**** BEGIN LOGGING AT Mon Jul 10 02:59:56 2006 Jul 10 03:07:10 03lenehan 07org.oe.dev * r467047c9... 10/packages/liblockfile/ (liblockfile-1.06/glibc-2.4.patch liblockfile_1.06.bb): Jul 10 03:07:10 liblockfile 1.06: Fix for building with glibc 2.4. The local function Jul 10 03:07:10 eaccess conflicts with a function of the same name supplied wih glibc 2.4. Jul 10 03:07:10 So add a patch which globally renames this function within the library. Jul 10 03:53:13 feedback/discussion concerning the patch in bug 1157 would be welcome; it solves my libltdl build problems Jul 10 04:39:26 ~seen ixoth Jul 10 04:39:39 ixoth was last seen on IRC in channel #oe, 1d 7h 7m 22s ago, saying: 'goodbye everyone, have to go'. Jul 10 06:40:10 good morning all Jul 10 06:41:43 morning koen Jul 10 06:42:29 ~seen ixoth Jul 10 06:42:41 ixoth was last seen on IRC in channel #oe, 1d 9h 10m 25s ago, saying: 'goodbye everyone, have to go'. Jul 10 07:20:43 morning all Jul 10 07:24:55 morning all Jul 10 07:25:17 morning Jul 10 07:38:21 hey RP, Dirk and XorA Jul 10 07:40:00 koen: I think I found the cure to the alignment problems, expect some checkins later on Jul 10 07:40:10 XorA: ah, cool Jul 10 07:40:11 ~lart hardware failure Jul 10 07:40:11 * ibot eats hardware failure's liver with some fava beans and a nice chianti Jul 10 07:40:21 XorA: if possible, submit upstream :) Jul 10 07:47:11 koen: its more, use upstream versions not our ancient ones :-) Jul 10 07:49:30 heh Jul 10 08:10:24 03xora 07org.oe.dev * r210506ba... 10/packages/fontconfig/fontconfig_2.3.95.bb: (log message trimmed) Jul 10 08:10:24 fontconfig_2.3.95.bb : new version of fontconfig Jul 10 08:10:24 autoconf/Makefile magic internall seems to do the right thing in cross Jul 10 08:10:24 compile situations so this one doesnt need a matching fontconfig-native. Jul 10 08:10:24 I have left at DEFAULT_PREFERENCE = "-1" to allow for proper evaluation Jul 10 08:10:25 as this library is fairly core. Jul 10 08:10:29 This version seems to cure the masses of alignment errors I was getting Jul 10 08:15:20 http://www.linuxdevices.com/news/NS6437147703.html Jul 10 08:15:26 morning Jul 10 08:15:34 OZ is 'hacker friendly' now ;D Jul 10 08:18:31 now it just needs to be 'user friendly' ;) Jul 10 08:19:54 hrw|work: did you get my mail about the kernel-modules? Jul 10 08:20:43 koen: yes Jul 10 08:21:22 koen: nice idea Jul 10 08:31:03 koen: what do you think about OZ 3.5.4.1-r1 images for spitz then? Jul 10 08:31:16 good idea Jul 10 08:32:05 koen: how can i contribute to the user manual? Jul 10 08:32:16 zwelch: edit and send patches? Jul 10 08:32:24 zwelch: what hrw said Jul 10 08:32:50 Q: 'how can I contribute to OE?' A: 'edit and send patches' Jul 10 08:32:55 take #2: where can i find the sources of the user manual, against which i should base my patch? Jul 10 08:33:07 org.openembedded.documentation Jul 10 08:33:08 zwelch: in branch Jul 10 08:33:24 thank you Jul 10 08:33:44 see http://www.openembedded.org/viewmtn/branch.psp?branch=org.openembedded.documentation Jul 10 08:33:52 what is the best mechanism for conveying these patches, such that they'll get incorporated asap? Jul 10 08:34:27 submit to bugtracker and pester koen/zecke? Jul 10 08:35:02 yeah, or mail them to freyther@inf.fu-berlin.de AND koen@dominion.kabel.utwente.nl Jul 10 08:35:46 okay, now this is a fun question: which bug tracker gets the most attention? Jul 10 08:36:05 OE use OE bugtracker Jul 10 08:36:06 i.e. bugs.oe.org or hh.org/~bugzilla? Jul 10 08:36:22 ah Jul 10 08:36:44 k then, they aren't as redundant as they appear then, i suppose Jul 10 08:43:53 zwelch: OE has virtually nothing to do with hh.org, except that the mailinglist is hosted there Jul 10 08:44:10 koen: no plans to move ML to ltg? Jul 10 08:44:44 hrw|work: yes, but I need to pester nils for the mailman password Jul 10 08:45:58 hrw|work: We're trying to make it oe@openembedded.org Jul 10 08:46:26 koen: not oe@lists.openembedded.org? Jul 10 08:46:48 koen: so oe@openembedded.org could be 'contact us' email Jul 10 08:47:27 hrw|work: sounds like a good idea Jul 10 08:47:58 seems so silly that the two projects have fallen out Jul 10 08:48:02 shame Jul 10 08:49:06 ade|desk: @lists.openembedded.org show exactly what kind of address it is. commits@ devel@ Jul 10 08:49:41 indeed Jul 10 08:50:15 koen: and that oe@handhelds.org gets a copy of NEW bugs hosted on their bug tracking system... since i just subscribed today, i haven't concluded whether or not the OE system does the same thing but in a different time zone Jul 10 08:50:56 but having said that , wasn't one of the primary things for the list to have oe@ to be for dev and users else the devs don't answer user emails :/ Jul 10 08:52:36 and since i submitted a patch for libtool-native to the bugs.oe.org tracker, it was disappointing not to seem my bug getting the same level of nagging/attention ;) Jul 10 08:53:21 incidentally, who should i bother about getting that one included :) Jul 10 08:53:32 zwelch: we used to have bugsquashing days Jul 10 08:53:48 i would love to help organize such again Jul 10 08:55:01 there are other purposes such focused efforts can be used for: documentation, version bumping.... things that benefit from lots of hands taking a relatively hands-on approach to fixing issues that affect the entire tree Jul 10 08:56:40 zwelch: that patch is already in :) Jul 10 08:58:06 koen: then WHY no info in BT? Jul 10 08:58:21 hrw|work: ? Jul 10 08:58:28 where did it go in? :/ why didn't CIA report it here? Jul 10 08:58:33 it's looks mightily resolved over here :) Jul 10 08:58:51 ah. Jul 10 08:59:06 it's in monotone.openembedded.org - org.openembedded.dev Jul 10 08:59:13 (since 5 minutes ago) Jul 10 08:59:38 :-p Jul 10 08:59:48 nice work :) Jul 10 09:00:12 your bug reminded my to fix some stuff in an other branch Jul 10 09:00:53 j #openslug Jul 10 09:01:07 good morning eFfeM Jul 10 09:01:07 koen: 1.5M larger are spitz images Jul 10 09:01:19 hi koen Jul 10 09:01:32 hrw|work: right, my ipaq kernels build everything and the kitchensync :) Jul 10 09:01:49 so that is > 1.5M Jul 10 09:02:04 need to patch kernel too Jul 10 09:02:40 eFfeM: judging from 2.6.18-rc1 thumb is broken in 2.6.17, so that could explain the ucslugc problems Jul 10 09:02:59 hrw|work: if it's only 1.5M akita could have those as well Jul 10 09:03:17 I'm surprised CoreDump|home didn't think of this :) Jul 10 09:04:02 koen, strange... if there is asm in a file I can imagine it is incorrect, but I fail to see how a kernel (or C code) can break the code generation of a compiler Jul 10 09:04:30 koen: 1.5M tarball is larger - how much it will take on device? Jul 10 09:04:37 eFfeM: ah, it seems that are multiple problems Jul 10 09:04:46 btw is it ok if /etc/ld.so.conf does not contain /lib or /usr/lib ? Jul 10 09:04:58 eFfeM: iirc those are in by default Jul 10 09:04:58 koen do you have a url or so ? Jul 10 09:05:10 ah, ok Jul 10 09:05:13 eFfeM: see effem-log Jul 10 09:05:34 good. wasn;t that far yet (my day just started :-) Jul 10 09:05:53 koen: lets check akita.. Jul 10 09:06:03 eFfeM: http://kernel.org/git/?p=linux/kernel/git/pavel/zaurus.git;a=commit;h=5247593c9634309d1b9f7b549495b8e5ad521688 Jul 10 09:06:07 koen: is it possible that your last commit didn't get reported by CIA because it created a second head? after updating, mt update is giving me yours and xora's last commit as candidates... Jul 10 09:06:15 also looking at the root cause for Couldn't find shared library provider for libc.so.6 Jul 10 09:06:41 * zwelch is still learning monotone, so maybe he just needs to read more about its branch management Jul 10 09:06:43 zwelch: the merge should also be there Jul 10 09:07:29 is it possible for me to have synced in between your commit and the following merge? Jul 10 09:07:53 * zwelch would find that... undesirable, but potentially sane Jul 10 09:08:49 because sure enough, another pull fixed it :) Jul 10 09:09:01 :) Jul 10 09:09:22 koen, checked the url, this one basically causes system calls to work. However the problem I had was that invalid asm was generated. didn't even get to building ... Jul 10 09:09:45 eFfeM: that is indeed a different problem Jul 10 09:09:49 * zwelch sacrifices another small furry animal in the name of non-determinisms Jul 10 09:10:15 there were a few minutes between the commit and the merge Jul 10 09:10:44 there was a reference to PIC related code, but according so someone here this was already fixed a while ago Jul 10 09:10:52 might dig into this later Jul 10 09:11:04 03koen 07org.oe.packaged-staging * r349030d4... 10/classes/ (base.bbclass package.bbclass): base.bbclass, package.bbclass: shuffle around tasks to make staging run after packaging Jul 10 09:11:09 03koen 07org.oe.dev * r41c57842... 10/packages/libtool/libtool-native_1.5.10.bb: libtool-native: stage libltl files, courtesy Zachary T Welch, closes #1157 Jul 10 09:11:17 hehe Jul 10 09:11:35 I was about to kick the ciabot Jul 10 09:11:36 * zwelch supposes he should take this as further evidence that he lacks patience :) Jul 10 09:11:48 i am _so_ in the right career ;) Jul 10 09:18:39 hm, where am I supposed to find ipkg_build Jul 10 09:19:07 s/_/-/ Jul 10 09:19:14 eFfeM: ipkg-utils(-native) Jul 10 09:19:48 ok, ipkg whatprovides ipkg-build didn't find it, thanks Jul 10 09:20:18 koen: 2.5M of flash is used by kernel modules Jul 10 09:20:24 extra 2.5 Jul 10 09:21:08 that's a bit much Jul 10 09:21:46 it that case I'd say 'just ship usbnet, rtl8150 and pegasus' Jul 10 09:21:57 that should satisfy the usb-ethernet people Jul 10 09:22:11 and maybe the zd wifi driver Jul 10 09:22:22 I'm writing to oz-devel Jul 10 09:26:15 so, i'm discovering that refactoring .bb functionality into a new bbclass is non-trivial Jul 10 09:26:22 let's hope gmane doesn't take hours to deliver that message Jul 10 09:26:32 koen: want cc? Jul 10 09:26:42 hrw|work: nah, I can wait :) Jul 10 09:27:15 I dislike getting private copies of mailinglists I'm subscribed to Jul 10 09:27:28 ~lart people that use 'reply-all' Jul 10 09:27:29 * ibot stabs people that use 'reply-all' Jul 10 09:27:35 ko Jul 10 09:27:36 ibot: botsnack Jul 10 09:27:36 aw, gee, koen Jul 10 09:28:00 that's the downside of gmane, I can't set a reply-to header Jul 10 09:36:34 have been googling and learned that the msg Jul 10 09:36:36 NOTE: Couldn't find shared library provider for libc.so.6 Jul 10 09:36:41 is rather harmless Jul 10 09:36:52 however would like to know if there is a way around this Jul 10 09:38:09 koen: You could fix your mail filter rules ;-) Jul 10 09:38:17 koen, ipkg-utils did the trick for me, had to use --force-overwrite because both ipkg-utils and ipkg-link provide ipkg-link Jul 10 09:38:27 RP: that's a bit hard on nntp lists Jul 10 09:39:32 koen: Adding a filter to your email to filter any copies that were sent to the list address and your personal one isn't possible? Jul 10 09:40:09 sure that's possible Jul 10 09:40:31 that doesn't change the fact that people should click reply-all everytime they get mail :) Jul 10 09:40:35 ehm Jul 10 09:40:39 shouldn't Jul 10 09:40:55 koen: Personally, I find two copies useful as I don't read every message to every list I'm on Jul 10 09:41:02 hi zecke Jul 10 09:41:20 hey zecke Jul 10 09:41:34 koen: and if somene dislikes it, they do have the option of filtering... Jul 10 09:42:23 yeah, the problem is that some people do not read the list; a reply to all covers everyone, but reply to the list covers only those that follow the list Jul 10 09:42:38 yeah, I'll proably start sending mail with a non-existant address Jul 10 09:42:50 that said, i'll be sure to trim you from any threads :) Jul 10 09:43:03 or make a rule that mails reply-all people with "don't reply-all me" Jul 10 09:43:47 I usually press reply-all and remove all addresses that I know that are subscribed Jul 10 09:43:56 not 100% foolproof Jul 10 09:44:02 that becomes a general selection problem of "don't reply to koen if he will receive it through one of these means" Jul 10 09:44:09 RP: did you saw recent pavel patch for collie charging? Jul 10 09:45:03 and yeah, i use the same heuristic; the question is, how do we teach our mailers to do that? ;) Jul 10 09:45:43 hrw|work: The series of four patches which annoyed me as they were badly thought out? Jul 10 09:46:16 hrw|work: did you notice the locomo bits in 2.6.18-rc1? Jul 10 09:46:41 koen: If you send an auto response, I'd just filter it out ;-) Jul 10 09:46:53 :) Jul 10 09:47:33 RP: that one is not yet commented by you Jul 10 09:48:36 morning Jul 10 09:48:55 hrw|work: Right. I guess I should ack it. It stikes me as hacks on top of more hacks but what can I do... Jul 10 09:49:03 hi mickey|thesis Jul 10 09:50:38 hi mickey|thesis Jul 10 09:56:31 03xora 07org.oe.dev * rf3b9f2a0... 10/packages/freetype/ (4 files in 2 dirs): Jul 10 09:56:31 freetype-native_2.2.1.bb, freetype_2.2.1.bb : new versions of stable Jul 10 09:56:31 I did these when trying to track down problems with alignment errors so Jul 10 09:56:31 I might as well check them in. Jul 10 09:56:31 DEFAULT_PREFERENCE = "-1" for now as these are pretty core and could Jul 10 09:56:31 do with review. Jul 10 10:00:21 oooh Jul 10 10:00:49 mickey|thesis: what about 'pretty-core' as a name for the stable branch? Jul 10 10:01:06 koen: .dev .stable Jul 10 10:01:16 clean names are better Jul 10 10:01:49 RP: http://www.oesf.org/forums/index.php?showtopic=20424 Jul 10 10:01:59 anyone feels like sending the reminder about going readonly soon? Jul 10 10:02:39 zecke: good plan, I had forgotten Jul 10 10:04:07 koen: *nod* Jul 10 10:04:08 ;) Jul 10 10:04:24 zecke: i can do that. Jul 10 10:04:32 * mickey|thesis does Jul 10 10:04:52 mickey|thesis: oh well, we have until the 13th Jul 10 10:05:28 ok, then I'll send it on 13th Jul 10 10:05:53 mickey|thesis: the good thing is. We can exploit any stock Qt and Qtopia phone Jul 10 10:06:04 mickey|thesis: I have at least three different working mechanisms Jul 10 10:06:25 mickey|thesis: I'm sure Operators will love this, specially because patches have been send years ago... Jul 10 10:13:27 uh oh... Jul 10 10:13:32 btw. Jul 10 10:13:47 any connections to ROAD still open? I'm a bit worried it's still not out. Jul 10 10:13:54 hrw|work: replied Jul 10 10:15:34 mickey|thesis: a new thought for a phone gui: "say no to scrollbars in portrait mode" Jul 10 10:16:49 koen: ya, we need some dogmas. Did you read my post to ZUG in response