**** BEGIN LOGGING AT Tue Mar 02 02:59:58 2010 Mar 02 04:15:19 * Tartarus kicks firefox around a bit more Mar 02 04:15:54 Got all of the SOL if you don't have 'em deps sorted Mar 02 04:16:02 kicking libIDL and parallel make stuff about Mar 02 04:16:44 * mwester cheers Tartarus on; anything that starts up *slower* than Microsoft IE deserves a mighty thorough kicking. Mar 02 04:17:02 ha Mar 02 04:17:05 That said, woop Mar 02 04:17:08 libIDL stuff works right now Mar 02 04:17:21 w/o relying on pkg-config to have be on the host Mar 02 04:17:32 (since it's a global thing, before PATH is munged :( ) Mar 02 05:12:41 mozilla parallel make stuff is funky Mar 02 05:12:48 But I'm pretty sure it's actually happening Mar 02 05:13:25 funky in that it's either hidden but active, I'm pretty sure from another error I hit wrt IDL stuff Mar 02 05:13:34 Or force down to -j1 for certain subtargets (install, export) Mar 02 05:13:41 * Tartarus waits for a 'time' to finish Mar 02 05:37:35 17min for clean to compile Mar 02 05:37:57 * Tartarus takes threads out and confirms really, now Mar 02 06:05:31 Oh yes, definitely a change :) Mar 02 06:06:12 * Tartarus goes to test 3.5.5 now w/ his various changes Mar 02 06:31:40 And, since I'm crazy, this will go overnight, and the machine is in the other part of the house, i'll do a firefox 3.0.8 + uclibc build while eglibc + 3.5.5 chugs along Mar 02 07:28:14 morgen from Embedded World 2010, Nurnberg :-) Mar 02 07:29:00 bb in 8h Mar 02 07:45:10 hi Mar 02 08:42:24 03Martin Jansa  07org.openembedded.dev * rf2f69d9ae0 10openembedded.git/recipes/tasks/task-shr-minimal.bb: Mar 02 08:42:24 task-shr-minimal: add phoneui-apps-quick-settings Mar 02 08:42:24 Signed-off-by: Martin Jansa Mar 02 08:59:23 03Koen Kooi  07org.openembedded.dev * r5ba1e9d435 10openembedded.git/recipes/ekiga/ptlib_2.8.0.bb: ptlib: add 2.8.0 Mar 02 08:59:23 03Koen Kooi  07org.openembedded.dev * r5da93ca6d9 10openembedded.git/recipes/ekiga/ (ptlib/rgb16.patch ptlib_2.6.5.bb): ptlib 2.6.5: add RGB16 support Mar 02 08:59:24 03Koen Kooi  07org.openembedded.dev * rac799de190 10openembedded.git/recipes/ekiga/opal_3.8.0.bb: opal: add 3.8.0 Mar 02 08:59:24 03Koen Kooi  07org.openembedded.dev * rd63e3673cf 10openembedded.git/recipes/ekiga/ (ekiga/rgb16.patch ekiga_git.bb): ekiga git: update to 3.3.1 and add rgb16 support Mar 02 08:59:26 03Koen Kooi  07org.openembedded.dev * r35df67a0c1 10openembedded.git/recipes/qt4/ (qt4-embedded.inc qt4-x11-free.inc qt4.inc): qt4: adjust sed magic to recent QT to sanitize .pc files properly Mar 02 09:38:08 morning Mar 02 09:38:28 good morning Mar 02 09:40:38 vadmeste: how goes nhk? Mar 02 09:41:05 not very fine lol Mar 02 09:41:43 got image which boots to console? Mar 02 09:42:25 now I compiled a HelloWorld statically with my crosstool.. but it works with the prebuilt kernel image that is delivered by STEricc and not the kernel image generated by OpenEmbedded Mar 02 09:42:56 hrw, my machine is little old.. I don't prefer to compile the whole image Mar 02 09:45:24 Notice that the kernel image generated by OE works fine with the rootfs delivred by STEricc Mar 02 09:45:34 that sound strange Mar 02 09:52:25 i don't know but maybe the problem is related to MMU Mar 02 09:56:46 no.. the mmu is activated in the .config kernel file in OpenEmbedded Mar 02 09:57:10 maybe it's little endian or big endian problem... lool Mar 02 09:57:18 can anyone guess with me ? lol Mar 02 10:56:12 if I clean a native package that installs python stuff, it seems that packaged-staging removes the easy-install.pth file Mar 02 10:56:19 which makes all other python packages break Mar 02 10:56:26 *native python packages Mar 02 11:05:05 quick poll. How many aliases do you folks usually have at most in /etc/hosts on the target? awk 'BEGIN{i=0};{gsub("#.