**** BEGIN LOGGING AT Mon Aug 25 02:59:57 2008 Aug 25 05:49:31 What to do about the broken qt4-x11-free? Aug 25 07:09:48 morning Aug 25 07:51:09 good morning Aug 25 08:24:23 bonjour Aug 25 09:41:18 hi Aug 25 10:03:10 woglinde: moin Aug 25 10:03:33 hi all Aug 25 10:05:27 hi florian Aug 25 10:05:30 hi ant Aug 25 10:24:27 hi laibsch Aug 25 10:24:43 good morning Aug 25 10:26:07 hi Laibsch Aug 25 10:28:42 hi xora Aug 25 10:28:47 hi woglinde Aug 25 10:33:31 G'Morning Aug 25 10:42:29 moin Aug 25 10:42:51 hi there Aug 25 10:44:04 I need a bit of help from you Aug 25 10:44:24 I'm trying to figure out what happened to the angstrom branch Aug 25 10:44:47 it seems I can't access it anymore from mtn Aug 25 10:45:26 I've tried to find informations in the wiki, but without success :( Aug 25 11:10:28 hi zecke Aug 25 11:13:37 ho and bye Aug 25 13:05:47 florian: did you see my botmail? Aug 25 13:09:43 pb__: no. the bot needs a coffee i guess ;) Aug 25 13:10:23 so do I Aug 25 13:10:42 * mwester just put the coffee pot on Aug 25 13:11:08 pb__: But I know what you intended to tell me. Aug 25 13:11:49 pb__: fixed Aug 25 13:12:11 hi Aug 25 13:12:30 hi unomi1 Aug 25 13:12:38 could someone point me towards a small form factor board with 2 nics? Aug 25 13:13:03 hi flo_lap Aug 25 13:13:20 * polyonymous-work points unomil towards a small form factor board with 2 nics. Aug 25 13:13:29 poly: soekris 4501 Aug 25 13:13:33 alternatively, what is a good resource for current SBC offers? Aug 25 13:13:48 xlele, you meant unomil, perhaps. Aug 25 13:13:55 yes :) Aug 25 13:14:22 beautiful Aug 25 13:14:23 thank you Aug 25 13:14:24 I used it as vpn concentrator Aug 25 13:14:24 * polyonymous-work points unomil towards xlele. Aug 25 13:14:47 I need a bit of help from you Aug 25 13:15:10 I used oe for my soekris board last year Aug 25 13:15:36 now I'm trying with norhtec microclient junior Aug 25 13:15:47 it uses a vortex86 soc Aug 25 13:16:10 so I started from scratch with a new oe environment Aug 25 13:16:30 but I can't find the angstrom branch anymore using mtn Aug 25 13:16:42 is this normal? Aug 25 13:16:53 angstrom branch? was there such? Aug 25 13:17:45 yes, it was called org.openembedded.angstrom-2007.12-stable Aug 25 13:17:46 that is sexy Aug 25 13:18:14 unomi: are you speaking about the 4501? I can give you a lot of infos about it! Aug 25 13:18:17 xlele, I think I missed that. Well, I don't know much about current branches, I normally play with .dev Aug 25 13:18:47 xlele: try .stable Aug 25 13:19:04 poly: ok, but as far as you know, x86 arch is still supported? Aug 25 13:19:04 or .dev if you want the recent fixes Aug 25 13:19:08 sure Aug 25 13:19:14 no, I tried with the stable Aug 25 13:19:32 What did you try? Aug 25 13:19:35 but I think there are some glithches with the prefix Aug 25 13:19:38 Are you on a deb-based distro? Aug 25 13:19:43 I'd be surprised if x86 were discontinued. Aug 25 13:19:51 Yes I am! Aug 25 13:19:51 * Laibsch as well Aug 25 13:20:07 my problem is the following: Aug 25 13:20:30 xlele: Try http://blog.leggewie.org/?p=39 if you have problems with setting up the OE environment Aug 25 13:21:02 You likely want openembedded-angstrom Aug 25 13:21:57 http://pastebin.com/m74d0c7aa Aug 25 13:22:45 any ethernet powered devices ? Aug 25 13:22:50 laibsch: maybe it's a really stupid question (my excuses), but what's openembedded-angstrom? Aug 25 13:22:52 and you are sure that used to work? Aug 25 13:23:14 xlele: read the link to my blog I posted Aug 25 13:23:16 unomi1: some soekris can use POE Aug 25 13:23:17 