**** BEGIN LOGGING AT Thu Mar 02 10:59:56 2006 Mar 02 11:01:29 Who pushed them this time out of interest? Mar 02 11:04:10 good morning Mar 02 11:04:14 hi Bernardo Mar 02 11:05:19 hrw|work: what is oe's most up-to-date source mirror? yesterday I tried building svn for a friend, but couldn't find apr's 1.1.1 sources anywhere. Mar 02 11:05:46 Bernardo: oesources.org or familiar.handhelds.org Mar 02 11:06:46 ok, I tried familiar but it is missing apr, maybe oesources has it. Googling for apr 1.1.1 oly got me outdated hits, as all mirrors it found only had 1.2.2 Mar 02 11:08:40 koen: sent Mar 02 11:08:47 njs: thanks Mar 02 11:09:43 OE is based on Portage, right? Mar 02 11:09:46 err, BitBake* Mar 02 11:09:54 luke-jr_: *was* based Mar 02 11:10:09 ah, ok... so Portage roots are eliminated? ;) Mar 02 11:10:13 yes Mar 02 11:10:17 fun Mar 02 11:10:28 guess I'll write the BB plugin seperately then Mar 02 11:11:06 * luke-jr_ is working on a prototype pkg mgr for his OS and the Portage plugin is now functional ;) Mar 02 11:14:05 * koen fink install monotone-viz Mar 02 11:14:37 ~lart monotone merge Mar 02 11:14:37 * ibot whacks monotone merge with the cluebat Mar 02 11:14:59 hrw|work: the 0.26 merger is pretty sweet Mar 02 11:15:15 koen: will not fail when meld cannot run? Mar 02 11:15:50 hrw|work: they promised to implement a check for that Mar 02 11:16:12 OK, so blame me for that push. Now can't these revisions be simply disapproved? Mar 02 11:16:26 http://pastebin.com/579924 Mar 02 11:16:53 I 'love' monotone.... Mar 02 11:17:16 now I do not even know can I trust my database ;( Mar 02 11:17:46 hrw|work: just hit control-C Mar 02 11:18:07 hrw|work: it's in a transaction, if you cancel then your db will be left unchanged :-) Mar 02 11:18:48 njs: but how can I merge it? its headless machine >1000km from here and I do not like X forwarding.. Mar 02 11:19:10 Can someone help me to use the right solution for that push (which I'm sorry for) Mar 02 11:19:59 hrw|work: what merger do you want to use? emacs emerge, vimdiff, ...? Mar 02 11:20:00 koen: can you install kdiff3 on ewi? meld suxx with those '/tmp/mt.ASVEPH /tmp/mt.LJHF5L Mar 02 11:20:03 names Mar 02 11:20:14 njs: vimdiff is ok for me Mar 02 11:20:38 njs: monotone should check somehow does kdiff3/meld run at all Mar 02 11:21:02 hmm, it looks like vimdiff is tried before meld, actually Mar 02 11:21:11 if you have the string "vim" in your EDITOR variable Mar 02 11:21:59 so the simple thing to try is $ EDITOR=vim monotone merge Mar 02 11:22:36 this whole merge tool selection thing is just totally stupid, really. definitely need a saner approach. Mar 02 11:22:40 njs: ok Mar 02 11:22:53 koen: thx for kdiff Mar 02 11:23:04 hrw|work: can I help revert that push, or is the db definitely affected? Mar 02 11:23:22 damn Mar 02 11:23:30 building meta-sdk failed Mar 02 11:23:35 again, glibc doesn't compile Mar 02 11:23:43 FredyD: koen is our monotone expert Mar 02 11:24:10 koen: can I help revert that push, or is the db definitely affected? Mar 02 11:24:22 hrw|work: thanks Mar 02 11:26:07 I hope version 0.26 of monotone includes a "dry-run" option for all network operations, saying what would be done without actually doing it... Mar 02 11:41:42 <_law_> RP, http://pastebin.com/579949 corgi module doesnt load Mar 02 11:41:55 FredyD: it doesn't for network operations, though I think someone added one for merging operations yesterday. Mar 02 11:43:17 FredyD: the _theory_ is supposed to be that network operations are totally safe, and you should always feel free to do them. I need to figure out how exactly far off from that we are... Mar 02 11:43:26 _law_: looks a simple one to fix :), I was worried I'd introduced a logic bug. Mar 02 11:44:03 (I just sent koen the description of how to workaround those merges being annoying to do, so perhaps he will decide the right thing is just to do the merges) Mar 02 11:44:42 <_law_> lrg, ok do you need anything else from me ? Mar 02 11:45:40 _law_: no, but I can't fix as RP has latest corgi files. I'm sure he'll fix and patch soon. Mar 02 11:47:53 _law_: The good news is that this is a reall simple problem - corgi.c is missing a define by the looks of it Mar 02 11:48:10 hrw@bitbake:~/devel/oe/org.openembedded$ for i in `cat killrevs` ; do monotone --db=../OE.db db