**** BEGIN LOGGING AT Wed Dec 05 02:59:56 2007 Dec 05 07:41:45 good morning all Dec 05 07:46:57 koen: morning Dec 05 07:49:00 koen: morning :) Dec 05 07:54:00 wooooot.... Dec 05 07:54:48 morning Dec 05 07:57:01 * hvontres|home found a way to get the kernel to ignore the stupid sharp bootloader's command line :) Dec 05 07:58:03 XorA: http://gizmodo.com/gadgets/lego/almost+perfect-lego-costume-allows-man-to-get-anonymously-drunk-to-pieces-329629.php Dec 05 07:58:58 koen: awesome Dec 05 08:01:44 * CoreDump wonders why spitz's rescue FS is ext2 and not jffs* Dec 05 08:10:19 probably because jffs2 was a pile of crap in 2001 Dec 05 08:13:45 RP:ping Dec 05 08:17:05 RP: I updated http://bugs.openembedded.org/show_bug.cgi?id=3397 with a new patch set. It now filters out the evile cmdline :) Dec 05 08:33:10 * * OE Bug 3422 has been created by koen(AT)openembedded.org Dec 05 08:33:12 * * ERROR: task-base-dbg is listed in PACKAGES multiple times Dec 05 08:33:14 * * http://bugs.openembedded.org/show_bug.cgi?id=3422 Dec 05 08:33:22 * * OE Bug 3416 has been RESOLVED (FIXED) by Dec 05 08:33:24 * *  apex-env-1.5.8-r0-do_compile Dec 05 08:33:26 * * http://bugs.openembedded.org/show_bug.cgi?id=3416 Dec 05 08:33:34 * * OE Bug 3422 has been RESOLVED (FIXED) by koen(AT)openembedded.org Dec 05 08:33:36 * *  ERROR: task-base-dbg is listed in PACKAGES multiple times Dec 05 08:33:38 * * http://bugs.openembedded.org/show_bug.cgi?id=3422 Dec 05 08:33:46 * * OE Bug 3419 has been marked as DUPLICATE of bug 3288 by koen(AT)openembedded.org Dec 05 08:33:48 * *  qmake2-native is broken due to target dependance Dec 05 08:33:50 * * http://bugs.openembedded.org/show_bug.cgi?id=3419 Dec 05 08:33:58 * * OE Bug 3423 has been created by autobuild(AT)openembedded.org Dec 05 08:34:00 * * gcc-cross-4.1.2-r13-do_configure Dec 05 08:34:02 * * http://bugs.openembedded.org/show_bug.cgi?id=3423 Dec 05 08:34:06 * * Flood control triggered, clearing queue... Dec 05 08:34:10 * * [39] pending messages removed. Dec 05 08:37:44 03koen 07org.oe.dev * r4debfc11... 10/ (1 packages/glibc/glibc-package.bbclass): glibc-package.bbclass: merge LOCALE_UTF8_ONLY from poky Dec 05 08:37:49 03koen 07org.oe.dev * r7e45814c... 10/ (1 conf/distro/include/angstrom.inc): angstrom.inc: only use utf8 locales Dec 05 08:43:51 n8 all Dec 05 09:31:47 hi! Dec 05 09:32:31 hi :) Dec 05 09:54:56 morning Dec 05 09:55:10 koen: I see that you read my blog post Dec 05 09:55:52 yo hrw Dec 05 09:57:03 hi hrw Dec 05 09:59:16 good morning Dec 05 09:59:24 03koen 07org.oe.dev * r575e3741... 10/ (6 files in 2 dirs): linux-rp: add default_prefs to get poodle to use 2.6.17 Dec 05 10:00:35 http://imgs.xkcd.com/comics/python.png Dec 05 10:02:27 hrw: you need to fix your URLs on planet.linuxtogo.org Dec 05 10:07:08 the people at http://www.phoronix.com/scan.php?page=article&item=934&num=3 need to follow a course in plot making Dec 05 10:07:38 XorA: or find person which manage it Dec 05 10:07:51 ~blame koen Dec 05 10:07:52 * ibot blames koen (and Canada) for all the evil in the world Dec 05 10:09:41 koen: typical marketing "Its faster man, heres some pictures" Dec 05 10:12:00 XorA: http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2007-December/001062.html Dec 05 10:13:00 koen: cant think of any now, the x configuraition is correct, the feeds are currently wrong Dec 05 10:13:18 so bitbake -s give the correct output? Dec 05 10:13:24 koen: yes Dec 05 10:13:28 koen: and images build Dec 05 10:13:32 ok Dec 05 10:13:55 I think there was a window between two commits that allowed them to become unversioned for probably about 2 hours at max Dec 05 10:13:55 if noone speaks up I can start rsyncing the current build from scratch Dec 05 10:14:02 just happened to be the hours I tested Dec 05 10:16:01 hrw, RP: how safe is it to merge the qemu from poky into OE? Dec 05 10:16:24 hrw: thanks for LOCALE_UTF8_ONLY Dec 05 10:16:34 koen: qemu itself should be safe, it works ok in our tests. I'd hold off the QEMU_OPTIONS stuff though Dec 05 10:16:54 RP: oe's glibc-package.bbclass has some multilib fixes Dec 05 10:18:19 koen: ok thanks, I'll keep that in mind Dec 05 10:18:50 RP: what would you think of a qemu-build.bbclass that assembles a chroot from packaged-staging and builds crappy apps that run generated binaries? Dec 05 10:20:02 koen: I suspect we'll need such a class eventually but I don't know if using a chroot is going to work Dec 05 10:20:07 * koen wonders how to mark postinsts as arch-specific, machine-specific or non-specific Dec 05 10:20:21 koen: overrides ;-) Dec 05 10:20:55 my latest thought was an extra if clause in the preamble Dec 05 10:21:19 we have very few machine specific postinsts afaik Dec 05 10:21:30 blueprobe and dropbear are the only ones that come to mind Dec 05 10:21:31 Just don't