**** BEGIN LOGGING AT Fri Aug 22 02:59:56 2008 Aug 22 07:09:41 Anyone experienced in moving ROOT FS to SD (MMC actually) ? I created a Loop image through AltBoot on a old 64MB card. Can I then just remove the Root files from the Z (ram/rom) itself to free up space for ipkg etc.? Aug 22 07:12:52 * NetAmp using OpenZaurus 3.5.4 on Collie/5500 Aug 22 07:21:42 hi Aug 22 07:28:46 khem: it seems that pax-utils 0.1.13 no longer exists in the g3nt00 archive, which makes it and hence uclibc unbuildable. Do you happen to have a bb file for a newer version? Aug 22 07:35:19 morning Aug 22 07:36:01 morning Aug 22 07:38:44 morning Aug 22 07:40:53 morrning! ;-) Aug 22 07:45:00 pb__: hi, did you disapprove *locally* that broken commit? Aug 22 07:46:15 ant__work: ah, seems my push failed. let me try again. Aug 22 07:46:22 apparently monotone.oe.net was offline at the time Aug 22 07:46:34 yesterday was the usual hydra multihead Aug 22 07:49:19 hm, right Aug 22 07:49:52 try pulling now, you should get 7f7af4c3c8ca9d531b3dd11bbecd2f5f8861a7e7 Aug 22 07:51:30 bonjour Aug 22 07:53:22 pb__: thx. Now I'm in office and cannot pull. I'll post a final diff in 5-6 hours. Aug 22 07:54:16 hoping you'll be around for committing it later today ;-) Aug 22 08:12:57 ant__work: yah, I will be here Aug 22 08:13:04 no more holidays planned in the foreseeable future :-} Aug 22 08:13:55 :-( Aug 22 08:18:57 pb__, master your foreseeing skills ;-) Aug 22 08:32:06 * * OE Bug 4515 has been created by elena.valhalla(AT)gmail.com Aug 22 08:32:08 * * libgcrypt-1.4.1-r0 do_configure failure Aug 22 08:32:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4515 Aug 22 08:36:07 I'm seeing this regularly, am I the only one? - mtn: fatal: std::logic_error: database.cc:1,610: invariant 'I(expected_mid == actual_mid)' violated Aug 22 08:36:07 mtn: this is almost certainly a bug in monotone. Aug 22 08:48:06 * * OE Bug 4516 has been created by elena.valhalla(AT)gmail.com Aug 22 08:48:08 * * new recipe for tuxmath Aug 22 08:48:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4516 Aug 22 09:00:33 good morning Aug 22 09:01:06 * * OE Bug 4518 has been created by elena.valhalla(AT)gmail.com Aug 22 09:01:08 * * New bb for libsdl-pango Aug 22 09:01:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4518 Aug 22 09:04:07 * * OE Bug 4519 has been created by elena.valhalla(AT)gmail.com Aug 22 09:04:09 * * new recipe for tuxtype Aug 22 09:04:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4519 Aug 22 09:11:24 ~seen boris* Aug 22 09:11:26 woglinde: i haven't seen 'boris*' Aug 22 09:12:03 he's here, though :) Aug 22 09:12:41 woglinde, If I were ibot, I'd quote him for saying "morning" at 9:38 CEST today ;-) Aug 22 09:18:06 moin woglinde Aug 22 09:30:45 hi ant Aug 22 09:30:55 hi polyonmous Aug 22 09:31:47 hi woglinde Aug 22 09:33:15 hi pb Aug 22 09:33:35 mickey|zzZZzz: pong Aug 22 09:34:39 morning all Aug 22 09:38:00 hi rp Aug 22 09:44:25 hi Aug 22 09:44:55 hehe Aug 22 09:44:57 lol Aug 22 09:45:04 why did he quit Aug 22 09:46:41 does anyone already have built oprofile for oe ? Aug 22 09:46:45 re boris Aug 22 09:46:56 re woglinde Aug 22 09:47:01 boris I checked directfb update in Aug 22 09:47:04 how are you K? Aug 22 09:47:07 did you test it yet? Aug 22 09:47:24 test what ?sorry Aug 22 09:47:33 last directfb release ? Aug 22 09:48:32 yes Aug 22 09:49:07 on x86 yes Aug 22 09:49:14 but not on angstrom/oe Aug 22 09:49:45 do you have write the bb ? Aug 22 09:49:53 wrote Aug 22 09:50:10 * ant__work hopes someone is trying to revive CIA bots and resync gitweb... Aug 22 09:52:43 boris hm I pushed it to the wolfson mirror Aug 22 09:53:35 ~oemirrors Aug 22 09:53:35 oemirrors are at opensource.wolfsonmicro.com, or at monotone.nslu2-linux.org Aug 22 09:53:50 boris try mtn pull opensource.wolfsonmicro.com Aug 22 09:54:05 then you should recieve the 1.2.3 bb Aug 22 09:54:50 ok thank you very much :) Aug 22 09:55:29 ~seen khem Aug 22 09:55:32 khem is currently on #oe (2d 1h 36m 21s) #uclibc (2d 1h 36m 21s). Has said a total of 7 messages. Is idling for 8h 28m 4s, last said: 'Crofton: what does spartan have'. Aug 22 09:56:26 is there any way to access it without pulling all tree ? i'm already working on mine ... Aug 22 09:56:54 hm dont know if wolfson made a webfronted Aug 22 09:57:01 I am at work Aug 22 09:57:10 didnt check out oe here Aug 22 09:59:23 hm http://linux.onarm.com Aug 22 09:59:32 one contender more Aug 22 10:02:53 thank you woglinde :) Aug 22 10:08:15 boris hm? Aug 22 10:08:21 did you get it now? Aug 22 10:10:17 i checkout Aug 22 10:10:18 :) Aug 22 10:10:39 mtn --db=./OE.mtn checkout --branch=org.openembedded.dev is it the right way ? Aug 22 10:10:48 yes it seems :) Aug 22 10:10:53 