kill_rev_locally $i;done Mar 02 11:48:13 monotone: misuse: revision 9c1b32d8825be71ddfdcbe3ddfa31b4cc8bb2aa0 already has children. We cannot kill it. Mar 02 11:48:17 I fscked my db? Mar 02 11:51:40 ~ewi Mar 02 11:51:41 ewi is, like, ewi546.ewi.utwente.nl - main backup server for OE monotone. Hosts database snapshot: http://ewi546.ewi.utwente.nl/OE/OE.db.bz2 Mar 02 11:52:21 hrw|work: no, you will get the revs back on the next pull Mar 02 11:53:25 if anyone is interested: http://pastebin.com/579962 Mar 02 11:53:47 I will rather start from snapshot - when I try to merge it take stuff from few days ago to merge... Mar 02 12:02:31 koen: thanks for that explanation Mar 02 12:05:28 hm.. hydra again.. Mar 02 12:05:33 hrw@bitbake:~/devel/oe/org.openembedded$ mt head Mar 02 12:05:34 3b1583d918d96ac1fb50e5e2108eb7ec175f698d koen@openembedded.org 2006-03-02T10:30:04 Mar 02 12:05:34 601faaded8a6427e29dc8dc43433fe9be662c96c frederic@openembedded.org 2006-02-20T14:43:35 Mar 02 12:05:45 I will not touch it now Mar 02 12:32:10 morning Mar 02 12:51:06 <[lala]> RP: ping Mar 02 12:54:05 X-UTwente-MailScanner: Found to be clean Mar 02 12:54:08 nice header Mar 02 12:55:10 hrw|work: my university runs spamassassin over all mails passing through the mx Mar 02 12:55:32 which means I can catch ~85% of the spam via procmail Mar 02 12:55:48 thunderbird takes care of the other 15% Mar 02 12:55:59 and its good Mar 02 12:56:06 my new hosting works Mar 02 12:59:23 which means that few packages from .dev does not fetch Mar 02 13:06:59 [lala]: pong Mar 02 13:09:47 <[lala]> RP: i just tried out that hciattach patch i got from mh (bluez guy) ... but it didn't help much :-( Mar 02 13:10:08 <[lala]> RP: would be great if you could take a look at it Mar 02 13:11:46 [lala]: I'll certainly aim to but I have some other work which needs my attention at the moment I'm afraid :-/ Mar 02 13:12:38 http://www.hrw.one.pl/2006/03/02/new-hosting/ Mar 02 13:17:26 <[lala]> RP: yeah sure. it's not that urgent :-) Mar 02 13:30:57 building the meta-sdk seems to fail when bitbake is building dbus. with MACHINE="h3600" and DISTRO="familiar-unstable". I'm tring it now with jan2006prebreakage revision Mar 02 13:32:33 I hate this feeling of not actually understanding what I'm doing :) Mar 02 13:53:11 finally! Mar 02 13:53:20 glibmm 2.8.4 works Mar 02 13:53:58 koen: can you fix repo to be one headed? Mar 02 13:54:25 I can, but I won't Mar 02 13:56:25 * jnc wakes up Mar 02 13:57:10 embarassed, to punish me for submitting a bug, God woke me up this morning seconds before my alarm clock went off Mar 02 13:57:45 jnc: God is the name of your cat? Mar 02 13:57:48 i've learned my lesson though, not going to reopen the bug Mar 02 13:58:07 jnc: I WONTFIX it Mar 02 13:58:35 oh belguim, man. belguim! Mar 02 13:58:55 s/ui/iu/g ? Mar 02 13:59:00 *nod* Mar 02 14:17:38 03koen 07org.oe.dev * r09e8c8e8... 10/packages/glibmm/glibmm_2.8.4.bb: glibmm: update to 2.8.4 Mar 02 14:17:43 03koen 07org.oe.dev * r1156e7b3... 10/packages/gtkmm/gtkmm_2.8.3.bb: gtkmm: update to 2.8.3 Mar 02 14:24:33 ~hail offlineimap again Mar 02 14:24:35 * ibot bows down to offlineimap again and chants, "I'M NOT WORTHY!!" Mar 02 14:25:25 hrw|work: tried it on pda yet? Mar 02 14:26:21 no need for it Mar 02 14:27:25 * chouimat kicks the repos Mar 02 14:32:29 03koen 07org.oe.dev * r0c67a234... 10/packages/gtk+/gtk+_2.8.9.bb: gtk+ 2.8.9: ship update-icon-cache binary Mar 02 14:32:52 ooooh Mar 02 14:32:55 it's snowing again Mar 02 14:33:37 grr Mar 02 14:33:41 nothing compiles Mar 02 14:38:43 * koen just build a gpe-image from .dev Mar 02 14:40:00 RP: I'll push a slightly better working hal in the next few minutes Mar 02 14:41:25 I wonder what's wrong with my efforts then. :) Mar 02 14:41:47 I'm trying 'bitbake nano' for h3600 and familiar-unstable now Mar 02 14:41:58 Im missing gnu/stubs-32.h which seems to be an important file Mar 02 14:41:59 with -rjan2006prebreakage Mar 02 14:43:11 bah, bloomin gaim needs a remove of /usr/include as well Mar 02 14:44:20 nothing like getting the monthly invoices ready to be send to customers with a cat purring on your lap Mar 02 14:47:26 hey zecke Mar 02 14:47:35 hey zecke Mar 02 14:48:07 zecke: http://pastebin.com/579962 Mar 02 14:49:51 03koen 07org.oe.dev * r9a56d894... 