**** BEGIN LOGGING AT Fri May 11 02:59:56 2007 May 11 05:26:44 03hrw 07org.oe.dev * rd1543bd8... 10/ (8 files in 5 dirs): xfce: drop empty libxfce* dirs May 11 06:14:38 morning May 11 06:15:30 hrw: morning May 11 06:17:10 hrw: ipkg -d sd does not set symlink properly May 11 06:17:17 xjqian: it does not have to May 11 06:17:25 ipkg-link is a tool for making links May 11 06:17:40 hrw: sounds like i'm missin something very basic May 11 06:18:19 hrw: does oe use ipkg install in two steps May 11 06:20:19 why two? May 11 06:20:45 oe do "ipkg update;ipkg install somestuff;ipkg install image-stuff" May 11 06:20:45 hrw: so if i want to ipkg -d sd install , I have to do ipkg -link afterwards? May 11 06:23:02 yes May 11 06:25:04 gee! that's not very user friendly ... May 11 06:28:00 I know May 11 06:28:31 btw - I started gnuplot build May 11 06:28:39 I would put this info on the top of Angstrom FAQ, otherwise converting SHARP/Cacko users gonna CRY... May 11 06:29:24 hrw: btw, I have the gnuplot 4.2.0 recipe ready locally, tested working fine on my Tosa May 11 06:29:49 push to bugtracker then May 11 06:29:57 I will replace 4.0.0 with it then May 11 06:31:19 do you want to retain 4.0.0 May 11 06:32:29 the patches are a little different May 11 06:37:43 if 4.2.0 works then why to keep so old one? May 11 06:52:36 !oebug 2300 May 11 06:52:38 * * Bug 2300, Status: NEW, Created: 2007-05-10 23:44 May 11 06:52:38 * * xjqian(AT)gmail.com: gnuplot 4.2.0 recipe May 11 06:52:39 * * http://bugs.openembedded.org/show_bug.cgi?id=2300 May 11 07:02:33 good morning! May 11 07:03:00 Are the images I have built ever going to be cleaned out or are they going to accumulate over time May 11 07:05:56 Laibsch: on angstorm website? May 11 07:06:36 website? May 11 07:06:46 ok, which images? May 11 07:06:47 I am talking about the images in deploy May 11 07:07:00 they are not cleaned May 11 07:07:34 The packages themselves are moved to morgue, right? May 11 07:08:31 yes May 11 07:10:45 !oebug 2298 May 11 07:10:46 * * Bug 2298, Status: NEW, Created: 2007-05-10 16:48 May 11 07:10:47 * * bugs.openembedded.org(AT)rolf.leggewie.biz: login impossible on collie May 11 07:10:48 * * http://bugs.openembedded.org/show_bug.cgi?id=2298 May 11 07:11:14 hrw: Any idea? Or is this likely some kind of mistake on my side? May 11 07:11:42 no idea now May 11 07:11:47 udev started? May 11 07:13:05 I don't know since I cannot log in May 11 07:13:12 This is right after flashing May 11 07:13:22 I am basically dead in the water after that May 11 07:14:00 hrw: Have you pushed all necessary changes to build minimo to .dev? May 11 07:15:19 Laibsch: I pushed to poky. XorA did tests with OE and pushed May 11 07:15:36 Laibsch: then look at screen during boot May 11 07:16:47 Messages fly by so fast May 11 07:17:11 As I mentioned in the report I get lots of warnings about a read-only file system May 11 07:17:25 It looks like at least /dev and /var are ro May 11 07:17:38 udev is slow May 11 07:17:44 takes few seconds May 11 07:18:01 The whole boot process on the collie takes less than two now May 11 07:18:04 For me May 11 07:18:16 so it is broken May 11 07:18:24 I knew that much ;-) May 11 07:18:24 check usr/lib/ipkg/status May 11 07:18:31 hrw: I can't May 11 07:18:34 no login May 11 07:18:45 I guess the serial cable is going to arrive soon May 11 07:19:48 Laibsch: it is YOUR image May 11 07:20:00 cvs version of minimo? Well, now