**** BEGIN LOGGING AT Wed Nov 14 02:59:57 2007 Nov 14 04:31:51 Crofton: are you attending FOSDEM? Nov 14 05:43:53 Does anybody know why monotone.openembedded.org is down right now? Nov 14 06:16:53 Where can I get the list for all recently revision's build status? The buildbot site only show the last 20 builds.. Nov 14 06:19:56 sorry , I have asked in the wrong channel.. Nov 14 07:01:10 Has anybody been able to pull from monotone.openembedded.org lately? I have been trying for the last hour and I keep getting mtn: network error: failed to connect: Connection refused Nov 14 07:03:36 hvontres|home: hmm I am seeing same problem Nov 14 07:06:42 Khem: hmmm, thanks. I guess somebody needs to kick the server :) Nov 14 07:07:17 well, time to go to sleep. Nov 14 07:42:17 morning Nov 14 07:58:56 morning! Nov 14 08:01:23 monotone.openembedded.org is still down.. Nov 14 08:20:20 note that monotone.nslu2-linux.org is up, and also contains the org.openembedded.dev branch (but some others as well, so make sure you ask for exactly what you want from it) Nov 14 08:30:14 gm Nov 14 08:32:53 monring Nov 14 08:33:12 good morning all Nov 14 08:33:50 the tunnel between monotone.openembedded.org and the actual monotone server is back up Nov 14 08:34:10 * CM updates Nov 14 08:34:10 some incompetent sysadmin killed the university router last night Nov 14 08:39:00 koen: its working now thanks Nov 14 08:43:14 03mickeyl 07org.oe.dev * r12505e50... 10/ (1 packages/tasks/task-openmoko.bb): task-openmoko.bb: remove vera fonts. gtk+ already suggest DejaVu ones. Nov 14 08:43:19 03mickeyl 07org.oe.dev * r9430cfa5... 10/ (1 packages/openmoko2/openmoko-session2.bb): openmoko-session2: RREPLACE/RPROVIDE instead of RCONFLICT Nov 14 08:44:27 mickeyl: AAAAARGH! Nov 14 08:44:47 that just killed everybodies deploy dir Nov 14 08:45:09 you can't build an image with matchbox-common now anymore :/ Nov 14 08:46:07 oops Nov 14 08:46:41 XorA|gone's commit message from yesterday states why it is only RCONFLICTS Nov 14 08:47:25 Khem: I was looking at eglibc configuration the other day. Does it make sense to trim down eglibc depending on existing OE variables (I think some localization flags are present, not sure on exactly what their meaning is) Nov 14 08:48:01 likewise: the locale stuff needs overhaul it should be possible to do a cross-build of locales Nov 14 08:48:18 khem: overhaul in eglibc? Nov 14 08:48:43 in OE's way of building eglibc Nov 14 08:49:19 Khem: ok, I was first trying to understand what the configuration options are of (e)glibc. Never looked into that before. Nov 14 08:49:26 likewise: but right now I am seeing issues with compiling at -Os is broken :( on trunk so I am trying to fix that locally Nov 14 08:49:39 Khem: on what platform(s)? Nov 14 08:49:42 koen: can't we just add a PREFERRED_PROVIDER? because otherwise you can't build an om-image once you have built matchbox. Nov 14 08:49:58 likewise: the options are relatively new Nov 14 08:49:59 koen: or will REPLACE be enough and we just remove PROVIDE? Nov 14 08:50:04 mickeyl: you can't Nov 14 08:50:13 mickeyl: ipkg looks at everything in deploy/ipk Nov 14 08:50:22 likewise: I saw it on ARM Nov 14 08:50:31 we had the same problem with matchbox--2 vs matchbox- Nov 14 08:50:40 likewise: thats the only arch I tried though Nov 14 08:50:54 koen: hmm. bummer Nov 14 08:50:59 yeah Nov 14 08:51:09 khem: how did you see breakage? Do you run the tests on a target? Nov 14 08:51:20 it took a while to figure out why all angstrom x11 image wouldn't boot Nov 14 08:51:23 Khem: ah compile breakage, right? Nov 14 08:51:40 till someone notices it had mb-panel-2 instead of mb-panel Nov 14 08:52:09 likewise: yes Nov 14 08:52:56 likewise: I believe its mostly because of static inline and extern inlines Nov 14 08:53:01 03koen 07org.oe.dev * r3427857a... 10/ (1 packages/openmoko2/openmoko-session2.bb): disapproval of revision '9430cfa5a334ac688d23932a066f6495abfa4709' Nov 14 08:54:31 anyway time to sleep bye all Nov 14 08:56:20 http://www.pennergame.de/ref.php?refid=6219097 Nov 14 08:58:49 Khem: good night Nov 14 09:19:12 03koen 07org.oe.dev * r9934077f... 10/ (1 packages/openmoko2/openmoko-session2.bb): openmoko-session2: bump PR massively to weed out broken packages in deploy and existing feeds Nov 14 09:23:47 koen: what difference does that make compared to the two PR bumps I commited yesterday Nov 14 09:26:57 gmorning Nov 14 09:27:00 arm-linux-uclibc/include/linux/wireless.h:1061: error: '__u16' does not name a type Nov 14 09:27:00 | /OE/OpenEmbedded/toolchain/armv4t-kernel_2.6-staging/arm-linux-uclibc/qt2/include/qdatetime.h:51: warning: 'class QDate' has virtual functions but non-virtual destructor Nov 14 09:27:03 | /OE/OpenEmbedded/toolchain/armv4t-kernel_2.6-staging/arm-linux-uclibc/qt2/include/qrangecontrol.h:52: warning: 'class QRangeControl' has virtual functions but non-virtual destructor Nov 14 09:27:06 | make[1]: *** [wlanimp2.o] Error 1 Nov 14 09:27:08 | make[1]: Leaving directory `/OE/OpenEmbedded/tmp/work/opie-networksettings-1.2.3-r0/networksettings/wlan' Nov 14 09:27:14 any1 ever saw this error? Nov 14 09:27:59 XorA: the disapprove decreased PR Nov 14 09:34:59 nik0n_: #opie? Nov 14 09:35:48 nik0n_: what program are you building that includes wireless.h ? Nov 14 09:35:54 this is a kernel header Nov 14 09:36:05 * mickeyl afraid of the answer Nov 14 09:36:25 make[1]: Leaving directory `/OE/OpenEmbedded/tmp/work/opie-networksettings-1.2.3-r0/networksettings/wlan' Nov 14 09:36:33 so, not wellenreiter :) Nov 14 09:36:39 *phew* Nov 14 09:37:06 then again, i think i had a hand in the wlan site survey dialog in networksettings as well ;) Nov 14 09:37:11 somewhere it should read Nov 14 09:37:17 "powered by Wellenreiter technology" Nov 14 09:37:23 iirc Nov 14 09:37:34 mickeyl: opie-networksettings-1.2.3-r0 Nov 14 09:38:12 patch it Nov 14 09:38:20 should include before including wireless.h Nov 14 09:38:29 gonna give it a try Nov 14 09:39:05 reparsing all my recipys atm ... +wait+ Nov 14 09:52:32 lumag: where you are from? Nov 14 09:53:32 mickeyl: i have linux/types.h always before linux/wireless.h any other ideas? Nov 14 09:53:44 hrw: Russia :) Nov 14 09:54:00 lumag: russia is big Nov 14 09:54:03 ooh, you're building for uclibc Nov 14 09:54:07 didn't see that Nov 14 09:54:14 no idea whether this has the same types Nov 14 09:54:15 St. Petersburg Nov 14 10:00:13 hrw: St. Petersburg Nov 14 10:00:59 thx Nov 14 10:19:07 koen: oh, emails finally arrive, I see mickeyl did what I did/undid :-D Nov 14 10:20:48 hello Nov 14 10:20:56 somebody working on navit? Nov 14 10:21:12 in oe there is the 0.0.1 release (not working) Nov 14 10:21:35 I compiled in some way from CVS and appears to work Nov 14 10:21:49 major problem for me: needs speechd Nov 14 10:22:02 speechd is also not in OE... Nov 14 10:22:18 I'm a beginner: two CVS is too much for me!!! Nov 14 10:22:36 I fear I could mess up the dependencies... Nov 14 10:22:57 guys, I'm talking about NAVIT, th GPS software open source Nov 14 10:23:19 http://navit.cvs.sourceforge.net/navit/navit/ Nov 14 10:23:43 some help for creating a working .bb? Nov 14 10:23:53 TIA Nov 14 10:34:54 XorA: I've sent a mail to the angstrom lists showing how a user can work around the RCONFLICTS thingy and a terse description to oe-devel about hte disapprove Nov 14 10:35:49 What is the strategy for version specific dependencies? Is it ok to add a DEPENDS="autoconf-2.13" and is there any way to specific a dependency for let's say libiconv version 1.9 or older (I guess libiconv-1.9+ is not supported). Nov 14 10:43:14 good morning Nov 14 10:43:37 hrw: fixed - thanks for the hint Nov 14 10:44:49 florian: common bug Nov 14 10:45:25 hrw: yeah, only a few people notice this... well the other link there is better ;) Nov 14 11:01:02 03mickeyl 07org.oe.dev * r4691fac1... 10/ (3 files in 2 dirs): navit: add 0.0.2 Nov 14 11:02:10 Danke mickeyl Nov 14 11:02:46 cvs Makefile is 0.0.3 Nov 14 11:07:54 mickeyl, would you pls look for speechd -> flite ? Nov 14 11:08:53 I fear it could be too heavy for let say a Zaurus Nov 14 11:14:41 mickey, sorry to insist... Nov 14 11:15:10 but I don't think the navit compile-fix patch apply on ver. 0.0.2 Nov 14 11:15:43 look at src/fib-1.1/Makefile.am Nov 14 11:16:23 prevoiusly was a libtool .la Nov 14 11:19:07 hmm Nov 14 11:19:09 this needs more work Nov 14 11:19:17 it calls binaries in the build process Nov 14 11:19:26 lt-osm2navit: cannot execute binary file Nov 14 11:19:29 i don't have time for that now Nov 14 11:19:48 I see, thanks for the moment Nov 14 11:20:07 it's probably just a small Makefile fi Nov 14 11:20:07 x Nov 14 11:20:12 i'll try to look at it later Nov 14 11:20:28 they changed in cvs after 0.0.2 Nov 14 11:20:49 just a couple of days ago... Nov 14 11:21:11 BTW the programs seems promising! Nov 14 11:31:07 mickeyl, ping Nov 14 11:32:34 Crofton|home: pong Nov 14 11:32:56 do you think we will get two tables? Nov 14 11:33:07 i have no idea, really Nov 14 11:33:15 I'm trying to talk sakoman_ into coming to FOSDEM with some gumstix stuff to show off :) Nov 14 11:33:28 hehe, yeah, we need room for that Nov 14 11:33:55 Maybe we should see how interested he/they are and have him request a table next to us ? Nov 14 11:34:09 ya, might be a good plan b Nov 14 11:34:26 gumstix (the company) does not usually do trade shows Nov 14 11:34:36 but, this is a low overhead trade show Nov 14 11:34:37 well Nov 14 11:34:47 except for usd being worthless :) Nov 14 11:34:56 hehe Nov 14 11:36:13 florian: ping Nov 14 11:36:52 morning Nov 14 11:44:50 mickey|bbiab: pong Nov 14 11:45:04 florian: you got mail Nov 14 11:46:42 mickey|bbiab: ok, you have an answer :) Nov 14 11:49:30 mickey|bbiab: isn't is conceptually confusing to have a window with buttons *and* panel applets that both do the same thing? (http://scap.linuxtogo.org/files/eb838b77fab8df226b5c5cd7d48ed171.png) Nov 