to amrein? Jul 10 10:17:05 mickey|thesis: yes Jul 10 10:17:20 let's start a research institute ;) Jul 10 10:17:39 * kergoth yawns Jul 10 10:17:40 hey all Jul 10 10:17:45 hey kergoth Jul 10 10:18:00 yo kergie! Jul 10 10:18:02 isn't it fricking early in AZ? Jul 10 10:18:11 morning kergoth Jul 10 10:18:19 koen: *nod* I'm working on that. give me two years and then let's see. how far are you from graduating? Jul 10 10:18:22 fricking late, actually Jul 10 10:18:24 3:45am Jul 10 10:18:34 :) Jul 10 10:18:40 mickey|thesis: ~2.5 years Jul 10 10:18:51 hi all Jul 10 10:19:14 how is the setting for BINARY_LOCALE_ARCHES supposed to look if i want to have de, en, fr utf8 locales? Jul 10 10:20:34 florian_kc: You can't limit the locales that are generated atm - its all or nothing Jul 10 10:20:51 and we have fscked utf8 generation iirc Jul 10 10:20:53 uh Jul 10 10:20:54 RP: i thought you had a patch for that, no ? Jul 10 10:21:11 http://www.rpsys.net/openzaurus/temp/glibc_limit_locale-r1.patch Jul 10 10:21:18 sort of works :-/ Jul 10 10:21:31 iirc you can choose which locales end up in the image Jul 10 10:21:35 (its an updated verison of the really old one that was around) Jul 10 10:21:45 IMAGE_LINGUAS Jul 10 10:21:55 ah Jul 10 10:22:03 IMAGE, not IMG Jul 10 10:22:11 that's why it didn't show ip in grep Jul 10 10:22:15 up* Jul 10 10:22:42 mickey|thesis: no, not much Jul 10 10:22:49 * RP set that in poky recently :) Jul 10 10:22:51 use uclibc and you don't have these problems with generating :) Jul 10 10:23:25 * ade|desk lies through back teeth w.r.t ARM Jul 10 10:26:19 RP: okay, many thanks Jul 10 10:58:05 woglinde: hey Jul 10 10:58:10 he zecke Jul 10 10:58:20 zecke are you at the university? Jul 10 10:59:20 woglinde: tomorrow, I will be Jul 10 10:59:24 woglinde: how is your son? Jul 10 10:59:52 fine Jul 10 11:00:05 yesterday he has the first "wachstumschub" Jul 10 11:00:13 so it was a little bit stress Jul 10 11:02:14 zecke: hi ! Jul 10 11:02:37 woglinde: wachwhat/ Jul 10 11:02:38 ? Jul 10 11:02:45 hrw hm Jul 10 11:02:49 growsoemthing Jul 10 11:02:59 he needed more food to grow fast Jul 10 11:03:00 ah Jul 10 11:03:07 dontknow the english word Jul 10 11:03:27 woglinde: no problem - I do not know this kind of dictionary too Jul 10 11:03:29 growth spurt ? Jul 10 11:03:38 ade hm maybee Jul 10 11:05:01 but after that all was fine Jul 10 11:05:25 growth spurt is right Jul 10 11:05:32 will remember thius Jul 10 11:06:32 is it correct to report OpenZaurus bugs at bugs.openembedded.org, as noted at the #openzaurus topic? The page says to choose between OE and BB bugs only Jul 10 11:08:05 DPRoldan: Apparently so, and you should specify the version of OZ you are using in the bug entry form Jul 10 11:08:52 ok, thanks Jul 10 11:11:01 Why would I get "../sysdeps/unix/sysv/linux/bind.S:5: Error: symbol `__bind' is already defined" when compiling glibc within OE? Jul 10 11:14:13 I've deleted the build directory, downloaded a new OE.db and rebuilt from scratch, and I still get this error Jul 10 11:15:06 * KerwoodDerby sighs and goes to work Jul 10 11:15:12 kerwood which glibc? Jul 10 11:44:59 03xora 07org.oe.dev * rf3515be6... 10/packages/mplayer/mplayer_1.0pre8.bb: Jul 10 11:44:59 mplayer_1.0pre8.bb : remove dependancy on libungif as this causes an unwanted Jul 10 11:44:59 dependancy on X11. Jul 10 11:44:59 Fix the truetype font handling so we can have subtitles again. This was Jul 10 11:44:59 a silly thinko on my part. Jul 10 11:46:45 XorA: another thing to update? Jul 10 11:47:03 03hrw 07org.oe.dev * r08b99a62... 10/packages/opie-help-en/ (opie-help-en.inc opie-help-en_1.2.2.bb): opie-help-en: PACKAGE_ARCH set to 'all', added empty PN package which depend on help browser Jul 10 11:48:48 hrw|work: only if someone moans about no subtitles :-) Jul 10 12:06:41 monotone: multiple update candidates: Jul 10 12:06:41 monotone: 08b99a6246391bc67af5422397346b95d3d3b3e3 hrw@openembedded.org 2006-07-10T11:44:13 Jul 10 12:06:42 monotone: f3515be6618dd23c0a5917c27c9607c052d6df83 xora@openembedded.org 2006-07-10T12:02:06 Jul 10 12:06:48 * koen looks at hrw|work and XorA Jul 10 12:07:17 03hrw 07org.oe.oz354x * r1c094e82... 10/packages/mplayer/mplayer_1.0pre8.bb: mplayer: disabled GIF support - close #1147 Jul 10 12:07:21 03hrw 07org.oe.oz354x * r3ec85f41... 10/packages/opie-help-en/ (opie-help-en.inc opie-help-en_1.2.1.bb): opie-help-en: PACKAGE_ARCH set to 'all', added empty PN package which depend on help browser Jul 10 12:07:36 koen: looks like it was hrw's fault.... Jul 10 12:08:12 Zero_Chaos: the timestamps don't say when it was pushed to the server Jul 10 12:08:35 which makes blaming people a bit harder :) Jul 10 12:09:30 koen: I know, but CIA-2 just said hrw pushed an update.... which means it is more likely his fault. Jul 10 12:09:39 koen: what do those timestamps mean? Jul 10 12:10:01 the timestamps note the commit data Jul 10 12:10:15 and I was talking about .dev, not .oz354x Jul 10 12:10:25 ahh Jul 10 12:11:16 koen: fixed Jul 10 12:12:30 if I add a package to an image, let's say opie-image, and it's compiled but not included in the image, what am I doing wrong? Jul 10 12:14:56 katossi: what section are you adding it to? Jul 10 12:16:01 the real case is I added packages to task-gpe.bb, under gpe-task-apps Jul 10 12:17:46 katossi: hmm, never done it with gpe, but I know in opie-image files I had to add an app in several places. One to build, then one to install Jul 10 12:18:25 Zero_Chaos: but it should work as long as the app is in DEPEND and RDEPEND, shouldn't it? Jul 10 12:18:52 katossi: no Jul 10 12:19:59 Zero_Chaos: what did you change for the install part? Jul 10 12:20:19 katossi: I used opie, but let me take a quick look Jul 10 12:22:25 katossi: gpe stuff looks way different than opie, sorry, can't help :-( Jul 10 12:22:56 too bad, anyone with gpe experience? Jul 10 12:23:39 Offhand, does anyone know the key combination to boot to the recovery shell on a Z? Jul 10 12:27:05 katossi: make sure you bump the PR of the task package you're altering Jul 10 12:27:20 RP: D+B Jul 10 12:28:36 koen: what does that mean exactly? Jul 10 12:29:26 that you bump the PR of the task package you are altering Jul 10 12:29:56 merged, pushed Jul 10 12:30:37 what I mean is why? what changes does that produce? Jul 10 12:31:05 mickey|lunch: another dogma: "persons are first class objects" Jul 10 12:31:19 katossi: otherwise the old version will still be in tmp/deploy/ipk and new dependencies you added wont be used Jul 10 12:32:12 ohh! cached stuff always gives me headaches! so that's why sometimes everything behaves as if I had changed basically nothing Jul 10 12:35:09 koen: thx a lot!! saved me a lot of time Jul 10 12:43:10 hrw|work: thanks. Its a long time since I've needed that :) Jul 10 12:49:22 RP: what you fscked? Jul 10 12:50:48 hrw|work: No me, trying to help someone else work out why an image doesn't work. Of course I'd forgotten 2.4 can't read 2.6 jffs2 partitions... Jul 10 12:50:48 hello Jul 10 12:50:57 hi cbrake Jul 10 12:51:00 hey cbrake Jul 10 12:51:33 16883 hrw 0 0 335m 108m 20m S 15.8 14.4 40:08.10 firefox-bin Jul 10 12:51:45 updated zlib is also in 2.6.18rc1 :) Jul 10 12:51:46 WHEN firefox will stop leaking... Jul 