yes Aug 22 10:11:01 then pull Aug 22 10:11:05 i got it thank you Aug 22 10:11:09 'll try it ;) Aug 22 10:11:33 hm maybee you use the tar.bz2 Aug 22 10:11:54 because inital checkout can last one week Aug 22 10:12:33 ? Aug 22 10:12:51 i have a directfb_1.2.3.bb file Aug 22 10:12:55 what do you mean ? Aug 22 10:18:24 hm okay Aug 22 10:18:28 than all is fine Aug 22 10:18:38 pleas try to run it Aug 22 10:18:47 I only made sure its build Aug 22 10:20:49 :) Aug 22 10:21:31 do you think that if i copy this bb and the directory directfb-1.2.3 in my org.openembedded.dev tree, it will work fine ? Aug 22 10:22:07 yes Aug 22 10:22:28 hm maybee you have to copy directfb.inc too Aug 22 10:22:34 i don't need the same .inc that you ? Aug 22 10:22:42 i think, cause there is some diffs Aug 22 10:22:44 look if it differs Aug 22 10:22:52 if it differs Aug 22 10:22:55 copy it too Aug 22 11:51:45 polyonymous-work: about initramfs (oe or sonkei), look at /packages/kexecboot/linux-kexecboot.inc Aug 22 11:52:47 ant__work, but should it be there at all for spitz? (will look anyway) Aug 22 11:52:50 morgning, Laibsch Aug 22 11:52:56 moin Laibsch Aug 22 11:53:27 good morning! Aug 22 11:53:55 ant__work, are you on #hentges.net or why do you mention initramfs? Aug 22 11:54:31 I read the logs Aug 22 11:54:55 ant__work, ah. channel logs? Aug 22 11:55:02 yep Aug 22 11:55:17 heh, nice way of using irc :)) Aug 22 11:55:56 well, here in office we have stupid M$-powered terminals...I don't install any irc Aug 22 11:56:16 ah. How come you're on #oe? Aug 22 11:56:31 java Aug 22 11:56:35 irc2go Aug 22 11:56:58 ah. single-channel? Aug 22 11:57:14 well, during worktime, yes Aug 22 11:57:18 ;-] Aug 22 11:57:22 poor thing you are Aug 22 12:02:50 ant__work: might be easier for you to use on of the web based ssh clients, ssh into your box and use irssi :-) Aug 22 12:03:52 heh Aug 22 12:08:13 he, I'd have to convince my wife to boot Gentoo and not use OSX during day... Aug 22 12:09:25 ant__work, why, you can use command line client on OS X Aug 22 12:09:50 Anybody understand why SharpROM compilation is now failing as early as glibc? Aug 22 12:09:52 http://tinderbox.openembedded.net/public/logs/808610.txt Aug 22 12:09:55 I can, she can't Aug 22 12:10:02 | configure: error: compiler support for __thread is required Aug 22 12:10:34 ant__work, why should she? Aug 22 12:10:46 Hmm.. seriously, is anybody using sharprom? Aug 22 12:10:46 to start sshd Aug 22 12:10:56 ant__work, you can start sshd on OS X Aug 22 12:11:17 I believe that SharpROM is still in use by the majority of people Aug 22 12:11:26 Plus, we've got the code, so why let it rot? Aug 22 12:11:29 Oh. Aug 22 12:11:50 Just not the majority in #oe or other geeky corners ;-) Aug 22 12:11:51 Well, if the code isn't used it may well rot, but if it is, it's better to make it work, of course. Aug 22 12:12:14 I thought there's no non-geeky majority that is using zauri nowadays. Aug 22 12:12:22 I may well be mistaken, of course Aug 22 12:23:08 polyonymous-work: before I go, comment out that INITRAMFS_IMAGE = Aug 22 12:23:22 you only need to bitbake linux-kececboot Aug 22 12:23:47 if you want the initramfs with graphical bootmenu Aug 22 12:24:08 BTW it makes sense for spitz too, IMHO Aug 22 12:24:26 ant__work, I have already commented it out. Aug 22 12:25:08 ant__work, it may make sense for spitz if it works, right. I will look into it after I get the successfull built. I haven't built OE for use for almost a year :) Aug 22 12:29:48 it works, guaranteed Aug 22 12:30:01 ant__work, good to know. Aug 22 12:30:06 I'll submit a patch later about this kexecboot, BTW Aug 22 12:30:13 (topass the mtdparts) Aug 22 12:31:58 please check the rotation (FBANGLE) on spitz...thesing put 270 as default, but I use 0 (landscape) Aug 22 12:32:26 Well, we may get to it later today if I get working build and will have my spitz handy :) Aug 22 12:32:47 in that case you are the candidate for testing u-boot too :-] Aug 22 12:32:49 Right now I can ssh to my spitz, but its about an hour away from me :) Aug 22 12:33:27 Yes, I have the spitz and right now it's not critical to keep it working, as I also have got n810 :) Aug 22 12:34:31 btw as I already said to someone else, the Zaurus will *not* explode installing u-boot.. Aug 22 12:35:21 have to go now, bbl Aug 22 12:36:11 well, as I said I'm one hour away from it, so I can stand a moderate explosion.... Aug 22 13:10:27 anyone know about meta-toolchain not including libmpfr? Aug 22 13:19:10 re Aug 22 13:19:22 woglinde, i got an error : Task failed: Error: /uc200/oe/org.openembedded.dev/packages/directfb/directfb-1.2.3/./ts_lib_autotools.patch not found Aug 22 13:19:46 That means woglinde forgot to mtn