add the postinst for packages that don't need it Dec 05 10:21:55 we could run the rest thru qemu for supported archs Dec 05 10:22:35 ah, you mean postinst that have to run on device rather than just having problems with cross arch Dec 05 10:22:44 right Dec 05 10:23:37 RP -- thanks again for your help last night. Dec 05 10:23:44 RP -- got it working. Dec 05 10:23:48 cdm: np :) Dec 05 10:24:04 seems to be working well -- gonna have a lot more people using OE now. :) Dec 05 10:24:13 things like e.g. locales and gdk-pixbug can run in qemu for qemu supported archs Dec 05 10:24:18 koen: That will probably have to be some kind of blacklist Dec 05 10:25:11 koen: Pre booting images under qemu-system emulation might be easier for postinsts Dec 05 10:25:26 RP: my plan was to set QEMURUN=true in env and have the offending postinsts check that Dec 05 10:26:25 koen: its one solution... Dec 05 10:26:36 RP: I'm not familiar with qemu terminology, but running postinsts and compiling obnoxious software would be a nice feature Dec 05 10:26:57 not to mention site-info autogeneration :) Dec 05 10:27:21 koen: agreed, it just needs someone with some free time Dec 05 10:28:41 I'm probably going to have to slow down or risk burnout :/ Dec 05 10:28:52 good idea Dec 05 10:29:12 Hopefully over time we can help out in real-time Dec 05 10:29:20 burnouts are only OK in races :) Dec 05 10:29:26 * cdm has to get irc ports opened up at work Dec 05 10:30:02 productivity drops with 90% Dec 05 10:31:22 We are doing some interesting stuff around booting and faster boot times. Dec 05 10:34:21 * koen goes all zen and says "the fastest way to boot is not to boot at all" Dec 05 10:36:46 koen: no problem Dec 05 10:37:20 hrw: I hope we can iron out the updater.sh problems for all machine before the release Dec 05 10:39:29 o yes Dec 05 10:39:46 the fastest way to boot is to run /bin/sh Dec 05 10:39:47 :) Dec 05 10:40:43 cdm: especially on headless machine Dec 05 10:42:50 :) Dec 05 11:04:27 * koen looks at http://trac.webkit.org/projects/webkit/changeset/28389 Dec 05 11:07:23 * Crofton|home looks at http://etd.vt.edu/submit/index.html Dec 05 11:07:59 koen, know much about directfb? Dec 05 11:08:04 * * OE Bug 3426 has been created by xydarcher(AT)hotmail.com Dec 05 11:08:06 * * build angstrom with unecessary packages. Dec 05 11:08:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3426 Dec 05 11:08:17 I know that the directfb dudes needs to eat way more steak Dec 05 11:08:31 and according to zecke also should stop smoking pot Dec 05 11:08:38 heh Dec 05 11:09:04 pot smoking vegatarians Dec 05 11:09:26 the local vegatarian restaurant used drug profits for start-up capital Dec 05 11:09:32 allegedly Dec 05 11:12:01 hi rob_w Dec 05 11:12:44 do you have 5 mins for me? Dec 05 11:25:18 Hi, somebody knows why this error? Dec 05 11:25:20 FATAL: oe_runmake failed Dec 05 11:27:25 my crystal ball is in repairs ATM, can't help you Dec 05 11:27:29 ~smartquestions Dec 05 11:27:30 well, smartquestions is http://www.catb.org/~esr/faqs/smart-questions.html Dec 05 11:35:17 esr is so elitist Dec 05 11:39:27 hi Dec 05 11:41:05 hello Dec 05 11:41:16 likewise: you use cairo with the image backend on ixp4xx, right? Dec 05 11:41:56 koen: exactly Dec 05 11:42:03 Is there a way to change the execution order of start scripts ? Most of the apps use update-rcd with defaults and no number so they default to 20 Dec 05 11:42:50 steliosk: yes, by adding numbers to them :) Dec 05 11:43:07 steliosk: and talk to XorA, he looked into initscripts a while ago Dec 05 11:43:28 the current plan is to write a .bbclass that can work out script dependencies Dec 05 11:45:04 Would it make sense to follow another's distro's numbering ? i.e debian for example Dec 05 11:47:57 hi Dec 05 11:48:27 i wanted to ask, if oe is a community project, or is it controlled by any company Dec 05 11:48:34 could'nt find this information Dec 05 11:48:45 steliosk: we want e.g. X to come up as early as possible, debian completes all initscripts first Dec 05 11:48:49 paines: community Dec 05 11:48:53 re Dec 05 11:49:02 steliosk: but yet, it would make sense to look at it for inspiration Dec 05 11:49:38 XorA: okay. thanks Dec 05 11:50:26 koen : Maybe add a bb. file that would call update-rcd and set the numbering to what a specific distro is doing i.e debian-update-exec oe something like that Dec 05 11:51:11 or a bbclass would be better for this kind of task ? Dec 05 11:56:53 XorA: btw, image build time has been cut down due to the UTF8 changes Dec 05 11:57:04 less time for coffee now ;) Dec 05 11:57:15 koen: I only ever generate enlglish US and UK anyway Dec 05 12:00:30 heh Dec 05 12:00:43 "do you have a patch to crosstools for that toolchain" Dec 05 12:04:49 !seen Spyr1 Dec 05 12:04:50 Spyr1 (n=ian@i-195-137-94-162.freedom2surf.net) was last seen parting #oe 22 days, 12 hours, 11 minutes ago stating "{}". Dec 05 12:11:31 XorA: how goes psplash unbreaking? Dec 05 12:38:06 http://pastebin.com/m21160375 - new angstrom build-release.sh script (not tested at all) Dec 05 12:39:52 looks much nicer Dec 05 12:40:21 will be easier to add notification about success/failure of build etc Dec 05 12:40:49 seppuku/tinderclient? Dec 05 12:42:50 sort of Dec 05 12:43:25 echo "$target built for $machine
">>autobuilder.html Dec 05 12:43:31 for example Dec 05 12:47:28 echo "schedule build foo-image" || nc Dec 05 12:47:37 s/||/|/ Dec 05 12:49:52 if [ "$BUILD_CLEAN" != "" ] Dec 05 12:50:01 this is 8th line in proper way Dec 05 12:54:06 http://pastebin.com/m3b22813a Dec 05 12:54:08 another Dec 05 12:55:01 ignore indents of course Dec 05 12:59:08 koen: what do you think? Dec 05 12:59:20 looks good Dec 05 12:59:56 will you test it? Dec 05 13:00:17 commit it and I will test it Dec 05 13:02:53 ok Dec 05 13:03:09 ~seen jbowler Dec 05 13:03:11 jbowler was last seen on IRC in channel #oe, 667d 18h 48m 14s ago, saying: 'k'. Dec 05 13:03:26 2 years was a good estimate Dec 05 13:03:41 !logs Dec 05 13:03:41 Channel logs for #oe are archived at: Dec 05 13:03:42 http://hentges.net/tmp/logs/irc/%23oe Dec 05 13:03:43 Live-logs are available at Dec 05 13:03:44 http://hentges.net/tmp/logs/irc/livelogs/%23oe.livelog Dec 05 13:03:46 See ?? help-logs for usage instructions Dec 05 13:07:04 * * OE Bug 3426 has been RESOLVED (INVALID) by koen(AT)openembedded.org Dec 05 13:07:06 * *  build angstrom with unecessary packages. Dec 05 13:07:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3426 Dec 05 13:08:44 koen: pushed Dec 05 13:09:42 I should make a directfb image Dec 05 13:09:51 once we have a few more things building with it Dec 05 13:12:13 Crofton|home: indeed Dec 05 13:12:51 I need an image that uses the LCD on teh OSK Dec 05 13:13:00 the x11 ones are to large to flash Dec 05 13:13:15 I think the overwrite u-boot when you load them into RAM :) Dec 05 13:13:41 I will pull those images from the supported list in a minute Dec 05 13:14:25 how much flash you have? Dec 05 13:14:39 in OSK? 32M Dec 05 13:14:57 the image should fit in flash, but I think u-boot is loading u-boot in a "bad" place Dec 05 13:16:08 uboot puts the entire image in ram for some weird reason Dec 05 13:16:30 well, I copy image into ram via tftp, then copy from RAM into flash Dec 05 13:16:41 the tftp load fails for larger images Dec 05 13:16:57 for now, we will support the console images for angstrom Dec 05 13:17:20 until I (or hopefully someone else) gets interested enough to solve the problem :) Dec 05 13:17:24 you can select address for which uboot tftp load iirc Dec 05 13:17:47 Crofton|home: I hope CELF gives up access to the testlab Dec 05 13:17:49 yes, but I think the the problem is u-boot loads into the middle of RAM .... Dec 05 13:18:00 I'll read that email later Dec 05 13:20:19 global warming is over, we have snow today Dec 05 13:24:29 koen: I talked with Tim about this during ELCE, I can poll him for this. Dec 05 13:25:14 likewise: check you gmail :) Dec 05 13:25:32 Crofton|home: which tftp server is on your host? some do not cope with transaction id number roll-over. Dec 05 13:26:06 sweet, binary localegen no longer takes 2 hours Dec 05 13:27:05 whatever came with fedora Dec 05 13:32:06 for now, I'll keep it out of the distributed images until we figure out why it won't flash Dec 05 13:38:24 * hrw -> afk Dec 05 13:39:44 in the bootmisc script we run ldconfig to update the dynamic libary case. I believe this is not need it when using uclibc Dec 05 13:39:46 03hrw 07org.oe.dev * rfa7d169d... 10/ (1 contrib/angstrom/build-release.sh): Dec 05 13:39:46 build-release.sh: rewritten Ångström autobuilder script Dec 05 13:39:46 - one function for building Dec 05 13:39:46 - uclibc is switchable in one place Dec 05 13:39:46 - simple function to give info about success of build Dec 05 13:39:51 03crofton 07org.oe.dev * r13898b09... 10/ (1 contrib/angstrom/build-release.sh): angstrom/build-release.sh : Remove graphics support from omap5912osk until flashing large image problem solved. Dec 05 13:46:02 * koen reads http://www.compulab.co.il/x270em/download/x270-em-linux-doc.pdf Dec 05 13:46:15 " OpenEmbedded configuration files used to build the Angstrom Linux image." Dec 05 13:47:11 koen: Initiating world dominance Dec 05 13:47:18 sequence started Dec 05 13:48:41 likewise: http://www.linux4sam.org/twiki/bin/view/Linux4SAM/OpenEmbeddedAngstrom Dec 05 13:49:22 hopefully more companies will try angstrom when we have a (stable) release branch Dec 05 13:50:07 not having builds crap out because of changes in .dev is pretty key Dec 05 13:50:40 also being able to manage vendor changes Dec 05 13:51:08 