10/packages/hal/hal_0.5.7.bb: hal: update to 0.5.7 Mar 02 14:50:47 zecke: switch to normal MUA not that crappy i-mail Mar 02 14:51:02 koen: so if I pull the latest org.openembedded.org and try to build meta-sdk with MACHINE="h3600" and DISTRO="familiar-unstable", it should compile? Mar 02 14:51:15 hrw|work: sorry, I guess it didn't create newlines again :} Mar 02 14:51:16 Cockroach-: no idea, familiar isn't maintained anymore Mar 02 14:51:16 * chouimat kicks the repo until org.openembedded.dev get one head as the result of the kicking Mar 02 14:51:28 koen: Oh I see. Mar 02 14:51:29 gosh I should switch my workflow... Mar 02 14:51:38 chouimat: just select the correct head and continue Mar 02 14:51:45 zecke: no - it only indented it wierdly Mar 02 14:51:51 koen: how? Mar 02 14:52:01 I'm really sorry Mar 02 14:53:31 zecke: no problemo Mar 02 14:53:43 * chouimat hates monotone Mar 02 14:54:11 I guess reading the docs is overrated and unneeded for most people Mar 02 14:54:39 Documents are fine, as long as they exist and are found :) Mar 02 14:54:48 with monotone it's not a problem though Mar 02 14:54:55 koen: bahj Mar 02 14:56:53 koen: nm Mar 02 15:10:04 OT: did I sound too mean again? Mar 02 15:11:31 zecke: only in the bit that said all our devs are too stupid to use DSCMs Mar 02 15:12:25 zecke: it's funny that people like a undocumented scm (git) over a documented one (monotone) Mar 02 15:13:55 Well, given that nobody reads docs before trying something, I think they would prefer the thing that worked better right out of the box Mar 02 15:14:50 CosmicPenguin: and I now have an extra excuse to not write OE docs Mar 02 15:15:15 "linus doesn't write docs, so nobody has to" Mar 02 15:16:36 We don't need no stinkin' documents Mar 02 15:17:45 real programers does not write docs - it was hard to write so it should be hard to understand Mar 02 15:24:05 hmm Mar 02 15:24:24 So, somebody who uses multimachine - how do you make sysvinit-inittab work? Mar 02 15:24:57 CosmicPenguin: it's on the 'to clean' list before switching machines Mar 02 15:25:08 CosmicPenguin: you'll notice that tslib-conf has the same problem Mar 02 15:25:30 CosmicPenguin: bitbake doesn't know how to handle machine-dependent *sub* packages Mar 02 15:25:49 koen: right - thats why I was confused Mar 02 15:26:00 CosmicPenguin: bitbake -c clean matchbox-panel task-bootstrap meta-gpe libtool-cross meta-opie gpe-image opie-image sysvinit tslib base-passwd prism3-support opie-button-settings Mar 02 15:26:00 koen: I'm thinking it must work for you guys, but how? :) Mar 02 15:26:14 * CosmicPenguin writes that down Mar 02 15:26:15 that's probably overkill, but it works :) Mar 02 15:26:20 I have few more Mar 02 15:26:51 CosmicPenguin: http://ewi546.ewi.utwente.nl/tmp/build-angstrom.sh Mar 02 15:26:57 that's the script I use Mar 02 15:27:12 pivotroot-image (spitz) zaurus-updater initscripts keymaps base-files Mar 02 15:27:33 koen: ah, cute Mar 02 15:27:55 http://ewi546.ewi.utwente.nl/tmp/hrw/build-oz.sh is mine for OpenZaurus Mar 02 15:28:43 CosmicPenguin: it's still too much 'unbreak me' kind of actions for my taste Mar 02 15:29:13 koen: we probably need something like MULTIMACHINE_UNFRIENDLY = 1 Mar 02 15:29:27 koen: its better then six bitbake; rm -rf /tmp actions Mar 02 15:29:54 ahah, I remebered gaim's compile on amd64 host fix I needed to push Mar 02 15:40:00 The MULTIMACHINE_UNFRIENDLY idea is a good one Mar 02 15:41:26 hi Mar 02 15:41:30 CoreDump|home: hi Mar 02 15:43:32 hey CosmicPenguin Mar 02 15:43:42 yes? Mar 02 15:43:57 aaargh Mar 02 15:43:59 ~lart me Mar 02 15:43:59 * ibot installs a bad bootloader on koen and turns koen into a brick Mar 02 15:44:04 hey CoreDump|home Mar 02 15:44:20 heh Mar 02 15:44:48 ~lart openssl/gnutls Mar 02 15:44:48 * ibot cuts openssl/gnutls into thin stripes Mar 02 16:03:29 03hrw 07org.oe.dev * ra04fd7da... 10/packages/ (8 files in 2 dirs): Mar 02 16:03:29 changed SRC_URI in my packages to ewi instead of hrw.one.pl Mar 02 16:03:29 - I moved with hosting and have transfer limits Mar 02 16:04:42 yay! Mar 02 16:04:48 avahi works Mar 02 16:07:50 http://pastebin.com/580295 Mar 02 16:17:18 03coredump 07org.oe.dev * r09145be3... 