**** BEGIN LOGGING AT Mon Mar 19 02:59:57 2007 Mar 19 03:03:49 kergoth: Can you push the patch from bug 518 upstream? I see you are author of tslib. Mar 19 03:04:49 !oebug 518 Mar 19 03:04:50 * * Bug 518, Status: NEW, Created: 2005-12-10 05:39 Mar 19 03:04:51 * * mardy(AT)users.sourceforge.net: \[PATCH\] tslib: work with different screen resolution Mar 19 03:04:52 * * http://bugs.openembedded.org/show_bug.cgi?id=518 Mar 19 03:07:38 night Mar 19 03:24:21 updates to the Sox package, ready for review when someone has a moment http://bugs.openembedded.org/show_bug.cgi?id=2019 Mar 19 05:12:52 03justinp 07org.oe.dev * r28fe31c4... 10/ (4 files in 2 dirs): evas: fix -dbg packaging Mar 19 05:12:56 03justinp 07org.oe.dev * r41b64fef... 10/ (5 files in 2 dirs): imlib2: fix -dbg and -dev packages Mar 19 05:13:00 03justinp 07org.oe.dev * r9433910d... 10/ (1 packages/efl/edje-native_0.5.0.037.bb): edje: fix edje-config name Mar 19 05:13:04 03justinp 07org.oe.dev * r8d108554... 10/ (1 classes/e.bbclass): e.bbclass: remove include hack, add -dev package Mar 19 05:13:08 03justinp 07org.oe.dev * r3d8929bf... 10/ (1 packages/e17/e-wm_0.16.999.037.bb): e-wm: fix -dev and -dbg packages Mar 19 05:15:46 03rwhitby 07org.oe.dev * r19c613da... 10/ (3 files in 3 dirs): slugos-init: improve the handling of the ready led for the nas100d (i.e. stop it blinking when everything is up) Mar 19 08:11:46 what's the standard way to fix a CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/libelf error? Mar 19 08:11:55 this happend while compiling ltrace Mar 19 08:17:04 never mind, the standard way is to fix the configure.ac :-) Mar 19 08:25:07 03schurig 07org.oe.dev * r04df4bfb... 10/ (4 files in 3 dirs): ltrace: fix configure.ac to not search in /usr/include for header files Mar 19 08:32:27 03ifaistos 07org.oe.dev * r412ed26a... 10/ (29 files in 17 dirs): Mar 19 08:32:27 Initial commit for a number of oplinux and oplinux-uclibc specific configs for base-files, netbase, sysvinit and Mar 19 08:32:27 initscripts Mar 19 08:45:29 03ifaistos 07org.oe.dev * r480cd70e... 10/ (6 files in 3 dirs): Mar 19 08:45:29 packages/php/php_5.2.0.bb : Make-over of the php 5.2.0 package. Mar 19 08:45:29 It builds PEAR and mysql support by default Mar 19 08:45:29 Provides seperate packages for each php function i.e php-pear, php-cli etc instead of a monolithic (and big) php Mar 19 08:45:29 package. Mar 19 08:45:30 Additional functionality (GD support, other databases) can be enabled easily. Mar 19 08:45:32 Closes bug #1992 Mar 19 08:57:42 good morning Mar 19 09:09:40 moin Mar 19 09:10:41 morning Mar 19 09:14:56 03ifaistos 07org.oe.dev * r7ed4682c... 10/ (1 packages/gcc/gcc-cross_4.1.1.bb): packages/gcc/gcc-cross_4.1.1.bb: Bump PR Mar 19 09:23:10 morning all Mar 19 09:23:59 hey RP Mar 19 09:26:59 good morning, XorA and RP Mar 19 09:27:24 ping apaulsen, slapin Mar 19 09:31:41 I think I have nano compiled! Mar 19 09:31:42 XorA: Can you do me a favor? It seems that moving require glibc-package.bbclass to .inc was creating problems: http://oz.leggewie.org/wip/286930242424d10b2503eb290dbc31c7975bab74.patch Mar 19 09:31:49 XorA: The other two patches which I had applied first do not trigger any problems it seems. Would be nice if you (and anybody else for that matter) applied http://tinyurl.com/3awwdu and http://tinyurl.com/2qufql to try and verify that they don't trigger the problem you saw earlier. Mar 19 09:32:14 keesj: Congratulations. And welcome to OE. Mar 19 09:32:37 thanks Mar 19 09:32:39 Laibsch: your theory agrees with conclusion I was coming to on friday Mar 19 09:32:45 Laibsch: I shall try the patches now Mar 19 09:33:26 What conclusion is that? Mar 19 09:33:52 Laibsch: that it was order dependand on require glibc-package.bbclass and inherit autotools Mar 19 09:34:24 OK Mar 19 09:37:00 trying to build with your patches applied now Mar 19 09:37:07 XorA: It seems order played a factor here. But it cannot be only order of require glibc-package.bbclass and inherit autotools since their order did not change AFAICS Mar 19 09:37:12 XorA: Thanks. Mar 19 09:38:25 Laibsch: its not tryig to build 2.3.2 Mar 19 09:38:32 Laibsch: so seems to be ok Mar 19 09:38:59 hm? 2.3.2 would not be OK Mar 19 09:39:12 What distro? angstrom should be 2.5 Mar 19 09:39:22 Sorry, missed the NOT Mar 19 09:39:26 Laibsch: heh heh Mar 19 09:39:46 Laibsch: but someone bumped gcc PR so it will take me a few mins to properly test Mar 19 09:39:53 Nice, thank you. I will wait for confirmation from one other person before I recommit the first two. Mar 19 09:40:18 It usually takes me one or two hours. So I can wait ;-) Mar 19 09:41:24 one of my xeon cores is at 100% Mar 19 09:42:01 Well, I guess that one got a little more oomph than the 600 MHz Via C3 Mar 19 09:42:10 just a little :-) Mar 19 09:42:11 morning all Mar 19 09:42:23 But I think I am ahead on the power-consumption front ;-) Mar 19 09:42:24 this is the first job Ive ever had where my work machine out powers my home machine Mar 19 09:42:42 you do13_ Mar 19 09:42:48 s/you/yo/ Mar 19 09:42:52 XorA: which job is that? Mar 19 09:42:59 good morning do13 Mar 19 09:43:03 hi XorA, Laibsch Mar 19 09:43:26 Laibsch: working at Wolfson Mar 19 09:43:40 Laibsch: Liams slave :) Mar 19 09:44:19 do13_: Did you receive my mail? Mar 19 09:44:39 do13_: It does not sound too bad :-) Mar 19 09:44:46 Laibsch: Yes. It's still on my list Mar 19 09:44:55 !oebug 1877 Mar 19 09:44:57 * * Bug 1877, Status: NEW, Created: 2007-02-15 18:17 Mar 19 09:44:58 * * bugs.openembedded.org(AT)rolf.leggewie.biz: linux-rp-2.6.20 will not build for collie (do_patch fails) Mar 19 09:44:59 * * http://bugs.openembedded.org/show_bug.cgi?id=1877 Mar 19 09:45:34 Somebody posted a patch in the BTS. If that one works all you need is to verify. Mar 19 09:46:58 Laibsch: ok. I have to refresh my collie build dir Mar 19 09:47:26 Does it look OK at least from outer appearance? Mar 19 09:50:41 Laibsch: I need a closer look at this Mar 19 09:50:49 OK Mar 19 09:51:24 Laibsch: does it compile and boots? Mar 19 09:51:36 To be honest I have not tested. Mar 19 09:51:50 I do not know the person and I do not know what the patch does Mar 19 09:51:57 So I refrained for now. Mar 19 09:52:03 :) Mar 19 10:05:53 hello everyone Mar 19 10:06:07 he pdet Mar 19 10:08:03 i want to find a petroleum enginer on the internet Mar 19 10:08:36 where can i find them Mar 19 10:08:46 pdet: and you search in #oe because you think the channel is named "Oil Equipment" ? :-) Mar 