14 11:53:42 drop some functionality from panel applets and power menu will have sense Nov 14 11:55:19 * koen discovers one can use gladeXML files with etk Nov 14 11:56:48 koen: absolutely. ten points for your usability class :) seriously though, the PM functions will be ripped out of the applets. Applets will only be there once the respective component has power to configure the said component. Nov 14 11:57:38 having icons in the neod menu would help as well Nov 14 11:57:47 fluff Nov 14 11:57:49 :D Nov 14 11:57:53 BT applet should also show what is going on Nov 14 11:58:10 mickeyl: with icons I could also use it upside down :) Nov 14 11:58:11 blink on transmission, rotate on searching etc Nov 14 11:58:24 or to put it in old-people language: without reading glasses ;) Nov 14 11:58:30 mickeyl: look at your SE phone during using BT Nov 14 11:58:34 ya, a lot of things can be done Nov 14 11:58:40 just by whom.... Nov 14 11:59:06 mistic openmoko software team? Nov 14 12:05:22 hi all ! Nov 14 12:05:37 hi steliosk Nov 14 12:05:59 * steliosk is the proud father of 3.250Kgr baby boy :) Nov 14 12:06:39 steliosk: cool! congratulations! Nov 14 12:06:48 steliosk: congrats Nov 14 12:07:15 congrats!! Nov 14 12:07:34 congrats Nov 14 12:07:45 mickeyl: Do you know what the plan for the railway strikes is? :-} Nov 14 12:08:01 any photos yet? Nov 14 12:08:15 florian: thursday to saturday right now Nov 14 12:08:28 florian: that's why i'm hoping for next week being ok Nov 14 12:08:53 Thanks guys ! Nov 14 12:08:55 steliosk: Congratulations, also to the proud mama. Nov 14 12:09:01 mickeyl: phew... I heard something about unlimited strikes yesterday. But yes, sounds like you could be lucky... Nov 14 12:09:30 unlimited are in france Nov 14 12:09:46 cool France is now closed :) Nov 14 12:10:51 steliosk: congrats! Nov 14 12:11:14 steliosk: GRATULATIONS! Nov 14 12:11:59 hrw, take notes :) Nov 14 12:12:19 Crofton|home: ;) Nov 14 12:12:59 steliosk: how will you name him? Nov 14 12:13:07 Crofton|home: hehe Nov 14 12:13:21 "Angstrom" Nov 14 12:13:54 hehe Nov 14 12:14:27 steliosk: congrats! Nov 14 12:15:43 hrw : We are going to name him after my father , Spiros, but my wife is also named after the same Saint so they will have the same name day :) Nov 14 12:16:10 * chouimat gets ready for the 1h45 commute :( Nov 14 12:16:14 Angstrom would have been *very* tough to sell Nov 14 12:16:18 :) Nov 14 12:16:18 :) Nov 14 12:16:44 maybe a nickname until he gets larger :) Nov 14 12:17:43 Crofton|home: "stoprunninginthehouse" is a nice one :) Nov 14 12:18:29 Crofton|home : The best i can do is bring him at the next OEDEM in order to show that OE is "child's play" now Nov 14 12:19:27 hrw : When the time comes for yours make sure you get to be in the delivery room. You won't forget the experience for the rest of your life Nov 14 12:19:49 steliosk: I will Nov 14 12:20:02 steliosk: OE child and oedem? Nov 14 12:22:05 hrw : yeah he has to be immersed into Open Source early, so it will not get snatched by the "dark side" :) Nov 14 12:23:45 steliosk: be careful it might backfire on you :) Nov 14 12:25:03 heh only 8500 emails to download... Nov 14 12:26:14 http://www.mcrd.ath.cx/kitten.jpg Nov 14 12:48:28 Hmm... a group of OE kids would be cool for next OEDEM :-) Nov 14 12:49:06 florian: OEDCEM? Nov 14 12:49:34 OpenEmbedded Developer's Children European Meeting Nov 14 12:52:06 hehe Nov 14 12:53:11 but children usually also means mothers so it will be families meeting even Nov 14 12:55:27 yes indeed... :) Nov 14 12:57:03 florian: how old children you have (if at all) Nov 14 13:00:01 hrw: Currently one, 21 months... Nov 14 13:00:42 Ania says that with such 'old' child you can go even without wife/mother for few days Nov 14 13:02:56 re Nov 14 13:03:16 hrw: Yes, with some at least... works pretty well with ours. Nov 14 13:03:21 hi Jin^eLD Nov 14 13:03:28 hey florian Nov 14 13:09:08 hrw,: OEDCEM is what happens after too many beers ;) Nov 14 13:22:42 lrg: you and your wifes.... ;D Nov 14 13:23:10 hrw: ex wives now :) Nov 14 13:23:30 btw - wifes or wives? how is proper in en-gb? Nov 14 13:23:37 ~spell wifes Nov 14 13:23:39 possible spellings for wifes: wife's wives waifs wides wafers waif's waives whiffs wife fifes wiles wines wipes wires wises woofers Waves waves woofs Wise fies wise wafer's waivers whiff's woes woof's Wes Wis fiefs ifs liefs Wise's fife's life's wile's wine's wire's wives's woofer's vies wees wive Wolfe's Wis's waiver's woe's Fe's Lief's fief's Nov 14 13:23:49 wafers:D Nov 14 13:24:10 hrw: English was never my best subject at school ! Nov 14 13:24:30 wives Nov 14 13:25:01 lrg: neither mine Nov 14 13:25:41 Certainly not mine either. But my French was even worse. Nov 14 13:26:53 but I had only few years of english and it was years ago Nov 14 13:27:17 argh... Nov 14 13:27:27 ~curse neo and its lack of power Nov 14 13:27:39 May you be