10/packages/altboot/ (4 files in 2 dirs): altboot: Releas of the first stable version: 0.0.5. Tested on Akita and Collie Mar 02 16:22:42 is a DNS configuration guru around? Mar 02 16:34:03 is the ugly head back in OE? Mar 02 16:37:32 koen: the mtn repo is broken again Mar 02 16:37:39 I know Mar 02 16:37:45 see my post to oe@ Mar 02 16:37:56 ugh, I'm not subscribed Mar 02 16:38:03 * NAiL probably should be Mar 02 16:38:23 http://www.handhelds.org/hypermail/oe/68/6823.html Mar 02 16:38:34 ty Mar 02 16:38:35 njs gave us a step by step solution Mar 02 16:38:57 http://pastebin.com/579962 Mar 02 16:39:20 i want my own head too! Mar 02 16:39:28 hey mithro Mar 02 16:39:33 back in oz yet? Mar 02 16:39:38 I like your reply ;) Mar 02 16:40:23 nope Mar 02 16:40:30 3 more days Mar 02 16:40:33 NAiL: http://www.handhelds.org/hypermail/oe/68/6826.html second point Mar 02 16:49:08 Yeah, when's the time for voting? :) Mar 02 16:50:38 It seems to me like there is a fundamental problem here, if you Mar 02 16:50:38 cancel/revoke/remove a head, why is it getting re-pushed? Mar 02 16:50:38 Shouldn't monotone know that you are pushing a revision which has Mar 02 16:50:38 already been pushed (even though it has now been cancel/revoke/removed)? Mar 02 16:52:17 mithro: the 'fix' was to use kill_rev_locally, which just removes the revs your the local db Mar 02 16:52:45 koen: so you are doing something outside the revision control system? Mar 02 16:53:00 yes Mar 02 16:53:07 hence your problems Mar 02 16:53:13 no Mar 02 16:53:25 our problems is still people who don't read Mar 02 16:53:43 koen: no, your problem is that I shouldn't have to read to use the revision control system Mar 02 16:54:01 you shouldn't Mar 02 16:54:31 re: the multimachine business, that's a bug in OE's handling. We should really have something that marks a target as having subpackages that are machine dependent... Mar 02 16:55:08 * RP will give that one some further thought as it irritates me too... Mar 02 16:55:10 if I push something to the database which isn't good, only one person should do something to fix it and it should be fixed Mar 02 16:55:18 mithro: the issue is it has ben killed (just read your email) Mar 02 16:55:53 koen: Good news re hal and avahi. I have an updated hal package and a list of things that need fixing in it to at least make it run but haven't had time to acutally fix yet Mar 02 16:56:06 zecke: so the head will keep reappearing until all the revision disappears from the universe Mar 02 16:56:33 right :) Mar 02 16:56:38 hence its the wrong solution to the problem Mar 02 16:56:53 mithro: well no. It is a social issue Mar 02 16:56:54 mithro: until frederic merges Mar 02 16:57:22 mithro: which is a matter of copy/pasting from njs' comments Mar 02 16:58:34 zecke: the social issue is unsolvable, you will never get everyone to read the email before pushing Mar 02 16:59:37 and even then you have race issues Mar 02 16:59:56 and infact the revision control system should mean that it isn't a problem Mar 02 17:00:33 * koen points to http://pastebin.com/579962 Mar 02 17:01:38 koen: still has race conditions i believe? Mar 02 17:02:01 no race conditions, just a bug in the lca algo Mar 02 17:02:50 I tried it in monotone 0.26 and the problem wasn't there Mar 02 17:03:00 just one merge of the kernel defconfig Mar 02 17:03:56 koen: so if everyone does a merge before push you don't have a problem? Mar 02 17:04:12 correct Mar 02 17:04:28 the problem here was caused by someone merging everyday, but not pushing Mar 02 17:05:04 which confused monotone 0.25 Mar 02 17:05:38 as njs points out, a look with monotone-viz would solve it, or the shell command Mar 02 17:06:36 cu Mar 02 17:06:41 cu Mar 02 17:06:54 but merge and push arn't 1 atomic operation are they? Mar 02 17:07:33 hence you have a race condition Mar 02 17:08:53 the race will always be there Mar 02 17:08:55 by design Mar 02 17:09:02 then that is a problem Mar 02 17:09:11 it isn't, actually: http://venge.net/monotone/downloads/codecon-2006.pdf Mar 02 17:09:24 it garantees that 'push' will *always* work Mar 02 17:09:39 koen: will a pull always work? Mar 02 17:09:48 it will Mar 02 17:10:10 the only problem arises when you do 'monotone update' Mar 02 17:10:25 that will say 'multiple heads' Mar 02 17:10:53 so you do 'monotone heads' and 'monotone update -r' Mar 02 17:11:21 so why is there not a default head? Mar 02 17:11:50 hi, folks. I just now, I bitbake bootstrap-image. And "ERROR: function do_patchcmd failed" while patching the mtd-utils-native. Mar 02 17:12:13 sounds like your bitbake is too old Mar 02 17:14:22 And I retried the "bitbake bootstrap-image", It seems is skip compiling "mtd-utils-native" and it start compiling "busybox" stuff. Mar 02 17:15:11 My question is , how can I continue compiling mtd-utils-native ? Mar 02 17:16:05 update bitbake, remove cache, try again Mar 02 17:17:17 anyway I got to go Mar 02 17:17:40 It still seems to me like there is something strange going on here Mar 02 17:24:45 koen, yes. The second time, I did as you said. Can I continue compiling mtd-utils-native ? Mar 02 17:25:06 leoncamel: bitbake -i Mar 02 17:25:20 then type 'rebuild mtd-utils-native' Mar 02 17:25:51 the old version of bitbake downloaded a wrong version against which the patches don't apply Mar 02 17:27:15 koen, thanks. I am trying now .. :) Mar 02 17:27:49 zecke, RP: any new release of bitbake in sight? Mar 02 17:28:28 leoncamel: RP and zecke fixed some really deep bugs a while ago, so the latest OE needs the latest bitbake from svn Mar 02 17:34:12 koen, I already update bitbake "BitBake Build Tool Core version 1.3.3.3, bitbake version 1.3.3.2" Mar 02 17:35:38 koen, but It still failed during patch mtd-utils-native: here is the log: Mar 02 17:35:43 koen, ERROR: function do_patchcmd failed Mar 02 17:35:59 koen, ERROR: log data follows (/stuff/tmp/work/i686-linux/mtd-utils-native-20050801-r0/temp/log.do_patchcmd.23321) Mar 02 17:36:45 20050801? Mar 02 17:37:13 mtd-utils-native-20060223: completed Mar 02 17:37:19 that might be the problem :) Mar 02 17:37:39 koen, "| Patch patches/add-exclusion-to-mkfs-jffs2.patch does not apply (enforce with -f)" Mar 02 17:38:02 old skool Mar 02 17:38:10 koen, hmm. shall I do monotone pull ? Mar 02 17:38:20 leoncamel: yes Mar 02 17:38:27 koen, thanks . :0 Mar 02 17:38:31 koen, :) Mar 02 17:41:12 leoncamel: after pulling do a 'monotone update -ra67386ccbaf4538441ab5d8769e018207ab0c744' Mar 02 17:42:09 koen, what's that mean ? Mar 02 17:42:19 that it will update to a certain revision Mar 02 17:42:22 03koen 07org.oe.dev * rbad62b43... 10/conf/distro/angstrom-2006.9.conf: angstrom-2006.9: fix stupid error in feed uris Mar 02 17:42:36 leoncamel: otherwise you will hit a stupid error I'm fixing Mar 02 17:42:41 koen, hmm. the lastest version is not stable ? Mar 02 17:42:50 that is the latest version :) Mar 02 17:43:29 koen, you mean a67386ccbaf4538441ab5d8769e018207ab0c744 is the lastest stable one ? Mar 02 17:43:45 yes Mar 02 17:43:58 koen, :) thanks. Mar 02 17:52:53 03koen 07org.oe.dev * ra67386cc... 10/ (13 files in 5 dirs): Mar 02 17:52:53 explicit_merge of '09145be36a071f7c39a4d421da54e87a449cecc1' Mar 02 17:52:53 and 'bad62b438dd2c5c176aff6068068f022faa98dbb' Mar 02 17:52:53 using ancestor '' Mar 02 17:52:53 to branch 'org.openembedded.dev' Mar 02 17:53:38 heh, that revision sounds ominous: 'bad...' Mar 02 17:56:02 mreimer: actually that was the 'good' revision :) Mar 02 17:56:13 hey koen Mar 02 17:56:24 hey mreimer Mar 02 17:56:41 how does your shiny new gcc4/gtk 2.8/softfloat image behave? Mar 02 17:56:59 mreimer: pretty good Mar 02 17:57:12 if i'm building multiple packages out of a single bb file, can i have do_install_append_PACKAGE ? I'd like to append just one of them. Mar 02 17:57:19 I'm trying to make a scap of dia right now :) Mar 02 17:57:41 koen: nice. does it use significantly more memory, or feel slower/faster? Mar 02 17:58:10 I switched from mtd to sd Mar 02 17:58:17 that makes it feels *a lot* faster Mar 02 17:58:22 koen: what device is this on? Mar 02 17:58:25 c700 Mar 02 17:58:28 32mb ram Mar 02 17:58:31 pxa250 Mar 02 17:58:34 yeah, jffs2 is slow Mar 02 17:58:37 wow, small machine Mar 02 17:58:46 HopsNBarley: yes. Mar 02 17:59:09 HopsNBarley: grep a bit in the OE tree to see examples of this being done Mar 02 17:59:21 HopsNBarley: I think util-linux is one of the packages using this Mar 02 17:59:33 i did, but i'm having a problem... thanks. BTW no love on the arp cache.. Mar 02 17:59:34 NAiL, HopsNBarley: do_install should run before do_package Mar 02 17:59:35 I'm pulling right now, so I won't bother looking ;) Mar 02 18:00:08 koen: This is a do_install_append function for a package that is being split off the main package Mar 02 18:00:45 NAiL: And how would that work? Mar 02 18:00:56 NAiL: do_install doesn't know about the PACKAGES variable Mar 02 18:01:28 unless you have a nasty way of splitting the package Mar 02 18:01:40 in the latter case you might consider using multiple .bbs Mar 02 18:02:19 ah Mar 02 18:02:31 * NAiL was thinking of postinst Mar 02 18:02:47 i thought so :) Mar 02 18:02:53 HopsNBarley: Just ignore my comment ;) Mar 02 18:02:59 i though all the functions could be adapted for individual packages, ala the pkg_postinst_util-linux-{losetup,unmount,mount,...