Isn't it supposed to build toolchain first? Aug 25 13:23:31 poly: indeed! Aug 25 13:23:45 Laibsch: thanks, I will! Aug 25 13:23:57 xlele: I don't think you are using plain angstrom, are you? Aug 25 13:24:05 Try glibc first if uclibc fails Aug 25 13:24:25 Laibsch: unfortunately Aug 25 13:24:31 xlele: and you are sure this setup used to work? Aug 25 13:24:31 I need to use uclibc Aug 25 13:24:44 Well, try to get your compile environment in order first Aug 25 13:24:49 and worry about the rest later Aug 25 13:24:52 in my previous environment, I succedeed in making it work Aug 25 13:24:53 is what I'd do Aug 25 13:25:08 Maybe you should try .dev Aug 25 13:25:18 but, just for my understanding Aug 25 13:25:21 With openembedded-$X you can easily switch between the two Aug 25 13:25:28 could it be that you have some excessive assume_provideds ? Aug 25 13:25:52 xlele, the branch might not have angstrom in the name Aug 25 13:26:19 poly: I just imported my old local.conf file, without changing anithing else... Aug 25 13:26:20 rahter jsut .stable, then you select angstrom in local.conf Aug 25 13:26:36 xlele: pastebin your local.conf Aug 25 13:26:42 xlele, read it then. Aug 25 13:26:43 sure Aug 25 13:26:48 or pastbin :) Aug 25 13:26:58 pastebin forever! Aug 25 13:27:44 http://rafb.net/p/atlPtD90.html Aug 25 13:27:48 http://pastebin.com/m19a25dd2 Aug 25 13:27:53 xlele: That is against .dev Aug 25 13:28:13 flo_lap: thanks :-) Aug 25 13:28:14 Looks to me like the vortex machine was dropped in .dev and I guess there must have been a reason for that Aug 25 13:28:23 Maybe you are out of luck or ahead of some work Aug 25 13:29:18 and this is my vortex86.conf: http://pastebin.com/m605059f0 Aug 25 13:30:00 This is something you cooked up yourself? Aug 25 13:30:12 Laibsch: sorry, but what do you mean exactly? Aug 25 13:30:17 yes, it is Aug 25 13:30:37 OK, then I guess you will need to understand what is wrong with it ;-) Aug 25 13:30:43 :D Aug 25 13:30:46 I wouldn't be able to help you with that Aug 25 13:30:55 ok, let's assume Aug 25 13:31:08 But I think taking a good look at that machine config is your best bet Aug 25 13:31:12 I use x86.conf; I have the same problem Aug 25 13:31:34 what's libc for gcc, anyway? Aug 25 13:31:51 * Laibsch does MACHINE="x86" bitbake nano Aug 25 13:32:41 polyonymous: good point. AFAIK glibc/uclibc is set by some other variable in Agnstrom Aug 25 13:32:46 xlele: yeah i saw, but the ones that have POE are all single NIC? Aug 25 13:32:52 LIBC_MODE or something Aug 25 13:33:02 ANGSTROM_MODE Aug 25 13:33:07 uclibc Aug 25 13:33:18 yes, already in there Aug 25 13:33:22 I just saw that now Aug 25 13:33:29 do you need the last line? Aug 25 13:33:40 in my local.conf; that's how I tell bitbake to use uclibc under angstrom Aug 25 13:33:41 Try commenting it out and see what happens Aug 25 13:33:58 PREFERRED_PROVIDER_virtual/i486-angstrom-linux-uclibc-libc-for-gcc = "uclibc" is not it Aug 25 13:34:07 no, I added the last line trying to solve my problem but it didn't help... Aug 25 13:34:08 that is Aug 25 13:34:10 ANGSTROM_MODE = "uclibc" Aug 25 13:34:15 which you already have Aug 25 13:34:28 yes, that's correct Aug 25 13:34:37 xlele: The x86 compilation is running fine here Aug 25 13:34:38 but seems this is not my prob Aug 25 13:34:49 let me try! Aug 25 13:35:41 just the same error! Aug 25 13:36:10 on which os are you working? maybe debian has something wrong? Aug 25 13:36:12 xlele wipe your tmp, maybe, it should be pretty much empty so far Aug 25 13:36:20 ok Aug 25 13:36:32 debian works fine. Especially