19 10:09:13 yes Mar 19 10:09:20 ROTFL Mar 19 10:10:56 heh Mar 19 10:11:17 what does oe mean here Mar 19 10:13:03 are you asleep Mar 19 10:13:23 pdet: "Open Embedded" - see openembedded.org Mar 19 10:14:53 what is the embedded Mar 19 10:15:37 pdet: Good one. Got me smiling ;-) Mar 19 10:18:38 good morning everyone Mar 19 10:20:22 is there a kind of 'clobber' mode, which will replace ipk files in deploy/packages if they already exist? Mar 19 10:20:47 so that I don't have to keep increasing my PR in development Mar 19 10:21:22 hello,Laibsch Mar 19 10:21:38 morning all Mar 19 10:22:34 hey Ifaistos Mar 19 10:22:41 could i talk with you privately Mar 19 10:23:29 pdet: yes Mar 19 10:23:29 anyone here knows about qmake integration in OE ? does the $(OE_QMAKE_CC) variable still exist ? where is it defined ? Mar 19 10:25:21 JustinP: I was wondering if your commit http://www.openembedded.org/viewmtn/diff.psp?id1=28fe31c4c998f1ca6f7f78aa9a0bb7dd71684039&id2=41b64fefd1f7cd87d2cd809bd28c0ab62bd6b0b4 fixed Mar 19 10:25:24 !oebug 1829 Mar 19 10:25:25 * * Bug 1829, Status: NEW, Created: 2007-01-29 14:12 Mar 19 10:25:26 * * stv(AT)kirkconsulting.co.uk: imlib2-loaders and imlib2-filters .debug packages not moved into -dbg package Mar 19 10:25:27 * * http://bugs.openembedded.org/show_bug.cgi?id=1829 Mar 19 10:26:11 Looking at it, I now wonder if it does not break packaging. The first debug package will get all debug files which I think is not what you intended. Mar 19 10:27:04 This analysis is just from looking at the bb file. Have not figured out yet how to compile imlib2-loaders. Mar 19 10:50:45 does anyone know if the kernel sources are get packaged when the kernel is build ? Mar 19 10:55:39 Laibsch: NOTE: package openmoko-devel-image-1.0: completed Mar 19 10:55:50 Nice, thanks. Mar 19 10:57:38 morning Mar 19 10:58:48 good morning Mar 19 10:59:02 hrw: Any objections to adding openmoko to http://www.openembedded.org/user-manual&dpage=section_variable ? Mar 19 10:59:20 add them Mar 19 11:00:10 hrw: Why are there gpe and x11 sections? Mar 19 11:00:23 Seems like one or the other would suffice. Mar 19 11:00:41 gpe is environment. x11 is platform Mar 19 11:01:14 Laibsch: gnumeric is x11 but anthing based on libgpewidget is gpe Mar 19 11:01:43 That does not come out so clear from the description. Mar 19 11:02:07 thats how I split Mar 19 11:02:08 Not that it would help but X11 also has the word environment Mar 19 11:13:45 hrw: Will you have time to look at Mar 19 11:13:49 !oebug 1568 Mar 19 11:13:50 * * Bug 1568, Status: NEW, Created: 2006-11-10 16:25 Mar 19 11:13:51 * * bugs.openembedded.org(AT)rolf.leggewie.biz: nunome fails to build after latest QT changes Mar 19 11:13:52 * * http://bugs.openembedded.org/show_bug.cgi?id=1568 Mar 19 11:13:59 ? Mar 19 11:16:18 Laibsch: I lack opie build here Mar 19 11:16:52 OK, I will build and export the results for you to take a look at. Mar 19 11:22:06 k Mar 19 11:29:09 03ifaistos 07org.oe.dev * r786f6df7... 10/ (2 files in 2 dirs): packages/linux/linux-efika-2.6.20: Add missing patch file Mar 19 11:36:32 good morning all Mar 19 11:39:11 hrw: http://oz.leggewie.org/wip/nunome-1.0.1-r2/ is the work dir in tmp for .dev/openzaurus/collie. If you prefer another distro or machine, I can provide that. Mar 19 11:39:14 wb koen Mar 19 11:39:28 hey Laibsch Mar 19 11:40:56 hi koen Mar 19 11:41:07 hey hrw Mar 19 11:41:09 koen: which PBook you have? Mar 19 11:41:27 15" core2duo one Mar 19 11:41:42 my girlfriend now has the 12" g4 Mar 19 11:42:10 koen: I'm thinking about buying 12" laptop for conferences. Mar 19 11:42:31 koen: thinkpad x31/x32, but few guys suggested 12" pbook. Mar 19 11:43:03 the tp is a bit thinner Mar 19 11:43:05 hrw: how about an X series Thinkpad? They can be had cheaply off ebay once they get out of lease. Mar 19 11:43:27 I use the X24 myself. So does Florian. Mar 19 11:43:55 Laibsch: "thinkpad x31/x32" as I already wrote Mar 19 11:44:03 yep Mar 19 11:44:29 I'm currently upgrading from the X24 to X31 Mar 19 11:44:36 hrw: I know. Our communication collided mid-air. Mar 19 11:44:40 ;) Mar 19 11:44:56 florian: what is your source? Mar 19 11:45:06 more or less than GWG? Mar 19 11:50:19 Laibsch: The X24 was from lapstore.de, but i would not recomment them... the first they sent was just broken. The X31 is from some other small reseller i found at ebay... Mar 19 11:50:36 03Laibsch 07org.oe.documentation * rb511f706... 10/ (1 usermanual/reference/var_section.xml): var_section.xml: add openmoko sections. Fix some typos. Mar 19 11:50:38 03Laibsch 07org.oe.documentation * r4503b786... 10/ (1 usermanual/reference/var_section.xml): var_section.xml: remove entrf for "applications" from list of SECTIONS that need an update Mar 19 11:51:11 koen : Hi. welcome back :) Mar 19 11:52:03 Ifaistos: hey Ifaistos Mar 19 11:53:04 florian: Hehe, I also purchased from lapstore.de I bought two X20 about two years ago. One of them was nice. The other had a few glitches that weren't major but that in sum I was not willing to tolerate. It took them about half a year to replace it and refund my money. I think we went throught several cycles. I would not recommend them either. Mar 19 11:53:21 The X24 I got from a private seller. Mar 19 11:53:35 Laibsch: its more or less the gwg maximum. but the ones sold by "asgu10" i would not recommend. On my one wifi and bluetooth are missing and it refuses to turn on without ac connection... but it looks like new. Mar 19 11:54:09 Is that the seller you purchased the X31 from ? Mar 19 11:54:23 Laibsch: "asgu10" Mar 19 11:54:55 So you would not recommend this guy, either? Mar 19 11:55:07 this guy = the one you got the X31 from Mar 19 11:55:40 Laibsch: We bought several times at lapstore - not a single time without trouble, but usually we succeeded quite fast. Mar 19 11:56:26 * Laibsch dreams of an X4* but cannot remember the exact model name. Mar 19 11:56:36 It seems it went downhill after that. Mar 19 11:57:23 Laibsch: right... even if it looks unused (even the fan is clean) and the battery is very good the missing features and bad communication are showstoppers. Mar 19 12:00:26 man.... Mar 19 12:00:32 it's snowing..... Mar 19 12:01:10 XorA, I have made some minor modifications to gtk-2.10.inc so that it will work for gtk+-directfb (package I have ported) as well. I would like you to review my modifications (gtk-2.10.inc) before I push them, if you don't mind.. Mar 19 12:04:51 Gerrath_svolpe: can do Mar 19 12:14:31 03florian 07org.oe.dev * rceb91959... 