**** BEGIN LOGGING AT Sun Sep 11 02:59:56 2005 Sep 11 03:06:06 hi Sep 11 03:09:52 03rwhitby 07org.oe.dev * r23e9c03c... 10/packages/ftpd-topfield/ftpd-topfield_0.6.0.bb: ftpd-topfield: updated to 0.6.0 Sep 11 03:09:57 03rwhitby 07org.oe.dev * r5565d716... 10/ (5 files in 2 dirs): openslug-packages, ucslugc-packages, openslug-native: added 'file' Sep 11 04:00:16 03rwhitby 07org.oe.dev * rb0b409a4... 10/packages/ftpd-topfield/ftpd-topfield_0.6.1.bb: ftpd-topfield: updated to 0.6.1 Sep 11 04:00:21 03koen 07org.oe.dev * rf87c918c... 10/packages/libschedule/libschedule_0.15.bb: packages/libschedule/libschedule_0.15.bb: set tarball suffix to bz2 Sep 11 04:20:54 NOTE: Fetch cvs://anonymous@anoncvs.gnome.org/cvs/gnome;module=libmimedir Sep 11 04:20:54 /usr/bin/env: cvs: No such file or directory Sep 11 04:21:26 looks like you haven't got cvs installed Sep 11 04:21:28 i can't finde "libmimedir" in the cvs - where it's gone? Sep 11 04:22:13 NOTE: package libmimedir-0.0+cvs-20050911-r0: task do_fetch: completed Sep 11 04:22:21 hmm... Sep 11 04:22:28 i'll check my environment again Sep 11 04:29:56 oh yes. my cvs install was corrupt Sep 11 04:30:02 thanks for pointing this out Sep 11 06:22:49 hi Sep 11 06:23:07 i've got a problem while doing a "bitbake opie-image" Sep 11 06:23:27 it seems that now it uses gcc 4 Sep 11 06:24:00 and gcc-4 fails in do_compile Sep 11 06:24:15 | checking for stdint.h... (cached) yes Sep 11 06:24:15 | head: `-1' option is obsolete; use `-n 1' since this will be removed in the future Sep 11 06:24:15 | checking for main in -lm... configure: error: Link tests are not allowed after GCC_NO_EXECUTABLES. Sep 11 06:24:15 | make: *** [configure-target-libstdc++-v3] Erreur 1 Sep 11 06:24:15 | FATAL: oe_runmake failed Sep 11 06:58:13 Good evening Sep 11 07:16:20 Titeuf: set PREFERRED_VERSION_gcc-cross = "3.4.4" in local.conf Sep 11 07:16:43 Titeuf: thats been broken for a few days, not sure why/who (could prolly find it in the logs if you wanted to open/assign a bug about it) Sep 11 07:16:55 or use a DISTRO which sets it Sep 11 07:17:00 yah Sep 11 07:17:16 but if he was doing that he wouldn't be asking ;) Sep 11 07:18:06 * Disconnect thinks "change distributions and/or hw platforms" is a bit of a rough answer for "someone seems to have broken something" Sep 11 07:18:20 it isn't broken Sep 11 07:18:36 and it wasn't a question either Sep 11 07:19:24 if you are 1337 enough to not define a DISTRO you should be 1337 enough to fix the problems stemming from that Sep 11 07:19:43 * Disconnect defined openzaurus. its broken. Sep 11 07:20:10 Try openzaurus-3.5.4, it works Sep 11 07:20:14 and if you are 1337 enough to use an unversioned distro you should be 1337 enough to fix the problems stemming from that Sep 11 07:21:31 * Disconnect points out that he DID fix the problem. (and the - sketchy - docs say nothing about versioned distros to begin with, so.. if you are following TFM you'll get F'd..) Sep 11 07:28:45 morning alll Sep 11 07:29:08 hi RP ... Sep 11 07:29:55 morning Sep 11 07:31:10 03koen 07org.oe.dev * r38885fa2... 10/conf/local.conf.sample: Sep 11 07:31:10 conf/local.conf.sample: add a note on DISTRO and INHIBIT_PACKAGE_STRIP Sep 11 07:31:10 local.conf.sample contained a wrong default which was masked by Sep 11 07:31:10 http://bugs.openembedded.org/show_bug.cgi?id=319 Sep 11 07:47:45 03koen 07org.oe.dev * rae5e978c... 10/packages/gpe-mini-browser/gpe-mini-browser_cvs.bb: packages/gpe-mini-browser/gpe-mini-browser_cvs.bb: fix typo in PV Sep 11 08:02:02 morning kergoth Sep 11 08:13:55 *yawn* good morning Sep 11 08:15:21 morning mickeyl Sep 11 08:16:47 <_alwin_> hi Sep 11 08:22:57 hey Sep 11 08:37:00 hello Sep 11 08:40:08 can someone tell me how bitbake finds the version of something to build when there are more versions in a package directory? like with glib, gtk+-2.8.3 wants glib>2.7.1 to be built, but glib2-0_2.6.4 got built.. Sep 11 08:42:53 I can probably just set preferred_version_glib-2.0 in local.conf.. but how does bitbake know to build glib2.0_2.6.4 when I don't set any version variable? Sep 11 08:57:55 Delta_: with familiar and openzaurus http://monotone.vanille.de/viewmtn/getfile.py?id=d93d5d5cd929176e24bad5bc3e5c5a6548e11eb6&path=conf/distro/preferred-gpe-versions-2.7.inc manages gtk stuff Sep 11 08:59:05 <[cc]smart> RP: is bug 41 / reducing builtin IO schedulers still sth. in the pot for planned changes ? Sep 11 09:05:26 [cc]smart: yes, I just keep forgetting to sort it out Sep 11 09:19:21 hi all Sep 11 09:19:41 morning all Sep 11 09:19:48 (er.. morning most, rehi some ;) ..) Sep 11 09:20:25 anyone know what this error mean when I "bitbake hostap-modules" : make[2]: arm-linux-gcc-2.95: Command not found Sep 11 09:21:35 but the arm toolchain path is already exported: /usr/local/arm/2.95.3/bin:/usr/kerberos/bin:/usr/local/bin:/bin:/usr/bin:/usr/X11R6/bin:/home/ztester/bin Sep 11 09:22:33 koen, I see.. I added the