**** BEGIN LOGGING AT Sun Mar 30 02:59:56 2008 Mar 30 06:23:09 qtopia already run s great on a neo. beter than anything else Mar 30 06:42:12 03xjqian 07org.oe.dev * re451e643... 10/ (8 files in 3 dirs): Mar 30 06:42:12 wxbase: fix broken, unify Mar 30 06:42:12 * upgrade 2.6.2 to 2.6.4 (upstream old stable, DEFAULT) Mar 30 06:42:12 * upgrade 2.7+cvs to 2.8.7 (upstream new stable) Mar 30 06:42:12 * NOTE: upstream only provide mwg build targets for > 2.7 Mar 30 06:42:13 * 2.8.7: patch Makefile non-sense Mar 30 08:33:11 03xjqian 07org.oe.dev * r290d13f5... 10/ (5 files in 2 dirs): Mar 30 08:33:11 wxwidgets: fix packaging/staging. Mar 30 08:33:11 * upgrade 2.6.3 to 2.6.4 (old stable) Mar 30 08:33:11 * add 2.8.7 (new stable, DEFAULT) Mar 30 08:33:11 * fix packaging/staging following wxbase, unify Mar 30 10:43:14 good morning! Mar 30 11:50:52 hi all Mar 30 11:56:22 terminfo seems missing from rootfs in .dev Mar 30 12:00:57 Whos is going to attend the Embedded Systems Conference here? Mar 30 12:39:04 * * OE Bug 4134 has been created by  Mar 30 12:39:06 * * embryo-native-0.9.1.042+cvs20080330-autobuild Mar 30 12:39:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4134 Mar 30 13:06:14 03xjqian 07org.oe.dev * r129bb5af... 10/ (3 files in 3 dirs): tosa asound.state: fix default values Mar 30 13:17:39 03xjqian 07org.oe.dev * r7a5e2e80... 10/ (8 files in 3 dirs): Mar 30 13:17:39 gnuplot: fix configure Mar 30 13:17:39 * remove unnecessary depends (wxwidgets, etc) and configure option Mar 30 13:17:39 * fix SRC_DATE in 4.3.0+cvs Mar 30 13:17:39 * bump default to 4.2.3 Mar 30 14:47:46 ~seen thesing Mar 30 14:47:49 thesing was last seen on IRC in channel #opensimpad.org, 1d 17h 7m 41s ago, saying: 'bis bald'. Mar 30 14:52:04 * * OE Bug 4135 has been created by  Mar 30 14:52:06 * * claws-mail-3.3.0-autobuild Mar 30 14:52:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4135 Mar 30 14:52:16 * * OE Bug 2158 has been REOPENED by Mar 30 14:52:18 * * angstrom collie fails to shut down Mar 30 14:52:20 * * http://bugs.openembedded.org/show_bug.cgi?id=2158 Mar 30 15:13:05 * * OE Bug 4136 has been created by jeesmann(AT)j01.itscom.net Mar 30 15:13:07 * * Akita: CF-cards not recognized Mar 30 15:13:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4136 Mar 30 15:20:31 03likewise 07org.oe.dev * r2e450cce... 10/ (1 classes/base.bbclass): base.bbclass: Enable 'overwrite' flag for unzip (to match tar behaviour). Broke rebuilds when disabled. Mar 30 15:46:43 Hi HopsNBarley, will you be attending the Embedded Systems Conference by any chance? Mar 30 15:47:19 hi likewise - i am, in fact, possibly doing some job hunting (-; you coming over? Mar 30 15:49:10 HopsNBarley: I'm trying to, I am being shipped to Vegas for http://www.nabshow.com/, but I'ld love to get a few days of the Linux classes as well. Small chance I can arrange that though... Mar 30 15:50:03 if there is anything you want me to examine, i'd be happy to do that. Mar 30 15:53:12 HopsNBarley: I am interested if the classes are videoed and put online, https://www.cmpevents.com/ESCw08/a.asp?option=C&V=1&SB=3 Mar 30 15:53:50 which class(es) are you interested in? Mar 30 15:55:29 349, 369, 431, 451, (450, 470) Mar 30 15:56:22 Have to be in Vegas up to and including wednesday, driving to San Jose will take me, what, 8 hours? Mar 30 15:57:24 yeah, but you could fly in about an hour. Mar 30 15:57:32 several flights daily. Mar 30 15:59:20 500 classes? Mar 30 16:07:17 likewise, southwest has some one-way fares at just $44. (wow!) Mar 30 16:25:05 * * OE Bug 4137 has been created by pieterg(AT)gmx.com Mar 30 16:25:07 * * ipkg: pre/post control files not removed ( remove_obsolete_maintainer_scripts) Mar 30 16:25:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4137 Mar 30 16:26:04 * * OE Bug 4138 has been created by  Mar 30 16:26:06 * * midori-0.0.17-autobuild Mar 30 16:26:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4138 Mar 30 16:53:35 HopsNBarley: thanks, yes, I am considering. Mar 30 16:55:25 likewise, it'd be no problem to pick you up and take you over to the conference. i'm heading over anyway. Mar 30 17:05:44 HopsNBarley: I need to see what I would miss out in Vegas, I think I know by April 4th. Thanks for the offer! I'll let you know if I plan to go. I could take a hotel that's on route for you. Mar 30 17:05:59 to miss out *on*, right? Mar 30 17:06:35 english idiomatic advice? Mar 30 17:07:17 i'd use either "would miss in vegas", or "miss out on in vegas" - so yes (-; Mar 30 17:16:18 hi Mar 30 17:25:49 HopsNBarley: tnx :-) Mar 30 17:55:46 cya - need some serious gaming. Mar 30 19:42:06 03hrw 07org.oe.dev * rc01723e9... 