reincarnated as a Windows XP administrator, neo and its lack of power ! Nov 14 13:27:44 I need to dig neo batteries charger Nov 14 13:28:09 it charge them in less then hour (neo needs ~night) Nov 14 13:29:28 If the om team would accept the patches that were written, then the neo would no longer a) discharge the battery when shut down, b) charge at a paltry 50ma instead of the correct 100ma :D Nov 14 13:31:11 there is one good thing in nokia - their crap^Wdevices are so popular that 3rdparty accessories beats their own by light years Nov 14 13:31:44 * mwester is hoping the N800 isn't crap, because he's bidding on one on eBay. Nov 14 13:32:08 mwester: how much for it? Nov 14 13:32:42 They seem to be selling for about USD 215 + about USD 15 for shipping. (average prices) Nov 14 13:33:07 * mwester is bidding lower, of course :) Nov 14 13:33:34 Given the USD, that translates to about 2 Euro then? Nov 14 13:33:45 *Sigh* Nov 14 13:36:58 hrw: I've another patch for tosa :) Maybe instead of bugging I should apply for commit access? Nov 14 13:41:00 mwester: kernel patches? Nov 14 13:41:44 lumag: bug me Nov 14 13:41:50 hrw: ok Nov 14 13:42:15 mwester: I can apply kernel patches to OpenMoko kernel Nov 14 13:47:19 linux-openmoko kernel is funny Nov 14 13:47:42 vanilla tarball + openmoko patchset + OE patches + vanilla updates Nov 14 13:52:33 XorA: I know you and mickeyl have been trying on getting the patch backlog addressed; I'm just grousing and complaining that it seems that for every hour of technical work with OM, there's 4 hours of work to get a patch accepted. :( Nov 14 13:52:43 The 50ma/100ma is a uboot patch. Nov 14 13:53:15 The other patches are under discussion, and I'm pretty confident that they'll get into the next snapshot if that's still on track :) Nov 14 13:54:40 mwester: I get lost in a see of crappy bugzilla mails and lose track Nov 14 13:54:48 50/100 or 500/100? Nov 14 13:56:16 hrw: unless instructed otherwise (by observing a "smart" device on the USB), the neo is supposed to charge at 100ma. Apparently there is a constant that is set incorrectly in the PM chip by uboot, and it only charges the battery at 50ma. Nov 14 13:56:25 a Nov 14 13:56:26 When booted, it should charge at the full 500ma Nov 14 13:56:26 h Nov 14 13:56:28 h Nov 14 13:56:53 And the gsm chip draws about 40mA, according to SpeedEvil's testing Nov 14 13:58:05 The patch to shut down the GSM modem when the neo is properly powered off is one that I hope will make it into the next snapshot -- that will help somewhat, because it is unrealistic to think that people should reflash u-boot. Too dangerous,IMO. Nov 14 13:59:16 mwester: yeah I saw that gsm patch, I think that should go in Nov 14 14:01:08 hrw: http://bugs.openembedded.org/show_bug.cgi?id=3295 Nov 14 14:02:48 lumag: does it allows to disable single ports too? Nov 14 14:02:54 yup Nov 14 14:03:04 lumag: like 'disable wifi, enable port'? Nov 14 14:03:06 coool Nov 14 14:03:18 it provides two sysfs attributes: one per port. 0 is external, 1 is wifi Nov 14 14:04:02 lumag: bump PR when you add new stuff Nov 14 14:04:18 hrw: ok Nov 14 14:04:36 wooo, in only a couple of days. tosa gets fixed :-) Nov 14 14:05:23 and improved Nov 14 14:05:50 pushing now Nov 14 14:06:05 so now just the graphics corruption problem? Nov 14 14:06:23 XorA: there is still one regression: sometimes I'm getting an oops during resume, while ipaq-sleep is reading /proc/interrups Nov 14 14:06:44 lumag: and wifi still needs to be disabled on suspend? Nov 14 14:06:49 lumag: its not a regression when you made 2.6.23 work at all to start with :-) Nov 14 14:07:17 hrw: didn't tried. Nov 14 14:08:24 XorA: it wasn't too hard. Nov 14 14:09:12 * mwester wonders when exactly firefox became so much like IE. Nov 14 14:11:21 hrw: btw, Rolf contacted me about serial cable and since then I didn't receive any news from him. Nov 14 14:13:06 you sent him address? Nov 14 14:13:09 yes Nov 14 14:13:20 he can be very busy and hard to catch sometimes Nov 14 14:14:14 ok Nov 14 14:14:45 lumag: you have usb host cable? Nov 14 14:15:23 yeah. But not the original one. Nov 14 14:17:25 lumag: want another tosa for testing? Nov 14 14:18:23 hrw: probably no. I hope mine isn't yet _that_ broken :) Nov 14 14:18:51 ok, just ask because we have 2 spare ones Nov 14 14:19:04 yup. I saw on ML. Nov 14 14:19:23 we will need to pester mickeyl to test with sleeve Nov 14 14:22:13 lumag: how good/bad does angstrom iamges are now on tosa? Nov 14 14:23:07 hrw: I'm not flushing new images since June. just ipkg upgrade'ing. And it seems good. Nov 14 14:23:21 gpe? Nov 14 14:23:53 Yeah. Probably I'll flush new image this weekend. I get too many JFFS2 errors :) Nov 14 14:24:17 lumag: do you know updater.sh internals? Nov 14 14:24:44 hrw: I can look into it. Nov 14 14:25:11 lumag: in bugtracker we have 'united' updater.sh - first version works on all devices except tosa Nov 14 14:25:20 second version