add :) Aug 22 13:22:52 boris mom Aug 22 13:23:38 yes woglinde ? Aug 22 13:23:57 boris put this one under files in the dir http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=blob;f=packages/directfb/files/ts_lib_autotools.patch;h=ced710478660cd419f8f9b029964f91dfcf7e6cd;hb=HEAD# Aug 22 13:24:36 * polyonymous-work prefers cgit to gitweb :) Aug 22 13:24:45 i found this file in org.openembedded.dev/packages/directfb/files/ts_lib_autotools.patch, so maybe i just have to move it to directfb-1.2.3 dir ? Aug 22 13:24:49 * woglinde wonders which old checkout boris has Aug 22 13:25:05 i make checkout this morning ! Aug 22 13:25:06 hm no Aug 22 13:25:12 if it is under files Aug 22 13:25:14 did you also pull? :) Aug 22 13:25:18 bitbake find it Aug 22 13:25:20 also alos :) Aug 22 13:25:21 also Aug 22 13:25:45 but indeed, bitbake would see it in files... Aug 22 13:25:49 dont know what you did wrong at the mom Aug 22 13:25:55 mom ? Aug 22 13:26:00 files is for generell patches Aug 22 13:26:01 your momma! Aug 22 13:26:10 i know that Aug 22 13:26:18 foo-version-dir is for version special patches Aug 22 13:26:27 i guess :)) Aug 22 13:26:31 * woglinde scratches head Aug 22 13:26:46 * polyonymous-work scratches woglinde's head goes out for a smoke. Aug 22 13:27:08 I dont smoke Aug 22 13:27:17 and I dont like it Aug 22 13:29:40 in fact woglinde, locate gives me result for this patch, but it don't exist any more , i do updatedb :) Aug 22 13:30:02 aeh Aug 22 13:30:15 -_- Aug 22 13:30:32 woglinde, it was me who smoked, no worries. Aug 22 13:31:13 polyonymous-work ah okay ;) Aug 22 13:32:23 ok i got it in wolfson tree and copy it...not a cleany way but it works :p Aug 22 13:33:50 thank you woglinde, i will test if it works, maybe this afternoon, or monday :) Aug 22 13:34:08 i'll say to you Aug 22 13:35:23 boris thanks Aug 22 13:49:33 Laibsch: sounds like the glibc configuration for sharprom is wrong, probably trying to use nptl or some such. Aug 22 13:49:50 the error you pasted means that your compiler doesn't understand __thread, which is correct for a non tls platform. Aug 22 13:50:50 OK, thanks, pb__ Aug 22 13:51:16 AFAIU, the Sharp compile is binary only, right? Aug 22 13:51:24 so, where can I tweak things? Aug 22 13:51:47 probably on glibc side? Aug 22 13:59:52 woglinde, for directfb-examples, you must use now 1.0.1 or 1.2.0 Aug 22 14:00:00 1.0.0 is not provided any more Aug 22 14:00:07 so fetch failed ^^ Aug 22 14:01:46 just by renaming directfb-examples_1.0.0 ->directfb-examples_1.2.0.bb, it gone on :) Aug 22 14:11:09 boris I didnt update examples or fusion yet Aug 22 14:11:33 ok Aug 22 14:12:17 do you think it will work with an older fusion ? Aug 22 14:14:04 woglinde: are you actively using directfb and fusionsound? Aug 22 14:20:55 boris no Idea Aug 22 14:21:00 do you need fusion? Aug 22 14:21:10 I always thought fusion is optional Aug 22 14:24:28 fusion is needed to build directfb Aug 22 14:24:32 i think Aug 22 14:24:43 yes it's needed, as direct Aug 22 14:24:50 but just used for sound, i think Aug 22 14:24:52 not sure Aug 22 14:25:08 i'm a beginner on directfb Aug 22 14:25:16 so if someone has advices :p Aug 22 14:28:27 anyone on that would care/know about me adding a page to the OE wiki? Aug 22 14:31:13 kgilmer: sure, I care and would be glad if people contribute :-) Aug 22 14:31:49 boris fusion is not needed to build directfb Aug 22 14:32:10 svople I cared about the packages Aug 22 14:32:13 cbrake: thx :) well I just created an account. Aug 22 14:32:22 but didnt use it a long time Aug 22 14:32:47 Unless I get yelled at I'm just going to write a brief howto on writing a recipe for a java library in the dev section. Aug 22 14:33:36 I'm wondering if I just start straight off or there's some process by which the content is reviewed or commented on before it's public? probably not but it's best to ask... Aug 22 14:35:28 kgilmer: several of us on the doc team review changes and try to keep things organized, but there is not workflow beyond that at this time. Aug 22 14:36:08 ok cool. so I'll just dig in then cbrake. I'll post here when I have something I think is decent. Aug 22 14:40:55 kgilmer: go right ahead Aug 22 14:40:59 contributions welcome Aug 22 14:41:11 there is a very well-maintained java page already Aug 22 14:41:15 by thebohemian Aug 22 14:41:49 Maybe you want to link your stuff from there or just add it (the page has gotten rather large IMHO, it could be broken into pieces) Aug 22 14:42:51 oh? good to know Laibsch Aug 22 14:42:53 where is it? Aug 22 14:43:13 oh Aug 22 14:43:20 i see :) i searched for Java and there it is. Aug 22 14:43:53 kgilmer hm ah I thought you were awere of this site Aug 22 14:47:07 this