 14:59:57 Most of works of art become precious after the artist dies. So I wish you long years of having no recognition as an artist :) Jun 10 15:00:08 koen: awesome! Jun 10 15:00:24 koen: could you try to cramfs and see if it is a lot smaller? Jun 10 15:00:50 zecke : It should get to about 60% of that Jun 10 15:01:51 zecke : scratch that.... it's tar,gz not jffs2 Jun 10 15:02:19 zecke: no cramfs in the kernel Jun 10 15:02:40 good morning :> Jun 10 15:02:52 hey Jin^eLD Jun 10 15:02:59 koen: well marekting. If you compare rootfs size you don't say something about bootable kernels? Jun 10 15:03:04 Jin^eLD: you need to use SRDATE, not DATETIME Jun 10 15:03:07 koen: you say a complete and minimal rootfs... Jun 10 15:03:22 zecke, koen : To reduce more you need to get stuff of uclibc/busybox basically Jun 10 15:03:30 koen:oops... so my versioning was screwed up? Jun 10 15:03:33 it is sooo hot... here Jun 10 15:04:00 speaking of SRCDATE where does it affect checkout? Jun 10 15:04:03 can any native english speaker spare a second for my uni assignment? Jun 10 15:04:10 Jin^eLD: yes, but I don't think it's related to your ipkg bug Jun 10 15:04:24 polyonymous: I suspect inside the bitbake fetchers Jun 10 15:04:27 oh ok.. bu tSRCDATE will also do a current timestamp? Jun 10 15:04:38 koen, I couldn't find it, where do they live? Jun 10 15:04:43 Has anyone any "rule of thump" on the number of threads per cpu/core on bitbake 1.8 ? 2xcpu/cores is it ok ? Jun 10 15:06:09 I've seen FETCHCOMMAND_cvs = "/usr/bin/env cvs -d${CVSROOT} co ${CVSCOOPTS} ${CVSMODULE}", but couldn't find where CVSCOOPTS is set... Jun 10 15:07:38 Jin^eLD: the fetcher gets the "last midnight UTC" revision, so DATETIME would be wrong Jun 10 15:08:07 uuh.. but my SRCDATE is set to "now", I want each build to check out the latest code Jun 10 15:09:03 ah found it Jun 10 15:09:27 will it be reset internally? else my PV may end up as 1.0.1+svnnow, which would not work, right? Jun 10 15:11:39 steliosk: Look at the amount of time your machine spends idle and increase until it doesn't ;-) Jun 10 15:12:03 * RP uses 2x, 3x or 4x :} Jun 10 15:13:27 RP: I have a dual Opteron machine (4 core total) doing builds and recording the results, but from the fist set i got it looks like its not as fast as i would expect it to be Jun 10 15:13:59 koen: its getting called as svnnow then, that can't be right.. Jun 10 15:14:13 steliosk: I sell CPUs specialised in accelerating m4 Jun 10 15:14:15 why not? Jun 10 15:14:23 I do not see, though where SRCDATE makes it into ;date= and I'm under an impression that it doesn't. On the other hand I'm under the opposite impression because it's used for the whole opie thing Jun 10 15:14:29 you set SRCDATE=now, so you get want you set Jun 10 15:14:48 * zecke thinks SRCDATE is broken for svn :} Jun 10 15:14:59 zecke, it isn't for cvs, though? Jun 10 15:15:04 koen: well yes, but I meant "can't be right" in the logical sense of the word :) I want to be able to upgrade the package each time a new build is avaialable Jun 10 15:15:09 zecke : That would make me, your fist client :) Jun 10 15:15:11 and each build checks out the latest code for that package from svn Jun 10 15:15:28 I would think tha I need some proper versioning to be able to do ipkg upgrade Jun 10 15:15:42 steliosk: IO bound machine? Jun 10 15:15:44 thats why I had the PV as "1.0.1+svn${DATETIME}" Jun 10 15:15:56 polyonymous: I don't know, I'm not really sure it is really broken for svn Jun 10 15:16:04 but you were saying that this is not