10/ (1 packages/gpe-conf/gpe-conf_0.2.5.bb): gpe-conf: add 0.2.5 Mar 19 12:14:35 03florian 07org.oe.dev * rebdd90a7... 10/ (8 files in 4 dirs): gpe-login: add 0.88, drop some unused scripts Mar 19 12:15:07 koen: it just starts here too Mar 19 12:29:09 ~lart nokia for rs-mmc Mar 19 12:29:10 * ibot whacks nokia with the cluebat for rs-mmc Mar 19 12:29:45 y Mar 19 12:30:48 can we use a local ipkg feed (on a nfs server for instance) instead of a URL ? Mar 19 12:32:18 yes Mar 19 12:32:25 file:// url Mar 19 12:42:09 thx hrw Mar 19 13:00:43 florian: snow is kinda scary after spending a week in 30+ degrees ;) Mar 19 13:02:43 heh Mar 19 13:08:02 it has been snowing here lately too Mar 19 13:08:26 I was kinda hoping the whole global warming stuff would make sure the spring came earlier Mar 19 13:08:57 It was looking good too. Then suddenly we went from 12C to -somethingC and bucketloads of snow Mar 19 13:11:02 * koen hopes for proper snow Mar 19 13:12:47 koen: did you enjoy driving in a buggy? Mar 19 13:12:54 zecke: yes! Mar 19 13:13:28 koen: did you try to kick z??? out of the buggy? :} Mar 19 13:13:43 no, zack is my 'homey' now Mar 19 13:14:02 hehe :) Mar 19 13:14:27 to tell you the trueth, I have never talked with him :) Mar 19 13:23:57 Crfoton is in Colorado skiing in 55(F) weather Mar 19 13:25:00 ~convert 55 f to c Mar 19 13:26:23 around 14C Mar 19 13:27:58 thx Mar 19 13:28:32 It is supposed to snow for the next few days, but we are leaving today Mar 19 13:31:03 re Mar 19 13:31:09 koen: heh indeed :-) Mar 19 13:40:35 Laibsch: still need me to test those patches? Mar 19 13:43:04 morning Mar 19 13:45:49 apaulsen: yes, please Mar 19 13:46:07 Laibsch: I'll try to look at that, thanks. I was just fixing what I saw from insane.bbclass while building e for angstrom. Mar 19 13:46:19 Laibsch: I'll let you know how it goes Mar 19 13:49:33 Laibsch: I've closed the bug for now. I'm not sure whether there should actually be multiple -dbg packages or not. Mar 19 13:49:44 (I don't even know what the damn things are for) Mar 19 13:51:40 insane.bbclass rocks :) Mar 19 13:51:48 finally people are starting to fix stuff Mar 19 13:52:25 koen: hehe that was the plan. Make people believe insane.bbclass Mar 19 13:52:26 XorA, here is the patch file: http://www.pastebin.ca/401459 I exported EXTRA_OECONF so I can change in it what I need in the gtk+-directb.bb and then search it in the gtk-2.10.inc.. I think it is clear what I'm doing by looking at the patch but just ping me if it is not.. Mar 19 13:52:34 koen: and then change it to ask for money transfer :) Mar 19 14:06:45 Laibsch: I cleaned and built gtk+ as well as gpe-image with your patches now. It all went smoothly! Mar 19 14:09:22 NOTE: package gcc-cross-4.1.2: started Mar 19 14:09:26 * koen crosses fingers Mar 19 14:09:51 LD_PRELOAD + LD_BIND_NOW roxx Mar 19 14:10:47 hrw: LD_BIND_NOW with a broken glibc makes fun :) Mar 19 14:11:17 zecke|study: s/fun/work Mar 19 14:14:01 Gerrath_svolpe: I dont think you need that export, and I would just all set GDKTARGET in gtk*.bb with a default to x11 Mar 19 14:14:14 Gerrath_svolpe: but thats a personal style choice Mar 19 14:25:31 ~seen RP Mar 19 14:26:01 rp is currently on #gpe (4d 8h 52m 36s) #oe (4d 8h 52m 36s) #openmoko (4d 8h 52m 36s) #openzaurus (4d 8h 52m 36s) #handhelds.org (4d 8h 52m 36s). Has said a total of 11 messages. Is idling for 5h 2m 51s, last said: 'morning all'. Mar 19 14:26:33 apaulsen: Thanks for confirming. With positive feedback from three people I feel confident to recommit those two patches now. Mar 19 14:35:15 koen: what are the -dbg packages for? Mar 19 14:35:32 JustinP: they contain the debugging symbols Mar 19 14:35:38 (as it says in the manual) Mar 19 14:37:54 koen: where? Mar 19 14:38:47 JustinP: when we need to use GDB against your code and need debugging symbols Mar 19 14:38:59 koen: ah-ha....it's in recipes_packages Mar 19 14:39:51 koen: is there a convention for whether sub-packages should have their own -dbg packages? Mar 19 14:40:12 I think subpackages should have their own -dbg packages Mar 19 14:40:28 but I never bothered to do that Mar 19 14:40:34 not sure of the name mapping to IRC nicks, but whanks for taking a look at http://bugs.openembedded.org/show_bug.cgi?id=2019 now could someone elaborate on what particular syle issues exist for these recipes, since the oe-stylize script doesn't properly handle the case of required sub-documents Mar 19 14:45:34 ~lart 16MB flash devices Mar 19 14:45:35 * ibot offers 16MB flash devices some herring Mar 19 14:45:41 hillct: ordering of the fields (which is what the URL is about) Mar 19 14:46:16 it was meant as a pointer to useful information. Nothing to scold or anything. Mar 19 14:46:19 ~lart Quebec Cityguy for being a moron Mar 19 14:46:19 * ibot takes a rusty axe and swings it violently, taking Quebec Cityguy's head off for being a moron Mar 19 14:46:26 ordering of the fields is based on the imported document Mar 19 14:46:36 twisted.web.sux ;D Mar 19 14:46:39 nice name Mar 19 14:46:41 hillct: An opportunity to learn Mar 19 14:46:52 hi Laibsch hrw and * Mar 19 14:46:59 howdy chouimat Mar 19 14:47:30 hillct: A lot of old bbs did not follow any style guide which is why it was invented in the first place Mar 19 14:47:32 * chouimat tries to find a FreeBSD 4.11 iso (only have a 4.8 and 4.10) Mar 19 14:47:37 there's nothing to be learned. The ordering is exactly as was used for the previous recipe, and the tar recipe, from which the imported stup style originated Mar 19 14:47:42 hillct: Take the opportunity now to clean it up Mar 19 14:47:55 fine Mar 19 14:47:58 hillct: It appears you still have a lot to learn then Mar 19 14:48:03 * mwester has never before found himself referred to as "*", but oddly enough finds it feels almost more personal than "everyone" ;) Mar 19 14:49:02 chouimat: having fun with City Hall are we? Mar 19 14:49:15 * mwester knows he *had* freebsd 4.11 somewhere, and goes to check... Mar 19 14:49:16 lesson for hillct: stuff in OE is not holy. Actually a lot of it is ugly and in dire need of a clean-up ;-) Mar 19 14:49:44 Laibsch: like glibc??? Mar 19 14:49:52 yes Mar 19 14:49:57 sincerely, that is a mess Mar 19 14:50:21 But we will cut the head off anyone breaking stuff while cleaning it ;-) Mar 19 14:50:28 Laibsch: did you ever figure out what was wrong with your attempt a unification ? Mar 19 14:50:31 mwester: found the iso on ftp-archive.freebsd.org but the ftp server