preferred-gpe-versions-2.7.inc to my distro config.. that should do.. Sep 11 09:24:35 abm_y4k: touch conf/local.conf Sep 11 09:24:51 (you renamed the toolchain binaries, did you?) Sep 11 09:25:12 hi mickeyl Sep 11 09:25:15 I already created the conf/local.conf Sep 11 09:25:22 hi pb_, how are things Sep 11 09:26:32 well, here in Cambridge it is cold, dark and wet. but, other than that, things are pretty good. Sep 11 09:26:40 mickeyl: yea, I cp /usr/local/arm/2.95.3/bin/arm-linux-gcc /usr/local/arm/2.95.3/bin/arm-linux-gcc-2.95 instead of mv /usr/local/arm/2.95.3/bin/arm-linux-gcc /usr/local/arm/2.95.3/bin/arm-linux-gcc-2.9 Sep 11 09:27:55 cold, dark and wet sounds like here :) Sep 11 09:28:17 pb_: same here Sep 11 09:28:26 mickeyl: I followed the wiki : http://oe.handhelds.org/cgi-bin/moin.cgi/ZaurusKernels but just use cp instead of mv Sep 11 09:28:37 will this cause te problem? Sep 11 09:28:38 hi mickeyl Sep 11 09:28:50 abm_y4k: good. you shouldn't use cp, but anyway. then just touch conf/local.conf and let it reparse and let it pick up your new PATH Sep 11 09:28:51 hi RP Sep 11 09:29:08 i guess 'which arm-linux-gcc-2.95' gives a valid answer? Sep 11 09:30:12 I get this : /usr/local/arm/2.95.3/bin/arm-linux-gcc-2.95 Sep 11 09:31:09 so it should be fine right? but why bitbake report this file doesn't exists? Sep 11 09:36:10 i guess you already ran bitbake before adjusting PATH Sep 11 09:36:28 then while it constructed the metadata cache it picked up your PATH without the new entry Sep 11 09:36:42 (note that bitbake modifies your PATH while it's running) Sep 11 09:36:56 touch'ing conf/local.conf invalidates the cache and lets it reparse Sep 11 09:37:06 (this is why I just said touch conf/local.conf 1h ago :) Sep 11 09:44:38 mickeyl: We had what looked like a successful test of borzoi 2.6 on the forums Sep 11 09:47:12 RP: excellent. so machine detected works like expected? Sep 11 09:47:30 mickeyl: it would appear so Sep 11 09:48:50 mickeyl: I've pushed the detection patchs into mainline and included tosa Sep 11 09:50:54 RP: very good. have the backlight things been finally eliminated yet? Sep 11 09:51:12 backlight things? Sep 11 09:51:38 You mean the linearity? Sep 11 09:51:58 03koen 07org.oe.dev * r427dc4e4... 10/conf/distro/familiar-0.8.3.conf: conf/distro/familiar-0.8.3.conf: prefer Opie 1.2.1 Sep 11 09:52:11 RP: no, the problems lardman had Sep 11 09:52:37 mickeyl: No, I've not had a chance to deal with that. It works fine under opie Sep 11 09:53:05 I've been spending all available time on sorting out and getting the core spitz changes into mainline Sep 11 09:53:20 I've done that so now I'm debugging the init rc.S problem Sep 11 09:54:03 mickeyl: I got a lot of feedback from rmk which meant rewriting a few of the patches Sep 11 09:55:27 RP: yep, i've seen a bit of it on lkml Sep 11 09:55:46 mickeyl: Even with the init problem, opie is looking good on spitz 2.6 :) Sep 11 09:56:00 good to hear :) Sep 11 09:57:05 hi ! Sep 11 09:59:29 I've also not had any time to look into the bl problems I'm afraid Sep 11 10:00:36 nm. i've read that cpufreq works good now? Sep 11 10:01:17 it seems to work okay for me, and hrw was only complaining about a slow touchscreen after suspend Sep 11 10:01:31 retouch the conf file Sep 11 10:01:36 it its working now Sep 11 10:01:40 thankx mickeyl Sep 11 10:02:01 The touchscreen does nasty timing things with the hsync line Sep 11 10:02:19 RP: I've come across another (minor) kernel pm issue - if the battery level gets low, the bl is turned down, then the batt level goes up again, bl goes up again - lucky I'm not epileptic ;) Sep 11 10:02:47 RP: might be worth altering the level at which the bl is turned back up again Sep 11 10:03:27 good lord ! libschedule-0.15.tar.gz is back. haaaa... thanks to the person who placed it in one of the sources feeds. :) Sep 11 10:03:28 RP: I'll have a fiddle with it next weekend when I'm back Sep 11 10:10:30 lardman: That code probably does need some tweaking - I generally never run my battery down that much Sep 11 10:11:18 lardman: How's the rest of the pm/battery code performing? Sep 11 10:14:21 RP: Seems fine other than that - I only noticed it as I daren't touch the bl setting when I was running GPE - I thought a contact had come loose, the flashing was pretty random until the battery got really flat then I was more worried I'd burn it out Sep 11 10:15:33 We should probably check it works off the averaged value and not raw data... Sep 11 10:16:53 How long is the period over which the average is calculated though? I didn't think it was that long Sep 11 10:17:21 ftpd-topfield failed to build: Fetch cvs://anonymous@cvs.sourceforge.net/cvsroot/puppy;method=pserver;module=libtopfield;tag=LIBTOPFIELD_0_6_1 failed. no such tag LIBTOPFIELD_0_6_1 Sep 11 10:17:38 But yes, it does sound as if it's perhaps not long enough. Even if it were longer though, you;d still get it flashing