correct? so... how would I do it then? > Jun 10 15:16:13 zecke, where does it make into uri, anyway? Jun 10 15:16:21 * zecke thinks with some luck he can get paid to do OE work Jun 10 15:16:43 RP: no. looking at top i see i get something like 20-30% idle and no (or very little) IO waiting even when set a 3x threads Jun 10 15:16:49 polyonymous: date? check the BitBake manual. There are two/three ways to set date,revision Jun 10 15:17:06 steliosk: Sounds like room for more tasks then ;-) Jun 10 15:17:44 zecke, in URI or via vars? Jun 10 15:17:56 well, maybe I indeed should read manual. Jun 10 15:18:09 although that feels weird :) Jun 10 15:18:37 RP: I thought that by 3x threads + j8 for make i can get the cpu's to be fully used.... but it does not look that way Jun 10 15:19:15 steliosk: I would like to know why it can't use all that power, it really should be capable of it Jun 10 15:20:14 IO bound? Jun 10 15:20:30 RP: not sure either.... but i have a feeling we will need 2 x the number of cores to see that Jun 10 15:20:35 koen : no Jun 10 15:22:10 The thing is i am testing it using bootstrap-image, which by nature has a lot of packages depending on the previous one, which i believe slow things down Jun 10 15:22:37 polyonymous: date;rev as part of URI Jun 10 15:23:00 zecke, yes, that I figured. I'm trying to figure if SRCDATE makes it into fetcher somehow or not. Jun 10 15:24:00 i.e at some point when all the unpacking and configuring has be done only a thread is actually doing something Jun 10 15:24:16 this is till gcc/glibc get build Jun 10 15:24:21 And I don't see that it does, although if it doesn't that would likely break opie build. Well, in fact it wouldn't, due to the lack of commits since SRCDATE, but it would break the very idea :) Jun 10 15:24:29 polyonymous: I have the feeling it broke with bitbake1.8 (SRCDATE) Jun 10 15:24:46 zecke, aha, so do I. Thanks for confirming it then. Jun 10 15:25:10 polyonymous: no conforming, I had the feeling while looking at musicpd Jun 10 15:25:32 zecke, confirming that you also have the feeling. I'm not going to sue you if you're wrong:)) Jun 10 15:28:02 /nick zecke|IANAL Jun 10 15:31:19 hi zecke Jun 10 15:31:55 hmm... "I, anal"? Jun 10 15:32:31 ah.. had to google for it. Jun 10 15:32:34 sure :) Jun 10 15:33:13 thanks God you ain't. Otherwise bipolar would hate you :-) (and not only bipolar) Jun 10 15:36:01 * * OE Bug 2466 has been created by jake@cogini.com Jun 10 15:36:03 * * Bitbake build for python-fam Jun 10 15:36:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2466 Jun 10 15:39:00 FAM? Jun 10 15:40:17 * polyonymous guesses it's file alteration monitor :) Jun 10 15:40:35 polyonymous: right, but check your date :) Jun 10 15:40:50 my date? Jun 10 15:40:56 hi all Jun 10 15:41:05 polyonymous: it is 2007 and people look into FAM Jun 10 15:41:14 ah Jun 10 15:44:24 Yup. File alteration monitor. To wait for changes in files. Jun 10 15:45:37 I also just finished getting python-gamin to build. Quite painful. Jun 10 15:46:11 Still need to test. Jun 10 15:47:12 reach: python-inotify is next? Jun 10 15:48:01 reach: #2466 looks good Jun 10 15:48:03 Could be. I have some existing code that uses fam. Jun 10 15:48:05 Like preserve chronological order? Jun 10 15:49:08 polyonymous: It runs on Debian Sarge, which doesn't support anything newer. Jun 10 15:49:23 reach: I'm sorry Jun 10 15:49:41 reach: you might want to add the md5 and sha256sum of py-fam to checksums.ini Jun 10 15:50:15 Ok. I will start doing that for my builds, I guess. Jun 10 