that is going to take a while ;-) May 11 07:20:03 so you should have tarball of it May 11 07:20:20 hrw: Well, that is an idea of course ;-) May 11 07:22:01 hrw: It seems the minimo check-out grabs all mozilla sources (such as chimera and camino). Couldn't it be restricted? May 11 07:25:29 you lack 1GB for sources? May 11 07:25:50 Laibsch: join us - larters of mozilla buildsystems May 11 07:26:21 A fucking gig? May 11 07:26:34 * Laibsch postpones the build May 11 07:26:38 987M is better? May 11 07:27:11 by a far margin ;-) May 11 07:27:20 no way to interrupt this May 11 07:27:25 :-/ May 11 07:30:16 hrw: http://developer.mozilla.org/en/docs/Mozilla_Source_Code_Via_CVS#Selecting_a_Project_to_Pull May 11 07:30:29 I guess May 11 07:30:39 SRC_URI = "cvs://anonymous@${CVSSVR}/cvsroot;module=minimo;tag=${BRTAG};date=${MOZDATE} \ May 11 07:30:43 instead of May 11 07:30:45 SRC_URI = "cvs://anonymous@${CVSSVR}/cvsroot;module=mozilla;tag=${BRTAG};date=${MOZDATE} \ May 11 07:30:52 should be fine, but I cannot test now May 11 07:31:07 /cvsroot/mozilla/ module=minimo rather May 11 07:31:31 OK, all the better. Shall I commit that change? May 11 07:33:45 good morning May 11 07:34:20 Laibsch: ONLY if it works May 11 07:34:34 hrw: I can only test this later May 11 07:34:48 I will postpone the commit May 11 07:35:28 on a related note; If minimo checks out /cvsroot/mozilla module=minimo and firefox checks out cvsroot/mozilla module=mail, will OE be smart enough not to download and store shared code twice? May 11 07:35:47 no May 11 07:36:07 SCM checkouts are per PN not SCM host May 11 07:36:11 * hrw -> breakfast May 11 07:36:36 03Laibsch 07org.oe.dev * r50409e32... 10/ (5 files in 4 dirs): ecj: add native package. Partly closes 2296. May 11 07:46:47 morning May 11 07:46:54 hi XorA May 11 07:52:28 good morning all May 11 07:56:21 hi XorA koen Bernardo May 11 08:05:03 303/1,440 this is gonna take a while :-) May 11 08:13:56 I suppose I can run standard gtk2 based gui programs in Angstrom-X11. Am I right? May 11 08:14:04 yes May 11 08:16:42 ***xjqian -> zzzZZZ May 11 08:48:48 03polyonymous 07org.oe.dev * re0d86bc1... 10/ (1 packages/efl/evas/pagesize.patch packages/efl/evas.inc): evas: fix build against recent kernel-headers, closes #2216 May 11 08:54:25 koen: mtn-0.34 has same on the wire protocol doesnt it? May 11 08:54:48 every mtn >=0.26 has the same wire protocol May 11 08:55:15 * XorA will upgrade wolfson mirror then while he is repairing it May 11 08:55:28 0.35 is out already :) May 11 08:55:41 piss, it was 0.34 last night when I looked May 11 08:55:42 (bugfix release) May 11 08:56:14 " - Private keys are now stored more safely, using file permissions." May 11 08:58:47 I will stay with 0.34 until debian package newer probably May 11 09:00:45 hrw: DEBUG: No providers of build target liburi-perl-native (for ['/home/koen/OE/monotone/org.openembedded.dev/packages/xfce/exo_0.3.2.bb']) May 11 09:00:51 shit May 11 09:02:50 pushing May 11 09:07:02 03koen 07org.oe.dev * r1a12dd64... 10/ (9 files in 5 dirs): xfce: drop some more deprecated dirs May 11 09:07:08 03koen 07org.oe.dev * r6173a453... 10/ (9 files in 4 dirs): xfwm: drop obsolete dirs May 11 09:07:17 thx koen May 11 09:12:52 03hrw 07org.oe.dev * r930528a7... 