on & off, just with a longer period Sep 11 10:18:39 As you say, we should also probably stop it turning off the limiter straight away Sep 11 10:18:51 It is a good and eye-catching signal that your battery is going flat though :) Sep 11 10:21:16 lardman: true :) Sep 11 10:24:02 can we close #207 ? Sep 11 10:24:15 what's with #169 ? Sep 11 10:25:32 ~lart people adding .bb files as attachements without stating PV Sep 11 10:25:33 * ibot installs PocketPC on people adding .bb files as attachements without stating PV's PDA Sep 11 10:28:26 Strange. The dev directory is empty after updater.sh untars to the microdrive :-/ Sep 11 10:30:35 odd Sep 11 10:32:52 Ok, my hdimage1.tgz definitely contains /dev/* devices. updater.sh runs "gzip -dc /mnt/cf/hdimage1.tgz | tar xf -". No dev/* devices are created. Why? Sep 11 10:35:06 * mickeyl clueless Sep 11 10:35:32 I changed it to tar xvf and it even mentioned dev/* files... Sep 11 10:35:35 RP: mounted with the nodev option? Sep 11 10:36:23 Just noatime... Sep 11 10:39:11 "mknod console 5 1" works :-/ Sep 11 10:40:13 I suspect a busybox tar bug related to device files... Sep 11 10:40:23 google suggests there have been some Sep 11 10:44:25 mickeyl: how much work would it be to add 'require ' to bitbake? Sep 11 10:44:55 koen: I'm not familiar with the parser / command interpreter, but I can't imagine it being much of a problem Sep 11 10:45:02 kergoth and pb_ may have more insights on that Sep 11 10:46:04 Any ideas how to work around a borken tar? Sep 11 10:46:18 does python have 'try foo ; die bar' ? Sep 11 10:47:52 koen: I'm not sure I understand what this "require" thing would do. Sep 11 10:48:07 error out if the file doesn't exist Sep 11 10:48:15 so, like "include"? Sep 11 10:48:22 but with an error on missing files Sep 11 10:48:28 yes Sep 11 10:48:38 s/error out/raise a parsing exception/ Sep 11 10:48:42 that wouldn't be too hard to arrange. Sep 11 10:48:59 yeah, what kergoth said Sep 11 10:49:29 RP: lets ship our own tar Sep 11 10:49:53 koen: you mean try: foo except BarError: dosomething ? Sep 11 10:50:17 mickeyl: yeah, like that Sep 11 10:50:25 mickeyl: I don't suppose you have a staically linked tar handy? :) Sep 11 10:50:35 RP: not offhand ;) Sep 11 10:50:54 RP: might check if the cacko updater set has one included Sep 11 10:50:59 or pdaxrom Sep 11 10:51:09 they ship a number of own tools as well Sep 11 10:53:44 They ship tar for some reason ;-) Sep 11 10:54:26 heh Sep 11 10:55:10 mickeyl: Shall I just add that binary to zaurus-updater for spitz/akita/borzoi? Sep 11 10:55:26 (providing testing shows it to fix the problem) Sep 11 11:01:02 RP: yes, please. Sep 11 11:01:32 sooner or later we will have our own installer, but until then... Sep 11 11:03:17 mickeyl: That day draws ever closer - I like having the updater scripts in OE :) Sep 11 11:06:16 mickeyl: Its a 1.4MB binary :-/ Sep 11 11:07:14 There was one used to untar that debian on Zaurus thing from a long long time ago Sep 11 11:07:36 People still use it, and I think the tar is still on the site Sep 11 11:07:52 This isn't a tar binary at all :-/ Sep 11 11:08:48 lardman: Do you have a link?> Sep 11 11:08:55 just looking Sep 11 11:09:09 http://www.oesf.org/forums/index.php?showtopic=14634&hl=gnu-tar Sep 11 11:09:18 doesn't look so good though if you read the post Sep 11 11:11:34 03mickeyl 07org.oe.dev * ra95e3e72... 10/packages/swig/swig_1.3.24.bb: swig: fix typo causing swig not to install its support files. Closes #321, thanks mreimer. Sep 11 11:11:39 03mickeyl 07org.oe.dev * r43c86569... 10/packages/linux/ (36 files in 2 dirs): remove older version of sharp pxa27x kernel, the new version works with all models Sep 11 11:11:43 03mickeyl 07org.oe.dev * rdb10c295... 10/packages/ (5 files in 2 dirs): Sep 11 11:11:44 add python-sword, python bindings for the sword document reader library Sep 11 11:11:44 patch courtesy Matt Reimer, thanks. Sep 11 11:15:20 lardman: I've found a copy on killefiz.de - I'll see how it works... Sep 11 11:17:15 bbl Sep 11 11:17:15 RP: cool Sep 11 11:19:40 This works much better... Sep 11 11:34:13 night all Sep 11 11:39:49 i still don't get how to create an initrd ('copy over all needed files' - but what are the needed files?) Sep 11 11:40:21 i'd appreciate any help Sep 11 11:44:12 did you look at the initrd bb? Sep 11 11:46:41 just did a search, couldn't find a initrd.bb (in /packages) Sep 11 11:55:49 is there any windowmanager except xfce in oe that is not meant for usage on a pda-like device? Sep 11 11:57:03 enlightenment? Sep 11 11:57:04 heh Sep 11 11:58:39 eh, I only have a 320mhz machine :) Sep 11 11:58:59 why not using xfce? Sep 11 11:59:30 you could go ahead and add something like metacity if you wanted. Sep 11 11:59:36 because I don't get xfce to work :( shame on me.. Sep 11 11:59:48 thats a good reason Sep 11 12:01:52 it compiles fine, but when it comes to do.rootfs it says it cannot satisfy the folowing