16:03:50 for some unknown reason ortp reciepe does not produce package... Jun 10 16:05:26 I am having a very strange problem with uboot-git. It builds fine on 32bit host but fails on 64bit host -> http://rafb.net/p/8CzBJR29.html Jun 10 16:06:44 hi steliosk. Jun 10 16:07:01 * * OE Bug 2467 has been created by slapinid@gmail.com Jun 10 16:07:03 * * ortp reciepe does not produce package Jun 10 16:07:04 How's the news about mesa cross compile problem? Anyone considering our fixes? Jun 10 16:07:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2467 Jun 10 16:07:11 antiix hi Jun 10 16:27:35 hi all Jun 10 16:28:20 hi florian Jun 10 16:41:47 zecke: Is there anything other than the task dependency info missing from the bitbake manual? Jun 10 16:43:06 RP: task dependency, multi tasking (BB_NUMBER), do we have the -g documented? Jun 10 16:45:47 polyonymous: python-pyqt4 builds for you? Jun 10 16:46:08 zecke: is there some particular order for the entries in checksums.ini? It seems to be almost sorted, but not quite... Jun 10 16:46:51 Laibsch, haven't tried yet. Jun 10 16:47:10 Laibsch, not only I'm not a qt programmer, neither I'm a python guru :))) But I think I can try nonetheless Jun 10 16:51:06 zecke: We have -g documented but the docs are now wrong :/ Jun 10 16:52:46 polyonymous: I thought it might have come across as a dependency in one of your builds Jun 10 16:53:01 I think it build fine last time I tried Jun 10 16:53:07 Laibsch, my only build upon qt4 so far was uqtreader :) Jun 10 16:53:13 But right now I got an error Jun 10 16:53:41 baking now Jun 10 16:53:56 oh god, it started to build tcl as a dep :) Jun 10 16:57:56 reach: It seems you have an interest in python. Maybe you care to take a look at bug 2468? Jun 10 16:59:30 03rpurdie * r828 10/ (4 files in 4 dirs): bitbake trunk/1.8: Update manual Jun 10 17:00:14 zecke (and/or anyone else): Any reason not to release bitbake? Jun 10 17:00:35 Laibsch: Sure, I will take a look. Jun 10 17:00:58 huh, sip? Jun 10 17:01:22 or is it some common python prefix? Jun 10 17:02:01 * * OE Bug 2468 has been created by bugs.openembedded.org@rolf.leggewie.biz Jun 10 17:02:03 * * python-pyqt4 fails do_compile Jun 10 17:02:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2468 Jun 10 17:03:23 polyonymous: Perhaps it is http://www.riverbankcomputing.co.uk/sip being used to create the bindings Jun 10 17:03:48 ah, thanks. I warned I don't know a thing about python :) Jun 10 17:04:11 Though, I didn't know a thing about qt as well until lately :) Jun 10 17:04:30 polyonymous: this may be a good way to find out what goes on under the hood :-) Jun 10 17:04:55 that happens to me all the time :) I learn way more that I wanted to know :) Jun 10 17:06:29 polyonymous: and in the order that the problems occur, not any logical order :-) Jun 10 17:07:54 I see a python-sip4_4.4.5.bb package. Perhaps it should be a dependency for the pyqt4 package. Jun 10 17:08:20 could be, can you try it, Laibsch ? Jun 10 17:08:36 ah my build failed already, I can try it myself Jun 10 17:08:51 My build still has 700 tasks to go... Jun 10 17:09:58 polyonymous: there is already a dependency there from the sip4.bbclass Jun 10 17:10:01 * * OE Bug 2469 has been created by pmiscml@gmail.com Jun 10 17:10:04 * * Add ability to create image consisting of exact package list ( by not pulling up any dependencies) Jun 10 17:10:05 * * http://bugs.openembedded.org/show_bug.cgi?id=2469 Jun 10 17:10:34 koen, hrw|gone: can we have quick turnaround on #2469? Jun 10 17:12:51 03rpurdie 