right? http://wiki.openembedded.net/index.php/Java Aug 22 14:47:17 how would I know this page is here without searching? Aug 22 14:48:12 strange question Aug 22 14:48:18 google will probably have it Aug 22 14:48:27 it probably has a couple of links here and there Aug 22 14:48:38 same as with every other page out there in the world Aug 22 14:49:07 sorry Laibsch, I mean when I go here http://wiki.openembedded.net/index.php/Main_Page Aug 22 14:49:15 what is so bad about searching Aug 22 14:49:17 ? Aug 22 14:49:32 kgilmer: I am sure you understand not every page can be linked from the main page Aug 22 14:49:43 and that would not help accessing stuff anyway Aug 22 14:49:56 a cluttered desk is not conducive to productivity Aug 22 14:50:06 neither is having a link to every page from the front page Aug 22 14:50:13 My assumption was that in the developer section, clicking on More would give me anything associated with writing recipes. Aug 22 14:50:41 i'm not saying it needs to be in the front page :) Aug 22 14:50:56 i click on Users -> More and Developers -> More Aug 22 14:51:03 in any case Aug 22 14:51:07 i'm glad it's here :) Aug 22 14:51:29 i will look and see if there's something I can do to help out. Aug 22 14:55:00 looks like my mini howto is still relevent, and i'll follow your suggestion Laibsch to link it from this page. Aug 22 14:56:51 boris_OmegA: there is a fusion portion build into directfb but like woglinde said the actual fusionsound is not needed by directfb Aug 22 14:57:55 ok :) Aug 22 14:58:06 thx Aug 22 15:00:30 CIA is just on vacation, or is it gone for good? Aug 22 15:00:44 well, the Java page is not strictly for devs only Aug 22 15:01:00 I agree it is a bit badly structured situation currently Aug 22 15:01:07 but, hey, it is a wiki Aug 22 15:07:48 Laibsch: should we add the Java page to the User and Dev categories? Aug 22 15:07:48 maybe Aug 22 15:07:48 Laibsch: I agree it should be browsable as it is so useful Aug 22 15:07:48 but this is a general problem Aug 22 15:07:48 how do we differentiate between user and dev pages Aug 22 15:07:48 if we end up putting all pages into both categories, we need to rethink categorization Aug 22 15:07:48 Laibsch: many things will apply to both Aug 22 15:07:48 Laibsch: nod Aug 22 15:16:41 Laibsch: I guesss for now it seems that all categories should be browsable from the main sections on the home page, so I added category:java to User and Dev. Aug 22 15:19:46 Would somebody be so nice to take a look at fixing libexif? Aug 22 15:19:52 I think it is beyond my skill Aug 22 15:20:15 The only other option I see is dropping tons of packages to nonworking which is what I have started doing Aug 22 15:20:40 morning Aug 22 15:21:09 yo kergoth Aug 22 15:21:10 hi kergoth Aug 22 15:22:55 * kergoth`work thinks packages should automatically move into nonworking if they have no active maintainer, after a certain period, personally. too easy to bitrot otherwise Aug 22 15:24:52 sounds reasonable at first, but how do you want to do that? Aug 22 15:25:00 implementation-wise? Aug 22 15:25:07 no idea :) Aug 22 15:25:10 mm, possibly. I would prefer to focus on trying to analyse and eliminate the causes of bit-rot in the first place. Aug 22 15:25:33 I guess it would help if people did pick up bitbaking world again Aug 22 15:25:39 I thought that koen was doing that Aug 22 15:25:39 it's terribly infuriating if you stop working on oe for a while, come back after a couple of months, and find that the stuff that you were previously using has either been deleted or no longer works. Aug 22 15:25:47 but apparently that stopped at least two years ago Aug 22 15:26:30 pb__: I understand. I wonder what we can do to prevent that. Aug 22 15:26:51 In that situation, I assume it was stuff nobody else was interested in or they would have noticed Aug 22 15:27:18 In that case, I assume the burden to resurrect it does indeed lie on the one who wants to use it. Aug 22 15:27:18 yeah, dunno. has anybody ever studied what actually causes recipes to become bit-rotten? Aug 22 15:27:23 I don't any way around that Aug 22 15:27:45 possibly, but my opinion would be that the burden of resurrecting it should lie on the one who broke it in the first place. Aug 22 15:27:50 pb__I can only assume that a) our QA routines have gotten more clever (this is the case for libexif, I think) Aug 22 15:28:07 pb__: in the case of a), it had always been broken Aug 22 15:28:13 nobody noticed, though Aug 22 15:28:29 that depends on your definition of "broken", I think Aug 22 15:28:35 b) some changes to core parts make things incompatible. Aug 22 15:28:36 i expect its a combination of reliance on upstream and the fact that despite our best attempts, our builds