which reminds me, I need to pick through the gumstix stuff again Dec 05 14:00:44 Crofton|home: that's why companies spend effort on choosing a stable revision, make it more stable, then release products from them. This takes long, and by the time of their release, they are massively behind, and none of their work can be ported over. Classical embedded linux/world syndrom. Dec 05 14:01:17 koen: did I hear stable release is planned? :-) Dec 05 14:01:30 sakoman_: for angstrom, yes Dec 05 14:01:38 for OE we need some more infrastructure Dec 05 14:01:58 We love angstrom -- a good start! Dec 05 14:02:10 But oe.stable would be really wonderful Dec 05 14:02:50 there will be an angstrom release branch before christmas Dec 05 14:03:07 sakoman_: under what email address are you reachable nowadays? Dec 05 14:03:19 03koen 07org.oe.dev * r845f49b0... 10/ (1 contrib/angstrom/build-release.sh): build-release.sh: remove console-image from ixp4xx builds Dec 05 14:05:26 koen: answered privately Dec 05 14:07:03 the mail should be in your inbox Dec 05 14:08:49 hmmm, now the ad on my blog is for a "Thesis Consultant" Dec 05 14:14:01 Crofton: Google is trying to tell you something? :) Dec 05 14:14:21 Crofton|home: How did it go yesterday? Dec 05 14:18:12 http://www.balister.org/~balister/card.jpg Dec 05 14:18:13 :) Dec 05 14:18:18 got the signatures Dec 05 14:18:28 some revisions, but allegedly not to bad Dec 05 14:18:37 I need to look at them later' Dec 05 14:18:53 congrats! Dec 05 14:18:57 thanks Dec 05 14:19:15 Crofton|home: congrats :) Dec 05 14:20:19 Crofton|home: hard work pays off, congratulations! Dec 05 14:22:05 * * OE Bug 3346 has been RESOLVED (FIXED) by likewise(AT)gmx.net Dec 05 14:22:07 * *  Upstream GCC C compiler bug Dec 05 14:22:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3346 Dec 05 14:22:17 * * OE Bug 2993 has been RESOLVED (FIXED) by likewise(AT)gmx.net Dec 05 14:22:19 * *  eglibc: export-fcntl2.patch no longer needed since svn rev 3468 Dec 05 14:22:21 * * http://bugs.openembedded.org/show_bug.cgi?id=2993 Dec 05 14:22:28 * * OE Bug 2163 has been RESOLVED (FIXED) by likewise(AT)gmx.net Dec 05 14:22:30 * *  glib-2.0 python variable expansion does not work Dec 05 14:22:32 * * http://bugs.openembedded.org/show_bug.cgi?id=2163 Dec 05 14:22:34 I'm still unable to make openmoko-devel-image from scratch (make clean). In do_install of pax-utils-native it tries to mkdir /usr/src/armtoolchain. Dec 05 14:22:41 * * OE Bug 2365 has been RESOLVED (FIXED) by likewise(AT)gmx.net Dec 05 14:22:43 * *  uclibc: can't handle reloc type 0x49 in lib '/lib/libc.so.6' for powerpc/angstrom/uclibc Dec 05 14:22:45 * * http://bugs.openembedded.org/show_bug.cgi?id=2365 Dec 05 14:22:58 Crofton|home: So you'll have all this free time now? :) Dec 05 14:23:11 heh Dec 05 14:23:17 my todo list is still long Dec 05 14:23:18 the /usr/src is owned by root and way outside $OMDIR Dec 05 14:23:31 and I need to convince people to pay me to make it shorter Dec 05 14:24:26 * RP knows this feeling... Dec 05 14:25:38 If an armtoolchain from another project existed in /usr/src, would OE find and try to use it? Dec 05 14:25:54 DukeOfURL: only if you told it to Dec 05 14:26:01 haven't Dec 05 14:26:11 is it in PATH? Dec 05 14:26:37 has anyone tried a build from scratch lately? Dec 05 14:26:44 Are you after OE to use it or not? Dec 05 14:27:03 DukeOfURL: Depends how you deinfe lately. I made on at the weekend... Dec 05 14:27:34 doing a build from scratch now Dec 05 14:27:40 yes, probably in $PATH Dec 05 14:28:35 boa (the webserver) needs an entry in the /etc/hosts file to work. The entry looks like 'hostname' 127.0.0.1 127.0.0.1 . I was thinking of adding it during the boa install but do you think it would be good to have an entry like that by default in the rootfs ? Dec 05 14:28:58 it maps the hostname to the localhost ip Dec 05 14:29:43 it could be generated by base-files Dec 05 14:29:52 steliosk: How about using the "localhost" name? Dec 05 14:30:12 That can break other software. Dec 05 14:31:39 broonie: hostname, localhost or both? Dec 05 14:32:17 hostname->127.0.0.1 mapping. Dec 05 14:32:51 right Dec 05 14:33:26 steliosk: I can't give you a concrete reason but it also makes me nervous Dec 05 14:34:20 The failure mode is that things that want to resolve the hostname to an IP address end up resolving the localhost address and using it where it is inappropriate to do so. Dec 05 14:34:27 * XorA shudders at the hostname -> 127.0.0.1 Dec 05 14:34:45 preferable solution, fix boa Dec 05 14:35:26 boa will only need this workaround while resolution of hostname is broken; unfortunately that's likely to be a common case for embedded kit :( Dec 05 14:35:32 * broonie nods XorA Dec 05 14:35:42 RP: Thank you. Removing the other armtoolchain from $PATH