10 12:51:54 hrw|work: when pigs fly Jul 10 12:51:59 hrw|work: when kill -9 firefox-bin Jul 10 12:52:57 16883 hrw 35 0 185m 56m 16m R 71.4 7.5 40:32.53 firefox-bin Jul 10 12:53:01 restarted Jul 10 12:57:32 grr fontconfig is not using cross toolchain Jul 10 12:58:05 ade|desk: the one XorA added today? Jul 10 12:58:26 2.3.91-r0 Jul 10 12:58:34 what did XorA add ? Jul 10 12:59:24 2.3.95 Jul 10 13:01:05 fontconfig-2.3.91/fc-cache/fc-cache.o: ELF 32-bit LSB relocatable, ARM, version 1 (SYSV), not stripped Jul 10 13:01:06 * XorA denies all Jul 10 13:01:17 "This version seems to cure the masses of alignment errors I was getting Jul 10 13:01:17 using previous versions." -- cos its using host toolchain and hence linking to host /lib/libc not the arm one he was expecting, sadly the symbols seem to almost match so not so easy to spot Jul 10 13:01:58 ade|desk: I doubt that very much Jul 10 13:02:09 http://ewi546.ewi.utwente.nl/tmp/fc-o.txt <- output of file Jul 10 13:02:17 hmm /usr/lib/gcc/i486-linux-gnu/4.0.4/../../../../lib/crt1.o: In function `_start': Jul 10 13:02:18 | ../sysdeps/i386/elf/start.S:109: undefined reference to `__libc_csu_fini' Jul 10 13:02:18 | ../sysdeps/i386/elf/start.S:110: undefined reference to `__libc_csu_init' Jul 10 13:02:18 | ../sysdeps/i386/elf/start.S:119: undefined reference to `__libc_start_main' Jul 10 13:02:18 | collect2: ld returned 1 exit status Jul 10 13:02:21 it uses the host toolchain to compile the stuff it needs to run Jul 10 13:02:33 thats hiting the wrong linker Jul 10 13:04:47 koen: sorry why is it using doing this for fc-case.0,fc-lang.o and fc-glyphname ? Jul 10 13:05:06 ade|desk: it runs those during build.... Jul 10 13:05:38 thats a bit naff Jul 10 13:05:44 it is Jul 10 13:06:12 I have moved OE to another computer. Now again, no more bitbakeing: http://pastebin.ca/84045 and http://pastebin.ca/84046 Jul 10 13:06:12 I completely removed everything in tmp and retried. The problem is still there. Jul 10 13:06:19 almost 3600 .bb files Jul 10 13:06:37 * koen remembers the sub 600 days Jul 10 13:06:49 koen: There are a couple of Japanese bb files waiting in the tracker if you want to push the limit. Jul 10 13:08:01 koen: i remember the days when kergoth was adding 2 or 3 files a day , so sub ~100 i guess Jul 10 13:10:23 * RP remembers the 2.6.7 zaurus kernel .bb :) Jul 10 13:10:42 RP: I remember 2.6.5 D: Jul 10 13:11:00 hrw|work: I'm not sure that had a .bb though ;-) Jul 10 13:11:15 hrw|work: I still have the patches... Jul 10 13:11:22 RP: http://oe-devel.bkbits.net:8080/openembedded/search/?expr=2.6.5&search=ChangeSet+comments Jul 10 13:11:34 "clean up 2.6.5 .oe file and apply Richard Purdie's patch" Jul 10 13:11:59 i miss BK Jul 10 13:12:01 ChangeSet 1.1165.1.2 Jul 10 13:12:01 upgrade kernel to 2.6.5-jl1 Jul 10 13:12:09 hi france Jul 10 13:12:17 Is binutils-cross-2.17 broken? Jul 10 13:12:19 hmm... cool. those ancient days Jul 10 13:12:31 and I remember days when OE images was not usable on collie... Jul 10 13:12:43 hrw|work: I must have gained commit rights for 2.6.7 which is what I'm confused with :) Jul 10 13:13:07 I like mickeyl's comment about someone needing to update openzaurus.bkbits.net :) Jul 10 13:13:54 where? :) Jul 10 13:13:57 Laibsch: they work fine over here Jul 10 13:14:19 Laibsch: you do need glibc 2.4 to use it Jul 10 13:14:44 * mickey|thesis looks at the All Changesets statistics Jul 10 13:14:46 mickey|thesis: Linked off one of the commits hrw posted - you were hinting someone should look after the kernel bk tree. I remember trying and deciding patches were easier Jul 10 13:14:55 RP: ah right Jul 10 13:14:58 The new machine has dapper. The pevisous one was Debian stable. So it should be recent enough. I'll check. Jul 10 13:15:30 mickey|thesis: http://oe-devel.bkbits.net:8080/openembedded/cset%401.1706.47.1 Jul 10 13:15:51 hehehe Jul 10 13:16:17 doh newer fontconfig is the same :( Jul 10 13:16:19 Laibsch: glibc 2.4 for your target Jul 10 13:16:25 RP: it took me a while to get accustomized to that patches workflow. Jul 10 13:16:43 RP: I love it nowadays though Jul 10 13:16:47 koen: or uclibc svn Jul 10 13:17:09 mickey|thesis: It took a while to stabilse but I think it was the right choice - look at the problems Pavel has with his git tree :-/ Jul 10 13:17:12 koen: I see. But shouldn't OE take care of building that? Jul 10 13:17:27 RP: yeah Jul 10 13:18:12 Laibsch: OE does what you tell it to do Jul 10 13:18:37 mickey|thesis: Its lovely to think that two years on we're in the position we're in :) Jul 10 13:18:40 RP: and how awful are his 'broken-out' diffs.. Jul 10 13:19:08 hrw|work: exactly Jul 10 13:19:19 NOTE: package binutils-cross-2.17-r0: task do_build: completed Jul 10 13:19:21 RP: no kidding. i can very well remember the pain I felt when me and kergoth were the only ones working on OE Jul 10 13:19:23 koen: I said bitbake qpf-unismall and bitbake kaffe-qt Jul 10 13:20:00 but no directory containg glibc in tmp Jul 10 13:20:02 mickey|thesis: I suspect kergoth will be getting back into things a bit as well which will be good :) Jul 10 13:20:18 03koen 07org.oe.dev * r11318244... 10/conf/distro/angstrom-2006.9.conf: angstrom 2006.9: use font tools without alignment traps Jul 10 13:21:21 RP: so every OE can take a monthlong holiday while mickey|thesis and kergoth`zzz are slaving away ;) Jul 10 13:21:32 for that real vintage development feeling Jul 10 13:21:59 heh Jul 10 13:22:27 * RP still needs that ellusive holiday Jul 10 13:22:28 only a month :( Jul 10 13:23:34 koen: When I say "bitbake glibc" it first tries to build binutils-cross-2.17-r0. Something wrong in the dependencies? Jul 10 13:23:58 Laibsch: no, something is probably wrong in your setup Jul 10 13:24:33 It was working before. Jul 10 13:24:45 Wait, I have an idea what I might have overlooked. Jul 10 13:28:05 you have updated Jul 10 13:36:38 koen: well thats one way to get review :-) Jul 10 13:36:58 XorA: yeah Jul 10 13:38:47 koen: My path to OE changed. I was thinking that maybe I had forgotten to change it somewhere. rgrep only returned one instance: MT/debug. Can that have an effect? I ask because I delete all of tmp again since there were some occurences in there as well of the old path. Jul 10 13:39:02 So it will be a while before I know if this worked or not. Jul 10 13:41:24 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Jul 10 13:47:19 mickey|thesis: thanks for spotting Jul 10 13:51:02 03koen 07org.oe.dev * re0113226... 