if you don't have dash as /bin/sh Aug 25 13:36:52 xlele: this is on debian Aug 25 13:37:23 no, I dont. moreover, I read somewhere that mawk (default in debian) can create issues, so I replaced it with gawk Aug 25 13:39:12 unomi1: 4511 should work with POE (but be shure to POE just on eth0!!!) Aug 25 13:40:15 ok, I wiped out my tmp and started bitbake with x86.conf... I'll let you know in a while Aug 25 13:45:39 done Aug 25 13:45:55 it does not depend on machine, unfortunately... Aug 25 13:46:30 I have the same error during /oe/org.openembedded.stable/packages/uclibc/uclibc-initial_0.9.28.bb, do_populate_staging Aug 25 13:47:24 the error is simple: Aug 25 13:47:26 make: i486-angstrom-linux-uclibc-gcc: Command not found Aug 25 13:50:43 clueless. Aug 25 13:50:50 * polyonymous-work is heading homewards Aug 25 13:51:47 xlele: has it actually built gcc-initial? Aug 25 13:53:04 if yes, you need to figure out why it isn't finding the resulting compiler. if no, you need to figure out why the dependency in uclibc-initial was apparently dishonoured. Aug 25 13:53:07 no, seems it didn't Aug 25 13:53:27 otavio: Is there any ZV-port driver for linux? Aug 25 13:53:58 flo_lap: no clue Aug 25 13:54:18 ~lart stupid irc clients Aug 25 13:54:18 * ibot nukes stupid irc clients with a single large nuke Aug 25 13:54:18 pb__: what if I try to build gcc-initial first? Aug 25 13:54:34 s/otavio/OT Aug 25 13:54:36 you could certainly try that, but it oughtn't to be necessary Aug 25 13:54:41 flo_lap: doh Aug 25 13:55:18 pb__: ERROR: Nothing PROVIDES 'gcc-initial' Aug 25 13:55:22 ???? Aug 25 13:56:35 uclibc 0.9.28 is old Aug 25 13:57:13 woglinde: I've tried just to see if 0.9.29 was my problem, but unfortunately it wasn't... Aug 25 13:57:58 xlele: oh, it's actually ${TARGET_PREFIX}gcc-initial Aug 25 13:58:14 or even virtual/${TARGET_PREFIX}gcc-initial Aug 25 13:58:30 pb___: yes, some people still use ZV :} Aug 25 13:59:38 pb__: what my $TARGET_PREFIX should be? Aug 25 14:00:19 pb__: I "feel" my problem is related in some way with a wrong $TARGET_PREFIX somewhere... Aug 25 14:00:57 i486-angstrom-linux-uclibc. apparently Aug 25 14:01:55 pb__: indeed! Aug 25 14:01:57 ERROR: Nothing PROVIDES 'i486-angstrom-linux-uclibc-gcc-initial' Aug 25 14:03:24 and if i bitbake virtual/i486-angstrom-linux-uclibc-gcc-initial Aug 25 14:06:03 make: i486-angstrom-linux-uclibc-gcc: Command not found Aug 25 14:06:38 seems I have a serious recursion issue! Aug 25 14:07:05 the pakage needs a binary generated by itself in order to be built! Aug 25 14:07:09 it seems Aug 25 14:08:00 can you tell what it was actually trying to do at that point? Aug 25 14:08:14 i.e. how far in the build process it had got before it crashed Aug 25 14:09:35 the task is do_populate staging Aug 25 14:09:37 http://pastebin.com/m1f6bb496 Aug 25 14:09:55 in the pastebin you'll find the bitbake output and the error log Aug 25 14:10:07 oh, that's uclibc-initial still Aug 25 14:10:12 yes :( Aug 25 14:10:35 hm, odd. can you try with bitbake -v, see why it thinks it needs to build that before gcc-initia; Aug 25 14:11:38 http://pastebin.com/m457d9f23 Aug 25 14:12:38 I can't see any clue, unfortunately... Aug 25 14:12:45 no, sadly that isn't terribly enlightening Aug 25 14:13:06 what happens if you try to build gcc-cross-initial by hand with "bitbake -b"? Aug 25 14:13:27 you can use bitbake -s to figure out what version to use if you're not sure Aug 25 14:13:28 let me try! Aug 25 14:15:32 bitbake -b uclibc-initial-0.9.28-r9 Aug 25 14:15:33 ???? Aug 25 14:15:47 no, I