allows openmoko-devel-image to proceed. Dec 05 14:35:54 well for angstrom you can always use hostname.local as the hostname as avahi is up and running Dec 05 14:37:05 boa doesnt happen to be a redhat project? Dec 05 14:37:09 * RP suspects relying on avahi being present is a bad idea Dec 05 14:37:45 boa is using gethostbyname to try and resolve things Dec 05 14:37:55 Right; it needs to cope with gethostbyname() failing. Dec 05 14:44:08 broonie : hmmm the only thing i can think of right now is to add a non caching dns like dnsmasq to run before boa and set up name resolution. I haven;t seen any patces to resolve the gethostbyname() Dec 05 14:44:24 are there any ? Dec 05 14:45:15 I've no idea about the boa codebase in particular, sorry. It should be fairly straightforward to write something, though. Dec 05 14:49:01 ~curse intel 855 chipset Dec 05 14:49:02 May the fleas of a thousand camels infest your most sensitive regions, intel 855 chipset ! Dec 05 14:50:23 intelfb does not works (bios tells that only ~800K ram for gfx is available), vesafb works nice but if you start X11 on intel driver then too often it means "goodbye framebuffer" :( Dec 05 14:50:53 and 1024x768 on 20" wide lcd looks a bit weird Dec 05 15:13:04 * * OE Bug 3427 has been created by likewise(AT)gmx.net Dec 05 15:13:06 * * Add LED classCPU activity trigger Dec 05 15:13:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3427 Dec 05 15:36:20 broonie : I think i found a way to solve the issue . By entering Servername "127.0.0.1" in the boa.conf it starts fine :) Dec 05 15:36:41 and works with whatever ip box has Dec 05 15:37:17 Ideally it'd be able to use that as a fallback if normal resolution failed. Dec 05 15:41:06 c Dec 05 15:55:01 bye! Dec 05 16:19:04 * * OE Bug 2124 has been RESOLVED (FIXED) by r.farina(AT)adelphia.net Dec 05 16:19:06 * *  Angstrom on akita can't login via gpe-login/xdm Dec 05 16:19:08 * * http://bugs.openembedded.org/show_bug.cgi?id=2124 Dec 05 16:21:23 hmm Dec 05 16:21:25 Crofton|home: ping Dec 05 16:26:39 re Dec 05 16:27:05 Hmm Dec 05 16:27:13 Can PR be overriden in local.conf? Dec 05 16:27:15 mickeyl: is expect part of python 2.5 or an external module? Dec 05 16:27:28 (I'm trying to add something like a build number to every package) Dec 05 16:27:43 koen: external, there's python-pexpect iirc Dec 05 16:28:08 Hi Dec 05 16:28:25 mickeyl: did you see xkcd.com today? Dec 05 16:29:03 koen: I see you made extra sure which kernel gets into poodle images :) Dec 05 16:29:38 koen: not yet. chekcing... Dec 05 16:29:42 koen: I was thinking of you when I saw that xkcd :) Dec 05 16:30:10 koen: hahaha, i love that Dec 05 16:30:12 :D Dec 05 16:31:59 It seems that bug 3363 is also hitting gcc-cross-4.2.2. Are you also suffering it? Dec 05 16:33:25 Not exactly the same problem, but fortran related though Dec 05 16:38:45 "Wow plays fine on my current machine, and it would be nice to play it on a laptop, but for $4500.00 I can pay a sherpa to carry my desktop where ever I go." Dec 05 16:39:36 koen: But what about the second sherpa for the Generator Dec 05 16:39:49 dunno, I don't play wow :) Dec 05 16:40:22 koen: heh. Real Geeks(tm) play nethack Dec 05 16:40:34 don't play that either :) Dec 05 16:41:01 I haven't played a computer game in like a year or so Dec 05 16:41:19 * koen checks is parallels vm now supports 3d Dec 05 16:41:23 ah, it does Dec 05 16:41:46 ~curse mac games not being universal binaries Dec 05 16:41:47 May the fleas of a thousand camels infest your most sensitive regions, mac games not being universal binaries ! Dec 05 16:41:50 koen: hehe... well the last game I played was Candyland with my 3 year old :) Dec 05 16:42:15 and I lost Dec 05 16:43:27 ^^ Dec 05 16:45:57 What should the variable ${layout_exec_prefix} expand to? Dec 05 16:47:22 It gives me nothing under an interactive bitbake session, and I think that's the source of my problem with gcc-cross, as --with-mpfr=${STAGING_DIR_NATIVE}${layout_exec_prefix} resolves to --with-mpfr=${STAGING_DIR_NATIVE} and hence, not able to find libmpfr Dec 05 16:50:22 does anyone know a cheap source for a serial/usb controlled poweroutlet? Dec 05 16:50:38 230V Dec 05 16:52:15 Probably replacing layout_exec_prefix with layout_libdir should be enough, but would be great to understand what I'm doing actually :-) Dec 05 16:52:36 I would like to find ATX PSU with more sata connectors Dec 05 16:58:44 * koen only has 2 sata disks Dec 05 16:59:01 well, 3, if you count the one in my laptop Dec 05 17:05:08 I have 2 Dec 05 17:05:23 and I am out of connectors because someone decided to put both on one wire Dec 05 17:19:40 koen@bitbake:~/OE/monotone/org.openembedded.dev/packages/linux$ tail -f ~/OE/build/autobuilder.log Dec 05 17:19:41 minimal-image (glibc) built for ep93xx Dec 05 17:19:41 