is buggy Nov 14 14:26:38 03dbaryshkov 07org.oe.dev * r402ffc9d... 10/ (3 files in 3 dirs): linux-rp: Tosa usb ports management - close #3295 Nov 14 14:27:31 Should I fix first version? Nov 14 14:28:44 http://bugs.openembedded.org/show_bug.cgi?id=1526 is bug Nov 14 14:29:10 second version has one syntax bug - better fix it and work on it Nov 14 14:29:58 ok Nov 14 15:09:38 XorA: I have bitbake changes in poky undergoing testing which let you lock down multi SCM packages like uboot-openmoko btw (and fix SRCREV_FORMAT for these cases) Nov 14 15:10:16 RP: cool, I have finally received a build machine so I can actually build stuff sometime this year Nov 14 15:12:12 | ERROR: The following packages conflict with openmoko-session2: Nov 14 15:12:12 | matchbox-common Nov 14 15:12:14 ;( Nov 14 15:12:25 openmoko-devel-image with up-to-date OE Nov 14 15:13:32 so something in openmoko-devel-image still depends on matchbox-common Nov 14 15:15:04 matchbox-wm Nov 14 15:15:38 hmm, guys.. my collegaue and me are using the same OE version/repo, readline builds fine for me but fails for him (do_patch does not work out), he is on debian/unstable I'm on Fedora 6, we both use quilt that was built by OE - there should not be any difference? Nov 14 15:18:21 any idea why this is happening? we even have the same bitbake version Nov 14 15:19:28 I'm trying to compile uiClibc 0.9.29 with gcc 3.3.4 for arm with linux libc headers 2.6.20 and i get the following error Nov 14 15:19:32 | Assembler messages: Nov 14 15:19:33 Jin^eLD: Something different between the host systems I guess. Its unusual for do_patch to fail though... Nov 14 15:19:34 | option `-mno-fpu' is deprecated: use either -mfpu=softfpa or -mfpu=softvfp Nov 14 15:19:37 | libc/sysdeps/linux/arm/mmap64.S:80: Error: undefined symbol ENOSYS used as an immediate value Nov 14 15:19:42 anyone got an idea please? Nov 14 15:20:05 i didn't find the place where he added -mno-fpu Nov 14 15:20:13 RP: yes.. its failing with a quilt message, i.e. quilt is printing command line args, indicating that it was invoked incorrectly Nov 14 15:20:13 nik0n_: using OE? Nov 14 15:20:22 Jin^eLD: pastebin please Nov 14 15:20:36 one sec Nov 14 15:21:01 hrw: yes Nov 14 15:21:08 hrw: http://pastebin.ca/773440 Nov 14 15:21:20 nik0n_: TARGET_FPU="soft" Nov 14 15:21:39 hrw: that generates the -mno-fpu ? Nov 14 15:22:03 nik0n_: check that var Nov 14 15:22:07 what should i add instead? or should i doint use TARGET_FPU? Nov 14 15:22:16 Jin^eLD: and both have quilt in same ver? Nov 14 15:22:33 nik0n_: 1st check which value it has on your system Nov 14 15:22:39 hrw: let me recheck... but since we use the same OE repo and quilt was built by OE I would assume so.. one moment Nov 14 15:22:51 check, not assume Nov 14 15:23:10 Jin^eLD: It could be using the host system's quilt, not the one it built Nov 14 15:23:28 Jin^eLD: that would of course be a bug but it would help track down the problem Nov 14 15:23:44 mhm.. in OE/staging we both have 0.45 Nov 14 15:23:57 and we also both have a standalone 0.46 installed on our system Nov 14 15:24:31 Jin^eLD: rename the system's quilt and see if the error changes? Nov 14 15:24:53 RP: I suggested that to him and he tried, but error remained the same Nov 14 15:25:48 Jin^eLD: I guess start adding debug to patch.bbclass then Nov 14 15:26:04 * * OE Bug 3295 has been created by dbaryshkov(AT)gmail.com Nov 14 15:26:06 * * usb ports management Nov 14 15:26:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3295 Nov 14 15:26:08 mhm.. ok Nov 14 15:26:09 thanks Nov 14 15:28:16 * * OE Bug 3295 has been RESOLVED (FIXED) by Nov 14 15:28:18 * *  usb ports management Nov 14 15:28:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3295 Nov 14 15:29:16 * * OE Bug 3119 has been RESOLVED (FIXED) by antti.andreimann(AT)mail.ee Nov 14 15:29:18 * *  Fakeroot 1.7.1 not available from Debian repositories Nov 14 15:29:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3119 Nov 14 15:29:28 * * OE Bug 3296 has been created by jeremy.laine(AT)bolloretelecom.eu Nov 14 15:29:30 * * python-core is missing struct.py Nov 14 15:29:32 * * http://bugs.openembedded.org/show_bug.cgi?id=3296 Nov 14 15:33:02 the quilt invocation is the same for both of us, what is quiltrc? Nov 14 15:33:09 maybe there is something different Nov 14 15:33:46 heh, I do not have a quiltrc Nov 14 15:33:51 he has? Nov 14 15:33:59 checking.. Nov 14 15:34:05 he also has no quiltrc Nov 14 15:35:33 hrw: i uncommented TARGET_FPU="soft" but still have the option `-mno-fpu' is deprecated: use either -mfpu=softfpa or -mfpu=softvfp error Nov 14 15:35:47 it happens with gcc 3.3.4 only, not with gcc 4.1.2 Nov 14 15:36:11 nik0n_: why 3.3.4? Nov 14 15:36:26 hrw: need to compile a 2.4 series kernel Nov 14 15:36:41 they dont boot when compiled with gcc > 3.3 Nov 14 15:37:09 you see uncompressing linux ............ and then they stop Nov 14 15:37:16 when gcc > 3.3 Nov 14 15:37:42 03koen 07org.oe.dev * rd30004f9... 