} Mar 02 18:03:07 NAiL: the evil heads are gone Mar 02 18:03:20 koen: thank you Mar 02 18:04:01 ah, so postinst does this override, but do_install does not. got it. any source tips for figuring these rules out? Mar 02 18:04:05 koen: thanks. Mar 02 18:05:20 HopsNBarley: it's about the split between building stuff and packaging stuff Mar 02 18:05:44 HopsNBarley: 'make install' doesn't know about how you want to package stuff Mar 02 18:06:24 In most cases, there's no need for do_install to know either Mar 02 18:08:38 03koen 07org.oe.dev * r16c0c1b0... 10/ (587 files in 252 dirs): Mar 02 18:08:38 explicit_merge of '601faaded8a6427e29dc8dc43433fe9be662c96c' Mar 02 18:08:38 and 'a67386ccbaf4538441ab5d8769e018207ab0c744' Mar 02 18:08:38 using ancestor '9c1b32d8825be71ddfdcbe3ddfa31b4cc8bb2aa0' Mar 02 18:08:38 to branch 'org.openembedded.dev' Mar 02 18:15:35 koen: so now everyone can safely sync even though they have the broken revisions? Mar 02 18:16:45 NAiL: yes Mar 02 18:20:00 koen: thanks fix fixing it Mar 02 18:21:19 JustinP: it was real hard, I had to copy and paste 3 command from njs' mail Mar 02 18:21:29 and click 'save' in kdiff3 Mar 02 18:23:29 http://www.computingunplugged.com/issuesprint/issue200602/00001725.html Mar 02 18:23:41 'One of the biggest problems we had was that the Zaurus platform was a combination of Trolltech's Qt Embedded and Lineo's embedded Linux platform' Mar 02 18:24:46 koen: oh. I wasn't aware that he'd "fixed" it ;-) Mar 02 18:25:06 Now there's an understatement Mar 02 18:26:06 heh Mar 02 18:27:13 03koen 07org.oe.dev * r39e8a324... 10/packages/dia/dia_0.94+0.95pre3.bb: dia: packaging tweaks and pkgconfig workarounds Mar 02 18:51:14 koen: hmmm... and i remember the state of gpe 6 years ago too... Mar 02 18:51:14 evening Mar 02 18:51:58 hey reenoo Mar 02 18:52:07 yo koen Mar 02 18:53:11 gotta love that article, first he talks about open source developers, Mar 02 18:53:41 then digs at TT for a "high" ($195, in fact) price of a comemrcial sdk Mar 02 18:54:03 as if open source developers sell software Mar 02 18:55:00 his price is wrong for the sdk Mar 02 18:55:35 didn't it used to be alot higher? Mar 02 18:56:21 nope Mar 02 18:56:49 thats the price of an sdk to develop qtopia commercial software Mar 02 18:57:03 sdk 'lite' they are now calling it Mar 02 18:57:31 ljp: ah. i was looking at a heavier weight license for qt3 ent. Mar 02 18:58:05 ya, desktop is more expensive Mar 02 18:58:12 ljp: the writer also says that gtk is gpl so alp must be completely open :) Mar 02 18:58:16 which is just so not true Mar 02 18:58:37 heh Mar 02 18:59:12 he also neglects to mention sharp pushed for release before it was really ready Mar 02 18:59:33 I haven't seen a megacorp not doing that Mar 02 18:59:38 if only i could tell the whole true story about all that Mar 02 19:00:45 ljp: go head :-) Mar 02 19:01:21 i wish i could Mar 02 19:25:02 ljp: why can't you? Mar 02 19:36:57 cause i like my job :) Mar 02 19:38:23 what's your job ljp ??? :) Mar 02 19:41:53 trolltech's qtopia community manager Mar 02 20:39:52 * koen looks for mickeyl Mar 02 20:40:07 * koen opens a bottle of Leffe Mar 02 20:40:17 whats leffe? Mar 02 20:40:22 kind of beer? Mar 02 20:40:30 woglinde: yes Mar 02 20:40:44 hm I like grolsch Mar 02 20:40:56 or oettinger Mar 02 20:42:42 woglinde: the grolsch brewery is ~3km from here :) Mar 02 20:43:06 he you lucky one *g* Mar 02 20:43:21 a right you are in twente Mar 02 20:43:25 ljp: I see.... Mar 02 20:46:47 I rember seeing the brewery at the trip to HAL 2001 Mar 02 20:47:05 that was in my backyard! Mar 02 20:47:15 although I arrived at uni just after HAL Mar 02 20:47:45 the weather was lousy Mar 02 20:51:59 re koen Mar 02 20:52:22 i like beer (-; Mar 02 20:52:29 who not Mar 02 