console-image (glibc) built for ep93xx Dec 05 17:20:52 nice Dec 05 17:21:29 build from scratch Dec 05 17:21:29 now give small rss for it and it will fly ;d Dec 05 17:21:40 :) Dec 05 17:21:59 koen: btw - do not import qemu from Poky yet - I disabled new version again Dec 05 17:22:44 koen: btw - ever heard of buildbot? you get www interface for builder automatically Dec 05 17:23:01 I don't like buildbot Dec 05 17:23:08 and I lack the skill to fix tinderclient Dec 05 17:24:02 I'd like to ask to someone to take a look at bugs: #3371 #3372 and #3373 (all them has patches included) Dec 05 17:24:24 I'd like to send some more patches that I've here but would be nice to have those in, first Dec 05 17:25:36 I have heard of buildbot Dec 05 17:25:47 I suspect you would like, if you are a python due Dec 05 17:25:53 make that python dude Dec 05 17:26:17 Crofton|work: I already uesd buildbot and it works really nice :-) Dec 05 17:26:26 I was able to run buildbot without going into python Dec 05 17:26:28 buildbots biggest flaw is that you have to tell it what to do Dec 05 17:26:38 Crofton|work: I've also used bitten that integrates fine on Trac too Dec 05 17:26:39 I think it is ok for simple things Dec 05 17:26:45 bitten? Dec 05 17:26:56 Crofton|work: yes. It's a plugin to Trac ... Dec 05 17:27:09 * Crofton|work makes a note to look into that Dec 05 17:27:12 Crofton|work: I think that buildbot suits better for OE but bitten is nice too Dec 05 17:27:39 Crofton|work: bitten.cmlenz.net IIRC Dec 05 17:27:44 koen: for your autobuilder you also have to Dec 05 17:28:35 hrw: but all clients continie building the buildmaster is offline Dec 05 17:28:43 which you can't do with buildbot Dec 05 17:28:52 it's just way to centralized for my liking Dec 05 17:29:01 koen: yes ... but buildmaster shouldn't be offline ;-) Dec 05 17:29:41 koen: well, in an ideal world it wouldn't even build itself but just coordinate it all Dec 05 17:30:27 the purpose of the OE tinderbox is that people can join in when they want Dec 05 17:30:35 you can't do that with buildbot Dec 05 17:31:51 I know Dec 05 17:31:52 koen: you'd need to provide an way to "register" on the master. But just it ... Dec 05 17:32:02 koen: wouldn't it be enough? Dec 05 17:32:26 otavio: buildbot tell slaves what do build. tinderbox listen and show what people build Dec 05 17:32:37 otavio: totally other things Dec 05 17:32:41 hrw: ah, in inversed Dec 05 17:32:52 hrw: inversed way of view things Dec 05 17:33:13 hrw: yes. master does coordinate it all Dec 05 17:33:24 buildbot is nice thing if you want to do repeatable builds Dec 05 17:33:35 tinderbox is just to show what people build Dec 05 17:33:36 So could someone take a look on those three bugs? Dec 05 17:35:43 ok Dec 05 17:37:13 otavio: #3371 doesn't apply Dec 05 17:37:50 b/packages/openrdate/ is not a valid path Dec 05 17:38:04 koen: I took care about 3371 Dec 05 17:38:57 03mickeyl 07org.oe.dev * r0fb6c6e1... 10/ (5 files in 4 dirs): python 2.5.1 enable and package ctypes module/package. closes #3368 -- thanks emdete! Dec 05 17:39:00 3372 doesn't apply as well Dec 05 17:39:27 hmm i guess i fixed more bug entries this week than together in one year Dec 05 17:39:37 koen: openrdate is a new path Dec 05 17:39:39 need to do this regularly Dec 05 17:39:52 otavio: and the patch creates the wrong path Dec 05 17:39:59 koen: using patch -p1? Dec 05 17:40:25 mickeyl: there was idea of BSW... Dec 05 17:40:43 otavio: why? patches are always in p0 format Dec 05 17:40:58 koen: not mines since I export them from GIT Dec 05 17:41:07 * koen isn't going to change his patch apply script Dec 05 17:42:00 koen: are you talking serious? Dec 05 17:42:06 otavio: we prefer 'mtn diff' patchhes Dec 05 17:42:06 yes Dec 05 17:42:28 hrw: I'm not confortable with monotone and very used to GIT Dec 05 17:42:41 hrw: would be too hard to apply it and commit? Dec 05 17:42:59 it just depend on how much time does developer prefer to spend Dec 05 17:43:09 hrw: that's crazy. I'm not asking anybody to change anything just to apply a diff and commit Dec 05 17:44:00 almost all patches inside of OE recipes are -p1 and koen wonders about this cuple of patches ... Dec 05 17:44:05 sorry but it looks insane Dec 05 17:44:20 patching file firefox-2.0.0.3/mozconfig Dec 05 17:44:20 can't find file to patch at input line 290 Dec 05 17:44:27 mtn diffs are p0 Dec 05 17:44:37 we want mtn diffs Dec 05 17:44:40 how's that hard? Dec 05 17:45:41 koen: to me, a lot. I don't ues mnt and neither has an update database of it. I export the patches from GIT and use stgit to remove them once them are included on mainline and that makes easy to handle with the patch queue while they aren't included mainline Dec 05 17:45:52 koen: I'm not used to it Dec 05 17:45:57 we use mtn for OE Dec 05 