**** BEGIN LOGGING AT Tue Jun 16 02:59:57 2009 Jun 16 05:06:47 i am trying to boot a console-base-image. it hangs after the message Configuring ppp. any idea what could be wrong? Jun 16 06:05:03 morning all Jun 16 06:05:25 booxter: I'm trying to compile qt 4.5 r0 Jun 16 06:06:12 maybe I need, on pc, qmake 4.5 installed ? Jun 16 06:08:22 recalcati_oe: qt sources shd come with qmake in the bin directory? Jun 16 06:09:50 the reason is that moc_qbuffer.cpp is create as qt 4.4.3 Jun 16 06:10:44 log http://www.mibbit.com/pb/tculXP Jun 16 06:11:05 i am trying to boot a console-base-image. it hangs after the message Configuring ppp. any idea what could be wrong? Jun 16 06:11:58 rzw9tj: stable/2009 branch ? Jun 16 06:12:17 no i think master branch Jun 16 06:13:35 any idea what could be wrong, i have no clue, im not a master in reading /etc contents ;-) Jun 16 06:13:49 recalcati_oe: please compile qt4 recipes found in .dev/.stable Jun 16 06:14:27 switch to stable one Jun 16 06:15:00 do i need to recompile all my stuff if i switch to stable one Jun 16 06:15:24 booxter: hi, I normally do like this. bitbake -b openembedded/recipes/qt4/qt4-embedded_4.5.0.bb Jun 16 06:15:50 commit d3f30fd19b2be315421377fcd10585ba30982647 Jun 16 06:19:53 maybe some dependency lacks ? http://www.mibbit.com/pb/HWkxLr Jun 16 06:21:27 I think that 4.5.0 depends on 4.4.3 qt4-tools Jun 16 06:25:17 recalcati_oe: fresh git copy? Jun 16 06:25:51 yes, I didn't modify by myself resipes Jun 16 06:26:35 now I'm compiling bitbake -b recipes/qt4/qt4-tools-native_4.5.0.bb Jun 16 06:26:43 recalcati_oe: there is no qt4-embedded-4.5.0 recipe in the tree. I think you use old git tree Jun 16 06:26:52 recalcati_oe: update it Jun 16 06:27:01 recalcati_oe: there is qt4-embedded-4.5.1 already Jun 16 06:27:33 I'm in stable/2009, you mean this branch ? Jun 16 06:28:11 recalcati_oe: ah, I see Jun 16 06:28:43 git diff http://www.mibbit.com/pb/xhqcXX Jun 16 06:31:58 I could get the only 4.5.1 recipe from dev branch Jun 16 06:32:20 recalcati_oe: ok, stable branch doesn't include lots of qt4 updates. I think we should push them from .dev... Jun 16 06:33:27 git checkout origin/org.openembedded.dev recipes/qt4 Jun 16 06:33:55 I'm not sure if I can checkout some files directly from remote branch Jun 16 06:34:04 recalcati_oe: there was massive reconstruction of recipes in .dev - by Jeremy Laine and I - you can try to copy qt4 directory (and some needed classes) from .dev to your overlay and try it there. If it works, send [STABLE] patch to ML Jun 16 06:34:28 recalcati_oe: I'm not a good gitter - no advice here:) Jun 16 06:34:49 recalcati_oe: but there were also some significant changes in classes... Jun 16 06:34:51 I do it, thx Jun 16 06:34:56 recalcati_oe: qt classes I mean Jun 16 06:35:17 I hope not to have dependencied outside qt4 Jun 16 06:35:29 recalcati_oe: recalcati_oe openembedded qt bbclasses I mean :))) Jun 16 06:36:19 recalcati_oe: some deps (like uicmoc and qmake) are grabbed by obsolete bbclasses. Update them too to reflect the new recipes layout Jun 16 07:13:23 booxter: strange, I need DirectFb to compile qt 4.5.1 . Not nice. Or there is some important reason for it? Jun 16 07:19:03 recalcati_oe: just a build time dependency Jun 16 07:19:14 recalcati_oe: qt4 embedded has directfb gfx plugin Jun 16 07:24:02 ok, now its compiling, thx Jun 16 07:24:15 good morning Jun 16 07:24:26 mpoullet: gm Jun 16 07:25:20 good morning Jun 16 07:26:10 recalcati_oe: sx: hi Jun 16 07:27:31 Does anyone know what's the name of the "pthreads (POSIX Threads)" package in OE?, I must add pthreads in my angstrom fs :) Jun 16 07:28:17 mpoullet, hi Jun 16 07:33:39 sx: it comes bundled with glibc Jun 16 07:33:51 good morning Jun 16 07:45:44 i switched from unstable to stable branch, i get Error, DISTRO_PR has changed (.2 to .1) which means all packages need to rebuild. Jun 16 07:46:09 do i really need to rebuild? Jun 16 07:52:08 good morning all Jun 16 08:01:24 rzw9tj: not absolutely _need_, no. you could either reset DISTRO_PR to .2 by hand, or just disable the error. Jun 16 08:02:24 I don't really know why that's an error, actually. It seems fairly bogus to me. Jun 16 08:02:28 I have trouble booting into the console-image. I'm using qemu. It boots up allright but it does not give a login prompt. Any thought of what might be wrong ? Jun 16 08:03:06 sgh: it could be almost anything. what's the last thing you see on the screen? Jun 16 08:03:47 good morning Jun 16 08:04:03 morning Jun 16 08:04:04 pb_: I search for "^upstream" when I look for undocumented patches, your recent commits will still show Jun 16 08:04:12 I haven't tried console-image myself, but base-image does work for me with MACHINE=qemux86, DISTRO=micro-uclibc Jun 16 08:04:18 BTW, tinderbox is about to move Jun 16 08:04:20 Laibsch: ah, you don't like the hashes? Jun 16 08:04:23 after we get a new database Jun 16 08:04:25 I guess I can remove them Jun 16 08:04:32 pb_: please do so Jun 16 08:04:37 okay Jun 16 08:04:47 I have written a page on the wiki on what I do Jun 16 08:05:03 search for "upstream" and you should find it even in crappy MW search Jun 16 08:05:34 Crofton, mickey|dinner: Can you update DNS again shortly? Jun 16 08:06:12 how do i rebuilt without building the toolchain Jun 16 08:06:21 BTW, tinderbox will restart from one. We can leave the old Tinderbox website up, but it should not accept any new submissions. I'd have loved to migrate the old logs, but decided against it. two reasons Jun 16 08:06:42 1) old TB stored data in postgresql, so it will be a bit of work to convert Jun 16 08:06:51 2) data is dropped after 100 days anyways Jun 16 08:07:10 I did not feel keeping this data for just three month justified the effort Jun 16 08:07:13 yeah, seems like moving the data would be more trouble than it's worth Jun 16 08:08:41 pb__: sory it is the minimal-image. The last thing I see is this : http://imagebin.ca/view/1H2DStw.html Jun 16 08:09:06 03Phil Blundell  07org.openembedded.dev * r658930dd64 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@new.openembedded.org:openembedded into org.openembedded.dev Jun 16 08:09:15 03Phil Blundell  07org.openembedded.dev * r05855dc272 10openembedded.git/recipes/ (3 files in 3 dirs): binutils, irda-utils, libxcb: remove extraneous hashes that were disrupting Laibsch's activities Jun 16 08:10:06 sgh: can you inspect the rootfs to see what's meant to be happening next? Jun 16 08:10:20 it might be getting stuck in some initscript, or it might just be that init is failing to launch a getty for some reason Jun 16 08:10:41 obviously, one thing to check is that inittab does actually call for getty on tty1 or whatever Jun 16 08:11:01 are you using DISTRO=minimal? Jun 16 08:14:04 Hi, me and my Via C3 box again Jun 16 08:14:22 