*","");if (NF>i){i=NF}};END{print i}' /etc/hosts Mar 02 11:06:00 (yes, it's on purpose that this includes the addr) Mar 02 11:07:27 blindvt`: usually no extras Mar 02 11:07:50 about the only thing I ever have in /etc/hosts is "127.0.0.1 localhost", and I'm not sure even that is really necessary anymore. Mar 02 11:08:13 I used to include the ipv6 entries, but "ip6-localhost" is so much more typing than "::1" that it seemed a waste of time :-} Mar 02 11:08:28 at home machine I have some but thats to avoid using dns server from other side of vpn Mar 02 11:08:42 and few for VM machines Mar 02 11:08:52 * blindvt` nods Mar 02 11:09:40 so a default of 3 should be ok. Thanks! Mar 02 11:26:33 hrw, my init stops with "Finished to configure packages.", I can't have a shell Mar 02 11:27:09 how to clean all packages but keep the native packages ? Mar 02 11:28:24 vadmeste: what HW you build on? Mar 02 11:28:36 nhk15 Mar 02 11:28:57 build machine I mean Mar 02 11:30:13 it's my laptop i686 core duo Mar 02 11:31:30 I did builds on slower machines Mar 02 11:31:55 it takes more than 7 hours Mar 02 11:31:58 vadmeste: do not be afraid of build full builds. Mar 02 11:32:12 vadmeste: night is usually longer then 7h so do build by night Mar 02 11:32:14 just base-image Mar 02 11:35:32 so it's not possible to clean all packages other the natives one ? :) Mar 02 11:36:30 it's a good feature especially for newbies Mar 02 11:37:03 newbie will try to build system itself as I did Mar 02 11:37:40 himself * Mar 02 11:38:00 for recipe in tmp/stamps/a*/*;do bitbake -cclean `basename $f|sed -e "s/\(.*\)\.do_fetch/\1/g"`;done Mar 02 11:38:17 something like this should clean all and armv5te Mar 02 11:38:30 * vadmeste sorry for the english.. don't know why english is still third language in his country Mar 02 11:38:35 emm Mar 02 11:39:47 thanks Mar 02 11:41:37 you don't think it's a good to add this functionality in OE ? Mar 02 11:42:23 no thoughts Mar 02 11:43:01 okey Mar 02 12:04:41 03Koen Kooi  07org.openembedded.dev * rf45f0d3495 10openembedded.git/recipes/pango/pango.inc: pango: convert pango.inc to use BBCLASSEXTEND for -native Mar 02 12:06:40 old packages should be remove when upgrading (if DEPENDS is manageable), shouldn't it ? Mar 02 12:07:28 there are many views Mar 02 12:08:00 i read the ML and the discussion, this here is related to a minor important recipe (libyaml) Mar 02 12:08:17 you maintain it? Mar 02 12:08:44 actually i look at the log and it is quite old (only rename from package/* listed) Mar 02 12:08:49 if yes then do what you feel right Mar 02 12:08:54 git log --follow filename Mar 02 12:08:55 just wanted to remove staging (since i use the recipe) Mar 02 12:09:23 I would do two steps then Mar 02 12:09:40 1. convert 2. upgrade (or add new) Mar 02 12:10:28 ok, did check packaged staging etc, and also could build DEPENDS since only 1 recipe is deps on it Mar 02 12:18:08 any reason why dbus has to have the x11 dependency, or in other words - why is x11 for dbus not optional? Mar 02 12:20:24 dbus-launch Mar 02 12:22:20 is that a requirement when using dbus or what is it? I am looking into the details right now Mar 02 12:22:41 its to do with tieing dbus sessions to x11 sessions Mar 02 12:23:07 its been so long I forget the details Mar 02 12:23:15 but on a system without x11 - that would not be needed? Mar 02 12:23:33 its a seperate package Mar 02 12:23:38 so I dont see the issue Mar 02 12:23:56 aah, so it just builds a lot more stuff which I do not have to install, I see Mar 02 12:24:10 I just saw it started to pull in x stuff and of course that scared me immediately Mar 02 12:24:17 common problem Mar 02 12:24:40 :) Mar 02 12:24:48 to build bluetooth enabled crap you need gstreamer so X11 so pulseaudio probably too etc Mar 02 12:24:52 we really should print that in big red letters :-) Mar 02 12:25:09 its no different in debian or redhat but people just dont seem to think Mar 02 12:26:01 ~curse classes/ for bb.data.getVar("sth", data, True) in one line, 'd, True' in other, 'd, 1' in other etc Mar 02 12:26:02 May the fleas of a thousand camels infest your most sensitive regions, classes/ for bb.data.getVar("sth", data, True) in one line, 'd, True' in other, 'd, 1' in other etc ! Mar 02 12:26:16 13:26 hrw@home:test$ cat tmp/legacy-staging.log Mar 02 12:26:16 /home/hrw/devel/oe/openembedded/recipes/fakeroot/fakeroot-native_1.12.4.bb