10/ (3 files in 2 dirs): liburi-perl: added 1.35 (needed by exo, only native one tested) May 11 09:15:01 03koen 07org.oe.dev * r6cf267ab... 10/ (4 files in 3 dirs): xfce: drop some more obsolete dirs May 11 09:15:11 03koen 07org.oe.dev * r846cac1d... 10/ (4 files in 3 dirs): xfdesktop/: drop May 11 09:36:45 03polyonymous 07org.oe.dev * rb34f7bef... 10/ (3 files in 2 dirs): libopieobex0: install missing picture May 11 09:41:27 Is there a way to use the revision of the SVN we have our OE rep to set DISTRO_VERSION ? May 11 09:45:47 hrw: ../exo-csource/exo-csource: line 105: /data/build/koen/OE/build/tmp/angstrom/work/armv5te-angstrom-linux-gnueabi/exo-0.3.2-r0/exo-0.3.2/exo-csource/.libs/lt-exo-csource: cannot execute binary file May 11 09:47:07 koen: not today May 11 09:47:23 ok, I'll file a bug so I don't forget May 11 09:47:47 I am ending work in next 2h and going to organize wedding May 11 09:47:55 MY^WOUR wedding May 11 09:48:26 * XorA wonders where the maemo source for minimo is hidden May 11 09:48:38 I know, I know May 11 09:48:51 garage.maemo.org -> minimo-maemo May 11 09:49:07 hrw: thats never been updated May 11 09:49:15 but exists May 11 09:49:26 hrw: I know is how I know its not the source May 11 09:49:48 you need to catch lamikr then May 11 09:50:10 or the indt guys May 11 09:50:16 antonio gomez iirc May 11 09:55:13 that's why I want the bitbake 1.8.2 dep in OE May 11 10:08:14 hrw: congratulations ! :-) May 11 10:11:45 27.10.2007 will be that day May 11 10:29:52 XorA: I discussed with hrw this morning about restricting the files for checkout of minimo. Did you catch that? May 11 10:30:09 ~seen ptitdav69 May 11 10:30:53 03koen 07org.oe.dev * r88cbce2e... 10/ (3 files in 2 dirs): bitbake: update svn to 1.9.0 and release to 1.8.2 May 11 10:31:00 Laibsch: i haven't seen 'ptitdav69' May 11 10:32:37 hmm .dev branch has multiple heads May 11 10:33:22 tkp: and the sky is blue May 11 10:33:36 not here it isnt May 11 10:51:12 Laibsch: no May 11 10:53:26 Laibsch: found it in backlog May 11 10:53:30 Laibsch: give it a try then May 11 10:54:11 I was just about to send it to you in a private message May 11 10:54:41 XorA: I don't have so much time in front of the computer today and my access is a bit restricted (i.e. 1G) May 11 10:54:52 XorA: I was wondering if you might give it a try as well? May 11 10:55:10 Laibsch: not today, got WM work to do May 11 10:55:18 Laibsch: maybe over the weekend May 11 10:55:36 Laibsch: 9am-5pm GMT time I lack time May 11 10:56:09 OK May 11 10:56:17 Let's work on this later, then May 11 10:56:23 see you May 11 11:28:39 polyonymous, did you submit your directfb paging patch to Directfb yet? May 11 11:29:04 svolpe_gerrath, upstream you mean? No, I didn't and I won't in two weeks at least, as I'm leaving shortly. May 11 11:29:45 polyonymous, yes upstream.. I can submit it if you like. I would like to reference you, what email should I use? May 11 11:30:19 svolpe_gerrath, sure, feel free to submit. You may reference me as a hacker@klever.net, but you don't have to give me credit for such a triviality. May 11 11:30:28 polyonymous, you mention newer versions of kernels in your patch post: starting with what version? May 11 11:30:39 svolpe_gerrath, I believe 2.6.20.something. May 11 11:30:53 polyonymous, what errors were you getting before the patch? May 11 11:31:00 svolpe_gerrath, you can raid changelogs for something like "sanitize asm/page.h" May 11 11:31:08 svolpe_gerrath, undeclared symbol. May 11 11:31:20 And the patch isn't supposed to break build for earlier kernels too. May 11 11:31:36 ahh, thats what I would have thought.. I'm using 2.6.20 so it must have been after that. I will check the changelogs. May 11 11:31:41 (I was talking about kernel changelogs) May 11 11:32:01 anyway, bbl. May 11 11:32:04 polyonymous, yes I figured you meant kernel changlogs. May 11 11:32:13 heh, just in case :) May 11 11:32:18 * polyonymous is gone for a while. May 11 11:47:29 cu May 11 11:51:02 df -h May 11 11:55:28 03pfalcon 07org.oe.dev * r5ae61fd4... 