think that's not what you asked me to do... Aug 25 14:16:13 not quite Aug 25 14:16:20 you want gcc-cross-initial, not uclibc-initial Aug 25 14:16:42 and you need to give "bitbake -b" the actual pathname to the .bb file you want to build Aug 25 14:17:07 ok, let me try again Aug 25 14:18:11 bitbake -b ../org.openembedded.stable/packages/gcc/gcc-cross-initial_4.1.2.bb Aug 25 14:19:38 ok, seems it is doing something... Aug 25 14:21:51 it worked! Aug 25 14:22:34 now, what can I do to make it work normally? Aug 25 14:23:38 I guess you'll have to debug why bitbake thinks that gcc-cross-initial depends on uclibc-initial. Aug 25 14:23:49 clearly it doesn't really, otherwise you wouldn't have been able to build gcc-c-i by hand. Aug 25 14:25:38 * pb__ bbiab Aug 25 14:27:32 thank you so much! I'll try to figure out why I have this strange behaviour... Aug 25 14:43:54 Guys, if you can help us out with filling in the blanks at http://wiki.openembedded.net/index.php/Talk:Monotone_to_Git_ID_mapping it would be greatly appreciated Aug 25 14:52:07 * * OE Bug 1505 has been RESOLVED (FIXED) by Aug 25 14:52:08 * * gsoap can not generate gsoap-dev correctly Aug 25 14:52:11 * * http://bugs.openembedded.net/show_bug.cgi?id=1505 Aug 25 14:52:34 Hi! Aug 25 14:52:47 qt4-x11-free is broken Aug 25 14:52:57 How can I fix it myself? Aug 25 14:53:57 Proton23, the qt overlord was around earlier, but he is gone now :( Aug 25 14:54:22 Do you mean woglinde or zecke? Aug 25 14:54:26 Laibsch: the link you sent me before Aug 25 14:54:45 are you the author of the .deb port? Aug 25 14:54:54 this is not a port Aug 25 14:55:04 sorry, package! Aug 25 14:55:05 It is just a bunch of scripts and dependency information Aug 25 14:55:20 you are then sure to have all necessary soft installed Aug 25 14:55:25 plus a working OE config Aug 25 14:55:29 that is the purpose Aug 25 14:55:30 Crofton|work: I asked on another channel but didn't get an answer yet Aug 25 14:55:35 have you tried it with angstrom? Aug 25 14:55:36 takes the hassle out of OE Aug 25 14:55:46 my setup seems to be broken in a strange way Aug 25 14:55:56 Proton23: I asked zecke about QT. Apparently, it was koen who last touched these recipes Aug 25 14:55:59 maybe I can try with your .deb and start again Aug 25 14:56:04 Maybe you want to ask him via mail Aug 25 14:56:12 Proton23, try the list Aug 25 14:56:19 xlele: It is what I personally use Aug 25 14:56:30 Crofton|work: Mailing list? Aug 25 14:56:37 right Aug 25 14:56:51 perfect! I'll give it a try! and thank you for sharing your work! Aug 25 14:58:23 Laibsch: so the bibake recipe is the one which is broken? Aug 25 14:59:02 Crofton|work: Thanks I will try it Aug 25 14:59:05 the QT something recipe Aug 25 14:59:12 Let me check out which one it was Aug 25 14:59:21 be sure to mention machine/distro Aug 25 14:59:39 packages/qt4/qt-embedded_4.4.1.bb Aug 25 14:59:52 Crofton|work: Ok thanks Aug 25 15:00:08 Crofton|work: if Proton23 is talking about the same problem, I don't think it would matter Aug 25 15:00:16 $PN-dbg depends on $PN Aug 25 15:00:25 Laibsch, always good to mention so people can try to reproduce Aug 25 15:00:29  Laibsch: Error message says something about packages/qt4/qt-embedded_4.3.3.bb Aug 25 15:00:30 but the complex qt recipe does not define $PN Aug 25 15:00:43 Crofton: In general, I absolutely agree Aug 25 15:00:48 and it never hurts Aug 25 15:01:02 distro/machine/branch is the info, I usually want Aug 25 15:01:10 exactly, and prevents the email asking for it Aug 25 15:01:13 yeah Aug 25 15:01:19 if I can't