Mar 02 12:26:22 nice ;) Mar 02 12:27:21 speaking of bluetooth, I heard its already submitted as a bug that usbhost machine features pulls in bluetooth support by force and you can't get rid of it? Mar 02 12:28:20 I have 8 usb bt dongles Mar 02 12:28:34 so any of my usbhost devices can also have BT Mar 02 12:28:47 why it should not get support for it if distro supports BT? Mar 02 12:29:01 prepare answer and I go eat soup Mar 02 12:29:05 * hrw -> food Mar 02 12:29:14 03Michael Lippautz  07org.openembedded.dev * r09bb1fc0bf 10openembedded.git/recipes/perl/libdbd-sqlite-perl_1.29.bb: libdbd-sqlite-perl: New recipe (Perl SQLite db backend) Mar 02 12:29:25 03Michael Lippautz  07org.openembedded.dev * r9fdf49ca44 10openembedded.git/recipes/libyaml/libyaml_0.1.2.bb: libyaml: Move to new staging Mar 02 12:29:57 hrw: I have a device that has usb but which due to its nature does not need any bluetooth support, so having bluez pulled into the rootfs just because I can attach a mass storage usb device is not really nice Mar 02 12:30:50 it's a samll box without a display that is running some dedicated software Mar 02 12:31:23 * XorA thinks hrw is making a usb microwave Mar 02 12:31:28 the stick can be used to automatically save log files, everything very simple Mar 02 12:50:32 Jin^eLD: i think it is even worse, it could well be that bluez4 pulls in gst which in turn pulls in pango and cairo Mar 02 12:50:45 indeed, it does pull in gst Mar 02 12:51:01 saw they get build if you build bluez, no idea if they end up in the rfs though Mar 02 12:51:05 so I ended using an own collection with a custom machine config where I drop usbhost from the featues Mar 02 12:51:20 doing the same for dbus x11 dependency now Mar 02 12:52:08 iirc even bitbake u-boot will result in gst and pango etc being build Mar 02 12:52:41 hmm no I don't think so.. why would u-boot depend on it? Mar 02 12:53:02 I Was building at91-bootstrap yesterday which uses u-boot iirc, and I did not have any gst/pango problems Mar 02 12:54:16 i might rememberd things wrong, Mar 02 12:55:11 btw anyone an idea how to handle this error: Mar 02 12:55:12 NOTE: Fetch svn://opkg.googlecode.com/svn;module=trunk;proto=http Mar 02 12:55:12 svn: Unrecognized URL scheme for 'http://opkg.googlecode.com/svn/trunk' Mar 02 12:56:13 I think I had this a year ago, but I do not have that tree around anymore to see how I solved it Mar 02 12:57:10 Jin^eLD: had a quick peek at u-boot indeed it does not depends on gst, it depends on libgfortran :-) Mar 02 12:57:19 :) Mar 02 12:57:44 gcc-cross seems to like it Mar 02 12:58:15 btw my svn is fairly recent Mar 02 13:00:40 well as gcc is also a fortran compilar Im not that surpsied Mar 02 13:11:31 XorA: i know Mar 02 13:14:50 Jin^eLD: think I found the svn issue, i rebuild a new one but the libs were not in the ld path Mar 02 13:15:46 oh.. Mar 02 13:15:55 ok that's something different from what I had back then Mar 02 13:16:15 Jin^eLD: task-base package does not fetch BT, task-base-extended does Mar 02 13:16:35 at91bootstrap is not related with uboot rather for building Mar 02 13:16:41 at91b loads uboot Mar 02 13:17:18 Jin^eLD: btw - which device you use? own one (which cpu?) or base for now on some devboard? Mar 02 13:17:20 hrw: there is some extra code in task-base.bb that looks at machine features and then enables bluetooth Mar 02 13:17:52 I know, I wrote it Mar 02 13:18:42 hrw: well - how do I get rid of BT support "the clean way"? :) Mar 02 13:20:51 hello: what is the status of qt embedded 4.6.2 in OE ? I have seen there is a recipe, but is someone using it ? Mar 02 13:21:13 Longfield: angstrom and shr is Mar 02 13:21:21 mickey|office: appears to be from his blog Mar 02 13:24:45 Jin^eLD: remove usbhost from machine - it adds only usbutils + few kernel modules Mar 02 13:24:54 I think that this is easiest Mar 02 13:25:41 hrw: thats what I did, but I had to put the machine conf into my overlay Mar 02 13:25:56 which is not a problem but still adds to configuration Mar 02 13:25:58 shit. my desktop gives me 'bus