is currently full ... I'm trying to see if I can use dragonfly as an update for the product Mar 19 14:50:35 hvontres|poodle: yes Mar 19 14:50:52 someone has ext2 kernel module for nokia770? Mar 19 14:51:29 Laibsch: it's always interesting when coming into a project, who's actually interested in contributions and who's just interested in throwing their weight around. It's a facinating dynamic isn't it? Mar 19 14:51:33 the last commit (take a look at http://oz.leggewie.org/wip) that moved require *bbclass to the .inc was the culprit Mar 19 14:51:38 hvontres|poodle: no a business partner that decided to vanish in to thin air ... and I have all the support call for his stuff (no physical and ssh access to the servers, and guess what no password too) Mar 19 14:51:52 hillct: So you wanna throw your weight around? Mar 19 14:52:04 my point is this Mar 19 14:52:14 hillct: Spare me your point Mar 19 14:52:18 I am not in the mood Mar 19 14:52:31 clearly not Mar 19 14:52:37 Laibsch: I see you went to #openwrt again ... tsk tsk ;) Mar 19 14:52:40 Discuss this with your cat (sorry, not meant to sound harsh, but I am not too intereste) Mar 19 14:53:02 someone: something like 'readkey' in sh? Mar 19 14:53:07 when I make a comment on a bug report, actual useful information is included, and it's always clear I've fully read the material attached ot the bug report before commenting Mar 19 14:53:37 hillct: I don't like talking to people who implicitly accuse me of not contributing and furthermore not only not contrributing but distracting others who want to do so. Mar 19 14:54:09 hillct: Are you making another claim my info was not useful? Mar 19 14:55:21 I'm just saying your comment contained insufficient detail, as the style guide covers a broad range of stylistic elements the vast majority of which are absolutely corrrect in the attached recipes Mar 19 14:55:28 hvontres|poodle: the last commit (take a look at http://oz.leggewie.org/wip) that moved require *bbclass to the .inc was the culprit. It has to do with ordering it seems. Moving stuff to .inc broke the order. Mar 19 14:55:52 anyway, Laibsch thanks for looking at the bug report at all. I'll get the recipes cleaned up when I have time Mar 19 14:56:03 Laibsch: opps..nothing quite like cleaning up a house of cards :) Mar 19 14:57:09 hillct: I was not making a comment as to that particular bug only. But making sure you know about good docs. Sort of like "give a man a fish" and "teach him to fish". You understand now why "There was nothing to be learned" from you did ring alarm bells here? Mar 19 14:57:44 hvontres|poodle: If you are bored think about how to clean glibc ;-) Mar 19 14:58:23 Ack! Best to leave glibc alone -- that magic is dark and unholy, and best not disturbed ;) Mar 19 14:58:35 cleaning glibc? easy ... just need a few guns and a few address Mar 19 14:59:15 hehe make the gcc build system sane, now that is a challenge Mar 19 14:59:34 Laibsch: awww, and here I thought you were going to be the brave soul that waded deep into the bowels of hell to clean up glibc :) Mar 19 15:02:14 Laibsch: If you are still in a spelunking mood, here is a potential gotcha I found in glibc-package.bbclass: http://rafb.net/p/jTg6MH68.html Mar 19 15:02:41 Laibsch: you're reading WAY too much into this. Comments that provide more detail are more useful. Comments that do not come off as insulting, are useful. That's all there is to it. Mar 19 15:02:44 http://www.thelins.se/johan/2007/03/stop-insanity.html Mar 19 15:06:28 chouimat: hey, at least they added a *second* pointer....:) Mar 19 15:06:40 hehe Mar 19 15:08:14 | *** These critical programs are missing or too old: gcc Mar 19 15:08:17 :( Mar 19 15:09:50 03florian 07org.oe.dev * r17e16e0a... 10/ (1 packages/gpe-conf/gpe-conf_0.2.5.bb): gpe-conf: 0.2.5 replaces bl Mar 19 15:09:54 03hrw 07org.oe.dev * r58bcdff8... 10/ (1 packages/linux/mx31ads-kernel_2.6.19rc6.bb): mx31ads-kernel: remove bashisms - close #1611 Mar 19 15:09:59 03koen 07org.oe.dev * re2140c3f... 10/ (1 packages/gcc/gcc-cross_4.1.2.bb): gcc-cross: add 4.1.2 Mar 19 15:10:03 03mickeyl 07org.oe.dev * r86cc8bea... 10/ (1 packages/gsm/files/fic-gta01/gsmd): gsm: run in background until gsmd properly daemonizes Mar 19 15:10:30 I have PREFERRED_VERSION_glibc to 2.3.2 but oe used 2.5 I don't get it Mar 19 15:10:42 * koen hugs mickey|zzZZzz for updating .dev moko stuff Mar 19 15:11:44 koen: he is even doing that while asleep Mar 19 15:12:56 :-) Mar 19 15:13:02 koen: I think you and mickey missed this comment in gsmd: # User needs to symlink ${GPS_DEV} to the right thing Mar 19 15:13:14 just FYI :) Mar 19 15:15:10 * florian tries to make gcc-cross-sdk work again Mar 19 15:15:52 XorA, I was not sure the best way to do it, I thought it was better NOT to change all the existing files so that is why I did it that way.. But I see your point, I can just do export GDKTARGET = "x11" at the top of gtk-2.10.inc and then set it to "directfb" inside of my file. That is much better and will work well if someone down the line adds another gtk target :-) Mar 19 15:18:12 hi all Mar 19 15:20:12 hi mr_nice Mar 19 15:21:18 hi florian Mar 19 15:21:42 Gerrath_svolpe: I dont think you need to use export on the variable Mar 19 15:21:59 Gerrath_svolpe: include/require is the equivalent of #include Mar 19 15:28:43 XorA, then why when I do things like ${PN} does it return the name of the gtk-2.10.inc and not the file it is included in? Mar 19 15:33:27 Gerrath_svolpe: it doesnt seem to inside RRECOMMENDS_${PN} Mar 19 15:35:05 XorA, good point, I wonder what I was doing wrong.. I originally tried using ${PN} to determine what target to use but it kept returning the .inc file name.. I will try it again, but either way I plan to just use the variable and default it to x11 but set it to directfb in my file. Thanks for the help. Mar 19 15:36:44 koen: I have found MetaClasses.plist :) Mar 19 15:37:58 zecke|study: what does that do on osx? Mar 19 15:39:13 sound interesting Mar 19 15:39:23 koen: it tells iSync which phones can do SyncML :) Mar 19 15:39:57 koen: so hcid+sdptool with SyncML client and libsyncml.opensync.org and one can have Neo OSX syncing :} Mar 19 15:41:37 koen: http://the.taoofmac.com/space/HOWTO/Write%20iSync%20Phone%20Plugins Mar 19 16:00:27 03justinp 07org.oe.dev * r9e9a261d... 10/ (5 files in 2 dirs): imlib2: add -dbg packages for loaders and filters Mar 19 16:00:30 XorA, just tried the variable with out it being exported and it does not work.. it does work if it is exported. Mar 19 16:02:06 XorA, I think its because each section of a .bb file (staging (), deploy (), patch()) are treated differently since they make different files so if a variable is not exported then it can't be accessed by the different sections. Mar 19 16:10:58 XorA, Here is an updated patch: http://www.pastebin.ca/401888 once you give me the ok I will push it. Mar 19 16:13:39 Gerrath_svolpe: looks fine to me Mar 19 16:14:33 03ifaistos 07org.oe.dev * r7f8f2c49... 