07bitbake-1.6.8 * r829 10/ (9 files in 4 dirs): Release bitbake 1.6.8 Jun 10 17:13:44 03rpurdie * r830 10/branches/ (6 files in 6 dirs): Update branches after release Jun 10 17:16:51 RP: bitbake 1.8.6 scheduled as well? Jun 10 17:17:04 koen: If svn would play nicely Jun 10 17:17:36 You mean 1.8.4, right? Jun 10 17:18:32 03rpurdie 07bitbake-1.8.4 * r831 10/ (7 files in 3 dirs): Tag semi 1.8.4 Jun 10 17:18:45 yes :) Jun 10 17:19:24 03rpurdie 07bitbake-1.8.4 * r832 10/ (ChangeLog bin/bitbake lib/bb/__init__.py): Release bitbake 1.8.4 Jun 10 17:19:40 Stupid svn couldn't cope with doing both those steps in one commit :/ Jun 10 17:22:51 reach, then it doesn't work, because I baked python-sip4 and it started to build it Jun 10 17:24:07 and the errors are different Jun 10 17:24:22 polyonymous: what are the errors? Jun 10 17:25:28 reach, http://rafb.net/p/jRm6e019.html - I haven't even read it yet Jun 10 17:27:01 reach, and sip class depends on python-sip, not python-sip4 Jun 10 17:28:14 polyonymous: what version of gcc are you using? Jun 10 17:28:41 as if I know, 4.1.2, I think. Whatever is default for angstrom. Jun 10 17:29:17 polyonymous: it may be being more strict about scopes than an older version. Jun 10 17:30:10 reach, maybe. as I told you - I haven't yet looked into the problem and haven't even bumped into it - it's Laibsch's fault :)) I will look into it, though, as soon as I sort out a few opie-reader issues. Jun 10 17:31:01 polyonymous: there is a sip class and also a sip4 class. Jun 10 17:31:39 reach, aha, you mentioned sip4, so I tried this one, I don't know what pyqt should depend on, of course. Jun 10 17:31:40 polyonymous: I am having trouble downloading http://www.riverbankcomputing.com/Downloads/sip4/sip-4.4.5.tar.gz, do you know of a mirror? Jun 10 17:32:07 not found... no, maybe oe mirrors/ Jun 10 17:32:14 angstrom mirrors? I dunno Jun 10 17:32:28 should we bump the version? Jun 10 17:32:46 there's a 4.6 there... Jun 10 17:33:04 re Jun 10 17:33:23 polyonymous: Might work... Jun 10 17:35:29 ~lart glibc-isms Jun 10 17:35:29 * ibot takes a large goose feather pillow and swings it wildly in glibc-isms's direction, hitting glibc-isms and sending glibc-isms flying into the closet Jun 10 17:35:39 polyonymous: and I can get that one from the Gentoo mirror... Jun 10 17:35:51 also the way to go :) Jun 10 17:36:02 the world is full of possibilities :) Jun 10 17:36:24 hi Jun 10 17:37:39 polyonymous: found a mirror here: http://linuxtogo.org/~mickeyl/mirror/ Jun 10 17:37:51 * koen wonders how to hack without grep, sed and and google Jun 10 17:38:15 preferably with no computer. Jun 10 17:38:19 ;) Jun 10 17:38:53 do you think that using GPE v2.8 is a good idea? Jun 10 17:38:58 Damn. uqtreader works nicely, but the buttonprefs dialog doesn't seem to work on Z. It works on desktop tho Jun 10 17:39:00 ~lart broken autotools needing MALLOC_GLIBC_COMPAT=y Jun 10 17:39:01 * ibot squishes broken autotools needing MALLOC_GLIBC_COMPAT=y like a bug Jun 10 17:39:12 i've found it a little unstable Jun 10 17:39:26 03rpurdie 07org.oe.dev * r1e91f243... 