error' when I want to run anything Mar 02 13:26:21 Jin^eLD: that or task-base into overlay or create image recipe which base on task-boot Mar 02 13:27:38 hrw: sure it can be done and is not much of a problem, I was more surprised that there was no "default" way to configure it, similar to how one can just easily pick glibc or uclibc or whatever in the local conf Mar 02 13:29:16 Jin^eLD: task-base is not perfect but it was better then task-bootstrap which we used before. I think that getting something nicer would be very good thing. Mar 02 13:30:40 hrw: would it be an option to put the BT stuff into some FEATURES (distro or machine) ? Mar 02 13:31:03 i.e. have it separate from usbhost, it could still depend on usbhost but not the other way around Mar 02 13:31:20 Jin^eLD: it is separate Mar 02 13:31:55 well, not quite, because of the logic in task-base? Mar 02 13:32:05 if DISTRO_FEATURE contains 'bluetooth usbhost' and MACHINE_FEATURE contains 'usbhost' then task-base-extended (package) will have BT stuff added. Mar 02 13:32:23 remove bluetooth from DISTRO_FEATURES and you will not have bluez built Mar 02 13:32:35 ahaa, that would be the trick Mar 02 13:32:52 thanks, this is what I Was trying to find out, but missed somehow Mar 02 13:34:05 at least for task-base level cause other recipes can fetch bluez too Mar 02 13:35:19 right.. Mar 02 13:35:55 switching irc machine Mar 02 13:36:27 looks like time to resize rootfs again Mar 02 13:37:08 8.5/10GB on / used... and I use "/ /home" style Mar 02 13:59:46 EXT4_USE_FOR_EXT23 - nice thing for kernel space limited people Mar 02 13:59:57 hrw: thinking about doing a prototype of some sort of USER_FEATURES. any thoughts? right now its rather painful for a user to remove bluetooth.. Mar 02 14:01:50 Longfield: do you have an issue with it? Mar 02 14:02:36 kergoth: USER_FEATURES will overtake COMBINED/DISTRO ones? Mar 02 14:02:43 yeah, exactly Mar 02 14:03:03 * kergoth hasn't quite worked out the details, but we need to do something there Mar 02 14:04:58 zecke: well, not yet, building it now. But I have a system working fine with 4.6.0, but I needed a few patches that got in 4.6.1 (and not included by OE). I just wanted to see if some already had been using it for a while Mar 02 14:10:46 argh Mar 02 14:10:53 sheevaplug boots from microsd Mar 02 14:10:58 and that microsd started to die Mar 02 14:11:58 mmcblk0: error -110 sending read/write command, response 0x900, card status 0x900 Mar 02 14:12:02 shit Mar 02 14:17:09 hrw: new microsd and dd? ;-) Mar 02 14:18:19 so long for runnning Debian on sheevaplug Mar 02 14:18:39 erk Mar 02 14:27:06 SD card in sheeva is unusable as rootfs Mar 02 14:27:28 I can migrate to usb stick but how long will it work? Mar 02 14:28:56 but one thing is sure: jtag in miniusb rocks! Mar 02 14:34:44 hrw: usb stick in mine has been going solid for 6 months now Mar 02 14:36:39 hrw: use a 1.8" usb harddisk with sheeva Mar 02 14:37:34 13:41:41 up 61 days, 4 min, 2 users, load average: 0.00, 0.01, 0.00 Mar 02 14:37:36 eFfeM: I have 3.5" usb harddisk with sheeva Mar 02 14:37:52 1.5TB size which makes sheeva kernel die on fsck Mar 02 14:38:24 nevermind do I use angstrom or debian on it Mar 02 14:38:29 hrw: 3.5" is way too big and needs a separate power supply Mar 02 14:38:29 does it die with oom ? Mar 02 14:38:53 maybe split in a few partitions Mar 02 14:39:03 2.5 can be usb powered Mar 02 14:39:16 and is cheaper than 1.8 Mar 02 14:39:35 eFfeM: what are you doing using real measurements anyway? Mar 02 14:39:36 XorA: yes, but 1.8 is quite small (actually bumped into one a while ago) Mar 02 14:40:00 XorA: find me 1.5TB 2.5" in low price ;D Mar 02 14:40:13 XorA: some of my 2.5" enclosures need dual usb, probably for startup Mar 02 14:40:27 XorA: what do you mean with "using real measurements" Mar 02 14:40:43 eFfeM: it die with oops which is hard to catch because it is all what they say - no kernel output after 'Oops' Mar 02 14:40:56 eFfeM: your talking imperial, I thought you EU guys died when using imperial measures :-D Mar 02 14:41:56 XorA: for hard disks