10/ (1 packages/gift/gift_0.11.5.bb): packages/gift/gift_0.11.5.bb: Set dependancy correctly. Mar 19 16:21:32 Hello Mar 19 16:21:59 yo sirfred Mar 19 16:24:09 time for a reboot Mar 19 16:27:02 RP: ping Mar 19 16:58:46 03pH5 07org.oe.dev * r7aec7f88... 10/ (1 packages/python/python-pygtk2_2.6.3.bb): python-pygtk2-2.6.3: add old version to go with OpenMoko's gtk+-2.6 Mar 19 17:06:29 hvontres|poodle: pong Mar 19 17:07:16 03koen 07org.oe.dev * rabdebc06... 10/ (3 files in 3 dirs): gsmd: fix another typo in gsmd Mar 19 17:09:30 LOL interesting what you come up with when googling 'gsmd' Mar 19 17:28:10 lrg: http://tinyurl.com/38ypa9 make sure to drop by next weekend! Mar 19 17:29:27 lrg: http://tinyurl.com/2lr4mc this one looks better Mar 19 17:30:07 zecke|study: I've screwed up here :(, I thought LAC was in May this year :-( Mar 19 17:30:37 zecke|study: it's too late for me to book now :( Mar 19 17:30:43 poor boy Mar 19 17:31:12 zecke|study: it's too much on my todo list atm :( Mar 19 17:31:30 zecke|study: cant see the wood for the trees Mar 19 17:32:08 well, next time you will make it Mar 19 17:34:03 zecke|study: yep, I'll check earlier next year, iirc 2006 LAC was in May Mar 19 17:35:00 hi all Mar 19 17:35:33 RP: I have a quick question about the sharp bootloader. Does it return any of the memory start addresses or are those hardcoded in the kernel like the commandline. I am trying to see how we could adapt arm-linux-shim for use with linux-rp Mar 19 17:37:36 hvontres|poodle: I don't know offhand. You'd have to look at the 2.4 source dode Mar 19 17:37:38 code Mar 19 17:37:55 03svolpe 07org.oe.dev * rb674c662... 10/ (3 files in 2 dirs): Mar 19 17:37:55 gtk+-directfb-2.10.9 r0: Added directfb support to gtk+ Mar 19 17:37:55 modified gtk-2.10.inc to work with gtk+-directfb Mar 19 17:37:57 RP: ok. Mar 19 17:38:07 * hvontres|poodle shudders... Mar 19 17:39:53 hvontres|poodle: You probably want somewhere in arch/arm/mach-pxa/corgi.c (from memory) Mar 19 17:40:12 RP: Would that also hold for 2.6 kernels? Mar 19 17:41:52 hvontres|poodle: Yes, it would but the 2.6 kernel specifically doesn't look anywhere for strings... Mar 19 17:42:05 hvontres|poodle: poodle.c is probably what you want too, sorry :) Mar 19 17:43:12 RP: np. I know you disabled the cmdline from the bootloader. I am just trying to figure out what else the shim needs to pass to the kernel, if anything Mar 19 17:43:37 RP: I might look at it when I get home. Thanks Mar 19 17:53:18 RP: I can get patches to add armv6 and armv7 support to qemu, is it ok to add them to OE? Mar 19 17:53:40 koen: Feel free Mar 19 17:54:06 ok, I'll ask vivijim to send the patches Mar 19 17:54:54 * koen tries to figure out llvms buildsystem Mar 19 18:04:53 koen "Just type make software" ;) Mar 19 18:09:18 Is there a "make hardware" too? :-P Mar 19 18:10:37 NAiL: yes but it's a link to call dell :) Mar 19 18:16:08 pish fan on one of my home machines died Mar 19 18:24:56 cu Mar 19 18:52:23 wow Mar 19 18:52:35 http://llvm.org/docs/GettingStarted.html has all the steps needed for cross-compiling Mar 19 18:59:20 NAiL: 'make hardware huh? How about this 1.) make magic && make test && make install; 2.) Magic happens 3.) Proffit!!!!! Mar 19 18:59:49 I'd like some of that Mar 19 19:00:16 unfortunately, that particular software disappeared from the market in 2001 Mar 19 19:00:33 that's a shame Mar 19 19:01:21 is there an elegant method to get the find command to strip off a certain number of path elements in it's results? the man page doesn't seem to think so... Mar 19 19:02:55 hillct, do cd first ;-) Mar 19 19:03:53 psokolovsky_: yeah. was considering that. Not what I'd consider elegant. in a recipe I'd be starting from ${WORKDIR} always right? Mar 19 19:04:04 find ... | cut -d/ -f3- Mar 19 19:04:16 Alemo: yeah Mar 19 19:04:41 I'm using --exec in find, to generate install commands Mar 19 19:04:51 but I could pipe the results all over the place Mar 19 19:04:59 also not elegant, but it would work Mar 19 19:05:17 I guess my question is which would be the preferred method within a recipe? Mar 19 19:06:12 find sounds/en_US/MelanieTaylor -name *.wav -o -print -exec echo install '{}' \; Mar 19 19:08:38 I guess I can pipe through | xargs cut -d/ -f3 | xargs install -t targetDirectory Mar 19 19:08:44 or some such thing Mar 19 19:08:53 'evening all Mar 19 19:09:25 hey cyrilRomain Mar 19 19:09:35 it's possible to install source/* targetdir Mar 19 19:09:50 but is there a way to get it to do that recursively? Mar 19 19:09:50 I would use 'cd', it's easier than piping the names (and taking care of whitespaces in filenames). But I'm new in this recipe business... Mar 19 19:10:28 Alemo: you're right. I jsut want to make sure I'm not going to be yelled at for bad form in a .bb file Mar 19 19:11:09 is it critical that I cd back afterward? Are there any relative paths used within the recipe classes? Mar 19 19:11:29 or any other gotchas I should be aware of in this regard? Mar 19 19:11:38 hillct: Anybody yelling at you? Mar 19 19:11:46 not at the moment Mar 19 19:11:54 I would cd in a subshell. Mar 19 19:12:10 hillct: Anybody from OE yelled at you? Mar 19 19:12:27 Alemo: hmm... ok Mar 19 19:12:53 Laibsch: you seem rather upset. Is there anything wrong? Mar 19 19:14:12 I seem upset? Mar 19 19:14:38 If I were that would only because I sense you are twisting oother ppl's words Mar 19 19:14:42 But no I am not upset Mar 19 19:14:48 Laibsch: you seem to think I was talking about someone in particular Mar 19 19:15:18 I was trying to see if you had reasons to claim "OMG somebody yelled at me" Mar 19 19:15:58 Is that something I said? sounds like someone is twisting my words ;) Mar 19 19:15:58 OMG!!!!!one!!!one!!!one Mar 19 19:16:22 There he is, the OMG-master ;-) Mar 19 19:16:24 Laibsch: I think he is using "yelled" to mean "corrected"... Mar 19 19:16:32 koen not it's yay! blood! Mar 19 19:16:35 koen: Were you a bad boy and yelled at anybody? Mar 19 19:16:46 ;--) Mar 19 19:16:58 