10/conf/distro/angstrom-2006.9.conf: angstrom 2006.9: make the typo police happy Jul 10 13:56:17 morning Jul 10 13:56:52 hey chouimat Jul 10 14:06:10 hi all Jul 10 14:44:56 * mrz80 is back (gone 01:03:33) Jul 10 14:53:49 koen: thanks for updating e-wm, I wans't aware that we hade u-a for x-window-manager :-) Jul 10 14:54:23 JustinP: me neither, till I noticed coredump's checkin :) Jul 10 14:54:26 JustinP: its new stuff Jul 10 14:54:46 JustinP: I was too lazy to teach entrance new tricks Jul 10 14:55:56 hrw|work: I noticed a dependency problem with taks-bootstrap for oz 3.5.4.1 Jul 10 14:56:33 hrw|work: specifically, usbuils and oz-bootsplash are not built from a new tmp Jul 10 14:57:22 hmm.. they are in BOOTSTRAP_EXTRA_RDEPENDS so should be.. Jul 10 14:57:27 koen: If you have any kind of advice what I should be looking for. I'd be all ears. Compilation failed again. Jul 10 14:57:50 hrw|work: I have a fix, mind if I push? Jul 10 14:57:51 hrw|work: or could you rather a patch? (I just add DEPENEDS to match the RDEPENDS) Jul 10 14:58:24 JustinP: depends should not be needed rather... Jul 10 14:58:54 hrw|work: well...they weren't built. and IIRC the oz354x branch doesn't use the auto-RDEPENDS => DEPENDS thing.... Jul 10 14:59:33 my image built just fine after I added the DEPENDS Jul 10 14:59:49 then please patch it Jul 10 15:03:06 JustinP: I sometimes forgot about those DEPENDS<>RDEPENDS.. Jul 10 15:03:25 hrw|work is too busy costumizing twm on his Z ;) Jul 10 15:04:27 koen: developing opie-notes using qvfb under x/qt2 running twm is awesome Jul 10 15:04:44 :) Jul 10 15:04:49 koen: we need gvim Jul 10 15:05:36 hrw|work: puched Jul 10 15:05:49 hrw|work: sound cool :) Jul 10 15:05:50 thx Jul 10 15:06:01 hr I'm also noticing some DEPENDS problems with gpe-*something* relying on libtodo. I'll update that as well if I find a fix Jul 10 15:06:09 03justinp 07org.oe.oz354x * r1a6bbf6b... 10/conf/distro/openzaurus-3.5.4.1.conf: openzaurus-3.5.4.1.conf: add DEPENDS to match RDEPENDS Jul 10 15:06:18 JustinP: gpe-timesheet probably Jul 10 15:06:55 yeah Jul 10 15:16:03 hi Jul 10 15:17:49 CoreDump|home, mornin Jul 10 15:18:25 hey CoreDump|home Jul 10 15:19:42 anyone motivated to fix fluxbox NET_WM implementation ( sending _NET_WORKAREA on workarea rezise / xrandr rotation)? Jul 10 15:20:11 #fluxbox's op is an idiot or I'd go there Jul 10 15:20:22 CoreDump|home: tried openbox? Jul 10 15:20:39 steal code from matchbox/metacity/etc? Jul 10 15:20:46 cu Jul 10 15:20:59 koen: I tried making sense of it and failed =) Jul 10 15:21:03 hrw|gone: bye Jul 10 15:21:38 03justinp 07org.oe.oz354x * r5cbb5dff... 10/packages/gpe-timesheet/gpe-timesheet_0.30.bb: gpe-timesheet: add missing libtododb DEPENDS Jul 10 15:35:07 * jkp waves Jul 10 15:36:27 hi Jul 10 15:36:31 * JustinP beaches Jul 10 15:43:27 will OM/PI be implemented in the dev branch? Jul 10 15:44:27 unlikely. for the time being, we recommend opie-mail which is the "upstream" version of OM/PI. Jul 10 15:45:06 ok Jul 10 15:49:29 mickey|thesis: good thing that OE is mit, if it was gpl, there would have been a OE/PI ;) Jul 10 15:49:43 koen: heh. most likely, yeah :D Jul 10 15:50:02 some guys have a tendency to just drag everything GPL into their own tree... Jul 10 15:50:12 must be some fancy collector's idea or whatever Jul 10 15:50:40 *cough* MVista *cough* Jul 10 15:51:56 btw: http://www.openembedded.org/filebrowser/org.openembedded.dev/conf/machine Jul 10 15:51:59 more eye-candy Jul 10 15:52:15 less scare for managers ;) Jul 10 15:53:02 nice work Jul 10 15:53:26 wget plugin ; untar :) Jul 10 15:53:58 drupal plugin? Jul 10 15:54:22 goxboxlive: zautrix is behaving unethical Jul 10 15:54:34 mickey|thesis: yes Jul 10 15:54:39 ~zautrix Jul 10 15:54:42 koen: cool. i wish i had something like that for wordpress, i.e. for a better version of http://www.vanille-media.de/site/index.php/musician/chord-pro/ Jul 10 15:55:00 ~emulate zautrix Jul 10 15:55:02 this is exactly what the GPL is about. grab the stuff and run away Jul 10 15:55:22 mickey|thesis: I'm sure wordpress have similar functionality Jul 10 15:55:30 mickey|thesis: well. The GPL is there to stop unethical things like unnecessary forks Jul 10 15:55:40 zecke: is it? how so? Jul 10 15:55:58 mickey|thesis: the first version of the EMACS License demanded all patches to be send to RMS Jul 10 15:56:21 well. i guess it boils down to who judges the necessity of a fork. Jul 10 15:56:26 zecke: nope, the GPL was created to _enable_ forks Jul 10 15:56:56 don't like what upstream does? just fork the damn thing ;) Jul 10 15:57:00 CoreDump|home: oh well Jul 10 15:57:12 CoreDump|home: I have just finished reading a/the stallman biography Jul 10 15:57:40 CoreDump|home: I can assure you, the GPL was not written to allow stuff like Lucid EMACS Jul 10 15:57:55 I'm not familiar with that, sorry Jul 10 15:58:14 but w/o a GPL-type license, forks would be basically impossible Jul 10 15:58:23 CoreDump|home: RMS mastermind of the GPL, would find zautrix's behaviour unethical Jul 10 15:58:34 CoreDump|home: and RMS did not write the GPL to further unethical behaviour Jul 10 15:58:44 CoreDump|home: so your statement is just plain wrong Jul 10 15:58:55 mickey|thesis: should I send you the book? Do you have time for reading? Jul 10 15:59:17 zecke: no time atm. can you keep it until we meet later this year? Jul 10 15:59:34 mickey|thesis: sure, I think I will ask eilers Jul 10 15:59:50 ~praise _law_ for adding multi-wm support to gpe-login Jul 10 15:59:52 All hail _law_ for adding multi-wm support to gpe-login! Jul 10 16:01:31 /query ibot Jul 10 16:01:45 sorry Jul 10 16:02:19 /l33t_h4x0r_DDOS_scrip03 mr_nice_slacker Jul 10 16:02:24 oh sorry =) Jul 10 16:02:55 *g Jul 10 16:03:23 and my all-time-favorite: Jul 10 16:03:34 /identidy CoreDump|home some-password Jul 10 16:03:56 hehe Jul 10 16:07:32 grrr Jul 10 16:07:38 NOTE: package mysql-native-4.1.18-r1: task do_fetch: failed Jul 10 16:08:37 zecke: i unfortune have killed my working mq200 on your 2.6 Kernel by pass console=/dev/console to the kernel. 2.4 mq200 is working as normal. now i try to figgure out what console=/dev/console made with my SIMpad. I will report if i could find some answer. Jul 10 16:10:58 no idea Jul 10 16:11:46 zecke:np Jul 10 16:13:05 console=/dev/console looks bogus anyway Jul 10 16:13:14 console=tty1 makes more sense iMO Jul 10 16:29:47 I'm getting 'Error: symbol __bind is already defined' while compiling glibc. Jul 10 16:30:08 bronson hm you are not alone Jul 10 16:30:20 Is DISTRO=openzaurus-unstable not a good idea? Jul 10 16:30:38 bronson: did you google for this error? Jul 10 16:30:40 woglinde, well, that's something at least. Jul 10 16:30:46 zecke, I did. Found one thing relevant: Jul 10 16:30:56 bronson: is that umamumba or mumba platform? Jul 10 16:31:03 http://www.handhelds.org/hypermail/oe/71/7132.html Jul 10 16:32:31 zecke, umamumba? mumba? Jul 10 16:33:12 https://bugzilla.ubuntu.com/show_bug.cgi?id=20792 ?? Jul 10 16:33:53 bronson: well. You come here, specify no information and I think you expect an answer Jul 10 16:33:58 bronson: this can't work Jul 10 16:34:37 zecke, I'm trying to understand what you said. Jul 10 16:34:42 What's mumba? Jul 10 16:34:51 bronson: Is that your TARGET_ARCH? Jul 10 16:35:01 heh - umamumba.. :) Jul 10 16:35:16 zecke, no, it's arm. Jul 10 16:35:21 CoreDump|home: hey, may I message you? Jul 10 16:35:31 bronson: what version of gcc, glibc, binutils Jul 10 16:35:31 fire away Jul 10 16:35:44 bronson: did it ever work Jul 10 16:35:54 zecke, no, this is my first attempt. Jul 10 16:36:00 bronson: pst - what he's looking for is a full bug report. Jul 10 16:36:12 CosmicPenguin: may I message you? Jul 10 16:36:20 sure Jul 10 16:37:29 bronson: okay. Could you take a look and see what versions of binutils and gcc it build Jul 10 16:38:00 zecke, gcc-4.1.1, glibc-2.3.5+cvs20050627, binutils-2.17 -- all pulled in via openzaurus-unstable distro. Jul 10 16:38:25 (took