reproduce, then I start digging for local.conf or other things Aug 25 15:01:23 branch/rev :) Aug 25 15:02:03 Proton23: It is very likely to be the same problem Aug 25 15:02:07 .dev or .stable? Aug 25 15:03:19 !oebug 1844 Aug 25 15:03:20 * * Bug 1844, Status: NEW, Created: 2007-02-03 01:01 Aug 25 15:03:21 * * werner(AT)almesberger.net: libmutil0_svn.bb setup misses libltdl creation Aug 25 15:03:22 I think it's dev Aug 25 15:03:23 still relevant? Aug 25 15:03:23 * * http://bugs.openembedded.net/show_bug.cgi?id=1844 Aug 25 15:03:41 I can't even find anything resembling *mutil* Aug 25 15:07:06 * * OE Bug 1844 has been RESOLVED (WONTFIX) by Aug 25 15:07:08 * * libmutil0_svn.bb setup misses libltdl creation Aug 25 15:07:10 * * http://bugs.openembedded.net/show_bug.cgi?id=1844 Aug 25 15:10:06 * * OE Bug 2614 has been RESOLVED (INVALID) by Aug 25 15:10:08 * * MACHINE_EXTRA_RRECOMMENDS doesn't work Aug 25 15:10:10 * * http://bugs.openembedded.net/show_bug.cgi?id=2614 Aug 25 15:24:06 * * OE Bug 3035 has been RESOLVED (FIXED) by Aug 25 15:24:08 * * please add help2man to Debian openembedded-essential task Aug 25 15:24:10 * * http://bugs.openembedded.net/show_bug.cgi?id=3035 Aug 25 15:30:07 morning all Aug 25 15:30:36 RP, gm Aug 25 15:31:45 * RP is remembering why he stopped using rpm years ago :/ Aug 25 15:32:10 poor rp Aug 25 15:32:23 not a nice package system at all Aug 25 15:32:41 stands for RP Maaaaaaaaad Aug 25 15:32:43 :-D Aug 25 15:33:28 XorA: ;-) Aug 25 15:33:44 in your best General Melchitch voice of course Aug 25 15:33:50 At least there is nobody around to think I'm going mad as I swear to myself... Aug 25 15:34:52 I just found a Makefile that did "uname -i" and then added -mtune=i686 into CFLAGS if it matched i386 :/ Aug 25 15:35:04 heh Aug 25 15:35:04 rpm is full of internal uname calls too :/ Aug 25 15:35:21 and don't get me started on yum... Aug 25 15:35:27 but RPM is perfect, one of the OM employees says so Aug 25 15:36:49 XorA: does minimo build for you? Aug 25 15:37:19 Laibsch: not for years Aug 25 15:37:28 XorA: I have some new collegues who think rpm is great too :/ Aug 25 15:37:40 At least it doesn't rely on the output of pom. Aug 25 15:37:41 RP: yeah, I wondered why they fell out with ubuntu Aug 25 15:37:41 http://bugs.openembedded.net/show_bug.cgi?id=1391#c19 Aug 25 15:37:50 XorA: It built around a year ago Aug 25 15:37:53 I think Aug 25 15:38:42 Laibsch: I heard rumours a mozilla guy was looking at giving us a fennec recipe, so I have little interest in diving into old firefox 2 era code Aug 25 15:38:53 OK Aug 25 15:38:57 I absolutely understand that Aug 25 15:39:05 FF3 runs faster than minimo ever did anyway Aug 25 15:39:09 Can you move the recipe to nonworking then and close the bug? Aug 25 15:39:20 or drop it Aug 25 15:39:22 all it needs is someone to do a nice embedded theme for FF3 Aug 25 15:39:39 Well I heard from my OM colleagues it builds and works for them Aug 25 15:39:57 * XorA tries a build for a laugh Aug 25 15:41:02 XorA: http://tinderbox.openembedded.net/public/logs/829865.txt Aug 25 15:43:50 NOTE: package minimo-1_0.02+cvs20070626-r0.02: task do_compile: started Aug 25 15:44:05 so I get further than that Aug 25 15:46:35 minimo still builds in poky... Aug 25 15:50:07 * * OE Bug 3444 has been RESOLVED (WORKSFORME) by Aug 25 15:50:08 * * gpe-irc-0.08 QA issues Aug 25 15:50:10 * * http://bugs.openembedded.net/show_bug.cgi?id=3444 Aug 25 15:51:52 bonjour Gnutoo Aug 25 15:52:02 Genesis, salut Aug 25 15:53:03 Genesis, je suis vraiment desole mais je me rapelle plus de