Laibsch: I was in brazil last week, no time for yelling :) Mar 19 19:17:13 So, you are all refreshed to whip us up? Mar 19 19:17:15 * hvontres|poodle wishes he could play a valley girl "ohh my god" over irc Mar 19 19:17:18 * Laibsch hides Mar 19 19:17:43 Laibsch: sure ;) Mar 19 19:18:02 should we be scared that hvontres|poodle keeps recording of valley girls on his PC? Mar 19 19:18:32 hillct: actually, I don't have one on my pc... just in my head ... :) Mar 19 19:18:40 oh Mar 19 19:18:44 03pH5 07org.oe.dev * rc6af65fa... 10/ (1 packages/python/python-pygtk2_2.6.3.bb): python-pygtk2: fix staging Mar 19 19:18:58 well, there is the third possibility that you have one looking over your shoulder Mar 19 19:19:08 Hello Mar 19 19:19:16 * hvontres|poodle shudders Mar 19 19:19:24 valley girls? Mar 19 19:20:30 I'm running into trouble generating gpe-image jffs2 rootfs for c7x0 on openzaurus-unstable. It seems that mkfs.jffs2 is not being passed the eraseblock size. Mar 19 19:21:11 I was investigating, and I see that c7x0.conf requires zaurus-2.6.conf and zaurus-clamshell.conf. Any of them has a different IMAGE_CMD_jffs2 define. Mar 19 19:22:18 The one used is defined in zaurus-clamshell.conf that is required latter. But it lacks of eraseblock pad and little-endian args, if I understand correctly. Mar 19 19:22:21 Any hint about this? Mar 19 19:23:41 NOTE: package llvm-native-1.9: completed Mar 19 19:28:20 chouimat: http://en.wikipedia.org/wiki/Valley_girl Mar 19 19:35:15 hvontres|poodle: scary Mar 19 19:37:09 llvmc: ELF 32-bit LSB executable, ARM, version 1 (SYSV), for GNU/Linux 2.6.14, dynamically linked (uses shared libs), not stripped Mar 19 19:37:11 schweet Mar 19 19:40:44 koen: what will you use it for? Mar 19 19:40:58 to compile stuff Mar 19 19:41:00 As I understand Opie in the current OZ doesn't come with a libqte-mt, so QThread and QMutex classes are missing? So, what is the best way for multithreaded applications, just using pthreads directly? Mar 19 19:41:18 pH5: I'd like to dump gcc and switch to llvm Mar 19 19:41:50 koen: llvmc? Mar 19 19:42:10 koen: I want StrongTalk and squeak Mar 19 19:42:39 koen: cool. are you going to to gcc vs llvmc comparisons? Mar 19 19:42:52 pH5: when I get time, yes Mar 19 19:43:16 pH5: I could ask the indt dudes for their benchmarks :) Mar 19 19:43:18 (incr koen time) Mar 19 19:43:22 :t incr Mar 19 19:43:33 incr :: Person -> What Mar 19 19:44:19 koen: can llvmc create code that decompiles and recompiles automatically to increase performance? Mar 19 19:45:01 zecke: I asked that, and the answer is "yes" Mar 19 19:45:16 03koen 07org.oe.dev * r219a3642... 10/ (4 files in 3 dirs): llvm: add llvm-native 1.9 Mar 19 19:45:20 koen: outside of Apple? Mar 19 19:50:51 * cyrilRomain reads: "The LLVM JIT optimizations combined with the new multi-threaded OpenGL stack have yielded a doubling of the frame-rate in "a very popular MMORPG" (which is code for "WoW")." Woahow !!! Mar 19 20:06:02 ibot: kergoth Mar 19 20:06:03 somebody said kergoth was Chris Larson, an open source developer working for CSC in Phoenix, AZ. Core developer on OpenZaurus, OpenEmbedded/BitBake, and TSLib. Often the resident BOFH and SCM(svn) nazi, and often an ass. Newbies and the clueless beware. Email at clarson@kergoth.com. Mar 19 20:06:45 Phoenix !!! hahahah Mar 19 20:07:32 hi all Mar 19 20:08:11 hi ggilbert Mar 19 20:08:42 Heya Mar 19 20:09:20 * chouimat|away wish to be able to hang up the phone right now but for some professional reasons he can't Mar 19 20:10:15 ~koen Mar 19 20:10:17 well, koen is our regular image monkey Mar 19 20:13:59 mickeyl, koen: could you have a look at this patch? http://en.pastebin.ca/402237 Mar 19 20:14:13 I tried to put the GSM_ variables into /etc/defaults for gsmd Mar 19 20:14:32 and removed the uucp things that are only needed for cu, but not for gsmd, especially as long as it's run as root. Mar 19 20:14:55 and I think the stty should be removed as well. Mar 19 20:18:07 i like it a lot Mar 19 20:18:14 please apply that Mar 19 20:18:26 i won't catch up with it in OM unless we're out of the debug phase though Mar 19 20:19:33 mickeyl: great, I'll s/defaults/default/ and commit. Mar 19 20:19:40 yep Mar 19 20:20:08 ah while you're there Mar 19 20:20:12 s/uncommend/uncomment/ Mar 19 20:20:25 oh, ok :) Mar 19 20:27:17 03pH5 07org.oe.dev * r5af140b5... 10/ (7 files in 5 dirs): Mar 19 20:27:17 libgsmd: gsmd init script updates Mar 19 20:27:17 - split out machine configuration into /etc/default/gsmd Mar 19 20:27:17 - add machine configuration for magician Mar 19 20:27:17 - remove uucp/spool handling, it is not needed for gsmd Mar 19 20:46:46 pH5: thanks Mar 19 20:48:08 duh. simple solution ot my path problem. Move the files to the top of the source tree first... Mar 19 20:57:10 * XorA|gone hopes the oversized P4 fan keeps his athlon cool Mar 19 20:57:40 * RP keeps thinking of AMDs then I hear things like that... Mar 19 20:57:46 * mmp has athlon64-based laptop... what more to say... Mar 19 21:00:00 * koen is more worried about the scsi disks in the opteron buildbox Mar 19 21:01:14 * v8jlene is back Mar 19 21:04:49 koen: or about when the red-tape finally catches up with ewi-546 ? Mar 19 21:05:19 hvontres|poodle: yeah Mar 19 21:05:24 zecke: "./setup.py dist" should create the tarball? Mar 19 21:05:48 yes Mar 19 21:06:07 zecke: error: invalid command 'dist' Mar 19 21:06:12 sdist Mar 19 21:06:19 :) Mar 19 21:06:24 ./setup.py --help-commands Mar 19 21:06:27 koen: what is it now, almost 4 months since it was supposed to "go away" ? Mar 19 21:06:39 something like that Mar 19 21:06:53 ~hail burocracy Mar 19 21:07:11 172 days Mar 19 21:07:16 according to uptime Mar 19 21:07:31 zecke: I tried cat ./setup.py which was much less helpful ;-) (and then bdist which didn't do what I wanted) Mar 19 21:08:24 zecke: COPYING should be in manifest? Mar 19 21:08:42 if it is a GPL header then yes Mar 19 21:11:29 * lrg -> tv + red wine Mar 19 21:11:40 night all Mar 19 21:11:49 03rpurdie * r791 10/ (branches/bitbake-1.8/MANIFEST trunk/bitbake/MANIFEST): MANIFEST: Add COPYING Mar 19 21:12:06 'night lrg Mar 19 21:12:46 should DEPENDS list every native tool on which the build depends, regardless of how common? sed-native find-native etc? Mar 19 21:12:55 RP: don't forget trunk Mar 19 21:12:55 where's the line in that regard? Mar 19 21:14:47 hillct: My guess would be only tools that there are -native.bb files for, otherwise, how would OE build them? Mar 19 21:15:08 