pb__: no Angstrom.2008.1 Jun 16 08:14:27 After getting past gcc-cross (thanks again, pb_), it is now openssl that fails Jun 16 08:14:28 http://tinderbox.openembedded.net/builds/159314/ Jun 16 08:15:09 sgh: ah, I don't really know much about angstrom. I think the dudes who do hang out in #angstrom mostly. Jun 16 08:15:41 pb__: should I try the minimal distro ? Jun 16 08:15:59 sgh: "micro" is what I use, could try that one. Jun 16 08:16:18 or yeah, minimal might be worth a go as well Jun 16 08:16:44 Laibsch: that link seems to be for a libqpe-opie build failure Jun 16 08:16:55 qpeapplication.h:130: error: multiple parameters named 'argv' Jun 16 08:17:02 that's just weird, but, well, crazy opie Jun 16 08:17:55 pb__: will do. Thanks Jun 16 08:24:34 bitbake still wants to build bluez-libs 3.33 and 3.36 for me :-( Jun 16 08:33:11 pb_: you're right of course, that fails, too Jun 16 08:33:21 The link I meant to post was http://tinderbox.openembedded.org/builds/159273/ Jun 16 08:34:48 I think the solution is straight-forward, I just need a pointer in the right direction Jun 16 08:36:06 hello Jun 16 08:37:43 I have git pulled the stable branch yesterday, removed ./tmp completely, and bitbaked opie-image DISTRO angstrom-2008.1 for MACHINE h1940 . by the end I got opkg-cl consuming 100% of CPU and hanging in there until infinite. if i kill opkg-cl, bitbaking of course fails.. Jun 16 08:40:46 that is a known bug Jun 16 08:41:04 there are two ways to fix it Jun 16 08:41:27 a) properly, but that is hard, especially in the stable branch Jun 16 08:41:33 b) quick fix Jun 16 08:42:26 sledgeas: you probably want http://patchwork.openembedded.org/patch/716/http://patchwork.openembedded.org/patch/716/ which is the quickfix Jun 16 08:43:15 booxter: ping Jun 16 08:44:03 bitbake qt4-embedded . compiled succesfully inside stable/2009 but getting recipes/qt4 from dev branch. Now I try it Jun 16 08:44:25 recalcati_oe: good Jun 16 08:44:35 recalcati_oe: ping me when you will test it Jun 16 08:53:10 k Jun 16 08:53:10 ok Jun 16 08:56:50 htop Jun 16 09:01:25 thank you Laibsch, it worked flawlessly and w/o rebuilds Jun 16 09:02:10 htop Jun 16 09:02:53 sledgeas: can you please do "bitbake -g opie-image" for me and pastebinit the output of "grep blue *.dot" somewhere? Jun 16 09:03:26 surething Jun 16 09:13:20 Laibsch: http://pastebin.mandriva.com/13127 Jun 16 09:17:56 Laibsch: ahh, the looks like the same issue I have seen with Bluez4 building Bluez-data as part of it's own big Bluez4 package but tasks still prefering to build the old bluez3 bluez-data as it is a seperate package. Jun 16 09:19:22 DJWillis: yes, bluez generally needs a big clean-up Jun 16 09:19:47 I'm not interested in bluetooth, so I've just ripped it out of all my builds Jun 16 09:19:52 Laibsch: I looked into it, got scared, and ran away ;-). Jun 16 09:19:55 There are a lot of compilation issues, too Jun 16 09:24:29 amazing, i have assembled a new linux-2.6.21-h1940 (not patchworked it yet though) and oe has automatically chosen it (latest) for the opie-image h1940 machine; great! Jun 16 09:26:46 sledgeas: the reason opkg was hanging is that bluez is broken Jun 16 09:26:55 there are bluez4 and bluez 3.x packages Jun 16 09:27:08 bitbake had to build both for some reason Jun 16 09:27:30 and then they started fighting over who installs files they both want to install Jun 16 09:27:40 opkg is then asking you what it should do Jun 16 09:27:50 but you never answered ;-) Jun 16 09:29:09 Laibsch: you are right, I am being warned about several recipes providing the same bluez. !same about opkg[-gnupg etc.] (even though I have wrote PREFERRED_PROVIDER_opkg = "opkg" in local.conf) and another package, too Jun 16 09:29:34 no, that is related but not the same issue Jun 16 09:30:02 ok Jun 16 09:30:05 You could very well install a ton of packages all providing virtual-japanese-font for example Jun 16 09:30:10 they would not conflict Jun 16 09:30:50 because i am bitbaking for this specific machine h1940 prolly is a machine-based issue, since not all of the bitbakers have it(?) Jun 16 09:31:01 why there are 2 bluezes being emerged.. Jun 16 09:32:50 it is not machine-specific Jun 16 09:33:03 and there are a number of people reporting it Jun 16 09:33:24 there is also a bug report in the tracker Jun 16 09:33:33 because i had this bug since the beginning of may Jun 16 09:33:42 wow Jun 16 09:33:46 i mean Jun 16 09:33:54 could very well have been in there since beginning of may Jun 16 09:33:54 then the opkg-cl did not hang Jun 16 09:34:13 ok, we are talking about the opkg-cl hanging, though Jun 16 09:34:21 but still two bluezes were candidating for Jun 16 09:34:21 again, the warning are harmless Jun 16 09:34:30 ok, gotchya :) Jun 16 09:34:41 they are informative that you might want to look into the build more closely Jun 16 09:42:13 another problem: in http://projects.linuxtogo.org/pipermail/angstrom-distro-devel/2009-June/003306.html recommended commit http://gitorious.org/gumstix-oe/mainline/commit/ad4e2d75a9d07bce3ff23fffd51f34b39ef97f68 was to start providing udev-utils. I have extracted udev-utils .ipk manually (provided a file /sbin/udevadm) onto my SD card and then 'Starting udev' did not take too so long (only ~0.5-1 min instead of 3.5), but did not fix the main problem (my h1940 Jun 16 09:50:18 booxter: ping Jun 16 09:52:05 booxter: opkg install qt4-embedded works Jun 16 09:53:14 from branch/2009 I get git checkout org.openembedded.dev recipes/qt4 . then bitbake qt4-embedded, then bitbake package-index, then opkg install qt4-embedded, and now I have the browser working Jun 16 09:53:26 more test will follow with my application Jun 16 09:57:40 hrw|gone: how is your mega-push to the stable branch coming along? Jun 16 10:04:40 hi I am getting some strange helloworld problem: Jun 16 10:04:41 ipkg install helloworld_1.0.0-r0.1_armv7a.ipk Jun 16 10:04:42 Installing helloworld (1.0.0-r0.1) to root... Jun 16 10:04:42 Configuring helloworld Jun 16 10:04:42 file_move: ERROR: failed to rename //usr/lib/ipkg/status.tmp to //usr/lib/ipkg/status: No such file or dy Jun 16 10:04:42 Collected errors: Jun 16 10:04:43 Multiple packages (helloworld and helloworld) providing same name marked HOLD or PREFER. Using latest. Jun 16 10:04:45 Multiple packages (helloworld and helloworld) providing same name marked HOLD or PREFER. Using latest. Jun 16 10:04:51 does anyone know what does it mean by Multiple packages (helloworld and helloworld) providing same name marked HOLD or PREFER. Using latest.? Jun 16 10:12:35 good morning Jun 16 10:13:55 whitefox: is there really supposed to be //usr/ instead of /usr/? Jun 16 10:15:10 Additional / shouldn't do any harm. Jun 16 10:15:25 recalcati: maybe after my diploma (friday) I'll find some time to push qt4 