10/ (4 files in 3 dirs): May 11 11:55:28 ipkg 0.99.163: Apply patches to optimize internal algorithms. May 11 11:55:28 * Leading to 10x+ speedup on typical OE feeds (thousands of packages) May 11 11:55:28 * Still needs more testing, though developers' testing didn't show any issues. May 11 11:55:28 * Read patch comments if in doubt. May 11 11:55:29 * Closes #2244. May 11 11:55:32 03pfalcon 07org.oe.dev * r5d2b69c4... 10/ (1 classes/devshell.bbclass classes/patch.bbclass): May 11 11:55:35 devshell.bbclass, patch.bbclass: Check exit code of TERMCMD* commands to see if they were found at all. May 11 11:55:37 * Fixes #2274. May 11 11:55:39 03pfalcon 07org.oe.dev * r1bb8268f... 10/ (1 classes/devshell.bbclass classes/patch.bbclass): May 11 11:55:41 devshell.bbclass, patch.bbclass: Error on any non-zero rc from TERMCMD*. May 11 11:55:43 * Suggested by Richard Purdie. May 11 11:55:45 * Closes #2274. May 11 12:18:06 ljp: thanks it just arrived May 11 12:23:22 wow that new ipkg is fast May 11 12:27:47 indeed. May 11 12:29:15 so it went from 'slow and buggy' to plain 'buggy' May 11 12:33:13 :))) May 11 12:46:15 vivijim: you where the one struggling with debian source packages? May 11 12:48:01 zecke|assignment: yes May 11 12:48:14 at least, I believe so ;) May 11 12:48:20 vivijim: you still have that python exception? May 11 12:48:22 heh May 11 12:50:01 vivijim: shasum-native might be special as it has no sourcecode archive May 11 12:50:22 zecke|assignment: Everytime that I try the "inherit sourcepkg" I got that exception. For a While I'm trying to generate the dsc without inherit sourcepkg May 11 12:50:56 vivijim: but debugging, as you have found already, is a bit tricky. The best thing is to find the except that catches this exception and comment it out (yes pretty bad tactic) May 11 12:51:19 vivijim: you could try with ASSUMED_PROVIDED = "shasum-native" and see if it fails somewhere else May 11 12:51:41 vivijim: otherwise if INHERIT += "sourcepkg" is enough to reproduce it, I could look into it in about ~4 hours May 11 12:52:20 zecke|assignment: In my code I exclude the native and cross ones, but I believe that sourcedpkg is not doing that... May 11 12:52:55 I'll try that May 11 12:53:31 vivijim: to generate debian/rules you might need some support from bitbake to generate the scripts for you May 11 12:54:55 aynway back to the assignment May 11 12:55:17 zecke|assignment: sorry, I didn't get this "bitbake support". Why bitbake and not our classes doing that? May 11 12:55:55 vivijim: how do you want to generate the do_configure, do_compile, do_patch script to call them from debian/rules? May 11 12:56:23 wow R rocks (as the R in R already indicates) May 11 12:58:17 uhm got it! I didn't thing about yet... May 11 13:22:19 03koen 07org.oe.dev * rf5401022... 