toi...comment sait tu que je parle le francais? ubuntu-fr? Aug 25 15:53:18 (j'ai une petite memoire...) Aug 25 15:53:22 your host Aug 25 15:56:54 NOTE: package minimo-0.02+cvs20070626: completed Aug 25 15:56:59 so builds for the OM distro Aug 25 15:57:06 quite to my surprise :-D Aug 25 16:06:06 * * OE Bug 3515 has been RESOLVED (FIXED) by Aug 25 16:06:08 * * typo in MACHINE_FEATURES: ushbost (should be: usbhost) Aug 25 16:06:10 * * http://bugs.openembedded.net/show_bug.cgi?id=3515 Aug 25 16:06:57 mickeyl: good morning Aug 25 16:09:29 morning guys Aug 25 16:10:47 yo kergoth Aug 25 16:12:06 hey pb__ Aug 25 16:12:09 hi guys Aug 25 16:13:54 morning kergoth`work Aug 25 16:19:59 RP: what do you think about http://rafb.net/p/WfrYJq69.html Aug 25 16:20:36 ? Aug 25 17:03:40 Laibsch: I'm not RP, but that patch looks pretty unappealing to me. What's the reason for it? Aug 25 17:04:26 ewwww, I dont like that Aug 25 17:04:44 even a python function that hunted .debug directories would be better Aug 25 17:04:50 yeah, right Aug 25 17:05:24 though that still seems like it would be a slightly undesirable solution to what is (afaik) a non problem Aug 25 17:05:31 I'd be happy with anything Aug 25 17:05:48 But I spent a lot of time adding .debug entries which seems rather silly Aug 25 17:05:55 What exactly is the problem that you're trying to solve? Aug 25 17:06:02 I've yet to encounter a .debug dir that should have gone elswhere Aug 25 17:06:12 Im guesses .debug directories for plugins and the like Aug 25 17:06:15 And of course, it is always possible to override Aug 25 17:06:49 pb__The problem is unnecessary work and a saner default (from which you can deviate) Aug 25 17:08:05 Laibsch: I still don't quite understand the circumstances under which this unnecessary work occurs. Can you give some specific examples of situations where the current default is inadequate? Aug 25 17:08:29 if it's plugins, then the right place to fix them is probably in do_split_packages() Aug 25 17:08:43 pb__Try bitbake xmms-embedded Aug 25 17:08:49 But this is not specific Aug 25 17:09:13 We are adding stuff to $PN-dbg Aug 25 17:09:40 What we should be doing IMHO is define default for $PN-dbg broader and subtract from that in case it is necessary Aug 25 17:10:07 subtract: override the choice that the new default $PN-dbg would be making and redefine packaging Aug 25 17:10:54 pb__: We could do away with most of the customized $PN-dbg definitions Aug 25 17:11:29 The work spent for generating them is down the river, but there will be new situations coming up all the time requiring unneccessary work Aug 25 17:12:21 I'd be OK with XorA's suggestion of a python function to hunt for .debug dirs. Alas, I won't be able to write that Aug 25 17:12:23 okay, fair enough. Aug 25 17:12:44 will you be so kind to provide one? Aug 25 17:12:45 I guess my preferred option would be for package.bbclass to enumerate the directories for itself at the same point as it creates them. Aug 25 17:12:46 globs have got to be bad for performance as well Aug 25 17:13:39 since, more or less by definition, it knows exactly what files have the debugging information in them and can create a precisely accurate list. Aug 25 17:14:07 my second preference would be XorA's suggestion of just doing an os.path.walk() to locate all the .debug directories and stick them in ${PN}-dbg. Aug 25 17:14:40 fwiw, i think we need RE for FILES sooner than later Aug 25 17:15:05 the glob-based solution is a distant third because (a) it has an arbitrary depth limit, i.e. it won't find files