10/ (1 conf/sanity.conf): sanity.conf: Increase bitbake minumum version to 1.6.8 since earlier 1.6 releases don't work with .dev properly Jun 10 17:41:14 every gpe application exits with a SIGSEGV when i use the file-open dialog Jun 10 17:41:33 could it be something related to my platform? Jun 10 17:42:23 HEXaBYTE: unlikely, but i remember some gtk versions in oe had a broken file dialog Jun 10 17:43:08 the only version i really know to work fine is 2.6.10 Jun 10 17:43:47 =) Jun 10 17:43:47 the buttonprefs widget ignores all mouseclicks, hmm.. Jun 10 17:44:26 flo_lap, HEXaBYTE: http://bugzilla.gnome.org/show_bug.cgi?id=389603 Jun 10 17:45:31 polyonymous: poor boy, same version of Qt4? Jun 10 17:45:38 * chouimat is watching the Grand Prix ... huge accident Jun 10 17:45:52 zecke, not exactly. Jun 10 17:46:02 (4.3 on Z, 4.2 on desktop) Jun 10 17:46:12 polyonymous: if you have no other idea for debugging, install an EventFilter on the combobox Jun 10 17:46:24 koen, i believe that it isn't the problem that i have Jun 10 17:46:25 polyonymous: and the lineEdit() and other parts of it and see which events are coming by Jun 10 17:46:37 zecke, not combobox the whole lot of widgets don't receive keyclicks Jun 10 17:46:49 that bug is on GtkFileChooserDialog Jun 10 17:46:53 polyonymous: do you play with modality? Jun 10 17:46:58 I mean, there's a widget containing dialog, which is raised from widgetstack. Jun 10 17:47:21 Can non-dialog widget have modality? Jun 10 17:47:35 mine is on the simpler dialog in libgpewidget Jun 10 17:48:00 polyonymous: QDialog? Jun 10 17:48:20 zecke, non-dialog. It's a QWidget, in fact Jun 10 17:48:40 polyonymous: http://doc.trolltech.com/4.3/qdialog.html#modal-prop Jun 10 17:48:42 ah okay Jun 10 17:48:59 I'm comparing it with a working one (file selector) Jun 10 17:49:18 but the code is such a mess Jun 10 17:49:21 ~lart it Jun 10 17:49:21 * ibot puts it into a headlock and administers a mighty noogie, rubbing half of it's hair of Jun 10 17:49:38 silly ibot, "its" is spelt with no \' ;-) Jun 10 17:50:35 also s/of/off right? Jun 10 17:50:53 well, the 2nd "of" Jun 10 17:50:54 that too :) Jun 10 17:51:00 =D Jun 10 17:51:09 But that's a typo in his "database" or whatever it is. Jun 10 17:51:16 yeah Jun 10 17:51:19 The "its" issue is lack of intellect :) Jun 10 17:51:57 same w/ lose and loose, their and there Jun 10 17:52:11 your and you're hehe Jun 10 17:52:22 but that again would be the bug in database Jun 10 17:52:40 The thing is that he substituted "it", but didn't strip ' (of course) Jun 10 17:53:12 polyonymous: I doubt it, I bet it's a typo Jun 10 17:53:24 CoreDump, note, I typed '~lart it' Jun 10 17:53:28 ~lart it Jun 10 17:53:28 * ibot squeezes it till it turns blue like papa smurf Jun 10 17:53:41 well, there's no need of 's :) Jun 10 17:53:45 ah, no Jun 10 17:53:50 w I get it Jun 10 17:53:58 it definiely has something like %s's :) Jun 10 17:54:05 right Jun 10 17:54:11 heh Jun 10 17:55:32 ~hail gnu-configize Jun 10 17:55:40 * ibot bows down to gnu-configize and chants, "I'M NOT WORTHY!!" Jun 10 17:56:02 Isn't English fun? :-) Jun 10 17:56:25 I wonder if I should just upgrade qt on the desktop. I don't think I use many qt apps, anyway... Jun 10 18:04:25 gnu-configize? Jun 10 18:06:42 chouimat|tv: replaces things like config.sub, so you don't have to spend ages patching autofoo Jun 10 18:07:13 koen cool ... but I'm now 100% sold to cmake :) Jun 10 18:07:23 that u-boot not building on 64bit is weird Jun 10 18:07:48 the .o files produced on both hosts are exactly the same Jun 10 18:07:55 and the fail on the linker Jun 10 18:08:24 if i copy the 32bit cross linker on the 64bit host it links uboot fine... Jun 10 18:08:44 Thanks to whoever fixed projects.linuxtogo.org :) Jun 10 18:08:46 reach, polyonymous: Anything I should take a look at? I read from the backlog you were discussing a few things. Jun 10 18:08:59 hvontres|home: Are you sure it is fixed? Jun 10 18:09:06 hvontres|home: What was the