10/ (1 conf/bitbake.conf): bitbake.conf: set patch resolver to 'noop' to get the old abort-on-error back May 11 13:46:47 * Crofton needs to hack at libtool May 11 13:52:34 is anyone else using gtk+ 2.10.10 and noticed that gpe-clock and gpe-watch's analog displays don't work? May 11 13:55:25 what is the deal qith qemu-native and gcc-4? May 11 13:55:55 qemu is fragile and breaks if the compiler optimizes something May 11 13:56:23 I swtched machinr from efika to omap5912osk and now I am trying to build it? May 11 13:56:37 not sure why I didn't have this problem when doing efika build May 11 13:57:16 do I need to nuke tmp when chanign machine? May 11 13:57:37 I thought multli machine allows builds for different machines in the sam tmpdir May 11 14:00:46 Crofton: it does May 11 14:00:57 Crofton: you need 'apt-get install gcc-3.4' May 11 14:00:59 trying to disable binary locale May 11 14:01:05 heh May 11 14:01:44 I am in a meeting working on a proposal so we can get money to travel to Ireland and have a sdr workshop :) May 11 14:01:54 becuase we want to drink guiness May 11 14:03:35 lol May 11 14:03:51 why is drinking guiness important for sdr? May 11 14:04:49 bleh May 11 14:04:54 now I want some guinness May 11 14:07:30 ah pure genius May 11 14:07:31 I would be happy to get some cider May 11 14:07:44 * polyonymous is content with coffee. May 11 14:09:14 it isn't we are just trying t fund a beer drinking trip :) May 11 14:09:53 Crofton: sure, but what is the explanation? I could try the same then May 11 14:10:15 well May 11 14:10:26 we can't say guiness in the proposal May 11 14:10:51 just dont make mistake of the Guineas brewery tour May 11 14:11:07 hmm May 11 14:11:14 what is wrong with the tour? May 11 14:11:36 Crofton: its 20 Euro to see some videos of how they make guiness followed by a "free" guiness May 11 14:11:49 ok May 11 14:11:50 Crofton: youd be better off just buying 10 pints May 11 14:11:57 we'll watch out for that May 11 14:12:00 is newbie questions allowed here ? May 11 14:12:09 XorA: 10 pints for 20 Euro in Dublin? May 11 14:12:29 zecke|assignment: Im sure they can find a student bar somewhere May 11 14:12:43 sandworm, you can ask :) May 11 14:13:07 sandworm: the worse that happens is we curse your bloodline for life :-) May 11 14:14:13 Crofton: the rooftop bar is nice though especially if its stormy, 10 stories up in a greenhouse overlooking dublin May 11 14:14:53 ok. now i'm trying to build oe for cirrus ep9312, but bitbake world hangs on "NOTE: Running task 1 of 163 (ID: 45, /home/fun/software/embedded/tion/openembedded/org.openembedded.dev/packages/shasum/shasum-native.bb, do_fetch)" May 11 14:15:26 and silence.... May 11 14:15:48 sandworm: hmm try bitbake-1.6 coredump has a similiar issue but we have not yet found the reason... May 11 14:15:55 koen, what is NERG tijdschrift, May 11 14:16:55 bitbake 1.8.2-1 from debian testing May 11 14:18:31 sandworm: 16:15 < zecke|assignment> sandworm: hmm try bitbake-1.6 coredump has a similiar issue but we have not yet found the reason... May 11 14:20:09 sure, trying... now it rebuilds the cache... May 11 14:24:01 * wick720 is away: sl33p May 11 14:25:06 " Bitbake version 1.6.2 is required and version 1.6.0 was found" May 11 14:27:29 well, use the bitbake-1.6 branch from svn May 11 14:27:37 you don't need to install it... May 11 14:35:01 sandworm: I wonder if there is some stty flag that might help May 11 14:36:43 RP: I wonder where it is stopping, a strace by CoreDump showed a lot of SIGSTOPPED signals IIRC May 11 14:39:22 zecke|assignment: I think it was to do with background processes needing access to write to the tty May 11 14:39:32 zecke|assignment: If they don't get it, they SIGSTOP May 11 14:39:46 Ideally, we need to fix IO redirection properly... May 11 14:40:12 Crofton: I don't know the NERG magazine May 11 14:40:21 1.6 works. thanx. May 11 14:40:45 but it is a magazine? May 11 14:41:06 'tijdschrift' is dutch for magazine May 11 14:41:17 or rather 'time based release of paper version' May 11 14:41:26 ok May 11 14:41:55 our dutch is not very good May 11 14:50:54 Do you people use different oe database/directrories for different projects? May 11 14:51:43 what do you mean? May 11 14:52:13 in general no May 11 14:52:55 I have a few project running at the same time (one scale based) one s3c24xx based and I would like to start playing with oe in combo with the n800 May 11 14:53:14 what distro do you use? May 11 14:53:35 distro inside oe? May 11 14:54:06 yeah May 11 14:54:28 that depends moost of the time generic May 11 14:55:04 angstrom uses multimachine, that lets you build for several machines i the same tmp dir May 11 14:58:27 ok, but I could put my OE.mtn in a separate directory and perform multiple checkouts of org.openembedded.dev May 11 14:58:44 you don't need that May 11 14:59:25 or could I even have a single checkout of org.openembedded.dev and have multiple build/conf directories May 11 14:59:43 you could, but you don't need that :) May 11 14:59:46 (one single source /git download) May 11 14:59:56 worst case, you can rerun bitbake with a different tmp dir May 11 15:00:03 with DISTRO=angstrom-2007.1, you can just change MACHINE= and build May 11 15:00:45 this isn't maemo that needs a zillion differend scratchbox instances :) May 11 15:01:14 :p May 11 15:01:56 koen: I am really in early stage right here. I had x11 + touchscreen running on the tomtom go 500 last week. so I was very happy May 11 15:02:24 cool May 11 15:02:49 XorA, how much is guiness in Ireland? May 11 15:03:57 but angstrom-2007.1 + gcc 4.1 glibc 2.5 is possible? something recent? May 11 15:05:20 angstrom uses gcc 4.1.2 and glibc 2.5 by default May 11 15:05:35 Crofton last time I was in Dublin ~ 2years ago it was ~E3.5 May 11 15:05:48 but that Dublin prices May 11 15:08:10 hi, all! May 11 15:08:17 psokolovsky_, ping May 11 15:08:33 slapin: Hiya! May 11 15:09:46 Crofton: I think about 3E May 11 15:10:16 I will try angstrom again then! May 11 15:10:18 Crofton: but the GBP to EUR rate Im never quite sure on as it bounces around May 11 15:10:26 heh May 11 15:10:31 Hopefully we find out :) May 11 15:11:47 Crofton: invite your #oe friends May 11 15:11:58 I'll try :) May 11 15:12:17 First we need to get the proposal funded May 11 15:12:59 Crofton: Im only an hour from dublin May 11 15:13:21 by air? May 11 15:13:33 Crofton: yeah, direct Edin->Dublin is like 40 mins May 11 15:13:47 + 20 mins for bus from Airport to Dublin center May 11 15:30:26 XorA: sounds like an extended lunch hour trip ;) May 11 16:06:47 hi all May 11 16:11:02 hi thesing May 11 17:07:27 hi all May 11 17:12:41 I did not work with branches in svn before but it looks like the changes all apply in branches/multi_cameras May 11 17:12:45 http://svnweb.tuxfamily.org/listing.php?repname=xmoto%20(xmoto)&path=&rev=0&sc=1 May 11 17:13:01 oops May 11 18:15:21 koen: my openmoko-devel-image fails, should I use the MasterMakefile? May 11 18:16:27 builds fine