10/ (1 conf/distro/include/sane-srcrevs.inc): sane-srcrevs.inc: bump matchbox-keyboard to solve 'unable to find expat' problem Mar 30 19:47:45 hi, i checked .dev yesterday out, and make a fresh build, but now Packages.sig is missing for request of opkg Mar 30 19:48:19 whats wrong by me? any hints Mar 30 19:49:14 a opkg update fails, because of that Mar 30 19:53:51 morning Mar 30 19:54:10 Packages.sig missing is not critical error but warning Mar 30 19:54:57 morning all Mar 30 19:55:08 elo Richard Mar 30 19:55:45 * hrw sends fucks to maemo and xkb Mar 30 19:57:52 hrw: okay, thanks Mar 30 20:13:44 hrw: maemo is screwed up allright :) Mar 30 20:19:25 bye Mar 30 20:19:37 bye hrw Mar 30 20:49:28 hi flo Mar 30 20:49:40 hi all Mar 30 20:49:47 hi flo_lap Mar 30 21:10:22 night all Mar 30 21:11:07 nite mrdata Mar 30 21:29:15 hi thesing Mar 30 21:30:07 hi everybody Mar 30 21:34:06 hmm, well my first try of cvs2svn2git didn't explode too badly :} Mar 30 21:36:07 RP: did your gcc commit fix the libgcc issue? Mar 30 21:38:33 thesing: I don't have a fix for that Mar 30 21:38:47 thesing: I did just send an email to the oe dev list about it though Mar 30 21:40:33 thesing: and finding that much has taken way too long :( Mar 30 21:48:12 hi all... Mar 30 21:48:52 could anybody enlighten me with the following issue? gpe-mini-browser2 builds fine in OE until the linking stage where it fails linking to libstdc++ Mar 30 21:49:17 which is not an issue on a normal system. It works if I manually add the linker flag. Mar 30 21:49:30 in the gpe-mini-browser2 makefile Mar 30 21:50:55 Philippe: Whats it linking with? Mar 30 21:51:08 g'morning Mar 30 21:51:21 Specifically is it using gcc to link against a c++ lib? Mar 30 21:51:25 hi ant Mar 30 21:51:31 RP: I'm having a little problem with packaged staging Mar 30 21:51:41 I deleted all apart /pstage Mar 30 21:52:08 but there is ane error on zaurus-updater Mar 30 21:52:29 needs -c clean and -c rebuild apparently... Mar 30 21:53:21 yes, after this rebuild both images (console and x11) rebuild very quicly Mar 30 21:53:56 RP: very nice ! Mar 30 21:53:58 RP: webkit Mar 30 21:54:17 the webkit linking fails, because webkit uses libstdc++ Mar 30 21:57:39 RP: this is the output http://www.pastebin.ca/963899 Mar 30 21:58:05 so yes it is linking against a c++ lib. Mar 30 22:01:12 Philippe: There are two schools of thought on this. Some say that webkit shouldn't really need stdc++ and that is the problem that needs fixing. The other thought is that while webkit is still using c++, you should be linking with g++ instead of gcc Mar 30 22:01:38 Philippe: A really quick hack is to add an empty dummy.cpp file into the autofoo Mar 30 22:02:42 * ant is following the libstc++.so.6 / libc6 discussion, hoping this can help with insane gpsd deps Mar 30 22:03:39 xjqian: ^^^ Mar 30 22:04:25 ant: Use a recent gpsd and the problem goes away Mar 30 22:04:44 RP: it pulled libc6 last time... Mar 30 22:04:48 ugh Mar 30 22:06:09 RP: ok. That clears up some things. It is just weird that it does not happen on my normal working machine. Mar 30 22:06:41 RP: thanks. I'll think of a hack then Mar 30 22:06:53 Philippe: Its OE's staging that does it, libstdc++ is in cross and not staging Mar 30 22:07:29 Philippe: The advanatge of that layout is it exposes bugs like this Mar 30 22:07:30 RP: read your mail. So what do you suggest to do now? Remove this check from the configure file? I have some time to look into this. Mar 