fixes to stable... Jun 16 10:24:59 !oebug 3678 Jun 16 10:25:14 * * Bug 3678, Status: HTTP transfer failed :( Jun 16 10:27:27 CoreDump|cf-18: what is there left to fix for cdbot2? Jun 16 10:34:42 Laibsch: few other things hold me Jun 16 10:34:47 re Jun 16 10:35:41 booxter: good luck !!!!! Jun 16 10:35:52 booxter: I go on testing. Jun 16 10:44:38 recalcati: have fun ;) Jun 16 10:47:48 re Jun 16 10:47:55 Laibsch: sorry, had to go and visit a vicar for a while there. Jun 16 10:48:34 sure, don't worry Jun 16 10:48:36 regarding your openssl thing, I think you just need to modify do_configure() to treat "linux-viac3" the same as "linux-i586" in that big case statement Jun 16 10:48:54 You won't suddenly come back to find all OE problems resolved ;-) Jun 16 10:48:54 i.e. where it says: Jun 16 10:48:57 linux-i586) Jun 16 10:48:57 target=debian-i386-i586 Jun 16 10:49:12 change the first line to: linux-i586 | linux-viac3) Jun 16 10:49:32 heh, one can always hope Jun 16 10:50:25 I'll try that, thank you Jun 16 11:03:30 Laibsch: by the time you finish you will be an expert on autotools :-) Jun 16 11:04:41 I'm afraid so Jun 16 11:05:27 I wonder how icecc will affect this Jun 16 11:05:47 IOW, can I have an icecc build cluster of a mix of i686 and viac3 machines? Jun 16 11:10:39 Laibsch: I had to ask about that... Jun 16 11:10:49 Laibsch: how viac3 differ from ix86? Jun 16 11:11:00 Laibsch: it has 'viac3' arch? Jun 16 11:11:17 hardware-wise it's a tricky beast Jun 16 11:11:28 it is branded as an i686 Jun 16 11:11:35 and that is what uname -m reports Jun 16 11:11:51 it does fulfill all mandatory i686 specs Jun 16 11:12:00 it does not have cmov (whatever that is) Jun 16 11:12:28 gcc folks define the difference between i586 and i686 apparently as having cmov or not having cmov Jun 16 11:12:47 cmov is conditional move Jun 16 11:12:52 compiling stuff as i686 on viac3 therefore fails Jun 16 11:12:56 asm instruction Jun 16 11:13:06 Laibsch: I see. Jun 16 11:13:18 does that answer your question? Jun 16 11:13:36 pb or broonie may be more capable of giving you the details Jun 16 11:13:43 thats enough. Jun 16 11:13:47 I can regurgitate what I've learned in the past few days Jun 16 11:13:49 OK Jun 16 11:13:50 -march=c3 for gcc etc? Jun 16 11:14:00 I think so Jun 16 11:14:35 similar stuff is with crusoe and geode Jun 16 11:14:48 both are i586 but not intel-i586 Jun 16 11:14:49 you have that hardware? Jun 16 11:15:06 I really like the C3 Jun 16 11:15:19 a great little home server CPU Jun 16 11:15:53 I have progear with transmeta crusoe cpu and alix.1c with amd geodelx Jun 16 11:16:56 nice Jun 16 11:17:11 I had thought a few times about replacing the C3 with a Geode Jun 16 11:21:47 Laibsch: you can have a mixed cluster but you will need to treat everything like the lowest common denominator Jun 16 11:22:02 i.e. if any machines in the group are c3, you will need to compile everything for c3 even if all the other machines are full i686. Jun 16 11:22:31 OK Jun 16 11:22:35 i686 machines will run programs built for c3 just fine, there's no problem there. Jun 16 11:22:47 I will have to feed the other machines strictly from the C3, then Jun 16 11:36:44 hi could someone help me to get my helloworld to install in beagle? Jun 16 11:36:52 I am getting some strange problem Jun 16 11:36:56 does anyone know what does it mean by Multiple packages (helloworld and helloworld) providing same name marked HOLD or PREFER. Using latest.? Jun 16 11:37:25 normally means you are installing the same package again Jun 16 11:39:03 XorA does it mean it is already there? Jun 16 11:39:12 but when I run helloworld, it is file not found Jun 16 11:39:29 opkg list_installed | grep helloworld Jun 16 11:39:36 opkg files helloworld Jun 16 11:39:52 Is it possible to place a file (e.g. an jpg) in /home/root/ directory from within a recipe? Jun 16 11:40:09 tsjsieb: yes, but we wont accept that recipe Jun 16 11:40:20 XorA ok i get it Jun 16 11:40:25 it's just for testing purpose Jun 16 11:40:26 whitefox: cool Jun 16 11:40:47 tsjsieb: just install the file there and make sure /home/root is in FILES_${PN} Jun 16 11:40:54 Helloworld is my practice program. Jun 16 11:41:03 pb__: About the openssl recipe patch, I did try building all the bb files for diff. versions on my tree. you are right - the addition should go in the .inc file. I am not posting the patch on the list,as I was on digest mode so far, so my posting would create a new thread again :( Jun 16 11:41:09 i actually want to cross compile a real open source application. Jun 16 11:41:22 is there anyway to convert Makefile to recipe? Jun 16 11:41:33 how does it usually work? Jun 16 11:43:28 * XorA is hopeless at teaching, but is you app autotooled? Jun 16 11:44:17 whitefox: there are some examples in the recipes subdir in oe :-) Jun 16 11:48:31 is there any good articles about how to do the Makefile to recipe conversion Jun 16 11:48:33 whitefox: maybe you have already seen it, but chapter 8 in oe user manual - its the one of the most user-friendly and comprehensive guides I have ever read :) Jun 16 11:48:43 whitefox: there is no makefile to recipe conversion Jun 16 11:53:26 Laibsch, dns? Jun 16 11:54:00 I installed tinderbox on melo today Jun 16 11:54:23 I'm waiting for a dedicated database to be created, though Jun 16 11:54:54 But I guess it does not hurt to change DNS already Jun 16 11:55:25 can you please point tinderbox.openembedded.org to melo? Jun 16 11:55:46 140.211.169.165 Jun 16 11:57:12 do you have an idea when the db will be created? Jun 16 11:58:07 within a few hours Jun 16 11:58:33 once it is up, we will restart logging anyway Jun 16 11:58:42 so there is no need to wait Jun 16 11:58:52 and hold back DNS update Jun 16 11:59:11 The provider we have seems to be very slow anyway Jun 16 11:59:16 just checking both would happen around the same time Jun 16 11:59:28 let me get some coffee Jun 16 11:59:39 Last time, their authoritative DNS servers were still giving out the old address hours later Jun 16 11:59:52 normally, I would wait, too Jun 16 12:00:01 But in this case, it is not necessary Jun 16 12:00:22 There is a database in place now and it is running Jun 16 12:02:48 XorA: I use this receipe to test http://pastebin.com/d73e2be9b but when I look i de oe tmp rootfs folder, I don't see the am.jpg in the /home/root/ folder? Jun 16 12:04:47 tsjsieb: I think that should have worked, maybe I miss something obvious, my brain is a bit melted from typing documentation Jun 16 12:05:18 haha, o.k. thanks for looking. Jun 16 12:06:31 try FILES_${PN} = "/home/root/am.jpg" just in case Im an idiot Jun 16 12:07:02 I tested that aswell, but that didn't work. Jun 16 12:07:03 * * OE Bug has been RESOLVED (FIXED) by Jun 16 12:07:05 * * GNU_HASH in madwifi-ng-r3837-20080802-r1 