we still use imperial (and for a few other things too, I used to have a 66 cm TV, but the new one suddenly was 42" iso 1m07 or so Mar 02 14:42:26 most hd's are from abroad so guess that's where the sizes come from Mar 02 14:44:15 when I am emperor of the world metric shall be banned :-) Mar 02 14:44:33 hrw unaware of 2'5" 1.5T but 1T is eur 170 (Western Digital Scorpio Blue, 1TB SATA-300, 8MB) Mar 02 14:44:54 XorA: UK already lost that battle :-) Mar 02 14:45:30 eFfeM: we didnt, we use Imperial Mar 02 14:48:46 do you? last time i was in the uk the monetary denominators became rather metric Mar 02 14:49:08 * XorA orders 1pt at the bar, not 568ml Mar 02 14:49:15 * XorA orders meat in lbs Mar 02 14:49:20 sweets in quarters Mar 02 14:49:26 * eFfeM just orders a beer, always ok Mar 02 14:49:32 * hrw do same Mar 02 14:49:38 works in EU, works in UK Mar 02 14:49:47 * GNUtoo|oeee order a free beer Mar 02 14:49:56 *orders Mar 02 14:50:12 actually beer is the main advantageof imperial: 12% more compared to half liters Mar 02 14:50:15 milk in pts Mar 02 14:50:20 yuk Mar 02 14:50:35 I dont drink it in pints Mar 02 14:50:35 milk? fu! Mar 02 14:50:51 heh, you quaint northerners. down here we order milk in litres and meat in kilos. Mar 02 14:51:06 not that I personally drink a litre of milk in one sitting (or even in a week, probably) but still. Mar 02 14:51:31 * XorA is lucky to drink 1/2pt in a week Mar 02 14:54:19 * XorA wishes people would agree 35ml is saner for spirits than 25ml though Mar 02 14:55:13 spirits as alcohol? Mar 02 14:55:28 I usually get 0.5l vodka bottles if need Mar 02 14:55:47 0.7l and higher are usually also available but I do not know about smaller Mar 02 14:56:40 hrw: its just some pubs use 35ml measures and some 25ml Mar 02 14:56:45 * XorA preferes the ones with 35ml Mar 02 14:56:55 hrw, yeah smaller are the airline thingies Mar 02 14:57:37 eFfeM: or drinkers without money Mar 02 14:57:42 mickey|office: moin Mar 02 14:58:16 moin hrw Mar 02 14:59:21 * XorA spent 140 GBP on rum the other day :-D Mar 02 15:02:28 I'll go bye Mar 02 15:20:35 ~curse SD cards and SD controllers Mar 02 15:20:37 May you be reincarnated as a Windows XP administrator, SD cards and SD controllers ! Mar 02 15:22:13 hmm. Mar 02 15:23:53 hi, is there a way to tell bitbake inside of a recipe to not look for a checksum of single url? the download source i want to use does not provide a checksum... Mar 02 15:24:44 so add those checksums into recipe Mar 02 15:24:58 after fetch being done look into tmp/ dir Mar 02 15:26:19 * kergoth mulls over a prototype of postponed packaging Mar 02 15:29:31 kergoth: what's postponed packaging? Mar 02 15:29:37 (good morning, by the way) Mar 02 15:30:10 was thinking about testing doing the packaging at image creation time, only the bits necessary, for people not building feeds Mar 02 15:30:44 ah right, interesting. I guess that'd just be an adjustment to the task dependencies. Mar 02 15:31:28 yeah.. and pkgdata has all the real info, would need to make the pkgdata emission also store a reference to the do_install output in the persist_data or something Mar 02 15:31:42 one thing angstrom is hitting hard on is the amount of postinst that need to be run on target :-( Mar 02 15:31:48 :( Mar 02 15:32:03 yeah, those postinsts are a long-standing suckage. someone should really take that nettle by the horns and get rid of them. Mar 02 15:32:09 I need to take a look at them sometime Mar 02 15:32:12 calling it a day, cya tomorrow Mar 02 15:32:38 pb_: you are truly management now I see :-) Mar 02 15:32:46 absolutely Mar 02 15:39:12 03Koen Kooi  07org.openembedded.dev * ra32f13d02d 10openembedded.git/recipes/flex/flex_2.5.35.bb: flex: add 2.5.35 with new-style checksums, staging and class extension Mar 02 15:41:59 hmm. Mar 02 15:50:35 pb_, any opinion on http://patchwork.openembedded.org/patch/1625/ ? Mar 02 15:53:27 pb_, either that, a default of PN or we'll have sprinkle loads of name= into SRC_URIs Mar 02 15:53:49 s/have/have to/ Mar 02 15:59:54 Do we have virtual/libintl-native? Mar 02 16:00:04 Tartarus: no Mar 02 16:00:19 Tartarus: which causes issues with BBCLASSEXTEND Mar 02 16:02:16 And old-ass firefox, he Mar 02 16:02:17 h Mar 02 16:11:33 Tartarus: you asked for gcc 3.2/3.3 yesterday Mar 02 16:16:03 hrw: About, and then figure it out, kinda and posted the qemu thing Mar 02 16:16:07 Haven't checked that mbox yet tho Mar 02 16:18:39 Tartarus: iirc we added those in Poky just because we found it in different systems Mar 02 16:18:54 kind of that - do not remember details Mar 02 16:21:34 hrw: thanks for the hint with the checksum in the .bb files... finally i figured out... i didn't know that SRC_URI can be used to define checksums Mar 02 16:27:45 * Tartarus goes and confirms what deps are needed on ff 3.5.x vs 3.6 now Mar 02 16:39:43 Tartarus: You should use a recent qemu and drop the gcc 3.x checks ;-) Mar 02 16:43:38 hurray for dropping old qemu! :) Mar 02 16:43:57 :-) Mar 02 17:05:31 03Koen Kooi  07org.openembedded.dev * r9f1b3bdace 10openembedded.git/recipes/webkit/webkit-gtk_svn.bb: webkit-gtk: bump PR for flex-native change Mar 02 17:05:33 03Koen Kooi  07org.openembedded.dev * rf8b1ae61d0 10openembedded.git/recipes/gtk-webcore/midori_0.2.3.bb: midori: add 0.2.3 Mar 02 17:13:01 ok, who tried to make working system *without* busybox? Mar 02 17:20:05 03Marcin Juszkiewicz  07org.openembedded.dev * reaa2d77d79 10openembedded.git/recipes/bison/ (4 files): Mar 02 17:20:05 bison: merged Poky changes (BBEXTENDCLASS + new staging) Mar 02 17:20:05 Signed-off-by: Marcin Juszkiewicz Mar 02 17:20:08 03Marcin Juszkiewicz  07org.openembedded.dev * re329d5e021 10openembedded.git/recipes/file/ (5 files): Mar 02 17:20:09 file: merged Poky changes (BBEXTENDCLASS + new staging) Mar 02 17:20:09 Note that 5.04 (newest) is now default instead of 4.21. Mar 02 17:20:09 Signed-off-by: Marcin Juszkiewicz Mar 02 17:20:09 03Marcin Juszkiewicz  07org.openembedded.dev * rf2f02f6e47 10openembedded.git/classes/base.bbclass: Mar 02 17:20:09 base.bbclass: create tmp/legacy-staging.log file with names of recipes which need work Mar 02 17:20:10 Signed-off-by: Marcin Juszkiewicz Mar 02 17:20:23 hrw: legacy-staging.log.. nice. Mar 02 17:20:30 thx Mar 02 17:20:35 :) Mar 02 17:28:09 erk.. the TERMINFO env var can break ncurses-native Mar 02 17:28:10 fun Mar 02 17:36:53 03Tom Rini  07org.openembedded.dev * r31e9385425 10openembedded.git/classes/ (base.bbclass sanity.bbclass): (log message trimmed) Mar 02 17:36:53 qemu: Perform our sanity checks based on ENABLE_BINARY_LOCALE_GENERATION Mar 02 17:36:53 Perform qemu-related checks not based on if we're ARM but based on if Mar 02 17:36:53 we'll be using qemu for binary locale generation. Clarify what the Mar 02 17:36:53 first of these sanity checks does. Next, change the check for a Mar 02 17:36:54 provided qemu binary to be generic enough to work on all arches (and catch Mar 02 17:36:55 distribution or user built versions of qemu). While we're in here, correct Mar 02 17:38:39 03Martin Jansa  07org.openembedded.dev * r4798f0d51e 10openembedded.git/recipes/dri/libdrm_git.bb: Mar 02 17:38:39 libdrm: use the same OECONF, even when gta01 won't use glamo api Mar 02 17:38:39 Signed-off-by: Martin Jansa Mar 02 17:38:42 03Martin Jansa  07org.openembedded.dev * r5a1d574bba 10openembedded.git/recipes/linux/ (3 files in 2 dirs): Mar 02 17:38:42 linux-openmoko-2.6.32: add last 2 patches from DRM branch" Mar 02 17:38:42 Signed-off-by: Martin Jansa Mar 02 17:38:42 03Martin Jansa  07org.openembedded.dev * rcf44d7790c 10openembedded.git/recipes/shr/libphone-ui-shr_git.bb: Mar 02 17:38:42 libphone-ui-shr: drop etk dependency Mar 02 17:38:42 Signed-off-by: Martin Jansa Mar 02 17:38:43 03Martin Jansa  07org.openembedded.dev * r0da9a5aef6 10openembedded.git/recipes/xboard/xboard_4.4.1.bb: Mar 02 17:38:43 xboard: add checksums Mar 02 17:38:43 Signed-off-by: Martin Jansa Mar 02 17:38:44 03Martin Jansa  07org.openembedded.dev * r36550d9fa8 10openembedded.git/recipes/e17/e-wm_svn.bb: Mar 02 17:38:44 e-wm: limit illume-keyboard-flow.patch to rev 46549, it was applied in 46550 Mar 02 17:39:39 Signed-off-by: Martin