well, there's that... Mar 19 21:16:10 hillct: also, if you inherit someting like autotools, you are probably already getting a bunch of DEPENDS from that. Mar 19 21:16:29 ah. good point Mar 19 21:16:45 hillct: but of course YMMV, since I am still new to this whole OE thing myself. Mar 19 21:17:33 zecke: That would be "trunk/bitbake/MANIFEST" ;-) Mar 19 21:20:02 is there a syntax whereby I can specify the output directory when pulling sources from SVN, comperable to when done on the command line like this? Mar 19 21:20:03 svn co svn://svn.openpbx.org/openpbx-sounds/trunk/sounds/en_US/MelanieTaylor sounds Mar 19 21:21:16 and in the recipe, the src URL is svn://svn.openpbx.org/openpbx-sounds/trunk/sounds/en_US;module=MelanieT Mar 19 21:21:16 aylor;proto=svn Mar 19 21:21:23 RP: my athlon isnt doing too bad seeing as its fan was stopped all day Mar 19 21:22:05 XorA|gone: I've have expected a pile of twisted melted metal ;-) Mar 19 21:22:31 svn: Invalid change ordering: new node revision ID without delete Mar 19 21:23:12 RP: what? Mar 19 21:23:29 RP: yeah, I was surprised CPU survived, lucky linux is good in idle :-) Mar 19 21:23:45 RP: it was hot to touch, but not burning yet Mar 19 21:23:49 hillct: That wasn't for you, I'm just muttering about svn breaking for me... Mar 19 21:24:17 oh, svn issues here too Mar 19 21:24:33 good morning all Mar 19 21:26:28 All soltions online appear to start "check out a new working copy" :/ Mar 19 21:27:21 hi likewise Mar 19 21:29:27 03rpurdie * r792 10/ (3 files in 3 dirs): Prepare for release Mar 19 21:31:03 RP: let us switch to git, svn loves to destroy wc's Mar 19 21:32:25 zecke: I'm suddenly tempted having watched this explode :/ Mar 19 21:33:19 RP: for me it blows up on merges Mar 19 21:33:46 03rpurdie 07bitbake-1.8.0 * r793 10/ (. ChangeLog bin/bitbake lib/bb/__init__.py): Release Bitbake 1.8.0 Mar 19 21:33:52 RP: alternatively bzr can be used, sadly bzr fails to import our bitbake svn :) Mar 19 21:34:09 RP: Yeah! Cheers, should I order you a pizza? Mar 19 21:34:18 or can I use bitbake to order you a pizza? Mar 19 21:34:19 zecke: For me, I managed to break it with an "svn cp"... Mar 19 21:34:36 for me the variation of svn cp, svn rm breaks it :) Mar 19 21:34:44 zecke: bitbake can order pizza, even multithreaded, you just need the right metadata ;-) Mar 19 21:36:25 pizzahut.bb failed to fetch here Mar 19 21:36:51 likewise: did you forget to include tip.inc ? Mar 19 21:36:52 likewise: An error in insane.bbclass? ;-) Mar 19 21:37:47 does anyone know if it is possible to include a cpio.gz ramfs inside a uimage? Mar 19 21:38:43 basically I do not want to compile the image into the kernel, but use it with the initrd mechanism, but still have everything in one uboot uimage file :> Mar 19 21:38:47 any hints? :) Mar 19 21:38:54 Ah no, I see I should upgrade to 1.8.0 which support BakeOven(tm) features. Mar 19 21:39:38 Jin^eLD: write a .bb that depends on the kernel and initrd (how do you create it now?) Mar 19 21:40:15 likewise: I have a .bb that uses mkimage to create a uimage from the kernel Mar 19 21:40:23 but I have no clue how to include the initrd thing there Mar 19 21:40:31 Jin^eLD: Something else: Did you try to work on the wischip board with OE? Mar 19 21:40:40 likewise: yes Mar 19 21:40:50 I am using OE with it Mar 19 21:40:54 Jin^eLD: How do you create the initrd now? (I do not build initrd's currently). Mar 19 21:41:16 Jin^eLD: And you can exploit the full features of the decoder chip? Mar 19 21:41:18 likewise: I don't :) I am not using initrds at the moment, thats the thing; but for demo purposes I'd like to create one image to give away Mar 19 21:41:36 likewise: yes, i can, of course conversion of the software that comes along with the board is needed Mar 19 21:41:45 conversion meaning - creating bb files and so on Mar 19 21:41:50 Jin^eLD: Ah ok, but all the sources are there? Mar 19 21:42:04 likewise: yes, everything needed comes with the SDK Mar 19 21:42:05 but.. Mar 19 21:42:15 you will need to stick to the glibc and gcc vesions that they use Mar 19 21:42:19 I build the toolchain from scratch Mar 19 21:42:22 but I had to match the versions Mar 19 21:42:24 Jin^eLD: uh oh. why so? Mar 19 21:42:36 because their kernel does not play together with nptl and newer glibc Mar 19 21:42:54 zecke: Its made berlios. Just need an announcement now :) Mar 19 21:43:03 that was my first attempt - take the latest glibc and gcc 4.1.1, tmbinc helped me a little Mar 19 21:43:05 Jin^eLD: ok, but they should release the kernel patches, so you should be able to build it yourself (tm) Mar 19 21:43:12 and I got everything built for mipsel but could not use it Mar 19 21:43:25 likewise: they release the full kernel source that they are using Mar 19 21:43:29 the problem is - the kernel is quite old Mar 19 21:43:36 2.6.12-rc2 Mar 19 21:43:44 and I am not much of a kernel hacker... Mar 19 21:43:50 so I can not patch it up into 2.6.20 Mar 19 21:43:56 Jin^eLD: hmm, not that old, I would go from there. Mar 19 21:43:56 :) Mar 19 21:43:57 RP: tag the walls of your neighborhood Mar 19 21:44:14 Jin^eLD: I think I would be able to bring the sources forward, done it lots of times for many drivers. Mar 19 21:44:16 likewise: well.. I do not have any kernel hacking experience... Mar 19 21:44:24 that would be cool Mar 19 21:44:48 Jin^eLD: no problem, this is community work :-) I just need to get one of those boards. Last time, it was 25 kilo-EURos.... Mar 19 21:45:08 likewise: well, the other problem is that we will have to wait Mar 19 21:45:18 I think they did not do a public release yet Mar 19 21:46:01 else I would have tried to bring the board into OE for everyone's use Mar 19 21:49:16 but anyway, the good news is that it can be fully OE'd... if you get your hands on the board and the code and manage to patch it up to a more recent kernel - that would be really cool Mar 19 22:04:12 zecke: Thanks for the instructions, they were a help :) Mar 19 22:06:05 at least that, please send your updates so at some point we have a complete release guide Mar 19 22:07:25 This should be impressive if it works... pulling all the openpbx audio prompts out of SVN then converting them at build-time to all the available voip codecs Mar 19 22:07:36 makes for a really LONG build though... Mar 19 22:11:39 zecke: update sent. Mar 19 22:11:50 congratulation :) Mar 19 22:18:47 RP: And there was much rejoicing....yeahhhhh Mar 19 22:22:19 oh they have a webform nowadays Mar 19 22:42:58 