are still somewhat prone to side effects due to other packages and the native system Aug 22 15:28:41 but that's just a guess Aug 22 15:28:49 it probably is true that a large part of the bit-rot results from somewhat-arbitrary tightening of the rules on what is and isn't acceptable. Aug 22 15:29:26 I guess the solution to that would be something along the lines of debian's Standards-Version. Aug 22 15:29:49 pb__: an example of b) is that I cannot build popt-1.7 in angstrom and was just about to move it to nonworking, after all we have 1.14. I stopped doing that when it was pointed out to me that openmoko uses 1.7 and indeed it compiles in OM. Aug 22 15:30:01 I'm not sure why, maybe they use an older glibc or something. Aug 22 15:30:16 kergoth`work: yah, that's true. there certainly are some recipes lying around for which the upstream tarballs have just plain disappeared. those would obviously be candidates for moving to nonworking. Aug 22 15:30:19 versioning for classes might be helpful, we've talked about that before Aug 22 15:30:21 sometimes upgrading library or tool just broke some other package Aug 22 15:30:40 But in that case, you will see breakage you can't really do anything about in OE, I guess. We have COMPATIBLE_MACHINE, but we lack other vars to mark stuff as incompatible. Aug 22 15:30:46 new libtool, new binutils, new gcc ^^ Aug 22 15:31:29 Laibsch: I think your (b) is just one example of a more general issue, that DISTROs can select incompatible versions of different components. Those don't have to be "core" components, they could be anything that has dependents. Aug 22 15:31:44 I don't think there's anything to be done about that, nor do I think it's necessarily a problem that needs to be solved anyway. Aug 22 15:31:53 and often library or tool have to be upgraded becouse they are no more supported upstream Aug 22 15:32:10 i'm also beginning to think that the system shouldn't automatically pick the latest version if the configuration metadata isn't specifying one. having things just change out from under you is bad for reproducability Aug 22 15:32:19 bye Aug 22 15:32:21 yeah, I agree with that as well Aug 22 15:33:28 pb__: I'm dropping stuff left and right now. Do you think you might be able to take a look at libexif? Aug 22 15:33:32 reproducing a break can be a nightmare, couse you have many: arch, distribution, libc vs ulibc and so on Aug 22 15:33:36 Laibsch: sure. what's wrong with it? Aug 22 15:33:48 It breaks do_qa_staging Aug 22 15:33:55 Just build it and you should see. Aug 22 15:34:00 okay Aug 22 15:34:06 I am sure, it should be not too difficult to fix Aug 22 15:34:11 I'm not sure if I have an up-to-date .dev build tree on hand here, but I'll have a go Aug 22 15:34:14 But above my skill level Aug 22 15:34:28 I'll build it, so you can take a look Aug 22 15:35:32 diego_afk: I think reproducing breakages in OE works reasonably well Aug 22 15:35:38 fun from beijing: http://news.softpedia.com/news/The-Blue-Screen-of-Death-Follows-Windows-to-the-2008-Olympic-91809.shtml Aug 22 15:35:41 That is at least my experience in bug triaging Aug 22 15:36:23 It is usually found out rather quickly why a build fails, but not in all situations Aug 22 15:40:07 yeah, my build tree here is in a pretty dilapidated state. I can start a fresh one but I doubt it will be done bootstrapping by the time I go home. Aug 22 15:40:26 if you can show me the errors then I might be able to figure out what's wrong, otherwise I will look at it later when I have my laptop on hand. Aug 22 15:46:48 hi mrnice Aug 22 15:49:55 pb__: This is the strangest thing Aug 22 15:50:02 I can't reproduce this anymore Aug 22 15:50:17 Just yesterday, it still bailed out Aug 22 15:50:54 * Laibsch is becoming suspicious that possibly something else had spoiled staging and then thrown the errors for libexif Aug 22 15:52:26 yah, could be. Aug 22 15:54:00 Hi! I installed bitbake and openembedded as discribed in wiki http://wiki.openembedded.net/index.php/Getting_Started Aug 22 15:54:26 Proton23: very good Aug 22 15:54:28 Now if I try bitbake qt4-x11-free I get http://pastebin.com/d6df1303d Aug 22 15:55:15 If local.conf is needed: http://pastebin.com/d2042237f Aug 22 15:56:39 That sounds like you haven't set the right things in your environment. What does "echo $BBPATH" say? Aug 22 15:57:43 Actually nothing, but I'm sure I exported it Aug 22 15:57:52 so I did something wrong Aug 22 15:57:55 mom Aug 22 15:59:58 I don't know what I did wrong but now it seems to work :D Aug 22 16:00:01 Thanks Aug 22 16:01:38 Btw if he's ready... what have I done? Aug 22 16:02:09 meaning: Did I compile the qt-libraries? Aug 22 16:02:33 if that command completes without error, yes. Aug 