in directories deeper than five levels, and as XorA says it will be slower. Aug 25 17:15:20 er, that second thing should have been (b) obviously Aug 25 17:16:29 yeah, RE for FILES would be cute, though in cases like this it does still suffer from the risk of CVS core-style lossage Aug 25 17:16:58 admittedly that risk seems fairly remote in this particular instance but still not completely impossible Aug 25 17:20:04 I think this is a discussion for oe-devel Aug 25 17:20:15 its too much of a change for an irc decision Aug 25 17:20:57 RP thought something along the lines of adding RE_ Aug 25 17:21:06 with keeping Aug 25 17:21:18 so we could gradually move over Aug 25 17:35:06 * * OE Bug 4521 has been created by  Aug 25 17:35:08 * * libtorrent FTBFS Aug 25 17:35:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4521 Aug 25 17:48:36 hmm Aug 25 17:48:47 anyone know if fedora's FEver tool is open source? Aug 25 18:20:08 Anyone seen the problem where bitbake can't finish a fetch from CVS because the module gets checked out into 'P' rather than the module name? Aug 25 18:30:32 still trying to figure out how to use busybox switch_root to switch the root on an initramfs filesystem - don't really want to use kexec as that adds boot-time - anyone using switch_root successfully? Aug 25 19:31:56 Hi, I am wondering if there is a way to get OE to select a build target between a few possible targets without a configuration disaster. Anyone have any advice or could anyone point me at some docs? Aug 25 19:32:32 you're going to need to elaborate Aug 25 19:33:12 kholzer, do yu mean build for several different machines? Aug 25 19:33:33 like omap5912osk and beagleboard Aug 25 19:33:40 Basically. Aug 25 19:33:51 only with a couple of homebrew targets mixed in there. Aug 25 19:34:14 just change the MACHINE= Aug 25 19:34:27 and it will do the right thing (tm) Aug 25 19:34:45 for Angstrom (since it inherits multimachine by default) Aug 25 19:36:13 it reuses toolchains when possible :) Aug 25 19:36:56 Could I set some environment variable and get bitbake to change MACHINE dependent on the contents of the environment variable? Aug 25 19:37:23 set MACHINE in the environment even :) Aug 25 19:37:26 kholzer: MACHINE should work as long as you set it using ?= and not = in local.conf Aug 25 19:37:35 k Aug 25 19:37:53 I am pretty sure you will be happy, this is one of OE's strengths Aug 25 19:38:10 re-use across machines is easy Aug 25 19:38:30 cool. Aug 25 19:38:49 I do builds for about 4 machines at various times Aug 25 19:39:21 things like OSK, DV EVM, and sffsdr all use same toolchain, so you can make images real quick Aug 25 19:39:33 Can I get the output directory to change based upon what MACHINE is set to? Aug 25 19:39:56 multimachine takes care of that Aug 25 19:39:59 try it :) Aug 25 19:41:16 what do you mean by multimachine? punch in a list or something? Aug 25 19:41:54 multimachine is the class that handles keeping stuff straight on a per machine case Aug 25 19:42:02 so build with MACHINE=a Aug 25 19:42:11 then do another build with MACHINE=b Aug 25 19:42:17 and look in tmp/deploy Aug 25 19:42:20 k Aug 25 19:42:59 ok, nevermind. I see. Works. Aug 25 19:43:33 :) Aug 25 20:55:11 ~seen thesing Aug 25 20:55:14 thesing was last seen on IRC in channel #oe, 9d 20h 40m 31s ago, saying: 'time to sleep. Night all.'