problem, then? Jun 10 18:09:31 Laibsch, well, I don't know, I only know that if you bake sip4 errors change, but haven't got enough terminals to look deeper into it yet Jun 10 18:09:31 Laibsch: well, either that or whatever was sucking resources has quit Jun 10 18:10:08 hvontres|home: No, it was always something that only occurred occasionally Jun 10 18:10:20 hvontres|home: I think you probably rejoiced too early Jun 10 18:10:23 Laibsch: still compiling dependencies on my end... I couldn't download the distribution for PyQt-x11-gpl-4.0.tar.gz, so I am trying 4.0. Jun 10 18:10:28 er. 4.2 Jun 10 18:10:58 hvontres|home: koen said the machine was lacking RAM and swapping like crazy. It has 1 G and apparently a single mtn session does suck up about one third of that. Jun 10 18:11:07 4.2 is what is current for Gentoo, so 4.0 may have problems in general. Jun 10 18:11:16 ok, time to go Jun 10 18:11:44 reach: You advise to generally upgrade to 4.2? Jun 10 18:11:54 [ 5] 0.0- 5.0 sec 23.3 MBytes 4.65 MBytes/sec Jun 10 18:11:59 that's more than I expected Jun 10 18:12:12 koen: Doing what? Jun 10 18:12:15 for a puny avr32 board Jun 10 18:12:18 Laibsch: iperf Jun 10 18:12:29 on ltg.org? Jun 10 18:12:33 Laibsch: I have no specific info one way or the other, but it may work better. Jun 10 18:12:52 Laibsch: no, from my workstation to the board Jun 10 18:13:55 reach: I just tried "bitbake python-sip4" but it had already been built previously Jun 10 18:14:10 reach: So, I guess adding that dependency is not going to fix it. Jun 10 18:14:37 no, it's not, I told you - it changed errors, not fixed them :) Jun 10 18:15:12 Laibsch: I was thinking that it may be a gcc issue, based on the last error from polyonymous, it may be more strict about scopes or something. Jun 10 18:15:41 it would be worth to take a look at the code, not only errors, but I'll get completely lost in terminals then :) Jun 10 18:23:26 03coredump2 07org.oe.dev * raf98a46d... 10/ (1 packages/gspcav1/gspcav1_20070110.bb): gspcav1: Update SRC_URI Jun 10 18:29:32 Laibsch: I am blocked now building qt4-x11-free-4.1... it is using the system freetype instead of the bundled version. Getting late here, 3:30 AM Taiwan time... so I will take another look at it tomorrow if you haven't found a solution. Jun 10 18:30:08 er, 2:30. Like I said, getting late :-) Jun 10 18:43:32 how do i determine the correc eraseblock size for a jffs2 image ? Jun 10 18:44:05 rob_w: it is given by the hardware? Jun 10 18:44:19 which info do i need to lock at Jun 10 18:44:46 redboot fe shows FLASH: 0x50000000 - 0x51000000, 128 blocks of 0x00020000 bytes each. Jun 10 18:45:09 rob_w: can't remember, I would guess 0x00020000 Jun 10 18:45:33 i tryed 40000 now i try 20000 Jun 10 18:45:57 03Laibsch 07org.oe.dev * r64150943... 10/ (3 files in 2 dirs): Jun 10 18:45:57 gspcav1: update to version 20070508 to make it fetchable again. Jun 10 18:45:57 Otherwise untested, not even compilation test. Jun 10 18:46:50 Laibsch: I hope you kept the old version? Jun 10 18:50:54 zecke|math, still Node at 0x004dfa08 with length 0x0000082d would run over the end of the erase block Jun 10 18:51:53 hooray! After qt upgrade it doesn't work on desktop either. Jun 10 18:53:24 03coredump2 07org.oe.dev * r891a91ed... 10/ (1 packages/gspcav1/gspcav1_20070110.bb): gspcav1: Add known-good version from 20070110 Jun 10 18:55:34 rob_w: a) you are hopeless and fix the printk to mention the erase block Jun 10 18:55:43 rob_w: b) search your nand/nor driver Jun 10 18:57:30 * koen ownders why the