10/ (1 conf/distro/include/sane-srcrevs.inc): sane-srcrevs: add openmoko-messages2, bump some other things Nov 14 15:38:48 nik0n_: you can have separate kernel for userspace and other for kernel with OE Nov 14 15:39:21 you mean seperate gcc do you? Nov 14 15:39:31 yes Nov 14 15:39:42 so 4.1 for all and 3.3 for kernel Nov 14 15:39:56 gonna give this a shot, but its still annoying Nov 14 15:40:03 look at packages/linux/openzaurus-sa recipe Nov 14 15:40:14 or better opensimpad 2.4 Nov 14 15:40:23 is there an option for distr/local.conf to say: use gcc 3.34 for package xy and the default one for the rest? Nov 14 15:40:29 okay Nov 14 15:40:34 tia Nov 14 15:53:19 ok that is really weird now.. I have an account on his machine, I did test it manually Nov 14 15:53:24 quilt --version reports 0.45 in both cases Nov 14 15:53:30 same command line works for me but fails on his machine Nov 14 15:53:42 oh wait Nov 14 15:54:14 hm no, that was not it either.. hmm, so still this problem mhmm Nov 14 15:55:04 Jin^eLD, strace -e trace=file, then compare. gotta be something! Nov 14 15:56:16 ok let me try that.. Nov 14 15:59:52 HopsNBarley: that was a good idea Nov 14 16:00:05 staging/i686-linux/share/quilt/import seems to be missing Nov 14 16:00:29 (-; Nov 14 16:00:30 so I guess it's not readline's package fault, but quilt-native is somehow screwed Nov 14 16:01:41 03mickeyl 07org.oe.dev * r48d0aed8... 10/ (1 conf/distro/openmoko.conf): openmoko.conf: add neo1973 feed Nov 14 16:01:46 03mickeyl 07org.oe.dev * r3591fc91... 10/ (1 conf/distro/include/sane-srcdates.inc): sane-srcdates.inc: add python-emotion entry Nov 14 16:03:34 XorA: http://bp1.blogger.com/_2o81e3u4ZFU/RzsVygsOWVI/AAAAAAAAAGM/f7BHYjELhBU/s1600-h/telepathy.gif Nov 14 16:06:19 koen: nive Nov 14 16:06:27 that on n810 would rock Nov 14 16:11:00 koen: http://www.engadget.com/2007/11/14/nokia-n810-firmware-leaks-runs-on-n800/ Nov 14 16:20:20 hrw: yeah, you can get the MAC from dmesg :) Nov 14 16:21:44 koen: ? Nov 14 16:22:03 you need an n810 mac adress to download it Nov 14 16:22:20 I know Nov 14 16:22:39 but I do not have n8xx Nov 14 16:24:03 hrw: but how many n810 have oops and people put oopses in bugzilla :-) Nov 14 16:24:30 * mwester likes the way XorA thinks. Nov 14 16:24:37 ;D Nov 14 16:26:00 like how we found out about neo1973 :-D Nov 14 16:26:51 pffft, nokia fanboy ;) Nov 14 16:27:15 * XorA did a trick like that for idiot customers who wouldnt give java applets to us to debug, they would always send a log which had the whole applet in hex, quick bit of perl, and suddenly I have applet back Nov 14 16:34:29 * XorA wonders when n810 goes on sale Nov 14 16:40:46 * hrw do not Nov 14 16:40:49 buy all! Nov 14 16:41:04 lumag|gone: not after 770 Nov 14 16:41:57 :) Nov 14 16:43:56 I have 770 and its crap Nov 14 16:44:23 atleast I do not plan to spend any cash to have n810 Nov 14 16:44:39 * lumag|gone wanted to look on n810. Nov 14 16:44:51 hrw: hehe, in germany we say "Einem geschenkten Gaul schaut man nicht in Maul." Be glad that someone gave the 770 to you Nov 14 16:44:52 I don't like gadgets w/o qwerty :) Nov 14 16:45:04 so I take it the n8xx is still rather closed Nov 14 16:45:04 lumag|gone: who will look at tosa then? Nov 14 16:45:13 rofl :) Nov 14 16:46:19 zecke: in Poland we have similiar one Nov 14 16:46:55 'when you get horse, do not check his teeths' (in very free translation) Nov 14 16:47:06 right Nov 14 16:48:18 Anyone know of something more up to date than http://www.openembedded.org/wiki/PackagedStaging ? Nov 14 16:48:33 hrw: nearly same here: don't examine the teeth of the presented horse :) Nov 14 16:49:09 * lumag|gone is now really gone Nov 14 16:49:26 bye! Nov 14 16:49:58 english: "Don't look a gift horse in the mouth" Nov 14 16:51:20 did you note the *promised* battery life of n810? Nov 14 16:51:33 5 days "on-line" Nov 14 16:51:36 ??? Nov 14 16:51:54 14 days stand-by..reasonable Nov 14 16:52:13 music-player 10 hours...bad Nov 14 16:52:36 full-use 4 hours..nothing new under the sun! Nov 14 17:31:19 oh this is a sweet little git tool Nov 14 17:32:44 what tool? Nov 14 17:33:44 What is the name of the external sdk? Nov 14 17:34:17 http://people.linux-vserver.org/~doener/code-follow.py Nov 14 17:34:27 it basically lets you loko at a log of commits touching a certain block of code Nov 14 17:34:31 it repeatedly annotates, working its way back Nov 14 17:34:40 python ./code-follow.py drivers/block/elevator.c '/^void __elv_add_request/,/^}/' | git shortlog Nov 14 17:34:43 <3 Nov 14 17:36:16 doener? I get hungry Nov 14 17:40:07 quick bug report: package xbitmaps depends on itself ( DEPENDS += libxmu xbitmaps ). Nov 14 17:40:46 I'm getting weird mtn warnings from alsa stuff? Nov 14 17:41:12 circular dependancies and shit Nov 14 17:48:39 So does libtheora :/ Nov 14 17:51:05 Is it just me, or is a lot of stuff wrong in OE? Nov 14 17:51:22 Say, wifi-radar