me a bit to figure out how to answer that :) Jul 10 16:38:42 bronson: hehe Jul 10 16:39:59 anyone around having knowledge on usb audio? Tried to reinstall my audio dongle, thought I had done everything ok, but speakertest gives me Jul 10 16:40:02 Broken configuration for playback: no configurations available: Invalid argument Jul 10 16:40:02 Setting of hwparams failed: Invalid argument Jul 10 16:40:25 the card is there /dev/snd/* is there /proc/asound lists the card Jul 10 16:41:03 i don't need oss modules, do I? Jul 10 16:47:26 sorry need to head home Jul 10 16:48:13 bye zecke Jul 10 16:48:47 bye Jul 10 17:44:36 * JustinP generates his locales Jul 10 17:59:30 there is no way to build synergy against oz 3.5.4.1 default X ? Jul 10 18:01:01 m4gnet: last I checked you couldn't build synergy against diet-x11, no Jul 10 18:01:14 m4gnet: as long as you build full x it will work, though Jul 10 18:01:42 Id like to keep a light X. Jul 10 18:02:52 I tried to compile it on the Z, all dependencies where ok, but I had compilations errors :( Jul 10 18:05:19 hi kergoth Jul 10 18:28:36 m4gnet: yeah, there are pieces missing Jul 10 18:30:00 too bad synergy would have been nice. Jul 10 18:39:59 what new syntax has been added to .bb since i was around? 'require'.. anything else? Jul 10 18:40:40 kergoth: I'm not aware of any changes other than that Jul 10 18:40:54 k, just making sure the vim syntax file is up to date :) Jul 10 18:42:54 kergoth: .= an =. Jul 10 18:43:06 s/an/and/ Jul 10 18:44:29 hi koen Jul 10 18:44:38 those are raw concatenation, right? lacking the space seperator? Jul 10 18:45:14 kergoth: iirc yes Jul 10 18:45:18 k Jul 10 18:45:26 kergoth: pb_ updated the bitbake manual after adding those Jul 10 18:52:20 heh, i sure didnt know vim that well when i did this syntax file Jul 10 18:52:31 i should redo this now that i do evil things with vim on a regular basis Jul 10 18:52:52 kergoth: btw, do you have a monotone key already? Jul 10 18:53:05 i dont think so, i'll generate one Jul 10 18:53:07 where should i send it? Jul 10 18:53:14 i used to, but its probably long gone Jul 10 18:53:57 mickey@vanille.de and koen@dominion.kabel.utwente.nl Jul 10 18:54:08 let me check if your old key is in the .db Jul 10 18:54:41 it's not in the main db Jul 10 18:54:52 so kergoth@openembedded.org should be safe Jul 10 18:55:45 i could be crazy and generate one using a real email Jul 10 18:55:48 craaaazy Jul 10 18:58:18 it has to be @openembedded.org Jul 10 19:00:43 that's silly, at best. Jul 10 19:00:48 JustinP: take a look at http://www.openembedded.org/the-people-behind-openembedded Jul 10 19:01:11 now tell me all keys end with @openembedded.org Jul 10 19:01:50 koen: or nslu2 Jul 10 19:01:57 koen: I was told I had to have @oe Jul 10 19:02:49 I was just mentioning the policy that I was told back in the day Jul 10 19:03:19 s/policy/convention/ ;) Jul 10 19:03:30 * kergoth yawns, goes to get more caffeine Jul 10 19:03:41 I'd accept cris@n00b.net as a key for kergoth :p Jul 10 19:04:17 hahah Jul 10 19:04:29 hmm, does CIA in here not notify for the bitbake repository? Jul 10 19:04:41 kergoth: zecke messed up the trigger iirc Jul 10 19:04:48 kergoth: feel free to fix it Jul 10 19:05:10 k Jul 10 19:05:50 at any rate, svn up bitbake && cp -av bitbake/contrib/vim/* ~/.vim/ Jul 10 19:05:59 * kergoth glares at his buildd Jul 10 19:06:02 er, s/d$// Jul 10 19:10:34 k sent the pubkey Jul 10 19:11:25 craaaaazy Jul 10 19:11:40 :P Jul 10 19:12:01 feels good to be playing with oe and bitbake again. Jul 10 19:12:06 * kergoth cracks his neck Jul 10 19:12:56 * RP is waiting for the "who on earth did xyz" comments :) Jul 10 19:13:19 hahah Jul 10 19:13:29 kergoth is back in the house Jul 10 19:13:31 hide your newbies Jul 10 19:13:40 i already started the "alright, who broke XYZ?" ones a few weeks ago :P Jul 10 19:14:00 kergoth: true :) Jul 10 19:14:01 to which people respond "you can do that?" Jul 10 19:14:05 * kergoth rolls eyes Jul 10 19:14:29 kergoth: It was more "someone would do that?" :) Jul 10 19:14:37 hahah, true that Jul 10 19:15:29 kergoth: At least one of those problems is outstanding and needs a total rewrite of the dependency code (although the symptom is fixed in OE but not in poky) Jul 10 19:16:09 ow Jul 10 19:16:47 hi kergoth Jul 10 19:16:52 hey chouimat Jul 10 19:17:07 (the autoconf automake build order issue - turns out to be very painful in the bitbake) Jul 10 19:19:38 * koen updates http://www.openembedded.org/the-people-behind-openembedded Jul 10 19:20:00 CosmicPenguin: ray is free to come up with a new description for that page :) Jul 10 19:20:25 * chouimat swears ... Jul 10 19:20:47 koen: heh - I'll let him know Jul 10 19:20:49 but it sounds good to me Jul 10 19:22:24 NOTE: The key addresses are not (yet) E-Mail addresses, so please don't try to contact us over these addresses - they won't get delivered. .. change that to "NOTE: The monotone key addresses may or may not be valid email addresses. Please do not attempt to use them as such." or something like that. Jul 10 19:22:35 * kergoth wanders off for a bit Jul 10 19:26:26 done Jul 10 19:28:43 RP: OT: my poky builds do work :) Jul 10 19:29:58 bronson: did you get help? Jul 10 19:30:24 zecke, sort of... Jul 10 19:30:41 I'm re-doing with or.gopenembedded.oz354x instead of .dev. Jul 10 19:30:49 I think I should have started with that branch anyway. Jul 10 19:30:54 bronson: yes Jul 10 19:31:17 bronson: I'm still interested in finding the issue :) Jul 10 19:31:23 So, while I think the problem still exists, it's not blocking me... Jul 10 19:31:32 zecke: not locking down binutils version Jul 10 19:31:44 zecke, me too. I think I'm going to run into it again later on today... Jul 10 19:31:49 bronson: so if you find some time, I would appreciate if we look into it Jul 10 19:31:49 zecke: binutils >=2.16 only works with glibc 2.4 Jul 10 19:32:02 koen: oh well Jul 10 19:32:22 zecke: a patch exists, see launchpad Jul 10 19:34:05 koen, would a more proper fix be locking down glibc, or would it be applying the patch? Jul 10 19:34:39 I think applying the patch would be the best option Jul 10 19:35:26 zecke: excellent :) Jul 10 19:36:12 RP: every 20 builds I start from scratch Jul 10 19:38:28 zecke: Is that a sustainable machine load? Jul 10 19:41:38 RP: oh well, it is a P4 machine Jul 10 19:41:51 RP: if we get too high load, I will switch over to a dual xeon in the rack :} Jul 10 19:42:35 zecke: Sounds like a nice rack to have access to :) Jul 10 19:42:51 RP: I'm more scared this firewall facist will play with the firewall again... Jul 10 19:43:17 yeah we have some nice spare hardware Jul 10 19:43:39 * JustinP waits some more for locales to be generated Jul 10 19:43:47 zecke: the budget needed to be spent completely? Jul 10 19:44:33 koen: well, these are machines we 'maintain' or got lended Jul 10 19:45:31 "stress testing" Jul 10 19:48:01 zecke: ah Jul 10 19:48:13 zecke: like the sgi altix and sunfire next to ewi546 Jul 10 19:48:31 koen: oh we have some sunfire's as well :} Jul 10 19:48:31 mostly used for freakishly fast pdflatex sessions :) Jul 10 19:48:57 http://f10.putfile.com/7/19005024357.gif Jul 10 19:48:58 lol Jul 10 19:50:11 RP: I will start .oz354x and .dev builds as well Jul 10 19:51:29 zecke: what's the status of setting up tinderbox at oe.org? Jul 10 19:52:28 status: *busy* Jul 10 19:52:53 koen: I would like to get some major improvements for tinderbox first Jul 10 19:53:10 if we have a pressing need, we could do the setup now Jul 10 19:53:16 can the tinderbox page be 'embedded'? Jul 10 19:53:44 or do I have to annoy hrw again by writing atrocious css and php? Jul 10 19:54:49 you can put it in a frame? :} Jul 10 19:55:14 hrw: *sorry*, I'm stuck with html 3.x Jul 10 19:55:45 yeah and put a xhtml header on top :) Jul 10 19:55:47 yay, packaging locales now Jul 10 19:58:26 * zecke needs to hack some C# now :( Jul 10 19:59:16 c# is pretty fun to use Jul 10 19:59:37 * JustinP goes back to his PHP Jul 10 19:59:51 koen: yeah, sadly mono has issues with the System.Drawing class Jul 10 20:00:01 zecke: never used that Jul 10 20:00:20 zecke: I just patched dashboard to work with my LFS install :) Jul 10 20:00:36 'layout' in Windows Forms is scary Jul 10 20:00:41 worse than java Jul 10 20:00:49 zecke: and those were accepted upstream :) Jul 10 20:08:05 * RP ponders why evoultion needs 80% of my system's memory... Jul 10 20:09:29 RP: evil gnomes Jul 10 20:16:55 hi kergoth, zecke Jul 10 20:23:29 hmm Jul 10 20:23:33 slow news day Jul 10 20:23:47 * koen watches an item about crop circles Jul 10 20:24:26 hi Jul 10 20:24:50 hey Hrw|tosa2 Jul 10 20:27:11 dd checking floppies... Jul 10 20:28:03 hey, everone.. Does anybody know anything about this "ixoth" guy who posted here on saturday? Jul 10 20:28:18 I would like to compare notes on poodle sound stuff, Jul 10 20:29:12 ~seen ixoth Jul 10 20:29:27 ixoth was last seen on IRC in channel #oe, 1d 22h 57m 15s ago, saying: 'goodbye everyone, have to go'. Jul 10 20:29:35 I should really have a look at poodle sound as it shouldn't be that difficult to sort out if you understand ASoC... Jul 10 20:31:04 poodle use i2s insteasd of scoop iirc? Jul 10 20:31:32 RP: I looked at pxa_i2s.c and pxa_i2s_corgi form the 2.4 tree.. It looks like the two main diffenences are 1) Poodle uses locomo instead of scoop and 2) The headphones are either on or off, no micorphone stuff to mess with Jul 10 20:32:40 those files are the only diffenrence on the 2.4 side, BTW Jul 10 20:32:55 hvontres|work: I think poodle has a built in mic? Jul 10 20:33:38 hvontres|work: I guess the locomo bits are the problem? I know little about that bit of it :-/ Jul 10 20:37:13 ~lart kfloppy Jul 10 20:37:13 * ibot shoots kfloppy in the head Jul 10 20:37:33 +nite Jul 10 20:39:48 n8 Zero_Chaos Jul 10 20:58:03 someday I need to learn how to use VI better Jul 10 20:58:32 CosmicPenguin: I feel it from time to time... Jul 10 20:58:56 why there was noone familiar with Vi when I started using unix.... 2 years of pico... Jul 10 21:00:34 KerwoodDerby: teco? Jul 10 21:07:37 emacs Jul 10 21:11:02 RP: hrw-pcmcia-ids-r2.patch was sent to ide people? Jul 10 21:17:49 JustinP: But vi is ALLWAYS there :) Jul 10 21:20:24 hvontres|work: yeah. Thankfully I started with vi(m) ;-) Jul 10 21:21:28 JustinP: Lucky you.. started in '90 with vi and a dumb terminal..;( Jul 10 21:22:48 RP: I plan to sync .dev kernels with .oz354x ones Jul 10 21:22:57 RP: Was that podle sound info any good? I got kicked off right after I posted :( Jul 10 21:24:15 RP: 2.6.17+git is not synced at all with .17? Jul 10 21:24:37 hrw: Not by me Jul 10 21:25:12 ok Jul 10 21:25:33 hrw: Sorry, I mean the ide patch wasn't sent by me Jul 10 21:25:43 .17+git should be synced with .17 Jul 10 21:26:13 git does not have squashfs enabled for example Jul 10 21:26:24 hrw: They should be in sync Jul 10 21:26:54 hvontres|work: Yes, its useful to know. I'll set a poodle build going and see if I can get poodle a little further along Jul 10 21:27:32 RP: thanks for the help. Let me know if you want help testing :) Jul 10 21:27:52 03mickeyl 07org.oe.dev * r2d36b229... 10/packages/ (polypaudio/polypaudio_0.8.bb pulseaudio/pulseaudio_0.9.2.bb): Jul 10 21:27:52 polypaudio 0.8 upgrade to 0.9.2 Jul 10 21:27:52 NOTE: polypaudio is now pulseaudio Jul 10 21:27:52 first step in fine granular plugin packaging Jul 10 21:28:01 RP: so I will look at it Jul 10 21:28:36 hrw: If you see any other differences, feel free to sync them up as they should be the same Jul 10 21:28:58 RP: ide-cs patch can go directly to lkml? Jul 10 21:31:03 hrw: I guess so, maybe cc the pcmcia maintainer Jul 10 21:31:42 ok Jul 10 21:49:41 03hrw 07org.oe.dev * ra347a1a8... 10/packages/linux/ (8 files in 3 dirs): linux-openzaurus: enable USB Audio for 2.6.16/2.6.17 (taken from .oz354x) Jul 10 22:01:25 03hrw 07org.oe.dev * r6a2a6fd5... 10/packages/linux/ (9 files in 4 dirs): linux-openzaurus: added PCMCIA Id for 8GB CF card Jul 10 22:01:29 03hrw 07org.oe.dev * r10c4b592... 10/packages/linux/ (3 files): linux-openzaurus: fixed hrw-pcmcia-ids patch PV Jul 10 22:12:02 03hrw 07org.oe.dev * re6023258... 10/packages/corsair/ (files/corsair-024.patch corsair_0.2.0.bb corsair_0.2.5.bb): corsair: upgraded to 0.2.5 Jul 10 22:18:09 03hrw 07org.oe.oz354x * r3c0eda63... 10/packages/linux/ (3 files in 2 dirs): linux-openzaurus: added PCMCIA Id for 8GB CF card Jul 10 22:18:11 03hrw 07org.oe.oz354x * r3b51d223... 10/packages/corsair/ (files/corsair-024.patch corsair_0.2.0.bb corsair_0.2.5.bb): corsair: upgraded to 0.2.5 Jul 10 22:19:42 http://www.hrw.one.pl/2006/07/10/tosa-and-26-kernel/ Jul 10 22:19:43 cu Jul 10 23:20:48 anyone around that can help me debug a new bbclass? Jul 10 23:21:35 i tried to refactor things from using a .inc to a .bbclass, but 1) it didn't work as expected and 2) it didn't fail to give me any proper indication of #1 Jul 10 23:48:22 zwelch: I'll be in and out. If you want to pastebin it, I'll take a look. Jul 10 23:53:48 cbrake: actually, i'll probably just commit the changes to my project's BBFILES area: http://svn.minisplat.org/svn/src/trunk/build.d/dist/oe/ Jul 10 23:54:08 the existing .bb file for libsplutil is already there; i'll get the bbclass in shortly Jul 10 23:57:39 zwelch: ok, I need to put some kids to bed and will be back in a bit ... Jul 10 23:58:02 cool, no hurry, but i am looking forward to the help :) Jul 11 00:07:52 cbrake: okay, i just got my latest revisions into the repo. Jul 11 00:28:18 zwelch: so this all worked when you had the contents of the bbclass in the inc/bb file? Jul 11 00:40:45 cbrake: correct. r335 should reflects that working state Jul 11 00:40:59 and really, all i've done is some refactoring Jul 11 00:41:07 (from what i can tell by the rules) Jul 11 00:43:00 using the 'showdata' command, i can basically see that the svn bb files are only inheriting from base; they do not appear to pick up the autotools and minisplat bbclass definitions :/ Jul 11 00:44:09 i have tried this with both 1.4.2 and 1.5.0+trunk Jul 11 00:45:09 zwelch: hmm, have you tried cleaning the cache/time stamps, etc? Jul 11 00:45:23 i have nuked the cache many times Jul 11 00:45:32 what's the best way to nuke the time stamps Jul 11 00:46:42 rm tmp/stamps//*? Jul 11 00:47:02 * zwelch hasn't tried that, so here goes another build.... Jul 11 00:47:15 zwelch: yes, that should do it. There should be stamps for the package you are building. Jul 11 