22 16:02:41 you should find a pile of stuff in your tmp/ directory Aug 22 16:04:25 yes, I changed the directory, but that shouldn't be an issue, right? Aug 22 16:06:19 that should be fine Aug 22 16:06:35 I also choose openmoko instead of angstrom is this also correct? Aug 22 16:12:33 Maybe I should mention that that I wanna use it for openmoko... Aug 22 16:12:45 re ant Aug 22 16:12:59 hi Aug 22 16:13:30 pb__: let's go http://bugs.openembedded.net/attachment.cgi?id=9933 Aug 22 16:15:19 ant__: rock Aug 22 16:17:16 pb__: to you the honor to close #3738. thx Aug 22 16:17:48 great! Aug 22 16:17:52 thanks, guys Aug 22 16:18:46 polyonymous: Laibsch but I suspect that on spitz one could have issues with standard ROOT_FLASH_SIZE Aug 22 16:19:12 (only on SL-C3000) Aug 22 16:19:49 other machines categorized as 'spitz' have 128mb nand Aug 22 16:20:17 (SL-C3100, SL-C3200) Aug 22 16:21:44 hey, they are almost called like Mercedes... Aug 22 16:21:49 :-] Aug 22 16:53:17 <_hc> so I applied a patch to get linux-hackndev-2.6 building, and it rebuilt everything in stuff/org.openembedded.dev/tmp even though I already had everything else built in stuff/tmp Aug 22 16:53:25 <_hc> any ideas why that might happen? Aug 22 16:53:57 depends how you have TMPDIR defined. Aug 22 16:54:07 * * OE Bug 3738 has been RESOLVED (FIXED) by philb(AT)gnu.org Aug 22 16:54:08 if it's a relative path, and you ran bitbake from a different directory, that would explain this Aug 22 16:54:08 * * RFC: u-boot for zaurus c7x0 and akita Aug 22 16:54:10 * * http://bugs.openembedded.net/show_bug.cgi?id=3738 Aug 22 16:58:37 ant__: there you go Aug 22 16:58:39 * pb___ go home now Aug 22 16:58:53 thx, cu later Aug 22 16:59:28 <_hc> pb___: makes sense, but I never defined TMPDIR, where would I check that? Aug 22 17:01:07 * * OE Bug 4520 has been created by andrea.adami(AT)gmail.com Aug 22 17:01:09 * * boot: passing extra cmdline arguments Aug 22 17:01:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4520 Aug 22 17:05:28 woglinde: hi Aug 22 17:18:54 <_hc> what's the standard location for "tmp" then? is "stuff/tmp" the best place? Aug 22 17:34:53 it doesn't really matter, you can put it anywhere you want. Aug 22 17:44:54 <_hc> where do most people have it? or where does it go by default? It seems to have shifted back to stuff/tmp, perhaps when I tried out Poky, it somehow changed the TMPDIR setting. I restarted since trying Poky, and it's back in stuff/tmp/ Aug 22 17:47:11 I think the default is $PWD/tmp, but I'm not certain about that. Aug 22 17:48:37 pb__, yeah Aug 22 17:48:57 so if you run bitbake in a different directory, it builds the toolchain etc again Aug 22 17:49:14 for the first couple of weeks using OE, I was really stumped by that :) Aug 22 18:01:50 i usually set it in the local.conf so that i don't have that problem (different directory) Aug 22 18:02:26 same, hardcode the TOPDIR/TMPDIR, less headaches that way Aug 22 18:03:11 OT, but any of you guys know any good references on use cases, requirements, and specifically their use in agile or oss development? Aug 22 18:15:59 kergoth, get your ass in college .... Aug 22 18:16:38 heh, someday when i have money, perhaps Aug 22 18:16:57 in the meantime i'll be researching design and architecture and process on the web :P Aug 22 18:17:24 we used scrum at the first job i had at ti, but i don't remember all that much Aug 22 18:17:27 i wonder if i still have the book in MN Aug 22 18:22:01 ant__: which version of udev are you using for the c7xx images right now? Aug 22 18:24:06 * * OE Bug 4504 has been RESOLVED (FIXED) by Aug 22 18:24:08 * * freetype building for armv5te dies on libtool Aug 22 18:24:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4504 Aug 22 18:42:17 <_hc> ah, cool, my patch got accepted :) Aug 22 18:43:12 <_hc> so now I have another question: how do I build an rootfs image? I have built lots of *-image recipes, but they always end up as a .tar.gz Aug 22 18:43:21 <_hc> I want a .ext2 or .squashfs Aug 22 18:45:10 It's in local.conf, I believe Aug 22 18:45:20 hvontres|work: 092 Aug 22 18:45:24 <_hc> hmm, like what? Aug 22 18:45:30 It's in local.conf, I believe ---> it is Aug 22 18:45:34 _hc, have you read local.conf.sample ? Aug 22 18:45:41 <_hc> will do again Aug 22 18:45:58 <_hc> thanks Aug 22 18:46:00 IMAGE_FSTYPES = "jffs2 tar" Aug 22 18:46:15 put whatever you want instead of the tar Aug 22 18:46:15 valhalla, well I was almost certain, but I haven't been building OE for quite a while, so I'd rather sound unsure :) Aug 22 18:46:24 why instead? :)) Aug 22 18:47:43 because my brain is turning itself off and i can't find the english to explain "add whatever you want to the list" in less words :D Aug 22 18:48:16 heh :)) Aug 22 18:48:23 ant__: ok. One thing I noticed on 118 was that the regular mount command (not busybox) was generating alignment traps, so maybe that might be part of the problem with udev Aug 22 18:49:03 Has anyone tried mdev as a udev replacement? Aug 22 18:49:37 hvontres|work: here a log with 092 booting from SD, it looks there are issues too Aug 22 18:49:40 http://www.pastebin.ca/1181822 Aug 22 18:52:54 _hc: about moving TMPDIR: http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2008-July/002416.html Aug 22 18:53:27 only Poky can do it...