. Aug 25 21:09:25 hvontres|work: hey, how's going with kexecboot ? Aug 25 21:21:32 what woudl be responsible in an image for creating a symlink from /bin/bash to /bin/sh? the bash package doesn't do this Aug 25 21:47:52 hmmm Aug 25 21:48:46 moo Aug 25 21:48:53 he Aug 25 21:52:46 hi zecke , kergoth`work, woglinde Aug 25 21:53:51 RP: I had a look at all_names2, you learned typing at your summer job?! :) Aug 25 21:54:28 hey RP Aug 25 21:55:32 zecke: No, but I did learn that doing data manipulation at high speed made it less painless (I only used to have a certain amount to get through) ;-) Aug 25 22:03:07 zecke: You are aware that I took RP's work and added a few more things? Aug 25 22:03:12 It's in the wiki Aug 25 22:07:13 Laibsch: url? Aug 25 22:12:08 zecke: http://wiki.openembedded.net/index.php/Talk:Monotone_to_Git_ID_mapping Aug 25 22:12:48 It is now possible to work on this together and track the changes Aug 25 22:15:32 cool, and it is getting more complete Aug 25 22:15:40 this means I should look for the conversion bug? Aug 25 22:18:09 yes Aug 25 22:18:12 zecke: That would be good! Aug 25 22:18:19 that is probably your time spent most wisely Aug 25 22:18:28 everybody can pitch in with the list Aug 25 22:18:42 We might even want to request help on the ml Aug 25 22:19:01 I'll call it a day and go to bed a bit more early than as of late Aug 25 22:21:19 Laibsch, I think you need the sleep :) Aug 25 22:21:36 zecke, moving on from mtn would be good Aug 25 22:21:39 Why, my commits so bad lately? ;-) Aug 25 22:22:03 even if I still haven't really worked out git Aug 25 22:22:48 Laibsch, everyone needs a break now and then :) Aug 25 22:23:20 03  07org.openembedded.dreambox * rb8c385755d 10OE.dev/packages/zd1211/zd1211b-r222/ (ZD1211LnxDrv_2_22_0_0.patch zdiface.patch): zd1211b-r222: fix patches Aug 25 22:24:31 and "emily" is backup and running mtn2git again Aug 25 22:26:00 emily? Aug 25 22:26:13 you are naming machines after women? Aug 25 22:28:05 Crofton|work: women are named after my machines? Aug 25 22:28:52 Crofton|work: this machine sits at my old office and has "emily the strange" stickers on it saying "Get Lost"... this doesn't prevent people from shutting it down... Aug 25 22:31:31 03  07org.openembedded.dev * r9f7d440c38 10OE.dev/conf/distro/include/sane-srcrevs.inc: sane-srcrevs.inc: add SRCREV for tichy Aug 25 22:31:32 03  07org.openembedded.dev * ra820ced208 10OE.dev/: Aug 25 22:31:32 merge of '24a01b709a1340721c80ac90b928ab9d4e1ea0be' Aug 25 22:31:32 and '60842b466603b662ce4ff57d65e2d0ff713c018e' Aug 25 22:31:35 heh Aug 25 22:31:38 03  07master * r9f7d440c38 10OE.dev/conf/distro/include/sane-srcrevs.inc: sane-srcrevs.inc: add SRCREV for tichy Aug 25 22:31:38 03  07master * ra820ced208 10OE.dev/: Aug 25 22:31:39 merge of '24a01b709a1340721c80ac90b928ab9d4e1ea0be' Aug 25 22:31:41 and '60842b466603b662ce4ff57d65e2d0ff713c018e' Aug 25 22:31:46 waaaa!!!! CIA is back!!! Aug 25 22:31:54 hehe Aug 25 22:31:58 go cia Aug 25 22:32:01 (I can no longer commit in secrecy!) Aug 25 22:34:52 mwester: well, it is back until emily get's shut off again... which currently happens once a week :) Aug 25 22:35:21 zecke you should fire the facility mamanger Aug 25 22:35:34 woglinde: den auszubildenen? Aug 25 22:47:02 zecke: just hook up the power switch to the mains... that should get them to stop turning it off :) Aug 25 23:01:58 nite Aug 25 23:05:06 'night all Aug 25 23:05:08 03  07org.openembedded.dev * r4062e9bd97 10OE.dev/packages/xmms-embedded/xmms-embedded_20040327.bb: xmms-embedded: put .debug dir in $PN-dbg Aug 25 23:05:15 03  07master * r4062e9bd97 10OE.dev/packages/xmms-embedded/xmms-embedded_20040327.bb: xmms-embedded: put .debug dir in $PN-dbg **** ENDING LOGGING AT Tue Aug 26 02:59:57 2008