Jun 16 12:07:07 * * http://bugs.openembedded.net/show_bug.cgi?id= Jun 16 12:07:33 dpkg-deb -c test.ipk ??? Jun 16 12:07:41 to check the contants of ipk :-) Jun 16 12:12:26 hello Jun 16 12:12:59 what is the purpose of conf/distro/include/sane-srcdates.inc ? Jun 16 12:13:45 XorA: that gives this as output http://pastebin.com/d2106905b . Seems o.k. to me, or do I miss something Jun 16 12:14:04 and what is the purpose of conf/distro/include/sane-srcrevs.inc ? Jun 16 12:15:42 tsjsieb: I think you forgot to bump PR since you had a non working one Jun 16 12:16:04 mckoan: to be buildable srcrevs for svn/git packages Jun 16 12:18:19 XorA: could you please explain the meaning of "srcrevs" ? Jun 16 12:18:41 mckoan: the rev that is checked out Jun 16 12:19:33 XorA: and what happens with such rev ? maybe is a version check? Jun 16 12:19:51 mckoan: no, it is the rev that is checked out Jun 16 12:19:58 mckoan: simple as that Jun 16 12:20:16 XorA: rev = revision | rev = revert Jun 16 12:20:18 ? Jun 16 12:20:19 I did bump it :( and this ipk is the only one in deploy/glibc/ipk/armv5te/ Jun 16 12:20:27 mckoan: rev = revision Jun 16 12:25:13 mm, there is a r6 directory in de work dir, can that cause the problem? Or is only the ipk used on rebuild (wich is r7) Jun 16 12:25:44 tsjsieb: opkg should always install the latest Jun 16 12:26:19 XorA: so what are the differeces among PREFERRED_VERSION, SRCDATE, SRCREV ? Are they used for the same purpose of select the proper version of a package? Jun 16 12:26:25 Laibsch, done Jun 16 12:27:35 mckoan: PREFERRERED version selects a .bb file, SRCREV/SRCDATE select the revision or date to be checked out inside a bb file Jun 16 12:31:34 Crofton: cool, thanks Jun 16 12:33:22 XorA: so SRCREV/SRCDATE should be strictly related to a specific recipe (.bb), why is defined into include/sane* ? Jun 16 12:33:50 mckoan: so that distros can override, use different Jun 16 12:35:19 XorA: now is clear to me, thank you Jun 16 12:36:36 Xora, I found it, was fould at my site, it's working now (was a fault in the image receipe). Thnx! Jun 16 12:40:18 tsjsieb: awesome Jun 16 12:42:22 haha, more silly (shame shame) Jun 16 12:44:23 florian: good morning Jun 16 13:04:22 pb__: minimal-imaga for micro end up with this : | mkdir: cannot create directory `/home/sgh/openembedded/stuff/build/tmp/work/armv5te-oe-linux-gnueabi/util-linux-ng-2.15-r3/image/bin': File exists Jun 16 13:17:42 oh, yeah, you need a patch for that. I posted one on the mailing list a few days ago. Jun 16 13:18:36 http://lists.linuxtogo.org/pipermail/openembedded-devel/2009-June/011831.html Jun 16 13:21:15 I'm getting an Alignment trap on my samsung 2440 ARM board with some more complex applications Jun 16 13:21:31 like both dillo and web Jun 16 13:21:55 what can cause those? Jun 16 13:22:08 programming errors, essentially Jun 16 13:22:26 I did build them separately from the main image and install them with opkg Jun 16 13:22:44 still using bitbake though Jun 16 13:22:46 no, that wouldn't make any difference. Jun 16 13:23:07 but X works fine, and the kernel, etc... Jun 16 13:23:07 most likely, the application in question (or one of the libraries it uses) just has a bug. Jun 16 13:23:30 hmm Jun 16 13:24:32 both Dillo and Web are specifically made for embedded aren't they? Jun 16 13:25:20 dillo isn't. I don't know about Web. Jun 16 13:28:08 what browser is someone using successfully in OE? Jun 16 13:28:16 on an ARM >=9 Jun 16 13:28:30 or maybe <=9 Jun 16 13:37:31 josh: i'm using a modified version of the gpe minibrowser2 Jun 16 13:46:37 prpplague: iirc there are alignment traps in gpe-confd and keylaunch Jun 16 13:46:57 at least on armv5te Jun 16 13:47:28 ant_work: correct, those need to be corrected, however most of the other apps are working Jun 16 13:47:57 are alignment traps still common on ARM? Jun 16 13:48:13 No. Jun 16 13:48:21 I read an article from 1999 that talked about them, but nothing recent Jun 16 13:48:33 flo_lap: is there any improvement in gpe-confd? only in SVN? Jun 16 13:49:16 josh: pretty rare these days, and normally they get fixed pretty quick, however, gpe hasn't been maintained as strongly as it should have been Jun 16 13:49:49 there was a spike of ak Jun 16 13:49:59 activity, driven by gentoo guys Jun 16 13:50:17 couple of months ago Jun 16 13:50:33 ant_work: yea, a bunch that was related to our M8050 handheld Jun 16 13:50:46 ant_work: hopefully be another round of dev in the next 30-45 days Jun 16 13:50:56 great Jun 16 13:51:37 what is AK? Jun 16 13:51:51 is a typo ;) Jun 16 13:54:21 thebohemian: ping Jun 16 13:54:38 ant_work: pong Jun 16 13:54:44 thebohemian: hi ;) Jun 16 13:54:58 someone is getting crazy with java Jun 16 13:55:05 pls read OE ML Jun 16 13:55:22 I pointed him to the OE java how-to Jun 16 13:58:50 ant_work: yeah. i've seen that Jun 16 13:59:03 ant_work: I am at work to. I will answer him later on Jun 16 13:59:26 np, just to inform you about classpath... Jun 16 13:59:55 (btw, I'm totally java-unaware) Jun 16 14:12:17 ant_work: not really... you mean because if the alignment issues? Jun 16 14:12:45 how do I get git to send an email with a patch? Jun 16 14:13:26 florian: no, I'm more interested to the fixes to singular gpe-conf tools (eg. serial port) Jun 16 14:15:32 Crofton: man git-send-email Jun 16 14:15:53 hmm, not on this machine Jun 16 14:16:05 Crofton: aptitude install git-email Jun 16 14:16:31 sudo yum install git-email Jun 16 14:16:46 03Koen Kooi  07org.openembedded.dev * re33293bf4d 10openembedded.git/recipes/iputils/ (files/debian/fix-cmsghdr-offset-bug.diff iputils_20071127.bb): iputils 20071127: touch up debian patches Jun 16 14:16:46 03Koen Kooi  07org.openembedded.dev * r4dae7ed553 10openembedded.git/recipes/clutter/ (4 files in 2 dirs): clutter 0.9: update SRCREV, fix gtk-doc properly Jun 16 14:16:47 03Koen Kooi  07org.openembedded.dev * raa91f7575e 10openembedded.git/conf/checksums.ini: checksums: add more checksums Jun 16 14:16:48 03Koen Kooi  07org.openembedded.dev * rb7be1779d7 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Jun 16 14:16:49 enlightenment: bump SRCREV to latest rev before soname change Jun 16 14:16:49 ant_work: ah well, yes I'm the one to ask about gpe-conf :-) Jun 16 14:16:51 * This gets us up to date with the latest snapshot release Jun 16 14:16:53 * runtime tested on beagleboard/angstrom Jun 16 14:16:55 03Koen Kooi  07org.openembedded.dev * rd895433b6c 10openembedded.git/conf/distro/include/ (4 files): Jun 16 14:17:00 Angstrom: build with -ggdb3 by default Jun 16 14:17:02 "Level 3 includes extra information, such as all the macro Jun 16 14:17:04 definitions present in the program. GDB supports macro Jun 16 14:17:04 * florian agrees tha the serial tool would need soem love Jun 16 14:17:06 expansion