Jansa Mar 02 17:41:47 have a nice rest of day Mar 02 17:51:54 03Tom Rini  07org.openembedded.dev * re1b0916de8 10openembedded.git/recipes/eglibc/ (5 files): (log message trimmed) Mar 02 17:51:54 eglibc-package.bbclass: Add eglibc-pic, resync eglibc-dev with glibc, bump PR Mar 02 17:51:54 Resync what goes into eglibc-dev based on what we have in glibc-dev. This Mar 02 17:51:54 includes picking up the reset of the static libraries (many are missed Mar 02 17:51:54 otherwise and static linking in SDKs fails). We also add in a eglibc-pic Mar 02 17:51:55 package to catch all of the pic stuff (some of which would otherwise be caught Mar 02 17:51:56 in -dev). Mar 02 17:52:00 03Tom Rini  07org.openembedded.dev * r2d132de5ab 10openembedded.git/classes/glibc-package.bbclass: Mar 02 17:52:00 Add glibc-pic package Mar 02 17:52:00 glibc-package.bbclass: Add glibc-pic package Mar 02 17:52:00 By default, in our glibc builds we don't have these libraries. They may Mar 02 17:52:00 however come from various binary toolchains, or custom source builds. Mar 02 17:52:01 Signed-off-by: Tom Rini Mar 02 17:52:02 Acked-by: Khem Raj Mar 02 18:02:40 03Koen Kooi  07org.openembedded.dev * raadaa1b0bf 10openembedded.git/recipes/file/file.inc: file.inc: attempt to create proper file.inc that got missing in the earlier file commit Mar 02 18:06:28 03Koen Kooi  07org.openembedded.dev * rb540bf3525 10openembedded.git/recipes/bison/bison.inc: bison.inc: attempt to create proper file.inc that got missing in the earlier bison commit Mar 02 18:09:26 03Koen Kooi  07org.openembedded.dev * reae7f83b03 10openembedded.git/recipes/ (bison/bison.inc file/file.inc): bison/file.inc: INC_PR -> INCPR Mar 02 18:32:36 Anyone else wanna review the firefox changes I posted? Mar 02 18:39:01 hello, i have learnt that opie uses qt embedded for GUI rendering, qt embedded draws directly to the framebuffer. In opie when i am only in the terminal mode(without gUI and i try to run drawpad app. then i get the error no qt embeded server seems to be running Mar 02 18:39:01 how come it is utilising a server like X11, theorotically it shouldn't do that Mar 02 19:45:08 03Marcin Juszkiewicz  07org.openembedded.dev * r5ab6358096 10openembedded.git/recipes/file/file.inc: Mar 02 19:45:08 file: added forgotten files Mar 02 19:45:08 Signed-off-by: Marcin Juszkiewicz Mar 02 19:45:11 03Marcin Juszkiewicz  07org.openembedded.dev * rb7374a40f2 10openembedded.git/recipes/bison/bison.inc: Mar 02 19:45:11 bison: added forgotten files Mar 02 19:45:11 Signed-off-by: Marcin Juszkiewicz Mar 02 20:14:37 What's the correct "MACHINE" setting for openembedded/bitbake for the Gumstix Overo "Fire" module? Mar 02 20:16:10 Is it "overo.conf"? Mar 02 20:20:35 hmm, a buddy of mine is triyng to build openssl on his quadcore, and he gets an error "Makefile is older than Makefile.org, Configure or config.", my first guess was to disable parallel make in the recipe, but it is already disabled... :) any ideas? Mar 02 20:22:29 robtow: should be just plain "overo" Mar 02 20:23:11 03Sebastian Spaeth  07org.openembedded.dev * r732af689fa 10openembedded.git/ (2 files in 2 dirs): Mar 02 20:23:11 sane-srcrevs.inc: bump shr-related recipes to current state Mar 02 20:23:11 Signed-off-by: Sebastian Spaeth Mar 02 20:30:18 Jin^eLD: interesting does system have proper timing ... Mar 02 20:35:51 mlip2 - thanks! Mar 02 20:37:29 khem: was my first guess too, but as we kept digging it turns out he had an alias set for make with some custom MAKEOPTS, and I am pretty sure this was the problem Mar 02 20:37:31 testing right now Mar 02 20:38:10 I guess whitelisting variables is not enough to ensure a good environment :) you'll always find something new Mar 02 20:49:15 Jin^eLD: OK cool Mar 02 20:49:41 Jin^eLD: yeah once I had a dir in my path which was on a server across the globe Mar 02 20:49:47 and I thought bash sucked :) Mar 02 20:51:42 :)) Mar 02 20:51:52 it builds now, it was his alias thing :) Mar 02 21:01:34 * Tartarus begs for someone, that perhaps uses / cares about firefox to ack his patches :) Mar 02 21:01:43 