efika won't boot Jun 10 18:57:55 or rather, why it hangs during netboot Jun 10 18:58:26 koen : You need to set all the parameters right or it won't Jun 10 18:59:18 I think I need to put a fan on the heatsink of the radeon.... Jun 10 19:06:09 koen: do you think we cna activate switch_root in angstrom busybox, or is it time to create busybox-initrd recipe? Jun 10 19:06:23 both :) Jun 10 19:06:36 I've been meaning to add a busybox-static recipe for ages :) Jun 10 19:06:37 ;-) Jun 10 19:07:55 steliosk: http://dominion.kabel.utwente.nl/koen/cms/measuring-network-throughput-with-iperf Jun 10 19:09:44 the network is gigabit ? Jun 10 19:09:51 yes Jun 10 19:10:17 cat 5e or cat 6 ? Jun 10 19:10:57 the gigabit machines have cat6 Jun 10 19:11:52 The server is also ok then, with very few exceptions stock gigabit adapters don't go much higher than that Jun 10 19:12:07 the efika on the other hand is really low Jun 10 19:12:20 right, I don't want to fumble with mtu and jumboframes Jun 10 19:12:30 old kernel, debian rootfs on the efika Jun 10 19:13:02 I'm preparing to update to a recent kernel and OE rootfs Jun 10 19:13:22 I finally figured out why it only booted half of the times Jun 10 19:13:28 overheating agp card... Jun 10 19:13:31 ~lart ATI Jun 10 19:13:31 * ibot readies the nuke launcher and fires some rounds at ATI Jun 10 19:13:38 i would expect the efika to reach the max on 100mbit segment (i.e around 75/80 mbits) Jun 10 19:13:52 s/expect/expected Jun 10 19:14:13 it looks like 20-25% less than that Jun 10 19:14:14 the avr board impressed me Jun 10 19:14:33 I suspect we could squeeze an Mbyte/s more out of it Jun 10 19:15:58 i have been using avr (the 8 bit variety) since they came to market, and they really are fast chips with a very nice architecture. From the little i managed to read on avr32 it seems that it keeps on that Jun 10 19:17:29 Also they have some "dsp" functionality for mpeg and stuff so i can assume with optimization you could see a big performace difference from other arch's in the same frequency Jun 10 19:18:20 * koen reads http://www.softcatala.org/~jmas/bloc/pivot/entry.php?id=287 Jun 10 19:18:41 steliosk: atmel claims 30fps h264 at qvga Jun 10 19:18:53 I'm waiting for the stk to arrive to verify that :) Jun 10 19:19:22 heh i have seen a demo ;) Jun 10 19:27:01 Grand Prix de Montreal: 10 out and 1 dead groundhog Jun 10 19:28:52 RP: First build results.... Jun 10 19:28:56 zecke|math, the first widget added to widgetstack exhibits this weird behaviour - that is if I swap the widget with the next one it works fine... but the other of course doesn't. Jun 10 19:29:50 RP: With threads set to 0 and -j also set to 0, build time is 160minuts 22.836 sec Jun 10 19:30:18 RP: With threads set to 4 and -j set to 8, build time is 158min 11.913s Jun 10 19:31:00 RP: not that big difference... Jun 10 19:31:20 RP: now building with threads 8 and -j8 Jun 10 19:46:25 zecke, looks like half-a-bug in qt3 support and half-a-bug in opie-reader. I decided to fix it by rewriting to use non-deprecated qt4 QStackedWidget Jun 10 19:46:51 polyonymous: could you extract the code and send it to me (buildable as qmake?) Jun 10 19:47:00 polyonymous: or mail it directly to qt-bugsATtt.com? Jun 10 19:47:28 zecke, I said 'half-a-bug', because opie-reader behaviour was unnice, anyway. So I'm not sure if one should blame it on qt Jun 10 19:47:52 the qt problem was inconsistent behaviour. Jun 10 19:48:00 polyonymous: well if Qt4.2 and 4.3 behaved differently it is a bug Jun 10 19:48:11 In that sense, yet. Jun 10 19:49:02 zecke, I do not fancy the idea of hacking the extracted code from ground up... Jun 10 19:54:52 I don't think it's a good idea to have altboot installed by default. Jun 10 20:07:00 Package netbase wants to install file /etc/{protocols,services} But that file is already provided by package iana-etc Jun 10 20:13:45 oh well, basically, uqtreader is working on qt4-x11 now. Jun 10 20:14:40 03polyonymous 07org.oe.dev * r1eb1e872... 