over here May 11 18:16:48 can we get the makefile added to the list of "mind altering drugs"? May 11 18:16:59 koen: not in the .nl May 11 18:17:00 it seems to turn people into drooling retards May 11 18:17:31 koen: in germany you could argue "MasterMakefile helps Terrorist to carry out their work." and it will be forbidden :) May 11 18:18:09 koen: let me get the ubbot error once my machine responds May 11 18:18:26 didn't pH5 fixed that yesterday? May 11 18:18:49 let me update then May 11 18:19:53 zecke: http://www.openembedded.org/bonsai/view/rev/27875/ May 11 18:20:23 zecke: when are we going to australia? May 11 18:21:17 koen: dunno :) May 11 18:21:33 http://linuxdevices.com/news/NS2294059569.html <- bunch of PR crap May 11 18:25:37 zecke: this morning openmoko-devel-image built for an up to date .dev May 11 18:47:54 http://www.micetek.com/englishweb/product/new/PQIIproDevelopmentProducts/8313E-RDB.htm May 11 18:48:26 tsss, the IXP435 is still 100Mbit/s only... May 11 18:48:32 * koen waves to likewise May 11 18:48:43 hey all, waving back May 11 18:51:00 likewise: how affordable is "affordable" ? May 11 18:51:12 hey, I wanted to ask that! May 11 18:57:13 hvontres|poodle: the chip is cheaper, or as cheap as the ixp420. the board price ranges from $99 (Freescale workshops in the US) to $300 (from manufacturer) to EUR 300 (from .eu distributor) May 11 18:57:58 I ordered one board from the manufacturer May 11 18:59:57 * koen needs to ping [g2] about which kernel to use on the avila May 11 19:03:16 EABI/EB, here I come :) May 11 19:03:51 koen: ixp4xx-kernel has avila support, AFAIK May 11 19:04:15 likewise, I think we need to beat on libtool? May 11 19:05:18 Crofton: Not for the next hour, I am about to leave. May 11 19:05:37 Crofton: Will be back in ~1hr for sure. May 11 19:06:05 Crofton: what parts need beating? I'm a libnoob. May 11 19:06:51 I think we need to read the remarks on the debian rpath issue page and attempt to create a patch for libtool cross May 11 19:07:05 I think I may do this tomorrow morning May 11 19:07:48 hi koen May 11 19:07:51 hi likewise May 11 19:07:54 hey HopsNBarley May 11 19:08:05 hey HopsNBarley May 11 19:08:19 Crofton: at what time approx? May 11 19:08:40 my morning May 11 19:08:43 0900 May 11 19:08:47 just FYI - the latest gentoo glibc-2.5-r2 has reduced the fakeroot failrate from 90% to 10%. so we are getting there (-; May 11 19:09:11 Crofton: ok, I'll try to join in. What target/build do we work on to test? May 11 19:09:33 efika or omap5912osk May 11 19:09:40 HopsNBarley: define fakeroot failrate? May 11 19:09:44 they both have the same problesm May 11 19:10:14 Crofton: efika/angstrom for me then. May 11 19:10:20 wow, bitbake 1.8 is much faster than 1.6 in parsing files! May 11 19:10:49 Crofton: got efika deepsleep to work btw, processor and linux are halted, and SDRAM is kept alive. May 11 19:10:55 http://www.worldtimeserver.com/convert_time_in_US-VA.aspx?y=2007&mo=5&d=12&h=9&mn=0 May 11 19:10:57 neat May 11 19:10:58 koen: thanks, I'll go the 1.8 route to fix the do_rootfs prob May 11 19:11:02 Crofton: resume time is <0.3 seconds or so May 11 19:11:09 I think I have both tool chains at homenow May 11 19:11:24 T0mW: the patch polyonymous posted slow bitbake 1.6 down to a crawl, so 1.8 is the way to go :) May 11 19:11:37 Crofton: yup, together with your PicoPSU's I'll continue my WiFi-syncing car ctereo project May 11 