17:46:15 if you want to make it difficult for yourself by using git, that's your choice Dec 05 17:46:18 koen: and since I have not commit right, I thought a patch would be enough Dec 05 17:46:41 don't go whining to OE developers if they don't want to go to extra trouble Dec 05 17:47:19 koen: :) Dec 05 17:47:41 that's not the problem. If the patch is broken I can work it out and prepare a new one but just a p0 to p1 worring is crazy, imho Dec 05 17:47:57 hrw: let me check the firefox one Dec 05 17:48:42 otavio: its in already Dec 05 17:48:56 otavio: I had local change Dec 05 17:50:13 hrw: ah ok nice Dec 05 17:51:08 hrw: http://people.debian.org/~otavio/OE/patches/cmake-add-2-4-7 http://people.debian.org/~otavio/OE/patches/linux-rp-do-not-fail-if-the-ke and http://people.debian.org/~otavio/OE/patches/xkbdata-add-1-0-1 are ready too Dec 05 17:52:23 hrw: want me to report bugs for those too or you could apply them? Dec 05 17:52:30 i wrote my first true bbclass :') Dec 05 17:52:46 it was a bit hard , but it's full of joice :o) Dec 05 17:52:49 http://people.debian.org/~otavio/OE/patches/linux-rp-do-not-fail-if-the-ke is nice Dec 05 17:52:54 otavio: report Dec 05 17:53:24 this part of code I would like to have in kernel.bbclass instead ;) Dec 05 17:55:40 hrw: but it then would need to change linux-rp completely to use this from kernel.bbclass wouldn't it? Dec 05 17:56:18 yes Dec 05 17:56:25 but I am one of linux-rp maintainer Dec 05 17:56:26 s Dec 05 17:56:34 hrw: since koen had problems to apply #3372 patch, could you apply it too? Dec 05 17:56:47 hrw: OK. I could take a look and prepare something next days Dec 05 17:56:57 otavio: 3372 is in OE now Dec 05 17:57:07 hrw: not closed yet Dec 05 17:57:09 or not.. Dec 05 17:57:15 3372 is libxcb one... Dec 05 17:57:18 I missed it Dec 05 17:57:21 hrw: have you commited it? Dec 05 17:57:22 hrw: ok Dec 05 17:57:25 hrw: please do Dec 05 17:57:34 hrw: and thanks Dec 05 17:57:55 hrw: once I finish this prototype I can take a look on the kernel.bbclass change Dec 05 17:58:28 hrw: If you get a chance, could jyou take a look at Dec 05 17:58:57 hrw :http://bugs.openembedded.org/show_bug.cgi?id=3397 and let me know what you think? Dec 05 17:59:55 looking at kernel.bbclass I think other linux recipes would be affected, no? Dec 05 18:02:15 hvontres|work: packages/linux/linux-rp-2.6.17/poodle_serial_vcc.patch is nice. can you refresh it for git head and submit to lakml? Dec 05 18:02:22 03otavio 07org.oe.dev * r1a6625be... 10/ (15 files in 10 dirs): firefox: fix firefox 2.0.0.3 building - close #3372 Dec 05 18:02:22 03mickeyl 07org.oe.dev * r99925725... 10/ (7 files in 3 dirs): ecore|evas cvs update Dec 05 18:07:41 mickeyl: http://bugs.openembedded.org/show_bug.cgi?id=3418 - can you? Dec 05 18:10:21 *assigning* Dec 05 18:17:14 hrw: Is there a how-to for that somewhere? This is my first voyage into kernel stuff Dec 05 18:18:21 hvontres|work: linux/Documentation/SubmittingPatches Dec 05 18:18:43 hrw: thanks. Dec 05 18:18:53 christ the openwrt guys are almost as unpleasant as we are Dec 05 18:19:15 I was having a density related issue with trx versus bin files :) Dec 05 18:21:21 hrw: I'd like to try to get the kexec related stuff pushed soon so that the autobuilder can pick it up. It makes root-on-SD systems work pretty well Dec 05 18:22:32 Crofton|work: which device you want to flash? Dec 05 18:22:51 wrtsl54 Dec 05 18:22:54 I have it done now Dec 05 18:23:09 trx are bin + header iirc Dec 05 18:23:24 yeah Dec 05 18:23:47 and header is used by webinterfaces to figure out if it is ok to flash trx Dec 05 18:23:48 Crofton|work: install x-wrt web frontend Dec 05 18:24:06 I think I am going again Dec 05 18:24:09 Crofton|work: and some devices require that header (not wrt ones) Dec 05 18:24:14 kk didn;t work on the wrtsl54 Dec 05 18:24:29 x-wrt is great frontend for openwrt Dec 05 18:25:04 * * OE Bug 3368 has been RESOLVED (FIXED) by mlauer(AT)vanille-media.de Dec 05 18:25:07 * *  python 2.5 is not build with ctypes on arm Dec 05 18:25:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3368 Dec 05 18:25:21 I tend to get the routers set up then leave them alone forever ... Dec 05 18:33:15 hello all Dec 05 18:33:22 hi likewise Dec 05 18:34:56 hrw: hello, how's life in .pl Dec 05 18:35:15 likewise: good, thx Dec 05 18:35:29 likewise: but do not ask me about weather - I work at home Dec 05 18:35:53 hrw: for the reason of extremely bad weather? Or are you enjoying the sun? :-) Dec 05 18:36:02 Let me guess... Dec 05 18:36:25 snow or storms? Dec 05 18:38:48 likewise: and I have redecoration at home now Dec 05 18:38:52 hi Lorn Dec 05 18:39:23 hi Crofton Dec 05 19:02:13 03hrw 07org.oe.dev * r8d6f2bff... 10/ (1 packages/pimlico/contacts.inc): contacts: added libowl dependency - close #3377 Dec 05 19:02:29 03mickeyl 07org.oe.dev * r2d067cd2... 