depends upon python-pygtk2, which doesn't exist Nov 14 17:51:55 python-pygtk_2.something exists, which is obviously what wifi-radar needs, but wifi-radar doesn't know that, does it? Nov 14 17:51:58 :) Nov 14 17:53:06 B_Lizzard: it's a bug, so why not file one? Nov 14 17:53:13 Uh huh Nov 14 17:53:29 I get a whole load of those when trying to build "world" Nov 14 17:53:56 They're just *too* obvious to be missed, that's why I'm a bit "duh". Nov 14 17:54:08 I'll post the bitbake log Nov 14 17:54:08 B_Lizzard: well, somebody has to be poked to fix those issues. Nov 14 17:54:16 I understand. Nov 14 17:54:33 in the pygtk2 case, it got renamed to pygtk in the python2.5 transition. Nov 14 17:54:46 so wifi-radar is an oversight, obviously. Nov 14 17:55:40 Uh huh Nov 14 17:57:20 B_Lizzard: we need people to do world build and report breakage Nov 14 17:58:42 Would posting a pastebin so you can see the errors help for now? Nov 14 17:58:47 hehe Nov 14 17:58:58 If they're worth the bugreport Nov 14 17:59:29 B_Lizzard: A mail to the ml might make sense as well (and doing that regulary) Nov 14 17:59:42 regularly, that is a good point :) Nov 14 17:59:53 :) Nov 14 17:59:57 Noted. Nov 14 18:00:06 Sorry if I'm being annoying or something :) Nov 14 18:00:11 Here it is: http://pastebin.ca/773631 Nov 14 18:00:34 git and me do not match Nov 14 18:02:29 sorry guys, before launching "world" I'd like a clean bitbake "console-image" Nov 14 18:02:46 dunno you...but have nasty logs Nov 14 18:02:54 see bug 3276 Nov 14 18:03:15 perhaps my toolchain is misconfigured, but was ok till end october Nov 14 18:03:28 ok enough to compile Nov 14 18:13:56 03pH5 07org.oe.dev * r3bbd8cd0... 10/ (1 packages/xorg-data/xbitmaps_1.0.1.bb): xbitmaps: remove self-dependency Nov 14 18:14:00 03pH5 07org.oe.dev * re4cbb18f... 10/ (1 packages/wifi-radar/wifi-radar_1.9.6.bb): wifi-radar: python-pygtk2 got renamed to python-pygtk Nov 14 18:14:05 03koen 07org.oe.dev * r22ffaff4... 10/ (2 files in 2 dirs): desktop-file-utils: merge from poky Nov 14 18:14:09 03koen 07org.oe.dev * r94fb8804... 10/ (1 classes/insane.bbclass): insane.bbclass: merge dot desktop check from poky Nov 14 18:17:22 http://blog.haerwu.biz/2007/11/14/git-and-i-do-not-match/ Nov 14 18:35:24 anyone know much about pulse audio? Nov 14 18:36:17 bump bug #2157 (qemu vs. gcc 4.x) Nov 14 18:36:43 what's the "state of the art"? Nov 14 18:37:06 can it really be done oneday ? Nov 14 18:38:16 today every distro ships gcc 4 Nov 14 18:38:27 antvr: tell that to qemu developers ok? Nov 14 18:38:36 touched Nov 14 18:38:57 it is not our fault that qemu req gcc3 Nov 14 18:39:30 (Poky req gcc 3.x) ? Today ? Nov 14 18:39:37 and version which we have in repo finally works properly on x86 and x86-64 which was not true before Nov 14 18:39:51 antvr: apt-get install gcc-3.4 is so hard? Nov 14 18:40:10 antvr: my system use 4.2 by default and have 3.4 just for qemu Nov 14 18:40:21 I'm building on Core2-Dual and I'd be happy to compile overnight for debug Nov 14 18:40:34 but this gcc3 thing makes me crazy Nov 14 18:40:41 only for the locales! Nov 14 18:40:52 antvr: then disable generation of locales Nov 14 18:40:55 or FIX glibc Nov 14 18:41:06 very easy task :-) Nov 14 18:41:23 I'm not a developer Nov 14 18:41:23 if you say so;) Nov 14 18:41:38 antvr: then do us a favour: apt-get install gcc-3.4 Nov 14 18:41:54 since 2007 I'm on openususe 10.2 Nov 14 18:42:12 and it just works Nov 14 18:42:34 btw I have Leopard 10.4 on the other partition Nov 14 18:42:36 so what is a problem? Nov 14 18:42:50 but don't want to mess up my wife's photo works Nov 14 18:43:17 hrw: could you double-check if NOTE: Task failed: Error: /home/gordon/oe/org.openembedded.dev/packages/linux/linux-rp-2.6.23/./tmio-ohci-r9.patch not found or just the .dev is slow Nov 14 18:43:18 Opensuse is not like debian Nov 14 18:43:54 I compiled by myself gcc3.4.6 but failed miserably with the linkings Nov 14 18:44:04 xjqian: ops Nov 14 18:45:52 each end-user distro has a static "baselayout" Nov 14 18:46:11 probably I've chosen the wrong distro Nov 14 18:46:32 antvr: gcc can be anywhere - let binary name gcc-3.4 and it should work Nov 14 18:46:43 xjqian: pushing fix Nov 14 18:47:57 at work I have Gentoo and it's very easy with gcc-config Nov 14 18:48:09 dunno if Suse ever has that Nov 14 18:49:09 thanks hrw, I'll try later Nov 14 18:49:11 bye Nov 14 18:49:21 god! give me machine with opensuse to tests.. Nov 14 18:49:33 and fedora, rhel5 and other !debian Nov 14 18:49:49 4xquad will be enough ;D Nov 14 19:01:34 03hrw 07org.oe.dev * r5bba2762... 10/ (3 files in 2 dirs): linux-rp: fix tosa patch name (my fault) Nov 14 19:07:41 03hrw 07org.oe.dev * rf4cbdb09... 10/ (1 conf/machine/tosa.conf): tosa: stop preffering 2.6.17 kernel - 2.6.23 is up-to-date and have improvements Nov 14 19:46:08 03koen 07org.oe.dev * r9f2c0aa6... 