20:53:06 he but you are from usa there isnt good beer perhaps imported Mar 02 20:53:32 contrary to popular belief, there is excellent beer in the US! You have to look for it though... Mar 02 20:53:47 woglinde: I had a great amber beer in New Orleans Mar 02 20:53:51 for example, you could start in my brewing fridge in my garage (-; Mar 02 20:54:04 abita turbo dog? not bad. Mar 02 20:54:12 sorry but I dont think I will visit usa Mar 02 20:54:18 *g* Mar 02 20:54:26 args Mar 02 20:54:41 HopsNBarley: abita indeed :) Mar 02 20:54:53 lol Mar 02 20:55:04 the name is funny Mar 02 20:55:23 koen: i've had that in new orleans also. does it mean something auf deutsch? Mar 02 20:55:28 we have somehting like tannzäpfle Mar 02 20:55:46 HopsNBarley: I don't think it means anything Mar 02 20:56:01 dog is dog Mar 02 20:56:08 woglinde: dancing apples? Mar 02 20:56:18 koen lol no Mar 02 20:56:24 my favorites are european though. pilsner urquel, guiness, guerz, ... Mar 02 20:56:54 damnit dic.leo.org dont know tannzapfen Mar 02 20:57:05 try tannenzapfen Mar 02 20:57:18 oh thanks reeno Mar 02 20:57:41 HopsNBarley: the funny thing is that in the US they go all "munich style" and "dresden style" when describing beers Mar 02 20:57:43 koen means something like fir cone Mar 02 20:58:03 I still have to find a person from germany which knows what those mean Mar 02 20:58:20 oh my god last time I saw something carneval in usa Mar 02 20:58:22 woglinde: I thought so when reenoo mentioned tannenzapfen Mar 02 20:58:27 koen: keep looking heh Mar 02 20:58:38 koen oh okay Mar 02 20:58:49 koen: that's funny. i haven't heard that. do you mean like "crisp hop flavor" (usually meaning no hop flavor...) and similar rubbish? Mar 02 20:59:10 HopsNBarley: similar rubbish Mar 02 20:59:13 koen hm maybee dresden == pils munich == weizen or maas Mar 02 20:59:24 HopsNBarley: it was in the skymall magazine Mar 02 20:59:46 koen: a pillar of US culture i assure you. Mar 02 20:59:51 Dresden style would probably be the pils Mar 02 21:00:03 cosmic yes Mar 02 21:00:16 HopsNBarley: http://www.penny-arcade.com/comic/2005/09/26 Mar 02 21:00:22 night all Mar 02 21:00:26 i'm thinking that would be more Alt, no? americans get bavaria confused with pils. Mar 02 21:00:28 night lrg Mar 02 21:01:05 koen: thanks for the laugh!!! Mar 02 21:01:06 hops no, alt is more ruhrpott, duesseldorf Mar 02 21:01:07 HopsNBarley: no - definately not Bavarian Mar 02 21:01:15 iirc pils isn't even german Mar 02 21:01:36 it sorta bleeds over into germany, if i have my history correct. Mar 02 21:01:56 but yeah - pilsen-er - of pilsen, which is NOT germany. Mar 02 21:02:05 Dresden is what, all of 30 minutes from the Chech border anyway Mar 02 21:02:19 If that Mar 02 21:02:30 Cezch Mar 02 21:02:31 lol next fosdem every one should bring a crate of his favorite beer Mar 02 21:02:40 CosmicPenguin: i clearly have my germany geography all messed up. Mar 02 21:02:51 woglinde: they actually sell beer at fosdem :) Mar 02 21:03:00 * mhentges s favourite beer is Vitamalz =) Mar 02 21:03:04 http://en.wikipedia.org/wiki/Pilsner Mar 02 21:03:05 HopsNBarley: I have knowledge of this only becuase I'm in Dresden a few times a year on business... :) Mar 02 21:03:11 koen but I wasnt there Mar 02 21:03:15 is fodem always in brussels? Mar 02 21:03:31 HopsNBarley: yes Mar 02 21:04:55 i love the puppet bar there - is it still? Mar 02 21:05:31 I did encounter a very nice bar in brussels Mar 02 21:05:45 "the brewer of the grand place" Mar 02 21:05:53 decorated with kettles and stuff Mar 02 21:06:12 after half an hour the owner explained that it wasn't decoration Mar 02 21:06:22 they actually brewed it there :) Mar 02 21:06:25 koen: brewing paraphenalia? ah- brewing equipment! Mar 02 21:06:40 Heh Mar 02 21:09:12 here's my rig: http://www.morebeer.com/product_images/1/1874.jpg Mar 02 21:10:29 cute Mar 02 21:10:40 top or bottom fermentation? Mar 02 21:10:44 40L at a time. Mar 02 21:10:45 Heh Mar 02 21:10:52 I love those - they look so cool Mar 02 21:11:15 that's not the fermenter, just the cookers, etc. could do either top or bottom. Mar 02 21:12:08 there are beers that are fermented both top and bottom Mar 02 21:12:32 CosmicPenguin: they do a great job making these. i looked high and low and they were the best. Mar 02 21:13:08 koen: sure, though