when you use `-gdb3'." Jun 16 14:17:08 03Koen Kooi  07org.openembedded.dev * r8687813463 10openembedded.git/recipes/dvsdk/ (42 files in 3 dirs): Jun 16 14:17:11 dvsdk: sync with TI overlay (http://arago-project.org/git/people/?p=brijesh/arago-dvsdk.git;a=summary) Jun 16 14:17:15 * cleanups will follow later Jun 16 14:17:17 03Koen Kooi  07org.openembedded.dev * re2b659db22 10openembedded.git/recipes/powervr-drivers/ (17 files in 3 dirs): Jun 16 14:17:20 libgles-omap3: add support files for .08 release Jun 16 14:17:22 * actual .08 recipe will follow later Jun 16 14:17:24 03Koen Kooi  07org.openembedded.dev * rec8cee753f 10openembedded.git/classes/image.bbclass: image.bbclass: unbreak postinst status in the install_lingua method Jun 16 14:18:34 is openembedded-stable the best branch to work from? Jun 16 14:19:32 no Jun 16 14:19:36 .dev or stable/2009 Jun 16 14:23:09 hello Jun 16 14:23:23 how do I tell which head this is branched from? http://repo.or.cz/w/openembedded/mini2440.git Jun 16 14:24:04 I would like to know if it's possible to have the *.pem ssl certificates in an OE-based distro ? Jun 16 14:24:23 I would need them to connect to a WPA2 enterprise network Jun 16 14:28:46 we have wpa-supplicant... Jun 16 14:28:53 but I never tried such setup Jun 16 14:30:59 hrw: yes, wpa_supplicant works fine, but I cannot get it to authenticate with WPA2 enterprise because of the certificates Jun 16 14:32:09 that sounds like wpa-supplicant is just configured wrong. you might need to seek assistance from the hostap people. Jun 16 14:32:58 I have the same wpa-supplicant config on my laptop (doc given from hostap people), and it works fine Jun 16 14:34:32 ah.. Jun 16 14:34:48 we build with gnutls not openssl. maybe something was not compiled in Jun 16 14:36:41 in what way does it fail? does it try to do the handshake and get rejected, or does it not even try? Jun 16 14:50:47 I get a TLS: Peer certificate not trusted message, that's why I think the problem comes from the certificates Jun 16 14:51:29 I have copied the file from my Ubuntu laptop where it is valid Jun 16 15:15:04 That does rather sound like something is wrong with your certificate installation, yes. Jun 16 15:27:29 gnutls is used instead of openssl in OE by default ? Jun 16 15:29:07 iirc yes Jun 16 15:29:36 Longfield: ages ago when I was forward porting gnutls patch for sylpheed I found gnutls had no API for verifying certificates Jun 16 15:30:03 Longfield: it could be this is still the case, or that wpa-supplicant doesnt use any new API that has been added since I last looked Jun 16 15:30:23 yeah, it does look like recent wpa-supplicants (0.5.x and 0.6.x) are built against gnutls. 0.4.x was built with openssl. Jun 16 15:30:39 I don't know why that was changed offhand, it seems like a slightly odd thing. Jun 16 15:31:15 ideally I guess the gnutls/openssl thing would be a distro-controlled variable, it seems like a good candidate for a global policy setting. Jun 16 15:32:25 * XorA thinks its time to quit work Jun 16 15:32:58 if (gnutls_certificate_verify_peers2(conn->session, &status) < 0) { Jun 16 15:33:13 that's the check that wpa-supplicant uses. seems like gnutls must indeed have a verification api nowadays. Jun 16 15:33:21 pb_: awesome Jun 16 15:35:28 XorA|gone: you work? Jun 16 15:35:45 I have found out something about wpa_supplicant version, it looks like version 0.6.9 is at least required Jun 16 15:42:56 03Phil Blundell  07org.openembedded.dev * r4e0f698675 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@new.openembedded.org:openembedded into org.openembedded.dev Jun 16 15:42:59 mickeyl: good morning Jun 16 15:43:07 03Phil Blundell  07org.openembedded.dev * r73bdf9cc0c 10openembedded.git/recipes/busybox/busybox-1.13.2/micro/defconfig: busybox: fix typo in defconfig for micro Jun 16 15:43:15 morning pb_ Jun 16 15:44:51 hi mickeyl Jun 16 15:45:07 ah good, finally my nslu2 boots to a shell prompt. Jun 16 15:45:07 hey hrw Jun 16 15:45:09 how's work? Jun 16 15:45:12 pb_: congrats Jun 16 15:45:13 so much easier with a serial port :-} Jun 16 15:45:35 turned out that I had forgotten to make mdev the hotplug helper, so the npe firmware wasn't getting loaded. Jun 16 15:45:48 is there a standard initscript that's meant to take care of that? Jun 16 15:46:43 mickeyl: good Jun 16 15:46:51 mickeyl: had a time to update flyer? Jun 16 15:47:02 ~curse uboot env vars Jun 16 15:47:26 hrw: unfortunately not yet. i'm afraid all i have time for is change the logo. where do we have a vector-variant of it? Jun 16 15:47:32 and I still have 4976k out of my 6272k flash free :-) Jun 16 15:47:38 crap, that patch I sent is for stable Jun 16 15:47:38 * pb_ <3 micro Jun 16 15:49:04 mickeyl: contrib/logo Jun 16 15:49:35 k Jun 16 15:51:17 hmm Jun 16 15:51:17 i have no logo dir Jun 16 15:51:52 artwork sorry Jun 16 15:53:46 ah, righto Jun 16 15:53:56 changing the flyer will be easy Jun 16 15:54:13 mickeyl: I have the PDF for the FOSDEM up-to-date flyer, see feb 3 openembedded-dev Jun 16 15:54:14 changing the poster is tough due to the way the logo overlaps the text Jun 16 15:57:27 likewise: ah right Jun 16 15:57:34 ok, then lets this poster Jun 16 15:57:38 +use Jun 16 15:57:54 mickeyl: I can adapt it if you give me some time Jun 16 15:58:29 Did you find that email? It's called "FOSDEM poster final" Jun 16 15:59:14 likewise: it is poster, not flyer Jun 16 15:59:45 hrw: yes, so it's 50% of what you need? Jun 16 16:00:03 uf. I found print company which can make 2500 a5 in 3 days Jun 16 16:00:15 mostly want 10 days.. Jun 16 16:00:51 I can make a similar flyer if you give me one day, and the text? Just let me know if I can help. Jun 16 16:01:15 likewise: yeah, got it. it's nice, no need to change IMO. I'm adjusting the flyer now to sport the new logo Jun 16 16:01:18 likewise: contrib/marketing/ has older flyer Jun 16 16:01:28 if you have text change requests, please submit them _now_ :) Jun 16 16:01:32 mickeyl: push when done Jun 16 16:01:37 sure Jun 16 16:01:39 mickeyl: "6000 recipes" is one Jun 16 16:01:40 ah wait Jun 16 16:01:49 i'm boycotting oe.dev until we have the TSC Jun 16 16:01:57 i might make an exception for the contrib dir... Jun 16 16:02:00 *shrug* Jun 16 16:02:01 mickeyl: mail me then Jun 16 16:02:07 tsc? Jun 16 16:02:13 technical steering committee Jun 16 16:02:18 ok Jun 16 16:02:27 mickeyl: also update (C) year Jun 16 16:02:59 mickeyl: we also support rpm packaging now Jun 16 16:03:09 mickeyl: and ubifs Jun 16 16:03:24 mickeyl. hrw: I got to run, if I can do something, just push me an email. Jun 16 16:03:28 ok Jun 16 16:03:41 mickeyl: you accept openoffice.org formats? Jun 16 16:04:55 sure Jun 16 16:05:04 i can copy'n'paste that to Pages Jun 16 16:08:09 mickeyl: http://marcin.juszkiewicz.com.pl/download/ulotka-oe.odt