19:11:53 koen: is this for real? bitbake is now slaming through the build much faster! May 11 19:12:03 it is :) May 11 19:12:13 wow May 11 19:12:52 koen, that bad? I didn't notice much difference. May 11 19:12:53 Crofton: 1 hour leave is cancelled. wanna beat now? :-) May 11 19:13:19 oh, it is running parallel tasks! May 11 19:13:25 whos beating who? May 11 19:13:35 koen, not that I object going for 1.8 :) May 11 19:14:23 what is the right way to choose the xerver ? http://paste-it.net/2149/raw/ (I guess neither of them...) May 11 19:14:57 the second and last one combined May 11 19:15:08 PREFERRED_PROVIDER_virtual/xserver = "xserver-kdrive" tells OE to build kdrive May 11 19:15:24 XSERVER = "xserver-kdrive-fbdev" tells images to install the fbdev kdrive server May 11 19:15:34 (since kdrive builds multiple servers) May 11 19:16:23 Crofton: make it tomorrow, i'm tired :-) May 11 19:16:37 timtimred: gonna beat the new libtool up till it works May 11 19:25:23 03cbrake 07org.oe.dev * ref4e23b2... 10/ (1 conf/machine/compulab-pxa270.conf): May 11 19:25:23 compulab-pxa270 machine: add --no-cleanmarkers as we are using NAND May 11 19:25:23 add link to rootfs-summary image in deploy/images directory. May 11 19:31:12 like|dvd, I'll be watching the giro prologue tomorrow AM, but it will be on the computer, so I should be working May 11 19:34:59 like|dvd, i've got a strange problem, seen by only a few, where the fakeroot cache doesn't work right, and it gets a file/directory exchanged. I've been through the fakeroot code, and it looks good, so i think this is an odd interaction with fakeroot/glibc/kernel. May 11 19:35:24 like|dvd, it happens only some of the time. very strange. May 11 19:56:05 re May 11 20:19:20 03polyonymous 07org.oe.dev * re70ceff9... 10/ (5 files in 3 dirs): postgresql: added 8.2.4 May 11 21:02:14 * wick720 is back (gone 06:38:15) May 11 21:15:55 HopsNBarley: you told me about the fakeroot problem earlier, it's a real bummer if you encounter a bug that most of us cannot reproduce. May 11 21:17:19 noooooooo kidding. i even totally rebuilt my gentoo box with less agressive optimization. i've spoken to the fakeroot maintainer, and he's gotten a handful of similar reports, but nobody has tracked it down. May 11 21:17:25 i'll bet it's deep. May 11 21:20:47 HopsNBarley: At least it's a real bug and not your local.conf shooting you in the foot :) May 11 22:40:11 is there a way to figure out which version of a package is in the OE database? May 11 22:55:30 summatusmentis: cd packages/$appname ? May 11 23:09:53 CoreDump|home: are all the packages in the database there? May 11 23:10:27 usually yes May 11 23:10:33 ok, thanks May 11 23:41:31 * wick720 is away: f00d May 12 00:28:14 ok, I seem to have run out of room on my host machine, because of OE May 12 00:28:18 how do I clean up? May 12 00:30:59 hmm, rm -rf tmp/ would remove all compiled files and extracted sources May 12 00:31:40 that's the tmp in the top level? or in org.openembedded.dev? May 12 00:32:45 no, the tmp as defined by TMPDIR May 12 00:32:56 from local.conf May 12 00:34:47 ok, I didn't set one, but I found the default one May 12 00:36:15 CoreDump|home: how big will it get? May 12 00:36:43 depend an what you are building May 12 00:36:54 I'm not exactly sure for angstrom May 12 00:37:08 ~10Gb maybe as a guess May 12 00:37:48 hrm, alright, thanks **** ENDING LOGGING AT Sat May 12 02:59:56 2007