(Poki is evil:-) Aug 22 18:55:10 <_hc> what's the relationship between the Poky people and OE people? Aug 22 18:55:20 <_hc> isn't Poky just a subset of OE? Aug 22 18:55:32 for the most part, yep Aug 22 18:55:49 <_hc> kind of like Ubuntu and Debian? Aug 22 18:56:25 Strictly speaking, I'd say that intersection of OE and poky is approximately equal to both :) Aug 22 18:56:38 Yeah, I think that's close enough, _hc Aug 22 18:57:02 their metadata is a great deal more limited, they maintain a subset of th epackages the way most companies do, along with a custom distro and their own modifications and extra packages Aug 22 18:57:06 Poky is the next-gen OE Aug 22 18:57:12 IMHO Aug 22 18:57:23 talking about buildsystem Aug 22 18:57:29 not distro Aug 22 18:57:32 not really, it just happens to be the first to get the changes from the devs o-hand pays to work on it Aug 22 18:57:42 so..its beyond Aug 22 18:58:00 for -some- things it can be staging area for new changes Aug 22 18:58:01 OE comes always later Aug 22 18:58:06 and poky -is- the distro and metadata, as is OE Aug 22 18:58:09 perhaps you mean bitbake. Aug 22 18:58:24 no, lok just at the dir structure Aug 22 18:58:33 i'm familiar with it, i worked for o-hand for 3 months. Aug 22 18:58:35 RP plans to implement this in OE in future Aug 22 18:58:41 or not? Aug 22 19:00:04 I think the mlimachine thing is not yet 100% Aug 22 19:00:20 argh two dead keys in one line... Aug 22 19:00:32 mlimachine? Aug 22 19:00:38 multimachine Aug 22 19:01:53 but FWIW I only have a little experience, and with angstrom Aug 22 19:04:04 hvontres|work: funny link what you posted ! Aug 22 19:05:20 embarrassing Aug 22 19:43:06 Wow, unbelievable, after I dropped about 100 bb files or so that had unfulfillable deps and spending a stretch of about two days, "bitbake world" really build a few things now (before failing somewhere, but that was expected) Aug 22 19:43:27 I would have never thought that our meta-data contains so many obviously unfulfillable dependencies Aug 22 19:47:17 Laibsch, in all seriousness what is the ratio of working to non-working stuff Aug 22 19:48:44 Can't say Aug 22 19:48:55 It also depends on circumstances Aug 22 19:49:04 like what machine and distro you are building Aug 22 19:49:10 one example Aug 22 19:49:47 skip the example Aug 22 19:49:51 That is not about deps Aug 22 19:49:53 heh Aug 22 19:50:00 I'll give you a number in aminute Aug 22 19:50:29 # mtn status|grep dropped|wc -l Aug 22 19:50:29 211 Aug 22 19:50:40 IOW, I had to locally drop about 211 recipes Aug 22 19:50:58 maybe even more since I sometimes dropped complete directories Aug 22 19:51:15 I'll try and bring them back locally one-by-one and fixing their dependency Aug 22 19:51:36 can you publish the list to the email list Aug 22 19:51:48 along with the assumptions you are building with? Aug 22 19:51:57 it would be good data for people to know Aug 22 19:52:36 OK, I'll do that later Aug 22 19:54:17 Laibsch: How did you test this? Aug 22 19:54:35 Many things got broken by intrusive upgrades such as libtool. Aug 22 19:54:48 ...is what I notice on a daily basis Aug 22 19:57:57 * mwester is invisible Aug 22 19:59:07 at least until I commit something that becomes intrusive... then everyone notices. But there are more critisims than solutions offered in general, which makes it difficult for folks such as I to fix recipes. Aug 22 19:59:50 So there may well be things that work for MACHINE == NSLU2 or DISTRO == unslung|slugos just because it's safer for me that way! Aug 22 19:59:55 :shrug: Aug 22 20:00:26 (and I'm about to commit more of that type of change!) Aug 22 22:13:51 mwester: My take on this is "It's OK to break things. Best efforts is what counts IMHO. nobody is perfect. If something's broken, let's try and fix it" Aug 22 22:14:11 holding back stuff out of fear it might break something is worse than actually breaking something Aug 22 22:28:28 I agree with Laibsch: committing something that is in the right direction but not completely right yet, raises awareness and might quicker improve a piece of our metadata. Aug 22 22:29:43 after all that's what SCM is for :) Aug 23 