Jun 16 16:09:44 oki, will merge that into Jun 16 16:10:01 mickeyl: read that first - especially bolded part Jun 16 16:10:14 i don't know anything about canadian SDKs Jun 16 16:10:20 but i can include the term :) Jun 16 16:10:30 that what I meant ;) Jun 16 16:10:35 hehe Jun 16 16:10:53 * mickeyl kicks Pages for trouble with svg Jun 16 16:11:02 mickeyl: add 'even Montavista uses OpenEmbedded' ;D Jun 16 16:11:07 hahaha Jun 16 16:11:14 canadian sdk is built on Linux, but runs on windows, and generates code for ARM Jun 16 16:11:16 or s/Montavista/Palm Jun 16 16:11:17 for example Jun 16 16:11:47 other example is: built on x86-64 linux, runs on ppc-linux, generates arm-linux Jun 16 16:12:31 the person that makes an eclipse plugin that uses an OE compatible toolchain that runs on windows is a rich man Jun 16 16:13:05 ? Jun 16 16:13:15 that is a sellable product Jun 16 16:13:19 ~change 205 pln to eur Jun 16 16:13:25 i got some error trying that Jun 16 16:13:47 45 eur for 2500 a5 flyers Jun 16 16:13:52 heh, I thought you meant he would have to be a rich man to afford the amount of memory you'd need to run that configuration. Jun 16 16:13:57 heh Jun 16 16:15:30 the "canadian" terminology probably isn't something you'd want to include on a flyer, it's a little bit too obscure for that. Jun 16 16:17:36 * Crofton just noticed mckoan's linkedin status :) Jun 16 16:18:33 Crofton: ;DD Jun 16 16:19:23 by the way, does anybody know why dropbear_0.52 has DEFAULT_PREFERENCE -1? Jun 16 16:19:48 pb_: added, not tested Jun 16 16:20:51 probably Jun 16 16:22:03 Crofton: ;-) Jun 16 16:31:53 ah yes, untested. looks like mickey added it just before the start of his boycott. Jun 16 16:32:17 see u Jun 16 16:32:51 yes, added it some time ago Jun 16 16:32:55 has worked for me but i didn't have time to test more environments than just one Jun 16 16:32:57 so it was left on -1 Jun 16 16:33:01 fair enough Jun 16 16:33:06 I'll see if it works for me Jun 16 16:33:24 bbaib, house cleaning Jun 16 16:33:29 enjoy Jun 16 16:33:33 heh, thanks Jun 16 16:34:05 if you have more comments to http://linuxtogo.org/~mickeyl/misc/OE-Flyer.pdf, shoot right away. I can change more things later tonite Jun 16 16:35:36 mickey|cleaning: old logo in background... Jun 16 16:36:29 mickey|cleaning: 'feature highlights' frame text do not fit anymore. Jun 16 16:37:25 mickey|cleaning: nothing more from me Jun 16 16:38:35 some of the text could maybe use a bit of adjustment; there are a few bits of phrasing that look slightly incongruous as a native speaker. Jun 16 16:39:13 saying that flexibility to customize ... "makes you stay in control" sounds like you're going to be strapped to the keyboard with no way to stop being in control. Jun 16 16:40:32 "automated sanity" sounds rather amusing, it might be better to drop that and just say "automated regression tests at build-time and run-time". Jun 16 16:40:35 pb_: thats why I do not try to write such texts - my english is very non-native Jun 16 16:41:30 right, that's entirely understandable. Jun 16 16:43:31 in OH times I usually passed them though RP ;) Jun 16 16:44:23 * hrw be back in ~1-1.5h for some time Jun 16 16:49:15 03Felix Domke  07org.openembedded.dreambox * r9ed207c8b2 10openembedded.git/packages/dreambox/dreambox-secondstage.bb: dreambox-secondstage: dm500hd: we need dm8000-like staging Jun 16 16:49:16 03Felix Domke  07org.openembedded.dreambox * ra1cf5749a6 10openembedded.git/packages/ (4 files in 4 dirs): dm500hd support: some fixes Jun 16 16:49:16 03Felix Domke  07org.openembedded.dreambox * rd17d25bc3a 10openembedded.git/ (16 files in 11 dirs): add dm500hd machine support Jun 16 16:49:17 03Felix Domke  07org.openembedded.dreambox * r80a2dc4b0c 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of ssh://tmbinc@10.0.120.77/home/tmbinc/opendreambox/1.5/dm500hd/dm500hd/openembedded into org.openembedded.dreambox Jun 16 16:49:21 03Felix Domke  07org.openembedded.dreambox * r36f847aa74 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git://git.openembedded.net/openembedded into org.openembedded.dreambox Jun 16 16:49:24 03Felix Domke  07org.openembedded.dreambox * ra81c938aff 10openembedded.git/packages/dreambox/dreambox-dvb-modules.bb: dreambox-dvb-modules: update for dm500hd Jun 16 16:49:27 03Felix Domke  07org.openembedded.dreambox * r111a9e8e68 10openembedded.git/: Jun 16 16:49:29 Merge branch 'org.openembedded.dreambox' of git://git.openembedded.net/openembedded into org.openembedded.dreambox Jun 16 16:49:31 Conflicts: Jun 16 16:49:51 packages/dreambox/dreambox-secondstage.bb Jun 16 16:49:51 packages/initscripts/initscripts-opendreambox_1.0.bb Jun 16 17:16:47 03Phil Blundell  07org.openembedded.dev * rf4c63b6c9d 10openembedded.git/recipes/dropbear/dropbear.inc: dropbear: last micro change was a bad idea, revert it Jun 16 17:25:24 * pb_ go home Jun 16 17:25:25 later all Jun 16 17:44:48 Hi guys. I'm trying to build uclibc-initial_0.9.30 for mips and I get the following error: Jun 16 17:44:50 | CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/ Jun 16 17:44:50 | cc1: internal compiler error: in add_path, at c-incpath.c:371 Jun 16 17:44:50 Jun 16 17:45:24 There are some google results for this error, but I can't find a solution. Jun 16 17:46:59 The machine definition file I'm using is a modified db1200. I changed the TARGET_ARCH from mipsel to mips. Jun 16 17:48:49 BTW, the build goes fine for the original db1200 machine definition file. So, I guess it's something with mips. Jun 16 18:06:52 morning Jun 16 18:53:56 03ghost  07org.openembedded.dreambox * r5ad3e74115 10openembedded.git/packages/gstreamer/ (3 files in 2 dirs): update gst-plugins-bad and gst-plugins-good to latest release Jun 16 18:53:56 03ghost  07org.openembedded.dreambox * r193b7b8997 10openembedded.git/packages/enigma2/enigma2.bb: enigma2.bb: small changes for modem plugin Jun 16 18:53:57 03ghost  07org.openembedded.dreambox * r8300c68947 10openembedded.git/packages/dreambox/ (7 files in 2 dirs): add dreambox-modem-ppp-scripts package Jun 16 18:53:59 03Felix Domke  07org.openembedded.dreambox * rc0ba8f6ed3 10openembedded.git/: Jun 16 18:54:02 Merge branch 'org.openembedded.dreambox' of git://git.opendreambox.org/git/ghost/opendreambox-1.5 into org.openembedded.dreambox Jun 16 18:54:05 Conflicts: Jun 16 18:54:08 packages/dreambox/dreambox-secondstage.bb Jun 16 18:54:10 03ghost  07org.openembedded.dreambox * rfea57beb51 10openembedded.git/packages/dreambox/dreambox-modem-ppp-scripts/options: add user to options Jun 16 18:54:13 03ghost  07org.openembedded.dreambox * r4d7076cec3 10openembedded.git/packages/dreambox/dreambox-dvb-modules.bb: Jun 16 18:54:16 dreambox-dvb-modules.bb: new dm8000 and dm800 drivers Jun 16 18:54:18 changes for dm8000: Jun 16 18:54:19 fix for flickering video/osd via scart and sometimes green flashes via DVI Jun 16 18:54:22 changes for dm800: Jun 16 18:54:24 small stability fixes Jun 16 18:54:58 03ghost  07org.openembedded.dreambox * r62bffc74a3 10openembedded.git/packages/dreambox/dreambox-secondstage.bb: Jun 16 18:54:58 dreambox-secondstage.bb: mem initialisation fix for dm800 .. 