00:47:42 yup. i got 'em all; however, shouldn't a 'rebuild ' do that? Jul 11 00:48:42 zwelch: could also try putting some bogus code in the new bbclass to make sure it is getting parsed. Jul 11 00:48:50 * zwelch points at the oenote Jul 11 00:48:57 though that's not bogus :) Jul 11 00:49:12 zwelch: ahh, so do you see that? Jul 11 00:49:16 nope Jul 11 00:49:28 but it's not even in the 'showdata' output Jul 11 00:49:33 the inherit is silently failing Jul 11 00:50:07 however, i'm currently retesting it with the manual stamp cleaning Jul 11 00:54:59 cbrake: manual purge failed; status unchanged Jul 11 01:04:42 zwelch: well, I'm out of ideas other than cleaning tmp. You could run "bitbake -DDD -v" and see if that tells you anything. Jul 11 01:05:13 cool, that's very useful Jul 11 01:05:16 zwelch: if -DDD does not tell you, I'd probably starting instrumenting bitbake code :-\ Jul 11 01:05:22 hehe Jul 11 01:05:56 zwelch: well, good luck -- I'm done for the day :-) Jul 11 01:06:22 thanks for your help; i'm just glad for the sanity check. Jul 11 01:07:14 i guess the real bug here is that you can inheret nonexistant-class and it will silently ignore you Jul 11 01:07:43 also, bitbake doesn't complain if BBPATH contains non-existant paths Jul 11 01:07:45 :) Jul 11 01:13:45 woot! bbclasses work much better when they are found in the BBPATH and inherit'd correctly ;) Jul 11 01:24:25 zwelch: heh, i think thats a remnant of the fact that 'include' doesnt fail (by design), and inherit was syntactic sugar on top of include. nowadays itd probably be best off failing out Jul 11 01:25:42 kergoth: i figured it was something like that Jul 11 01:26:29 like the libltdl bug i had, i'll probably investigate and fix these two as well Jul 11 01:26:52 i don't see either of them in bugzilla Jul 11 01:27:15 sounds like you should fix that :) Jul 11 01:28:05 the combination of them was painful; a check for BBPATH would have clued me in faster, and the inherit bug would have been a non-issue Jul 11 01:28:31 but a check there is probably best, because tpyos happen ;) Jul 11 01:29:29 BBPATH should definately allow nonexistant dirs, as i can see use cases for that.. though it could warn, at least when -D Jul 11 01:30:05 * zwelch grumbles Jul 11 01:30:27 Patches Welcome, as tim would say Jul 11 01:30:30 ;) Jul 11 01:30:31 hehehe Jul 11 01:31:10 nah, i jest; bb/oe rock Jul 11 01:31:25 details like this are fairly inconsequential Jul 11 01:39:03 Maybe you guys have seen this one before and can give me some clues. My bitbake of openembedded.dev for a sptiz is dying on a stupid complaint about "__bind is already defined" when compiling glibc. Jul 11 01:40:02 I started the same thing up on a system in work, same os (Suse 9.3) and had no problem. Jul 11 01:43:22 This has been going on for 3 days. I've been all over the Wikis and websites and I got nothin'. WTF? Jul 11 01:44:06 ANY suggestion would be welcome... Jul 11 01:46:16 Kerwood_: i had that problem Jul 11 01:46:21 you are not alone Jul 11 01:46:41 zwelch: any ideas? Jul 11 01:47:01 well, i am fairly certain it is a configuration problem Jul 11 01:47:16 i.e. it went away after futzing with my PREFERRED_PROVIDERS Jul 11 01:47:32 zwelch: hmmmm Jul 11 01:47:53 here are settings that finally worked for me: http://svn.minisplat.org/svn/src/trunk/build.d/dist/oe/conf/ Jul 11 01:48:02 compare and contrast Jul 11 01:49:28 oh, i think i finally decided to nuke my entire tmp directory shortly thereafter; when getting my initial local.conf file in order, i had changed a libc provider and assume that some things had been built with the wrong libc Jul 11 01:49:59 in which case, that error may be a symptom of such toolchain incompatibilities/insanity, and a fresh start is the sanest path forward Jul 11 01:51:00 zwelch: Curious. I have no PREFERRED_PROVIDERS lines in my local.conf. I've nuked my tmp and build/sources directories twice and it had no effect. I'm going to put some PREFERRED_PROVIDERS in my local.conf (and nuke my tmp &c) and see what happens. Jul 11 01:51:02 Thanks Jul 11 01:51:21 well, you may have them set by the various includes that you pull in from DISTRO or MACHINE Jul 11 01:51:34 right Jul 11 01:52:02 but, i used the sample conf and had to add a number of lines, after watching the output. Jul 11 01:52:11 incidentally, i was targeting gpe-image Jul 11 01:53:48 Hmm...my failure occurs before any application buidling starts, so I don't think it has to do with WM choice. Jul 11 01:58:56 yeah, you're in a core task; that was more of an explanation of the abundant providers that i have Jul 11 01:59:36 really, those lines were added by watching the output process complain about my not having them set :) Jul 11 02:00:20 however, it was suggesting the form PREFERRED_PROVIDER_ = "" over the form i used Jul 11 02:00:42 since the effect is equivalent, i have presumed the forms are too (though i don't remember reading that anywhere) Jul 11 02:01:20 iirc, showdata translates them to the longer form Jul 11 02:01:40 * zwelch throws up his hands at all of the implementation details Jul 11 02:01:59 * Kerwood_ 's eyes glaze over Jul 11 02:07:41 you're telling me Jul 11 02:08:45 at the moment, i'm trying to understand the relationship between the PACKAGES produced by a package and the DEPENDS and RDEPENDS that should be used in other packages Jul 11 02:09:23 since package A produces A and A-dev, should i DEPEND on 'A-dev' and RDEPEND on 'A'? Jul 11 02:09:37 * zwelch is asking the room, not Kerwood_, per se ;) Jul 11 02:09:56 * zwelch is going to try it and find out Jul 11 02:11:09 zwelch: You are SO beyond my level in these matters... Jul 11 02:11:36 hehe, yeah, i'm already trying to write bb files and bbclasses for a suite of about a dozen packages Jul 11 02:12:09 i'm still a relative newbie, but kergoth and others can attest that i've been pestering this group for years ;) Jul 11 02:21:38 haha. so, my plan didn't work. nothing automatically PROVIDES A-dev Jul 11 02:21:45 * zwelch assumes he has some other bug Jul 11 02:46:44 is anyone using OE to build for the mingw32 target? Jul 11 02:47:48 i dunno, thatd be cool though. msys isnt exactly packed with applications Jul 11 02:47:57 the project that i'm working on builds for it (as do its deps) Jul 11 02:48:06 i have a script that auto builds a toolchain Jul 11 02:48:29 but i'd much prefer to integrate that support into OE Jul 11 02:49:59 the urls i referenced earlier can lead you to the win32 dist support, where those scripts live (a sibling directory of the oe/ path) Jul 11 02:50:15 really, there's not much to that... Jul 11 02:50:34 ... getting it fitting into the OE/bb world ... now, that's "non-trivial" ;) Jul 11 02:51:00 it would open the door to a whole other breed of potential contributors Jul 11 02:52:13 btw, i forgot to mention the scripts i have are for linux->ming32 cross development Jul 11 02:56:09 incidentally, i'd appreciate feedback for the latest revision of my bbclasses and bb files Jul 11 02:56:38 the three classes that i've created will make almost every .bb file either 2 or 3 lines long (two of which are inherits) Jul 11 02:56:51 they even appear to work ;) **** ENDING LOGGING AT Tue Jul 11 02:59:57 2006