10/ (3 files in 3 dirs): Jun 10 20:14:40 uqtreader: a lot of fixes and improvements Jun 10 20:14:40 - removed opie inheritance Jun 10 20:14:40 - replaced hardcoded dirs with sane variables Jun 10 20:14:40 - new and better patch that fixes quite a few issues Jun 10 21:14:37 Laibsch, basically, most of this python*4 stuff picks up qt2 specs with all the natural consequences... Jun 10 21:43:23 steliosk: Strange, I would have expected better results than that Jun 10 21:44:34 RP: Just ignore them. I had forgot to change the thread to the correct value between runs Jun 10 21:44:51 RP: with 4 threads it cuts buildtime to half Jun 10 21:45:07 RP: trying now with 8 Jun 10 21:46:15 night Jun 10 21:47:26 'night zecke Jun 10 21:47:38 steliosk: That's more like it :) Jun 10 21:51:04 * CoreDump wonders why his ssh is b0rked Jun 10 21:53:12 ~lart FU Berlin Jun 10 21:53:13 * ibot stuffs FU Berlin into a shiny new tin can and vacuum seals it Jun 10 22:34:11 Laibsch, pyqt supports qt up to version 4.2.3 Jun 10 22:53:01 03pfalcon 07org.oe.dev * r5dd95087... 10/ (1 packages/helloworld/helloworld_1.0.0.bb): helloworld 1.0.0: Remove duplicate content. Jun 10 22:53:05 03pfalcon 07org.oe.dev * r4e1403b6... 10/ (3 files in 2 dirs): Jun 10 22:53:05 busybox-static 1.2.1: First cut at static busybox. Jun 10 22:53:05 * Yes, it works. 877916 bytes with uclibc. Jun 10 22:53:05 * Defconfig is based on angstrom's one, with switch_root enabled. Jun 10 22:53:09 03pfalcon 07org.oe.dev * rc8599288... 10/ (1 classes/rootfs_ipk.bbclass): Jun 10 22:53:09 rootfs_ipk.bbclass: Add new option, PACKAGE_INSTALL_NO_DEPS. If set to 1, Jun 10 22:53:09 list of packages in PACKAGE_INSTALL installed as is, without any dependencies. Jun 10 22:53:11 * Useful for initrds, etc. Jun 10 22:53:13 * Closes #2469. Jun 10 23:05:30 03polyonymous 07org.oe.dev * rde4ea7e7... 10/ (1 packages/subversion/subversion_1.4.3.bb): subversion 1.4.3: removed DEFAULT_PREFERENCE of -1 and added do_stage Jun 10 23:05:34 03polyonymous 07org.oe.dev * r3a6b01b0... 10/ (4 files in 3 dirs): qsvn: made it build for the sake of additional qt4 testing Jun 10 23:09:23 03pfalcon 07org.oe.dev * r8fb7c19a... 10/ (1 packages/initrdscripts): initrdscripts: New package dir for initrd, initramfs, etc. recipes. Jun 10 23:09:28 03pfalcon 07org.oe.dev * r64152bb9... 10/ (4 files in 3 dirs): Jun 10 23:09:28 initramfs-nfsboot 0.1: Sample initramfs script to boot from NFS. Jun 10 23:09:28 * Need too add parsing of kernel cmdline instead of hardcoded params. Jun 10 23:09:33 03pfalcon 07org.oe.dev * r4c877301... 10/ (1 packages/images/initramfs-image.bb): Jun 10 23:09:33 initramfs-image: A sample initramfs image. Jun 10 23:09:33 * Shows all ingredients of a good initramfs image: custom init script run Jun 10 23:09:33 by statically linked busybox, with all of them installed by explicit list Jun 10 23:09:34 without any unneeded dependencies. ~500k. Jun 11 02:26:54 03coredump2 07org.oe.dev * rb80d9ebb... 10/ (3 files in 3 dirs): netbase: Correct a comment in /etc/network/interfaces **** ENDING LOGGING AT Mon Jun 11 02:59:57 2007