01:14:06 * * OE Bug 2752 has been RESOLVED (FIXED) by Aug 23 01:14:08 * * e-image (images/e-image.bb) is broken Aug 23 01:14:10 * * http://bugs.openembedded.net/show_bug.cgi?id=2752 Aug 23 01:30:05 * * OE Bug 3246 has been RESOLVED (FIXED) by Aug 23 01:30:07 * * libgemwidget DEPENDS incomplete - requires gnome-vfs and gnome-vfs-modules Aug 23 01:30:09 * * http://bugs.openembedded.net/show_bug.cgi?id=3246 Aug 23 01:35:05 * * OE Bug 3159 has been RESOLVED (FIXED) by Aug 23 01:35:07 * * iputils package set ${PN} package empty Aug 23 01:35:09 * * http://bugs.openembedded.net/show_bug.cgi?id=3159 Aug 23 01:43:06 * * OE Bug 2116 has been RESOLVED (LATER) by Aug 23 01:43:08 * * db1200 machine description broken Aug 23 01:43:10 * * http://bugs.openembedded.net/show_bug.cgi?id=2116 Aug 23 01:46:06 * * OE Bug 2397 has been RESOLVED (FIXED) by Aug 23 01:46:07 * * libmimedir package depends on -dev package Aug 23 01:46:09 * * http://bugs.openembedded.net/show_bug.cgi?id=2397 Aug 23 01:50:05 * * OE Bug 2308 has been RESOLVED (FIXED) by Aug 23 01:50:08 * * spitz looses time while suspended Aug 23 01:50:10 * * http://bugs.openembedded.net/show_bug.cgi?id=2308 Aug 23 01:56:05 * * OE Bug 2843 has been RESOLVED (WONTFIX) by Aug 23 01:56:07 * * unmet dependency on bluez-dtl1-workaround Aug 23 01:56:09 * * http://bugs.openembedded.net/show_bug.cgi?id=2843 Aug 23 01:58:05 * * OE Bug 2805 has been RESOLVED (FIXED) by Aug 23 01:58:07 * * Updating librsvg Aug 23 01:58:09 * * http://bugs.openembedded.net/show_bug.cgi?id=2805 Aug 23 01:59:05 * * OE Bug 2801 has been RESOLVED (INVALID) by Aug 23 01:59:07 * * foo Aug 23 01:59:09 * * http://bugs.openembedded.net/show_bug.cgi?id=2801 Aug 23 02:14:40 Laibsch: hi :) Aug 23 02:14:58 Laibsch: I got email from OE patchtracker ... no, I dont have rw access ;) Aug 23 02:15:13 maybe you should Aug 23 02:15:19 ping me after we switch to git Aug 23 02:16:42 Laibsch: I was out of OE for quite some time now ... I got back to tinkering with it just recently again, but I still like working on the kernel more Aug 23 02:16:58 Laibsch: Im sure there is one more competent person to get rw access, that is |miska| ;) Aug 23 02:17:02 you can do kernel work in OE Aug 23 02:17:20 OK, let's talk about that after the switch to git Aug 23 02:17:38 Laibsch: ok :) Aug 23 02:17:51 j/w when is that scheduled ? Aug 23 02:18:38 since git is really a good choice and it will probably make lot of things much easier Aug 23 02:18:50 Im looking forward to that change Aug 23 02:21:31 * mwester_ chuckles at Laibsch's joke Aug 23 02:22:12 which one is that? Aug 23 02:22:28 Marex: we should already be having it Aug 23 02:22:39 but we are a good project, we delay things ;-) Aug 23 02:22:54 mwester: which joke was that? Aug 23 02:23:16 Laibsch: you said it like there were more of those jokes ;D Aug 23 02:23:39 *that* joke - delays. Aug 23 02:23:56 OM-git and Duke Nukem Forever :D Aug 23 02:24:24 I am sure, we'll get there eventually Aug 23 02:24:56 mwester_: and eventually DNF will be buildable using OE :-) Aug 23 02:25:23 hehe! Aug 23 02:26:14 Marex: what about http://bugs.openembedded.net/show_bug.cgi?id=2415 Aug 23 02:26:14 Laibsch: I too am sure we'll get there. But one has to laugh about the delay, lest one otherwise become too frustrated. Aug 23 02:27:13 well delays has both it's pros and cons :T Aug 23 02:27:48 mwester: how about you help me fix bug or just plain close forgotten ones while you are in such a good mood? Aug 23 02:27:55 I'm off to bed Aug 23 02:27:58 good night Aug 23 02:28:01 Laibsch: there was some bbclass that selected correct libqte Aug 23 02:28:20 anyways, do you still see it or not? Aug 23 02:28:38 Laibsch: good night... I'm happily fixing bugs -- and probably introducing new ones that will make everyone very angry. :D Aug 23 02:28:40 Laibsch: no, since I dont see the OE installation that was on my harddrive last year ;-) Aug 23 02:29:00 Laibsch: I will recheck it in about a month from now (Im not joking here) Aug 23 02:29:05 mwester: I was of course talking about the ones reported in the tracker Aug 23 02:29:13 how about the stale bugs Aug 23 02:29:18 some fun ones to look at Aug 23 02:29:20 I have one more exam to go so that goes before anything and everything Aug 23 02:29:51 Marex: OK Aug 23 02:29:58 I'll close it as wfm for now Aug 23 02:30:03 you can always reopen Aug 23 02:30:08 Laibsch: ok Aug 23 02:30:19 if it compiles, then please do Aug 23 02:30:56 hm ... I have an idea Aug 23 02:31:05 * * OE Bug 2415 has been RESOLVED (WORKSFORME) by Aug 23 02:31:07 * * Vectoroids doesnt compie (fix included) Aug 23 02:31:10 * * http://bugs.openembedded.net/show_bug.cgi?id=2415 **** ENDING LOGGING AT Sat Aug 23 02:59:57 2008