30 22:07:59 RP: I see. So following the theory webkit-gtk is buggy right? Mar 30 22:10:48 RP: I could reproduce the zaurus-updater issue... | cp: cannot stat `updater.sh.c7x0': No such file or directory Mar 30 22:11:54 Philippe: well, it webkit meant to link against stdc++ or not? If it is, gpe-mini-browser is buggy, if it isn't, webkit is buggy Mar 30 22:12:21 ant: That doesn't surprise me, do_deploy in zaurus-updater needs tweaking Mar 30 22:12:57 thesing: Changing configure is one approach, yes Mar 30 22:13:27 thesing: I really don't know enough about where, when and why libgcc is required :( Mar 30 22:14:19 RP: it generates an huge error log: log.do_rootfs 1129446 (for nothing... (-;) Mar 30 22:14:54 RP: I will research about libgcc first. Mar 30 22:17:13 thesing: That would be good. I'm afraid I don't have a lot of time to look into it :( Mar 30 22:17:29 thesing: Was it only pthread showing problems with libgcc? Mar 30 22:18:01 thesing: If so it might be worth looking into what pthread is dynamically opening using something dl_open Mar 30 22:18:22 dl_open libraries aren't picked up by our shlibs code and need to be manually specified Mar 30 22:18:49 I don't know. Machine hang at boot. And didn't have a serial console without libgcc. Mar 30 22:19:12 thesing: boot it, then remove the lib and experiment? :) Mar 30 22:19:36 Will try. Mar 30 22:20:44 ant: Fix for do_deploy in zaurus-updater pushed Mar 30 22:21:07 RP: thanks for the quick fix Mar 30 22:21:57 RP: summary about libgcc http://gcc.gnu.org/onlinedocs/gccint/Libgcc.html Mar 30 22:22:10 03rpurdie 07org.oe.dev * ra61ed678... 10/ (1 packages/zaurus-updater/zaurus-updater.bb): zaurus-updater: Teach do_deploy about packaged staging Mar 30 22:23:32 thesing: perhaps find out which symbols are referenced from libgcc and what happens to these when libgcc isnt used Mar 30 22:49:16 RP: do these /pstage paths seem sane to you ? http://www.pastebin.org/26362 Mar 30 22:50:13 RP: some native packages are rebuilt, even from packaged staging Mar 30 22:50:53 RP: and curiously are in this fishy dir ./angstromglibc-oe-build-tmp-angstrom Mar 30 22:51:57 RP: pthread dlopens tries to dlopen libgcc Mar 30 22:53:54 RP: and my host system also doesn't have libgcc in NEEDED but has it installed. Mar 30 22:56:55 RP: I think it would be best do let glibc RDEPEND on libgcc. Mar 30 23:05:11 RP: for example by adding RDEPENDS += "${@['','libgcc']['ntpl' in '${GLIBC_ADDONS}']}" to glibc.inc Mar 30 23:06:45 thesing: Yes, I think the package with libpthread in needs to RDEPEND on it Mar 30 23:07:09 thesing: We need a nice big comment on this... Mar 30 23:08:19 RP: as far as I can tell libpthread is included in the glibc package. Mar 30 23:11:28 thesing: http://gentoo-wiki.com/NPTL Mar 30 23:15:22 ant: yes but it's included in glibc now. Mar 30 23:17:33 thesing: Support for some architectures added on, not maintained in glibc core. Mar 30 23:17:52 RP: should I go and add the line to glibc.inc with a comment bump all glibc PRs and commit? Mar 30 23:19:04 ant: Those packaged staging paths are all sane and correct Mar 30 23:19:30 ant: Go and read the mailing list posts, particularly the bit about -naitve and -cross packages Mar 30 23:19:55 ok, thx, I'll re-read Mar 30 23:19:57 thesing: For this one I'd post to mailing list first Mar 30 23:20:33 thesing: explain about the dl_open calls etc Mar 30 23:22:55 RP: the only think I can't explain is why libgcc was in NEEDED but isn't any more. But as it is not in NEEDED on std. x64 Debian it should be ok. Mar 30 23:34:15 RP: ok, I found the info in packaged-staging.bbclass, I was somehow thinking the tools could also come from package Mar 30 23:34:27 thesing: Maybe crtbegin needed something from libc or libgcc and was getting it from the latter rather than the former due to the flags? Mar 30 23:34:55 ant: The tools can but only if you don't change TMPDIR Mar 30 23:35:08 yes, now I get it Mar 30 23:40:01 good night Mar 31 00:00:35 Night all Mar 31 02:01:11 Crofton: hey **** ENDING LOGGING AT Mon Mar 31 03:00:02 2008