2nd #73 Jun 16 18:54:58 (this fixes problem that some dm800 hangs on powerup) Jun 16 19:20:10 mario-goulart: well; it is is indeed ugly Jun 16 19:20:39 mario-goulart: did you taken a look if you can spot from uclibc build log where it grabs the wrong info? Jun 16 19:20:53 mario-goulart: because maybe you just need to pass an extra param or like Jun 16 19:21:40 otavio: actually I don't understand what the problem is. Jun 16 19:21:59 mario-goulart: it includes files from /usr/include Jun 16 19:22:03 mario-goulart: probably in stagging Jun 16 19:22:12 mario-goulart: ahhhhh I think I know what is this Jun 16 19:22:16 mario-goulart: let me take a look here Jun 16 19:22:38 otavio: and the weird thing is that it works for mipsel. Jun 16 19:30:58 what the heck is wrongw ith rootfs_ipk, and why is there like 42 copies of all the installing messages from opkg? Jun 16 19:31:11 * kergoth grumbles, looking thorugh that log is a pain in the ass Jun 16 19:33:01 mario-goulart: it looks like something mangle the stagging files; did you look on the log what fails? If it is staggiong or compile? Jun 16 19:34:21 otavio: I'm not sure. Right now I'm trying to build using glibc. As soon as I finish this, I'm gonna try to look for more detailed information. Jun 16 19:34:38 mario-goulart: got it; ok Jun 16 20:04:35 kergoth: yeah, that's a pain Jun 16 20:04:47 gm again Jun 16 20:05:36 I kind of assumed that it was just some previously-unnoticed brokenness in opkg. Jun 16 20:19:51 kergoth: what's the intended way to deal with conflicts that occur during "git stash apply"? Git seems to stop applying my changes after the first conflict and I haven't figured out a way to get it to continue where it left off. Jun 16 20:22:06 pb__: hmm, not sure what's intended, but fyi, the stash is just a ref Jun 16 20:22:19 you can look at the reflog of it to see previous stashes, or treat its commit like a normal one Jun 16 20:22:29 git show refs/stash | git apply, or what have you Jun 16 20:23:46 oh, right, cool Jun 16 20:24:02 in that case I'll just pull out the (few) bits that I wanted to keep and ditch the rest, no need to worry about fixing the conflicts Jun 16 20:24:04 :-) Jun 16 20:25:44 re Jun 16 20:26:03 kergoth: my autotools master, I have a n00b question Jun 16 20:26:19 heh, i /may/ have an answer :) Jun 16 20:26:42 pb__: :) i was surprised when i realized it, but it makes sense, the stash command is just a wrapper around regular operations, for the most part Jun 16 20:27:19 florian: wb Jun 16 20:27:32 kergoth: right, yeah. just never occurred to me. Jun 16 20:27:42 kergoth: I am autotooling one simple app which require kernel headers and normal headers. it will be built with OE only rather but I want to make it good. how to add something like --with-kernel-includes= to configure.in? Jun 16 20:28:16 http://www.gnu.org/software/hello/manual/autoconf/External-Software.html should have some examples of AC_ARG_WITH usage Jun 16 20:28:22 :) Jun 16 20:28:28 gracias Jun 16 20:28:42 * kergoth has found the autoconf, automake, and make manuals to all be surprisingly good Jun 16 20:28:46 have a nice rest of day Jun 16 20:29:34 * kergoth thinks the call to cooker.cook() from main() in bitbake should catch a KeyboardInterrupt and exit nicely rather than exiting with a stacktrace Jun 16 20:29:57 kergoth: I remeber a time they were not really useful Jun 16 20:30:08 * kergoth thinks there are lots of places that don't handle keyboardinterrupt the way they should, really. we should stop using os.system() for shell task execution, for one, and catch keyboard interrupts of it and dont emit the BBINCLUDELOGS bits when you do it Jun 16 20:30:14 kergoth: yes I agree Jun 16 20:30:33 florian: well, they're more useful if you already know what you're looking for, as a reference, of course, :) Jun 16 20:30:47 but the rest isn't too bad either Jun 16 20:31:44 heh Jun 16 20:36:01 florian: so, what's the way ahead with this business of e.V. membership? it would be a shame for the thing to get stalled again. Jun 16 20:38:11 pb__: I'll try to avoid this happening again. Wrote a mail to te rest of the board about this issue yesterday. Still not a single reply :-( Jun 16 20:38:45 florian: doh. Jun 16 20:39:27 florian: I was thinking that it would be nice to amend the statutes to make them less ambiguous (and maybe to relax the requirements for membership), but I see now that changing the statutes can only be done at a full member's meeting, not by internet vote. Jun 16 20:39:35 so I guess that isn't going to be practical in the short term anyway Jun 16 20:40:48 pb__: Right... to be 100% sure that we are doing the right thing we could stay with the majority vote and define a comparably short voting time which should be valid. Jun 16 20:46:46 are the e.V fully translated and checked by someone who speak english and german? Jun 16 20:46:55 oops hello Jun 16 20:47:15 I'll check my mail Jun 16 20:47:40 damnit, someday I'm going to remember the format-patch + send-email syntax to send a patch to the list Jun 16 20:47:42 someday.. Jun 16 20:50:13 kergoth *g* Jun 16 20:51:51 florian: right, I guess that'd be the safest thing Jun 16 20:52:39 03Koen Kooi  07org.openembedded.dev * re34a3bfc42 10openembedded.git/conf/distro/angstrom-2008.1.conf: Jun 16 20:52:39 angstrom 2009.X: switch to glibc 2.9 Jun 16 20:52:39 * runtime testing on beagleboard, sheevaplug and om-gta01 reveals 2.9 to be as stable as 2.6.1 Jun 16 20:52:49 03Koen Kooi  07org.openembedded.dev * r179f4f5c34 10openembedded.git/contrib/angstrom/build-feeds.sh: angstrom feed buider: add dialog and unzip Jun 16 20:58:46 * kergoth works on some python usage cleanups in the bitbake codebase Jun 16 21:01:22 kergoth: git apply just says "error: No changes" when I feed it the contents of the stash. any idea what that's about? Jun 16 21:04:49 not offhand, that's odd. I'd do some sanity checks, do a git status to ensure bits from previous aren't hanging around, verify the diff you're piping isnt' empty.. :) Jun 16 21:04:51 * kergoth shrugs Jun 16 21:07:29 pb__: The contents of the stash - you mean a patch format stash? Jun 16 21:07:38 kergoth: yeah, I did "git reset --hard" to clean up my tree beforehand. diff indeed isn't empty. Jun 16 21:07:46 broonie: the output from 'git show refs/stash', specifically Jun 16 21:07:50 so yeah, a patch format stash Jun 16 21:08:24 Hrm, that ought to be OK for git apply Jun 16 21:08:57 well, can