other than khem of course Mar 02 21:06:40 Tartarus: how can I see a full copy of a file from a older commit ID Mar 02 21:06:53 Tartarus: git show commit-id gives me the diff Mar 02 21:10:01 Hmm Mar 02 21:10:04 I'm not that git-great Mar 02 21:10:09 But the hard way would be Mar 02 21:10:15 git checkout hash-before-the-diff Mar 02 21:10:25 Must be some git-tree-something to do just one file Mar 02 21:14:04 mhm, now he is getting the QA issues with .la files as described here http://www.mail-archive.com/angstrom-distro-devel@linuxtogo.org/msg02877.html, just that his TMPDIR is ok (not the same slash problem) Mar 02 21:14:24 I wonder what the differences between the systems are, it built just fine on my box Mar 02 21:21:34 ok... Mar 02 21:21:38 oops wring window :) Mar 02 21:37:03 which script run.do... takes care of sedding .la files for staging in native packages? Mar 02 21:37:42 the sysroot one? Mar 02 23:30:37 Good afternoon everyone. I have a quick question. I followed the steps given by Koen to build an angstrom image using bitbake for Beagleboard. After about 15 hours, the compilation ended successfully, but I don't seem to find the required files MLO, U-boot and uImage. Where should I be looking for them? Mar 02 23:31:38 Sri_, something like: oe/tmp/deploy/glibc/images/beagelboard/ Mar 02 23:32:11 not so sure about the MLO file though Mar 02 23:32:25 The only folders I got in my root OE folder are: angstrom-dev, build, download and openembedded Mar 02 23:33:15 i suppose the tmp dir usually used for building in has been put elsewhere. Perhaps you should ask in an angstrom channel? Mar 02 23:33:22 ok. inside angstrom-dev/, I have a folder called rootfs. Should this be corresponding to the new root file system for Beagle board that I need to install? Mar 02 23:33:42 Tried that... no one seems to be active there. :-( Mar 02 23:33:57 and looking in /tmp/deploy ? Mar 02 23:34:19 in /build of course... Mar 02 23:34:59 or /build/tmp/angstrom/deploy ? Mar 02 23:35:18 mc is your friend Mar 02 23:35:30 Sri_, its definitely a deploy dir you want. find . -name deploy -a -type d Mar 02 23:35:47 ant__ There is nothing even remotely related to Angstrom/OE in the /tmp folder. And the only thing inside build is: build/conf/local.conf Mar 02 23:35:58 let me try that Mar 02 23:36:32 build/tmp or tmp, not /tmp, Sri_ Mar 02 23:36:53 Yes.... there... I got it..... Mar 02 23:37:28 OE/angstrom-dev/deploy/glibc/images/beagleboard Mar 02 23:37:51 Thanks ppl... let me get some stuff up and running now.... Mar 02 23:38:24 Good day all of you. Mar 02 23:38:36 bye Mar 03 00:06:10 Tartarus: its git show :filename Mar 03 00:11:48 ah nice Mar 03 01:18:39 Hmm Mar 03 01:18:46 There is another stable planned for sometime, right? Mar 03 01:19:19 I'm starting to wonder if perhaps when we make the next stable, the next commits to .dev do some of the radical redesigns that've been needed for a long time Mar 03 01:19:50 Like reworking our "this is what we build against" stuff to work right away out of the box, no funky flags required with sysroot Mar 03 01:20:18 And "this stuff we build against" == "this stuff in our packages, no exceptions" (maybe modulo some munging) Mar 03 01:20:43 And some other stuff too, no doubt Mar 03 01:22:47 Tartarus: having a stable once a year is prolly ok Mar 03 01:23:08 Yeah, when is the next one is what I wanna know :) Mar 03 01:25:34 (And, if we have the stomach, so to speak, to do some massive breakage in dev, when there's a not stale stable) Mar 03 01:28:14 Tartarus: right, I think with git it should be easy to get in big things done on branch and pour into master without much breakage prolly Mar 03 01:28:36 I like the idea of keeping it building all the time as it helps in bisects Mar 03 01:33:41 "all" ? Mar 03 01:34:12 I think we might need to set a specific subset of stuff, make sure that works, push it Mar 03 02:04:42 I just did an initial OE/bitbake compile of a kernel for Overo. I got out of it uImage-2.6.29-r51-overo.bin. Isn't there a later kernel? **** ENDING LOGGING AT Wed Mar 03 02:59:58 2010