deps for task-xfce base: xfwm4-mcs-plugin etc.. Sep 11 12:03:27 try to manually remove the dependecy from the .bb file? Sep 11 12:03:46 but im far from an expert... Sep 11 12:04:18 I first have to find out what RDEPENDS exactly does.. Sep 11 12:05:10 Delta_: also make sure you don't have ipkg 0.99.153 built Sep 11 12:05:53 I have 0.99.152 Sep 11 12:06:25 ok, one problem less Sep 11 12:07:01 the xfce people won't win rewards for sane autotools usage Sep 11 12:07:28 although libexo isn't officially in the xfce core yet Sep 11 12:07:49 ~lart libexo Sep 11 12:07:50 * ibot hits libexo with an anvil and laughs with a contralto voice ... Haha Ha HA Ha Sep 11 12:08:20 ~lart quilt Sep 11 12:08:21 * ibot takes quilt to the vet for a "special" visit Sep 11 12:09:28 Should the borzoi updater.sh not be based on the spitz one? Sep 11 12:09:51 does RDEPENDS define all the packages that go into the rootfs image? Sep 11 12:10:50 no Sep 11 12:11:00 RDEPENDS defines what should be installed at runtime Sep 11 12:12:10 openzaurus.conf => openzaurus.inc. Thoughts? This *should* fix the "unversioned DISTRO" problem, right? Sep 11 12:12:39 JustinP: You need to ask mickey|away about that. I've already asked and he was thinking about it Sep 11 12:12:57 well it was solved for familiar with something else... Sep 11 12:13:02 * JustinP tries it Sep 11 12:14:37 JustinP: rename it to openzaurus-distro.conf Sep 11 12:15:56 JustinP: http://monotone.vanille.de/viewmtn/getdiff.py?id1=96f8aac9c83ed5752f79b627a828c8188722ce2a&id2=38885fa208a5b80dcfd041dd2ebcfa812948a589&fname=conf/local.conf.sample Sep 11 12:21:23 wouldnt/shouldnt an unversion distro default to last release? Sep 11 12:21:42 since that is what most users would expect Sep 11 12:23:16 i'm still deserately searching for any info about how to create an initrd Sep 11 12:24:39 there are lots of examples in packages Sep 11 12:24:40 i found info, but all of them are telling me to 'copy over needed files', which files are those? Sep 11 12:24:43 koen: don't say that evil word ;) Sep 11 12:25:31 Henry78, have you ever built an initrd outside of OE? Sep 11 12:25:44 no, just used automated tools Sep 11 12:26:42 http://www.linux-boot.net/InitRD/Howto/ Sep 11 12:26:49 might help as an outline Sep 11 12:27:09 okay, i'll dive into Sep 11 12:31:26 emte: that's the thing...it's really z misnomer. THe existance of openzaurus.conf lends itself to the belief that it's a valid DISTRO target Sep 11 12:31:29 it's not Sep 11 12:31:30 it's an inc file Sep 11 12:31:33 hence I'm renaming it Sep 11 12:32:39 03rpurdie 07org.oe.dev * r98f16aa0... 10/ (4 files in 4 dirs): Sep 11 12:32:39 zaurus-updater: The tar utility found on the Sharp inbuilt root filesystems used Sep 11 12:32:39 for reflashing the devices has bugs, one being that it won't create device Sep 11 12:32:39 files. We therefore need an external sane tar binary to successfully extract the Sep 11 12:32:39 root filesystem onto the device. This adds such a binary for spitz and borzoi. Sep 11 12:32:53 JustinP: check with mickeyl first Sep 11 12:33:05 since he's the actual release manager Sep 11 12:33:56 got my finger on the trigger.... Sep 11 12:34:02 this doesn't break anything Sep 11 12:34:12 that's not the point Sep 11 12:34:22 JustinP: Can you check bug 207 please - I'm presuming that is resolved now? Sep 11 12:34:52 koen: it's *not* a valid DISTRO target... Sep 11 12:34:54 :sigh:: Sep 11 12:35:05 JustinP: The point is you check with the maintainer of a file before making changes to it Sep 11 12:35:25 mickey|away: please comment on renaming of conf/distro/openzaurus.conf to openzaurus.inc. I've got a commit waiting if you approve. Sep 11 12:36:38 RP: sorry, I'm not sure. In fact, I'm not even sure if the auto-rotation works on startup right now Sep 11 12:36:42 * JustinP builds a GPE image Sep 11 12:38:53 morning Sep 11 12:38:55 hi XorA Sep 11 12:39:41 JustinP, its also a double edge, if there is not a generic that points to the last release then people will continue to build the image that doesnt work Sep 11 12:40:30 people just building apps dont rellay care what the version is as long as it works on the latest release Sep 11 12:40:31 DISTRO = "openzaurus" works once in a millenium or so :-) Sep 11 12:42:11 XorA: except that's not really a valid option....at all.... Sep 11 12:42:33 emte: I understand. But "we" have been telling people not to use unversioned DISTROs for so long.... Sep 11 12:42:36 only by current convention Sep 11 12:42:49 yes, a "choose the newest version" way would be better Sep 11 12:42:57 how so? Sep 11 12:43:01 JustinP: whats not valid about it? not sensible yes, but it appears to be valid Sep 11 12:43:03 but the current openzaurus.conf isn't meant for that either...so.... Sep 11 12:43:14 XorA: exactly. It appears to be valid but it's not. Sep 11 12:43:16 it takes months to get the newest to even compile Sep 11 12:43:37 and people try to build