always just pipe to patch and do the git adds yourself :) Jun 16 21:09:01 thats really odd though Jun 16 21:09:01 There's also git stash apply which does the same thing. Jun 16 21:09:29 broonie: yeah, I wanted to remove some bits from the patch before applying it, since git stash apply doesn't seem to handle conflicts very well Jun 16 21:09:35 kergoth: true enough, I guess that's what I'll do Jun 16 21:21:24 oh, I just noticed that "git show" is outputting this crazy combined diff format Jun 16 21:21:45 patch hates it, unsurprisingly, and I guess this is why git apply doesn't like it either. Jun 16 21:21:58 can I tell it to make a normal unidiff instead? Jun 16 21:22:09 ("git show -u" doesn't seem to have any effect, doh) Jun 16 21:23:58 try git show --pretty=raw Jun 16 21:24:19 or use git diff-tree on it directly, or use format-patch Jun 16 21:24:25 git format-patch -1 refs/stash Jun 16 21:24:34 i think, anyway Jun 16 21:32:55 !oebug 4711 Jun 16 21:32:56 * * Bug 4711, Status: RESOLVED (FIXED), Created: 2008-10-11 11:52 Jun 16 21:32:57 * * sander(AT)3v8.net: new bb recipe for Pingus, the free lemmings clone Jun 16 21:32:58 * * http://bugs.openembedded.org/show_bug.cgi?id=4711 Jun 16 21:33:11 Laibsch: fixed Jun 16 21:33:48 "git diff-tree refs/stash" and "git format-patch -1 refs/stash" give no output Jun 16 21:35:59 pb__: you need to tell diff-tree what to diff with. Jun 16 21:37:56 hmm, maybe format-patch got confused due to the commit's parent, not sure how stash sets up the relationships Jun 16 21:38:00 * kergoth scratches head Jun 16 22:49:41 * kergoth chuckles as he mulls over throwing together a proof of concept variable typechecker, opt-in via specifying a type in a flag Jun 16 22:52:23 hi kergoth , about proof of concept , do you know if someone build other system than linux using OE ? Jun 16 22:52:58 i've a bit of time this days , i wonder if i try rockbox build using OE Jun 16 22:53:07 long time not play with OE :) Jun 16 23:02:34 Genesis: i remember producing toolchains with it that targeted bare metal, and of course it can build u-boot and such, so there's nothing stopping you, most likely Jun 16 23:03:43 for sure , just i don't remember anyone publish something about a completely different OS build Jun 16 23:03:59 probably not Jun 16 23:04:22 i don't like the idea OE is linux distro , linux is a good OS builder , so it can build what you want Jun 16 23:05:55 it's possible that you'd want to use bitbake and not use OpenEmbedded at all, just make your own classes, or maybe use base only, and not pull over autotools and crap, just make your own little collection, pull over base.bbclass and bitbake.conf, or make your own, not sure if the distro/machine paradigm would still be useful there.. Jun 16 23:06:00 * kergoth shrugs Jun 16 23:07:36 i think i can wrap some interesting things in machine Jun 16 23:07:53 oki if i've time i'll do it :) Jun 16 23:08:14 :) Jun 16 23:08:26 I'd still really like to work the kinks out of building on Mac OS X and cygwin.. Jun 16 23:08:42 those not working well is a constant annoyance Jun 16 23:09:05 i wont work on macosx issue , i'd prefer change of work. Jun 16 23:09:19 ( build house , being farmer ) Jun 16 23:09:38 do we really need osX support ... Jun 16 23:09:44 it's on my todo... way down the list, but it's there Jun 16 23:10:04 my main daily use machine is my macbook :) but usually ssh to Linux boxes to do development Jun 16 23:12:02 i'd build on my mbp too Jun 16 23:12:15 i know at least a few others who would too Jun 16 23:12:43 currently do the same... ssh to a workstatio Jun 16 23:13:21 see linux devel working on propriatary OS makes me sad each time Jun 16 23:13:41 it's a failure to me. Jun 16 23:14:08 linux desktop is you know, ok, but i prefer a mac. Jun 16 23:14:14 linux is not linux developer ready , what's a crap. Jun 16 23:14:18 anyway, its embedded linux! Jun 16 23:14:46 you could embedded darwin , like iphone Jun 16 23:14:48 its not like we're fedora devs Jun 16 23:14:50 :) Jun 16 23:15:07 how a lot of french guys this time Jun 16 23:15:31 perharps someone of sagem Jun 16 23:16:02 i wonder how OE project on their STB prototype is now Jun 16 23:16:22 ptitjes, : bonjour Jun 16 23:17:16 Genesis: hello, I was not aware of that at all Jun 16 23:17:31 Genesis: however hopefully you are lucky my uncle works at sagem Jun 16 23:17:50 and I may ask him if ever he heard about that Jun 16 23:17:52 i work there last years , not good entreprise for innovation Jun 16 23:17:58 héhé Jun 16 23:18:00 sure Jun 16 23:18:10 velizy unit does openembedded for sure Jun 16 23:18:25 i was working on STB but at osny ( cergy ) Jun 16 23:20:27 montavista that use OE now it's very interesting , my boss 2 years ago thought i was crazy to not use montavista package "so more professional" for him ^^ Jun 16 23:20:48 ptitjes, : what do you do with OE ? Jun 16 23:21:05 SHR :) Jun 16 23:21:42 based on FSO Jun 16 23:22:01 ah i saw this on linuxfr Jun 16 23:22:07 you work on dead hw Jun 16 23:22:10 :> Jun 16 23:23:03 ptitjes, : do you come to hackerspacefestival of vitry ? Jun 16 23:23:17 when is this ? Jun 16 23:23:24 26-3 Jun 16 23:23:30 26-30/06 Jun 16 23:23:46 Genesis: FSO will soon work on HTC and Eten devices Jun 16 23:23:48 :D Jun 16 23:24:23 oki good , htc is affordable Jun 16 23:24:34 Genesis: the Linux desktop continues to disappoint me. where windows 7 and Mac OS X are both intuitive and easy, Linux is difficult and convoluted, or breaks with every upgrade. it makes me sad Jun 16 23:24:41 Genesis: oh so maybe I'll make a pass their (hackers...) Jun 16 23:24:48 * kergoth has mostly given up on it for the time being, maybe it'll get better Jun 16 23:25:07 ptitjes, : you should register not too late Jun 16 23:25:19 arf I'll see Jun 16 23:26:28 kergoth: I do agree for Mac Os being more intuitive Jun 16 23:27:07 however I would not say the same as for windows 7 Jun 16 23:28:26 I've been running it for months now, haven't had trouble with anything. lots of improvements to efficiency in the interaction, the new task bar is awesome, and there's lots of little fixes like not having to go through as many clicks to safely eject a usb storage device, and the nicer wireless interface Jun 16 23:28:33 * kergoth shrugs, could be personal preference Jun 16 23:32:50 but it's not GPL Jun 16 23:33:01 at work, i have vista buisness on my desktop *cries* Jun 16 23:33:04 it's not free , we can't go with it. Jun 16 23:33:22 there was a significant drop in productivity Jun 16 23:33:28 when that happened Jun 16 23:33:46 seriously, wtf Jun 16 23:34:33 roll on snow lepper Jun 16 23:36:28 good night Jun 17 02:27:20 hello all. how to i add my own files to tmp/rootfs that get added to the image? **** ENDING LOGGING AT Wed Jun 17 02:59:57 2009