10/ (12 files in 2 dirs): matchbox-wm: stop requiring mb-common Nov 14 20:44:04 * * OE Bug 3297 has been created by thebohemian(AT)gmx.net Nov 14 20:44:06 * * pulseaudio: build hal-detect module Nov 14 20:44:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3297 Nov 14 21:05:03 RP: bitbake trunk doesn't fail on rootfs generation errors, currently. I repeatedly ended up with images not containing essential packages without bitbake complaining. Nov 14 21:39:47 zecke: thinking about hacking on svn-fast-export.c.. any particular headaches you ran into with fast-import that is hould look out for? Nov 14 21:44:29 kergoth: trust the automatic rename detection of git (it wasn't available in earlier versions) Nov 14 21:44:38 kergoth: make sure to save the mapping in a goody way Nov 14 21:45:03 kergoth: for some 'nothing changed' revs in mtn git didn't create a revision so I ended up with invalid marks Nov 14 21:45:14 kergoth: so basicly common sense and the rest is monotone specific Nov 14 21:45:44 okay, cool Nov 14 21:45:59 nwo to figure out a more sane way than git-svn for handling tags Nov 14 21:48:05 hrw: you arent "adding files already known to scm", you're updating the index. git add happens tow ork because its basically git-update-index --add, but just a plain git-update-index works just as well. read up about the index if you want to understand it Nov 14 21:49:04 kergoth: then let git devs fix messages because they tell 'use git add' Nov 14 21:50:25 and git was never intended to be a super easy porcelain, thats why cogito used to exist, just git's caught up with it. someone could easily do a porcelain/wrapper around pull to call out mergetool Nov 14 21:51:22 it needed sometime I recognized I had to call git-mergetool and I have no idea how to get a file back into conflict mode... e.g. to undo git-mergetool Nov 14 21:51:24 i dunno why they say to git add, probably cause update-index is probably thought of as "plumping" rather than porcelain.. kinda weird Nov 14 21:51:48 I love git-rebase and git-add -i Nov 14 21:52:13 but if you see the update-index man page: "Each file mentioned is updated into the index and any unmerged or needs updating state is cleared." Nov 14 21:52:22 zecke: god no kidding, i use add -i so much to commit only certain hunks Nov 14 21:53:02 kergoth: now only splitting hunks up further... ah not that printk but the rest... Nov 14 21:54:25 zecke: looks like hg and svn fast export are in the fast-export git repo, maybe add mtn there too to keep em in one place. /shrug Nov 14 21:55:06 kergoth: it is not yet done ;) Nov 14 21:55:17 neither is svn. not sure about hg :) Nov 14 21:55:21 svn cant even handle tags or branches atm Nov 14 21:55:22 hehe Nov 14 21:55:39 kergoth: I was too lazy to compare the OE tree... I assume someone else will do it Nov 14 21:55:57 i use git log -S, git log --grep, and git grep so much. wonder if thers a vim git grep plugin yet Nov 14 21:56:28 kergoth: oh, and don't wonder if you can't git-push or git-clone your tree... git-repack -a -f -d (the -f is important) Nov 14 21:56:46 kergoth: and then some marks where gone :) Nov 14 21:57:32 ah Nov 14 22:01:34 03hrw 07org.oe.dev * r9187119a... 10/ (4 files in 3 dirs): ettercap-ng: added 0.7.3 - close #3298 Nov 14 22:01:43 03hrw 07org.oe.dev * r0e8df4b2... 10/ (4 files in 3 dirs): ettercap: drop 0.6 Nov 14 22:03:55 ~hail hrw Nov 14 22:03:56 * ibot bows down to hrw and chants, "I'M NOT WORTHY!!" Nov 14 22:13:04 * * OE Bug 3298 has been created by lukas.gorris(AT)gmx.de Nov 14 22:13:06 * * ettercap-ng - new verison of product ettercap Nov 14 22:13:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3298 Nov 14 22:32:21 hrw|tv, thank you very much for the add; gnite Nov 14 22:39:52 good night all Nov 14 22:58:04 * * OE Bug 3298 has been RESOLVED (FIXED) by Nov 14 22:58:06 * *  ettercap-ng - new verison of product ettercap Nov 14 22:58:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3298 Nov 14 22:58:16 * * OE Bug 3299 has been created by alessandro.iurlano(AT)gmail.com Nov 14 22:58:18 * * cvs recipe for navit Nov 14 22:58:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3299 Nov 14 22:58:58 how do i specify cflags for an autotools program? do i add do_configure_prepend() { export CFLAGS=...:$CFLAGS } ? Nov 14 22:59:13 ~lart Nov 14 22:59:18 er, ignore the colon Nov 14 22:59:30 CFLAGS="... $CFLAGS" Nov 14 22:59:46 EXTRA_CFLAGS += "foo" iirc Nov 14 22:59:59 kergoth, thanks Nov 14 23:02:53 Hey koen, you have some really nice pictures on your blog. Nov 14 23:03:00 Very nice takes there. Nov 14 23:03:02 :) Nov 14 23:44:05 * * OE Bug 3300 has been created by utx(AT)penguin.cz Nov 14 23:44:07 * * fix build of xserver-kdrive_git Nov 14 23:44:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3300 Nov 14 23:44:32 did we give utx commit access? Nov 14 23:46:31 anybody gotten ghc to work? Nov 14 23:46:36 (for haskell programs) Nov 15 00:00:47 kergoth: buildsystem master? **** ENDING LOGGING AT Thu Nov 15 02:59:57 2007