apps against a setup that doesnt work Sep 11 12:44:00 emte: then screw it. We leave it at versioned only. Rename the file so people get a big fat error when they try it. Perhaps a "file not found" would be best. Sep 11 12:44:26 sould just inherit the last relese Sep 11 12:44:30 release* Sep 11 12:44:39 renaming the file makes "openzaurus" an invalid option (in that it doesn't set TARGET_OS and the entire thing fails) Sep 11 12:45:02 ::shrug:: I suppose that would be ok as well, if that's what we want to do Sep 11 12:45:13 JustinP: familiar has the same generic .conf file Sep 11 12:45:15 regardless, the current openzaurus.conf is included from openzaurus-3.5.4.conf Sep 11 12:45:20 so it's an .inc, not a .conf Sep 11 12:45:26 it makes it "friendlier" for non-core developers Sep 11 12:45:27 JustinP: so this change should really go through oe@handhelds.org Sep 11 12:46:09 actually, familiar now has familiar-distro.conf Sep 11 12:46:27 IMHO anything which is not a valid DISTRO target should not be a .conf Sep 11 12:46:44 then it *cannot* be used as a value in DISTRO without it exploding Sep 11 12:46:56 currently you just get strange errors if you set it unversioned Sep 11 12:47:11 not really Sep 11 12:47:19 familiar builds fine unversioned Sep 11 12:47:36 or did i suppose Sep 11 12:48:39 reminds me i need to update monotone Sep 11 12:48:48 did someone say cpufreq is now working on Zs ? Sep 11 12:49:15 its on RP's list is all i knwo Sep 11 12:49:42 * XorA was reading up through last days worth of channel on quickskim Sep 11 12:50:08 XorA: lardman has a working patch Sep 11 12:50:47 RP: sweet, that'll max out the battery life Sep 11 12:51:07 I think its classed as mostly working as some bugs have been seen Sep 11 12:51:12 * XorA just discovered the joy on powernowd on his amd64 Sep 11 12:51:14 but its nice to know its there Sep 11 12:52:14 Why does IMAGE_FSTYPE="tar" create a tar.bz2 file? I've just asked oe@hh.org whether this can be changed to something sane... Sep 11 12:55:07 RP: it was changed to do that over a year ago Sep 11 12:55:39 actually from tar.gz to tar.bz2 Sep 11 12:55:52 koen: I know. I just think tar should actually give you a tar file ;-) (tar.gz and tar.bz2 can be added to give other options) Sep 11 12:57:20 In fact has anyone an objection to me adding a tar.gz target now? Sep 11 12:57:33 RP: no Sep 11 12:57:40 check if ext2.gz exists too :) Sep 11 12:57:46 I'll leave tar alone until I get feedback as that can potentially break things Sep 11 12:57:52 koen: it does Sep 11 12:58:27 when I want to be sure everything is cleaned up again, can I just remove the tmp/work dir or do I have to remove the whole tmp dir? Sep 11 12:59:36 zecke: we still need to vote for OE's most wanted feature Sep 11 12:59:46 koen: where and when? Sep 11 12:59:54 zecke: for the CELF people Sep 11 12:59:56 "make koen and zecke flithy rich" Sep 11 12:59:56 koen: an AI bb updater :-) Sep 11 13:00:17 wait I forgot something Sep 11 13:00:21 [x] "make koen and zecke flithy rich" Sep 11 13:00:25 :) Sep 11 13:00:39 03xora 07org.oe.dev * r7123616b... 10/packages/mozilla/ (3 files in 2 dirs): Sep 11 13:00:39 firefox: added the latest beta/rc 1.5b1 Sep 11 13:00:39 currently set at DEFAULT_PREFERENCE=-1 as it currently builds and runs but doesnt actually open a window on my OZ/GPE setup. Sep 11 13:00:55 koen: QA,uclibc and different init system are my top priorities Sep 11 13:01:01 putting staging under package management would be a good second option Sep 11 13:01:44 koen: Where is the voting taking place? Sep 11 13:01:45 hm I ran out of cheetos :( Sep 11 13:01:55 RP: we had a thread on the mailinglist a while back Sep 11 13:02:05 'vote' is a big word Sep 11 13:02:17 koen: I thought that had sort of stopped dead... Sep 11 13:02:22 it has Sep 11 13:02:35 and I don't want to loose the opportunity Sep 11 13:02:55 I would feel most happy if mickeyl,pb and kergoth could negotiate with that guy who explores Sep 11 13:02:56 Have you anyone in mind to do the work? Sep 11 13:02:57 options to help Sep 11 13:03:17 RP: we could contract montavista ;) Sep 11 13:03:34 RP: according to their wiki they would contract a 3rd party Sep 11 13:03:37 i.e not us Sep 11 13:03:49 zecke: haha ;-) Sep 11 13:04:06 koen: I thought they meant it would be a third party i.e. not them Sep 11 13:04:42 RP: I think it can be anyone, phils and mithros reasoning was people could stop loving what they do Sep 11 13:04:53 e.g what happened to kergoth at TI Sep 11 13:04:56 I suspect Openedhand might be able to offer its services... Sep 11 13:05:38 RP: right that could work, or something like ROAD GmbH Sep 11 13:05:54 but most importantly we need to define something that is good for OpenEmbedded Sep 11 13:05:55 Hi: Found new build dependency error while building familiar-0.8.3 image from scratch. gpe-calendar failed to build for following error: Package 'libsqlite0', required by 'libschedule', not