10/ (1 packages/efl1/evas_cvs.bb packages/efl1/ewl_cvs.bb): evas|ewl cvs explicitly enable the software-x11-16 engine now that it's more stable Dec 05 19:02:36 03hrw 07org.oe.dev * rc2307b4d... 10/ (1 packages/pimlico/dates.inc): dates: added libowl dependency - close #3417 Dec 05 19:04:46 bye Dec 05 19:16:02 hi! Dec 05 19:28:05 * * OE Bug 3371 has been RESOLVED (FIXED) by Dec 05 19:28:07 * *  openrdate recipe Dec 05 19:28:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3371 Dec 05 19:28:17 * * OE Bug 3373 has been RESOLVED (FIXED) by Dec 05 19:28:19 * *  fix firefox 2.0.0.3 building Dec 05 19:28:21 * * http://bugs.openembedded.org/show_bug.cgi?id=3373 Dec 05 19:30:04 * * OE Bug 3377 has been RESOLVED (FIXED) by Dec 05 19:30:06 * *  contacts-0.7-r1-do_compile Dec 05 19:30:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3377 Dec 05 19:30:16 * * OE Bug 3417 has been RESOLVED (FIXED) by Dec 05 19:30:18 * *  dates-0.4.4-r2-do_compile Dec 05 19:30:20 * * http://bugs.openembedded.org/show_bug.cgi?id=3417 Dec 05 20:06:37 re Dec 05 20:23:04 * * OE Bug 3428 has been created by andrea.adami(AT)gmail.com Dec 05 20:23:06 * * preferred xserver-kdrive-imageon for c7x0 Dec 05 20:23:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3428 Dec 05 20:51:46 Khem: you were right, I can't reproduce yesterday's problem on a build from scratch. Dec 05 21:13:03 ~seen RP Dec 05 21:13:07 rp is currently on #maemo (1d 18h 47m 49s) #handhelds (1d 18h 47m 49s) #oe (1d 18h 47m 49s) #openmoko (1d 18h 47m 49s) #openzaurus (1d 18h 47m 49s). Has said a total of 76 messages. Is idling for 6h 39m 41s, last said: 'steliosk: I can't give you a concrete reason but it also makes me nervous'. Dec 05 21:14:42 pH5: cool Dec 05 21:18:28 Khem: thanks anyway. sorry to have bothered you with that :) Dec 05 21:19:38 p Dec 05 21:19:54 pH5: no problem :) Dec 05 21:42:20 Hi. I trying to create small distro for my ARM-based hardware, but configuring of packet glibc-intermediate-2.5 fails because of missing arm-linux-gcc. gcc-cross-initial built succesfully. Can anybody help or point me proper documentation? Dec 05 21:46:00 jekhor: what happens if you perform bitbake "gcc-cross" Dec 05 21:55:02 likewise, building of glibc-intermediate-2.5-r8 failed with same reason Dec 05 21:55:41 | *** These critical programs are missing or too old: gcc Dec 05 21:56:06 and in configure's output: | checking for arm-linux-gcc... ccache arm-linux-gcc -march=armv4t -mtune=arm920t Dec 05 21:58:22 jekhor: did you build existing distro's before you tried this? Dec 05 21:59:46 yes, openmoko builds normally Dec 05 22:07:10 jekhor: I cannot help you any further, haven't tried specifying my distro yet. Dec 05 22:07:27 :) Dec 05 22:08:22 ok, I will read other simple distros configs... Dec 05 22:08:24 jekhor: Have you took one of the existing distro's as a basis? Dec 05 22:08:32 s/took/taken Dec 05 22:10:38 Yes, but I not sure :) It was some days ago. Dec 05 22:36:27 hi Dec 05 22:37:04 out of curiosity...how many GB of disk are needed for the complete tree Dec 05 22:37:14 ant built with autobuilder ? Dec 05 22:38:24 mmhh... koen is not here... Dec 05 22:40:52 in wiki/GettingStarted I read 35GB for world...actual? Dec 05 22:41:10 That wouldn't surprise me. Dec 05 22:45:28 Only for c7x0 minimal + console + altboot-console + x11 + x11-gpe + x11-pimlico = 8.4 GB Dec 05 22:45:32 [arklinux@localhost oe]$ du -sh Dec 05 22:45:32 8.4G . Dec 05 22:45:50 with sources Dec 05 22:48:47 BTW sources are only 361M Dec 05 22:49:42 all above only for c7x0 Dec 05 22:52:55 disks are cheap, and keeping some old tmp dirs can be useful Dec 05 23:07:17 newbie question: I'm seeing this message when trying to build minimal-image: Dec 05 23:07:19 ERROR: Cannot satisfy the following dependencies for task-boot Dec 05 23:07:23 base-passwd busybox initscripts modutils-initscripts netbase udev sysvinit sysvinit-pidof tinylogin Dec 05 23:07:35 morning all Dec 05 23:08:27 yet I've got ipk's for all those packages Dec 06 01:21:21 bye Dec 06 01:51:40 03mickeyl 07org.oe.dev * r84fabcc1... 10/ (4 files in 3 dirs): Dec 06 01:51:40 expedite: add hack patch to disable all unused engines since Dec 06 01:51:40 the PKG_CHECK_EXISTS macro seems to always return "yes" Dec 06 01:51:45 03mickeyl 07org.oe.dev * rc017ace9... 10/ (1 packages/e17/expedite_cvs.bb): expedite: add dependency on eet Dec 06 02:18:04 * * OE Bug 3248 has been RESOLVED (FIXED) by mlauer(AT)vanille-media.de Dec 06 02:18:07 * *  agg.bb needs a -D Dec 06 02:18:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3248 Dec 06 02:31:43 03mickeyl 07org.oe.dev * rd76f8fca... 10/ (1 packages/e17/expedite_cvs.bb): expedite: ship everything it needs in one package **** ENDING LOGGING AT Thu Dec 06 03:00:00 2007