03rpurdie * r794 10bitbake/lib/bb/parse/parse_py/BBHandler.py: BBHandler.py: Use 'TMPDIR/anonfunc' as T for anonymous functions Mar 19 22:46:43 good nite all Mar 19 22:58:50 I have some shell scripting badness on lines 12 and 13 where it appears things aren't being escaped properly. http://www.pastebin.ca/402531 It looks sane to me, but perhaps bitbake is doing something unexpected here. Here's the relevent script after substitution: http://www.pastebin.ca/402552 Mar 19 22:59:40 devshell is already in 1.6.6 as a task? Mar 19 23:00:43 hmm i get an invalid task exception Mar 19 23:01:26 paste it Mar 19 23:02:41 Error: InvalidTask event exception, aborting Mar 19 23:08:18 mickeyl: It should work ok in 1.6.x. I think I've just realised it will only work with -b in 1.8.x though :/ Mar 19 23:09:29 RP: :} Mar 19 23:09:44 RP: don't worry mate :) Mar 19 23:10:10 * RP might fix in 1.9.x and someone might backport it... :} Mar 19 23:10:38 zecke: Nobody mentioned this during testing so few people must use it... Mar 19 23:11:54 RP: right, and the few test it today :) Mar 19 23:12:03 03rpurdie * r795 10bitbake/ (4 files in 2 dirs): Add PE (Package Epoch) support from Philipp Zabel (pH5) Mar 19 23:12:13 ~hail ph5 Mar 19 23:13:01 This is where changes to OE get interesting since we can't depend on that yet... Mar 19 23:13:17 I guess it should get backported for 1.8.2... Mar 19 23:18:19 * RP wishes he could clean up and seperate the UI changes enough to checkin Mar 19 23:21:25 RP: yes, this is a obvious candidate for backporting Mar 19 23:26:20 hello all, have a problem compiling sqlite with MACHINE=native Mar 19 23:26:53 the error is: libc.so.6: undefined reference to `_dl_tls_get_addr_soft@GLIBC_PRIVATE' Mar 19 23:27:28 SirManiac: nm your libc.so.6 and see if this is defined Mar 19 23:27:41 ok Mar 19 23:28:36 03rpurdie * r796 10bitbake/ (ChangeLog lib/bb/build.py): build.py: Treat python functions the same as shell functions with respect to logging Mar 19 23:28:57 RP: I would even like to backport this to bitbake-1.6 Mar 19 23:29:05 RP: but we should declare EOL for 1.6 Mar 19 23:29:55 output of nm libc.so.6 | grep _dl_tls_get_addr_soft : Mar 19 23:29:55 U _dl_tls_get_addr_soft@@GLIBC_PRIVATE Mar 19 23:30:26 SirManiac: ask your distribution for a refund :) Mar 19 23:30:40 :), funny Mar 19 23:30:48 zecke: 1.6 is EOL as soon as 1.8.0 is proven slightly more and the OE.dev version is bumped Mar 19 23:31:34 zeake: I've found the funtion in the libc source code, I think it's only build with the confugure option --with-elf Mar 19 23:32:10 zecke: will try and build libc with that option and see if it works... Mar 19 23:32:10 SirManiac: which should be the case for every linux system for the last 13 years :) Mar 19 23:32:38 SirManiac: check if you have another libc.so somewehre Mar 19 23:32:49 zecke: yeah, thought so, but that option isn't in the configure script Mar 19 23:32:50 SirManiac: probably inside a tls directory or such Mar 19 23:33:06 which distro is that? Mar 19 23:33:09 zecke: in elf/tls.c Mar 19 23:33:24 more like /lib/tls/libc.so.6 Mar 19 23:33:26 zecke: angstrom and MACHINE=native Mar 19 23:34:08 we create a slightly broken libc then :} Mar 19 23:34:59 I'll go ahead and try adding the --with-elf option and tell you if that works Mar 19 23:41:17 03rpurdie * r797 10bitbake/lib/bb/event.py: event.py: Allow basic (ugly) UI event handler registration Mar 19 23:42:18 03rpurdie * r798 10bitbake/lib/bb/event.py: event.py: Add ParseProgress class Mar 19 23:44:03 'night all Mar 19 23:44:52 night Mar 20 00:56:18 anyone awake? Mar 20 00:56:29 hmmm? Mar 20 01:08:47 I'm still here Mar 20 01:25:29 psst, this link is broken. linked to in user manual. http://www.openembedded.org/user-manual&dpage=recipes_installation_scripts Mar 20 01:26:05 nite Mar 20 02:19:37 god, I hate stupid shit like that Mar 20 02:19:56 dealing with one bug for several hoursw Mar 20 02:20:22 turns out it was a trivial failure to escape slashes in an expanded path variable Mar 20 02:20:23 is there anybody who *likes* those types of bugs? Mar 20 02:20:52 people who are lazy and easily pleased with themselves? Mar 20 02:21:04 hehe Mar 20 02:21:27 "Ooh! Ooh! I solved it!" Mar 20 02:21:45 never mind that it was a stupid problem I never should have had to begin with Mar 20 02:22:14 seems like its those simple bugs always take inordinate amounts of time to solve Mar 20 02:22:25 yeah Mar 20 02:22:47 "could it be... ? No.. It wouldn't be that stupid..." Mar 20 02:22:49 I suppose that's why I like the stricter languages, like c++, there more chances the compiler will catch it. Mar 20 02:22:54 4 hours later Mar 20 02:23:04 "oh my god. how'd I miss that?" Mar 20 02:25:27 of course the package build has grown infianately more time consuming, so now I won't know if my fix worked, for another 45 minutes Mar 20 02:25:55 building glibc? Mar 20 02:26:04 openpbx Mar 20 02:27:11 it downloads several hundred MB of voice prompts for conversion to each of the 4 main voip codecs, for packaging Mar 20 02:27:16 very very slow Mar 20 02:27:28 ahhh Mar 20 02:27:56 does it redownload every time? Mar 20 02:53:14 yeah Mar 20 02:53:41 I can't reasonably expect to ceck something that big into monotone can I? Mar 20 02:54:19 the voice prompts are being updated relatively frequently anyway Mar 20 02:54:21 yeah, but isn't there a way to cache the downloads? Mar 20 02:54:28 oh Mar 20 02:54:59 yes Mar 20 02:55:08 once the new recipe is checked in Mar 20 02:55:18 wget has options that can check for modification dates, and not redownload when the date is the same Mar 20 02:55:21 or at least that's how I understand it works Mar 20 02:55:49 so does svn, in that regard Mar 20 02:55:55 check for changes Mar 20 02:55:57 still Mar 20 02:56:09 the conversion script doesn't Mar 20 02:56:53 ok Mar 20 02:56:58 next dumb problem Mar 20 02:57:18 remind me how I'm supposed to be setting shell (not bitbake) variables Mar 20 02:57:20 set relfile2 = `echo $relfile|sed -e s:\.wav$::` Mar 20 02:57:45 set isn't right Mar 20 02:57:52 not using set, isn't right Mar 20 02:58:05 I don't have a shell book readily available Mar 20 02:58:08 no set, and no spaces around the = Mar 20 02:58:15 really? Mar 20 02:58:19 spaces are the issue? Mar 20 02:58:23 * hillct tries that Mar 20 02:58:44 pretty sure shells don't do the space Mar 20 02:59:01 opie-rdesktop fails to build, how can I make bitbake pass it? Mar 20 02:59:41 testing now **** ENDING LOGGING AT Tue Mar 20 02:59:57 2007