found Sep 11 13:06:05 zecke: agreed Sep 11 13:06:23 I like the idea of staging under packaging control Sep 11 13:06:35 I also would like to see a new bitbake Sep 11 13:06:43 koen: could you summarize the thoughts we had and revive the thread? Sep 11 13:06:58 RP: hehe what is wrong with bitbake? Sep 11 13:07:26 sounds like kergoth is playing with the thought of starting bb-ng some time soon anyway Sep 11 13:07:29 zecke: faster, faster, faster! ;-) Sep 11 13:07:47 altho the CELF peopel would probably prefer you were using Eclipse Sep 11 13:07:51 zecke: ok Sep 11 13:09:57 wonder how well QNX did at thier conference this summer ... Sep 11 13:10:24 emte: oh they're still alive? Sep 11 13:10:29 i guess so Sep 11 13:10:57 i got an invite at $150 to it for using thier SA kit Sep 11 13:11:07 for ipaq Sep 11 13:11:43 there doesnt really seem to be very much left of the RTOS feild Sep 11 13:11:59 as far as large providers go Sep 11 13:12:56 then again i guess most people normally use one chip manuf Sep 11 13:13:12 and use thier RTOS tools Sep 11 13:13:29 i should play with orange one day i guess Sep 11 13:17:21 Orange as in the operator? Sep 11 13:20:20 zecke: http://pastebin.ca/22673 Sep 11 13:20:46 koen: sorry I've no browser on my text console :} Sep 11 13:21:18 RP: bitbake will be fast - at the day I start to understand what bitbake does Sep 11 13:21:39 Orange as in the Microchip RTOS Sep 11 13:23:00 Does anybody know how to fix gpe-calendar build with familiar-0.8.3? libschedule not finding libsqlite0. http://pastebin.ca/22675 Sep 11 13:23:37 zecke: I have the same problem - I need to spend some time on it... Sep 11 13:24:27 RP: you lost me Sep 11 13:24:57 zecke: I also don't understand bitbake. With time, I could do so and write a better version... Sep 11 13:25:17 You understand bitbake better than I do though ;-) Sep 11 13:26:54 RP: ah well, some people claim I never built something using OE, and don't understand how bitbake wors.. ;) Sep 11 13:27:59 zecke: ah, yes. In the same way I know nothing about zaurus kernels ;-) Sep 11 13:28:24 RP: yeah we're posers, we should stop selling crack ;) Sep 11 13:28:52 RP: but admitingly there are parts in the kernel and bitbake that just work, without further investigation Sep 11 13:28:52 :) Sep 11 13:29:21 zecke: indeed. I'm well aware of where they are in the kernel ;-) Sep 11 13:29:34 hehe Sep 11 13:29:45 I need to plug the ethernet cable from my cable modem Sep 11 13:29:48 be back later Sep 11 13:30:10 zecke: bfn Sep 11 13:33:03 well RP you could always ask kergoth what needs to be done for bb-ng Sep 11 13:33:36 i belive most of the python stuff is getting tossed Sep 11 13:35:22 zecke: yah, it would be good to get something going with CELF. Sep 11 13:35:51 and, if "something" means CELF giving a pile of cash to ROAD GmbH, that would be fine. Sep 11 13:37:31 mail sent, let's see what CELF does with it Sep 11 13:38:47 alsa i hate you~ Sep 11 13:55:02 Does anyone here know if it's possible, and if so, how hard, to interface a memory stick duo card? Sep 11 14:00:26 everything is possible Sep 11 14:00:45 the problem is the amount of information availible Sep 11 14:02:53 03mickeyl 07org.oe.dev * r53eb1350... 10/packages/efl/ (ecore/add-tslib-support.patch ecore-fb_20050907.bb): efl: add tslib support to ecore-fb (cvs) Sep 11 14:08:58 pb_: I can't help but wonder what I should call the uncompressed tar FSTYPE ;-) Sep 11 14:11:34 yeah, 3 airwires left in my pcb layout , I am soo good :) Sep 11 14:11:50 RP: "uncompressed tar"? Sep 11 14:14:13 'tar'? Sep 11 14:19:26 koen: I think you see the source of my amusement. Anyhow, I'll leave it alone for now and wait for several people to file bug reports... Sep 11 14:23:17 how about tape-ar ? Sep 11 14:23:22 since taht is what it is Sep 11 14:23:33 tape archive Sep 11 14:33:30 re Sep 11 14:33:40 wb zecke Sep 11 14:33:57 hey zecke Sep 11 14:36:15 * zecke just lost his laptop to his mother... Sep 11 14:36:57 your new ibook? that's rather sad. Sep 11 14:37:10 pb_: no, that was lost to Al... Sep 11 14:37:31 pb_: my old HP Omnibook is going into the hospital... Sep 11 14:38:26 you allowed to use wifi in a hospital? Sep 11 14:38:39 or is she playing games? Sep 11 14:38:56 emte: no, she got a dial up opportunity Sep 11 14:39:03 ah Sep 11 14:39:12 emte: or maybe they use DECT for it... Sep 11 14:39:32 your hospitals seem more advance then ours ... Sep 11 14:39:52 hehe that is a private station Sep 11 14:39:57 zecke: aha. Sep 11 14:40:14 actually it is the hospital I stayed roughly sixth month in a row Sep 11 14:40:29 emte: you got tv, radio, electronic bed, fridge, ac... Sep 11 14:40:34 emte: and internet Sep 11 14:40:37 wow Sep 11 14:40:48 electronic bed? that sounds eleet. Sep 11 14:40:57 lucky to get a bed here Sep 11 14:41:14 in under 6 hours anyway Sep 11 14:41:32 pb_: like in the simpsons... cloud up, cloud down... Sep 11 14:42:29 heh Sep 11 14:44:53 pb_: but there are good reasons to have control over your bed ;) Sep 11 14:45:19 zecke: I certainly wouldn't want anybody else to have control of my bed. Sep 11 14:45:32 pb_: certainly you don Sep 11 14:45:45 t want to ring a bell, to drink some stuff... Sep 11 14:46:01 you simply take the remote control ;) and move yourself into a suitable position Sep 11 14:46:14 'eject' Sep 11 14:46:14 you can control, X, Y and even Z Sep 11 14:46:22 hehe Sep 11 14:46:31 very good Sep 11 14:47:02 can you drive it from room to room around the hospital? Sep 11 14:48:26 not without modifying Sep 11 14:48:39 but AFAIK it has hooks to be mind controled Sep 11 14:48:40 too bad Sep 11 14:48:52 and they go straight to flying, driving is so yesterday Sep 11 14:49:20 though I guess it might cause a traffic hazard if all the patients are allowed to cruise around the corridors at high speed Sep 11 14:50:42 pb_: even the elete deserves natural selection ;) Sep 11 14:52:16 true enough Sep 11 15:14:30 03rpurdie 07org.oe.dev * r8886e2ff... 10/conf/ (5 files in 3 dirs): Sep 11 15:14:30 IMAGE_FSTYPE updates. Add new tar.bz2 and tar.gz image types so the compression Sep 11 15:14:30 format can be specified. Update tar references to tar.bz2. Update spitz and Sep 11 15:14:30 borzoi to correctly use tar.gz. ***Users of IMAGE_FSTYPE="tar" are advised to Sep 11 15:14:30 switch to tar.bz2***. tar is likely to become an uncompressed tarball at some Sep 11 15:14:31 point. Sep 11 15:16:22 'night all Sep 11 15:16:27 'night koen Sep 11 15:17:40 03rwhitby 07org.oe.dev * ra404b389... 10/packages/ftpd-topfield/ftpd-topfield_0.6.1.bb: ftpd-topfield: fixed the CVS tag Sep 11 15:28:48 03rwhitby 07org.oe.dev * r80332541... 10/packages/ftpd-topfield/ftpd-topfield_0.6.0.bb: ftpd-topfield: removed intermediate buggy version 0.6.0 Sep 11 15:34:39 mickeyl: ping Sep 11 17:14:16 * chouimat|tired is away: tv: the simpsons Sep 11 17:47:33 hi Sep 11 18:30:09 anyone home? :P **** BEGIN LOGGING AT Sun Sep 11 19:53:59 2005 Sep 11 20:01:14 hi anyone here? :) Sep 11 20:01:22 no Sep 11 20:01:26 hehe :P Sep 11 20:02:03 what is the min I need to clean up when I switch from 2.6 back to 2.4 when building opie-image Sep 11 21:27:37 * france is back (gone 36:27:20) Sep 11 21:45:10 * france is away: Away Sep 11 23:30:14 how to add iptable support in the borzoi kernel? Sep 12 00:43:47 morning all Sep 12 00:45:25 morning Sep 12 00:48:45 morning all ! Sep 12 00:49:27 hi alan Sep 12 00:52:20 who shall i thank for having ported lyx to OE ? Sep 12 00:54:29 http://esr.ibiblio.org/index.php?p=208 Sep 12 00:54:34 mickeyl Sep 12 00:54:37 iirc Sep 12 00:54:43 anyway - look into lo Sep 12 00:54:44 g Sep 12 00:59:39 morning hrw|work Sep 12 01:00:57 RP: cxK will use one kernel for all 3 models? Sep 12 01:08:06 hrw|work: Yes, but as the commandline will need to be different for akita and spitz+borzoi, we'll need two Sep 12 01:08:29 If the commandlines were the same, they could all run the same kernel Sep 12 01:17:23 aha Sep 12 01:17:44 spitz/borzoi will start from microdrive and akita from flash? Sep 12 01:18:46 mo0rning Sep 12 01:18:58 damn I need to get used to laptop keyboard Sep 12 01:22:33 hi XorA Sep 12 01:36:30 glurp... :( onca again and again, i need some help... Sep 12 01:36:30 http://pastebin.ca/22696 Sep 12 01:42:52 NOTE: package gcc-cross-initial-3.4.4-r2: task do_patch: failed Sep 12 01:42:53 ;( Sep 12 01:43:07 'morning all Sep 12 01:44:31 koen: gcc-cross-initial-3.4.4 should work out-of-box - right? | Patch gcc34-reverse-compare.patch does not apply (enforce with -f) Sep 12 01:46:36 guten tag, you crazy kids Sep 12 01:46:41 hey CP|Laptop Sep 12 01:46:48 hrw|work: afaik it should work Sep 12 01:49:25 hi crazypenguin Sep 12 01:49:31 * CP|Laptop is tired Sep 12 01:50:39 ok. I know what is wrong Sep 12 01:50:46 bzip2: Can't open input file /home/hrw/devel/sources/dl_dir/gcc-3.4.4.tar.bz2: No such file or directory. Sep 12 01:50:49 NOTE: package gcc-cross-initial-3.4.4-r2: task do_unpack: completed Sep 12 01:50:51 heh.. Sep 12 01:52:23 CP|Laptop: i tried to add libexo and thunar, but their buildsystem is just too weird for me to fix Sep 12 01:52:29 ~lart xfce people Sep 12 01:52:29 * ibot brandishes Excalibur! "With this sword, I vanquish thee, xfce people!" and lops off xfce people's head Sep 12 01:52:49 koen: oh, its not the standard autoffo stuff? Sep 12 01:52:59 s/autoffo/autofoo/ Sep 12 01:53:23 The core XFCE stuff was really easy to add because it obeyed all of the fd.o specs Sep 12 01:53:46 CP|Laptop: broken autofoo Sep 12 01:53:56 most of the .bbs only needed a SRC_URI, include autotools, and a staging Sep 12 01:54:18 yeah, I noticed that Sep 12 01:54:42 I guess that german dude didn't pay attention to that when he did libexo and thunar Sep 12 01:54:47 sad Sep 12 01:57:02 if you pass --with-jpeg=${STAGING} it add staging/lib and staging/include to cflags (as expected) but also puts -l/usr/lib/libjpeg.so there Sep 12 01:57:16 which obviously fails since I'm crossbuilding Sep 12 01:57:59 bah Sep 12 01:58:00 so I added the gnome 2.12 platform instead :) Sep 12 01:58:05 Heh Sep 12 01:58:11 that would have been worse, I would have thought Sep 12 01:58:18 03koen 07org.oe.dev * refe5432c... 10/packages/libschedule/libschedule_0.15.bb: packages/libschedule/libschedule_0.15.bb: switch back to .gz since the libschedule.pc in the .bz2 seems to be broken Sep 12 01:58:32 the platform was pretty easy, that's only a few libs Sep 12 01:59:15 hey people Sep 12 02:02:27 hey mithro Sep 12 02:02:48 http://www.giantshoulderinc.com/ <- product offered with OE **** BEGIN LOGGING AT Mon Sep 12 02:13:53 2005 Sep 12 02:14:20 koen: no DVB? Sep 12 02:14:35 I have no DVB here :) Sep 12 02:14:48 g2 did mention somthing about mpeg2 Sep 12 02:15:10 but I'm not sure if it was onboard on another minipci thing Sep 12 02:26:31 morning all Sep 12 02:28:49 hey ade|desk & Twiun Sep 12 02:28:59 yo Sep 12 02:29:03 happy monday morning. Sep 12 02:30:37 03koen 07org.oe.dev * rcdb411e2... 10/packages/abiword/ (abiword-2.3.6/cdump-hack.patch abiword_2.3.6.bb): packages/abiword/ : add abiword 2.3.6 and patches Sep 12 02:35:18 i'm still sutck into the same error... Sep 12 02:35:22 http://pastebin.ca/22696 Sep 12 02:36:00 i don't understand the "offline root mode: not running libgpg-error0.postinst" part... Sep 12 02:36:39 alan|home: that's not an error. It means that a .postinst script (which is written for your target) doesn't get executed on your build host Sep 12 02:37:13 schurig: ho... does it mean i miss a prog on my build machine ? Sep 12 02:38:00 alan|home: No. Again: that's perfectly sane. Your target machine is probably an PDA with some PXA processor. You cannot run program meant for this target on your build machine, which is probably an i386 box Sep 12 02:38:21 yes, that is right... Sep 12 02:39:07 alan|home: when you first boot your target, then some init script runs "ipkg configure" and then ipkg will run all those .postinst scripts in the real root, with the target, that wheren't run when you made your root filesystems Sep 12 02:40:15 ok ok, but i'm currently trying to build an image for my zaurus... i'm not even booting the zaurus... Sep 12 02:42:39 alan|home: and then it's ok that the .postinst scripts aren't run. Perfectly ok. Real, absolute, wonderously ok. Sep 12 02:42:48 morning Sep 12 02:43:08 morning mickeyl. Sep 12 02:43:14 mickeyl: moin! Sep 12 02:43:35 alan|home: That log file doesn't look quite right - can you file the do_rootfs log from build/tmp/work/gpe-image/temp and pastebin that please? Sep 12 02:43:37 hi mickeyl Sep 12 02:43:52 RP: ok. Sep 12 02:47:26 RP: i have several do rootfs logs. Which one is interresting ? .23132, .4184, .4782, .5798 ? Sep 12 02:47:43 alan|home: Just post the most recent Sep 12 02:48:24 RP : oj i'm stupid. It's writen on the message outputs... ^^' Sep 12 02:49:21 RP: http://pastebin.ca/22697 Sep 12 02:50:26 alan|home: look at line 809-810 Sep 12 02:51:19 alan|home: it somehow misses the right libgcc1 Sep 12 02:51:47 alan|home: Did you change compilers half way through the build? Sep 12 02:52:11 03mickeyl 07org.oe.dev * re1b7c2c0... 10/packages/qt/ (qt-x11-free_3.3.5.bb qt-x11-free_4.0.1.bb): qt: PROVIDE qt{3|4}-x11 as Qt is not upwards/downwards compatible across major releases Sep 12 02:52:16 03mickeyl 07org.oe.dev * reede0022... 10/classes/ (qt3x11.bbclass qt4x11.bbclass): qtx11 classes: DEPEND += qt{3|4}-x11 Sep 12 02:52:20 03mickeyl 07org.oe.dev * rc170fc61... 10/packages/ (libqanava/libqanava_0.0.3.bb lyx/lyx_1.3.6.bb): lyx, libqanava: catch up with new qt major release depends Sep 12 02:52:24 03mickeyl 07org.oe.dev * r014c1427... 10/packages/mythtv/ (6 files in 2 dirs): mythtv: remove explicit qt-x11-free dependency and some obsolete versions Sep 12 02:55:22 RP: my build missed several times. i had once to update my branch before restrarting the build. Can this be the origin for this ? Sep 12 02:55:57 OZ changed compilers a few days back Sep 12 02:56:52 arf... so what file should i erase and recompile ? Sep 12 02:57:21 everything gcc ? Sep 12 02:57:25 alan|home: If you have multiple versions of gcc in build/tmp/work, its highly likely the compiler change has broken things. The csl compiler was known to break things and best removed/not used Sep 12 02:57:49 alan|home: Sadly, with a problem like this, you might have to rebuild everything :-( Sep 12 02:58:04 noooooooooooooo.... :'( Sep 12 02:58:13 You probably have half a build directory compiled with one compiler and half compiled with another Sep 12 02:58:41 arf... Sep 12 02:59:13 The one consolation is that the csl compiled image wouldn Sep 12 02:59:13 never mind, let's build one again. Sep 12 02:59:22 wouldn't have worked anyway Sep 12 02:59:44 RP: maybe packages/gcc/*csl* should be moved to packages/gcc-csl **** ENDING LOGGING AT Mon Sep 12 02:59:56 2005