that perhaps is a more european way of describing lager and ale yeast. Mar 02 21:14:08 HopsNBarley: I've got a buddy that brews alot, and I keep trying to get him to buy one Mar 02 21:14:26 sire brew-a-lot Mar 02 21:14:33 s/sire/sir/ Mar 02 21:14:38 CosmicPenguin: is he doing all-grain? Mar 02 21:14:39 Mainly so he can brew more beer and so I can drink it Mar 02 21:14:49 HopsNBarley: He's starting to Mar 02 21:14:54 koen: sadly no Mar 02 21:15:26 CosmicPenguin: then he needs one. it's a terrific system. takes a few batches to "get your legs" with it. Mar 02 21:18:53 zecke did you drive to dresden? Mar 02 21:26:32 woglinde: yeah all the way through :) Mar 02 21:43:52 koen: I added some improvements to bitbakes git fetcher btw. Should be a bit more efficient now Mar 02 21:44:03 nice Mar 02 21:44:13 maybe I'll add cairo and xlib soon Mar 02 21:48:05 * chouimat is away: Out in the world to discover the different between man and woman ;-p Mar 02 22:04:22 hi zecke Mar 02 22:04:36 hey zecke Mar 02 22:05:27 Hmm. Sound on corgi is more broken than expected and its in Liam's code changes now, not mine :-/ Mar 02 22:05:27 I'm not here :) I belong to bed Mar 02 22:07:40 okay finally good nite for now :) Mar 02 22:07:48 'night zecke :) Mar 02 22:07:48 'night zecke Mar 02 22:08:02 dream of fft Mar 02 22:08:33 dft is much better Mar 02 22:08:55 koen: doc fft in your favorite matlab or octave window :) Mar 02 22:08:57 we are poor we only have fft Mar 02 22:10:09 well FFT is a fast implementation of the discreate fourier transformation :) Mar 02 22:11:42 any thing that sounds like a fart is crap ;) Mar 02 22:25:57 'night all Mar 02 22:27:07 'night koen|sleep Mar 02 23:38:41 * france is back (gone 42:50:32) Mar 03 04:01:36 * france is away: Away Mar 03 04:04:29 Hi guys, anyone know where I can get built kernel/bootstrap.initrd images for openzaurus 3.5.4-ish? Is it out yet? I've been trying to learn bitbake and built some programs, but found I didn't have the version I built for, and I don't want to spend 2 days attempting to build the release :) Mar 03 04:05:14 ebob9: set distro to openzaurus-3.5.3 and you can build for that distro, 3.5.4 isn't out yet Mar 03 04:06:49 ok, what program do I do that in? bitbake or svn or both? Mar 03 04:07:08 sorry, still learning this Mar 03 04:07:17 ebob9: set it in conf/local.conf Mar 03 04:07:31 ok Mar 03 04:07:56 thanks, I'll try that! Mar 03 04:18:35 ok, to build for openzaurus-3.5.3, do I need to pull another branch? I'm using org.openembedded.oz354fam083, is that correct? Mar 03 04:18:59 see #openzaurus, not here Mar 03 04:19:10 this is a developer channel, not a support channel Mar 03 04:19:30 okies, thanks Mar 03 04:57:29 ebob9: i was going to answer your question Mar 03 04:57:33 but then you left #oz Mar 03 05:19:27 ah, sorry jnc, I got a /msg with some info, seems to be working. Thanks though! Mar 03 05:19:48 ebob9: you are welcome :-) Mar 03 05:55:48 hey would some guys please work on integrating the new wpa_supplicant into the build root? Mar 03 05:56:00 and depreciate wpa_supplicant-0.2.6-nossl Mar 03 06:22:36 hi Mar 03 06:22:55 jnc: look at bugtracker - I added recent hostap stuff there Mar 03 06:23:04 jnc: I have no time to check it Mar 03 07:05:32 hey Mar 03 07:08:01 ~curse kmail/kio_imap badly Mar 03 07:08:02 May you be reincarnated as a Windows XP administrator, kmail/kio_imap badly ! Mar 03 08:21:19 Damn, I'm starting to get desperate :P Mar 03 08:23:44 I pulled the org.openembedded.oz354fam083 and tried to compile meta-sdk with MACHINE="h3600" and DISTRO="familiar-0.8.3". koen said yesterday that (at least with branch org.openembedded.dev) the "familiar-unstable" was not maintained anymore. Is this the case with oz354fam083 too? Mar 03 08:25:07 It's getting on my nerves, two days of trial and error, and I still don't have a environment where I could build a simple "hello world", cross compile it and install into an IPaq Mar 03 08:31:32 The road is long and dangerous Mar 03 08:39:03 ~change 1643 pln to eur Mar 03 08:39:09 1,643.00 Zloty (PLN) makes 432.138 Euro (EUR) (from http://www.xe.com/) Mar 03 08:39:22 132EUR extra for n770.. Mar 03 08:42:30 good morning all Mar 03 08:42:40 <[lala]> morning koen Mar 03 08:43:01 hey koen Mar 03 08:43:36 pH5: http://handhelds.org/scap/port.9075.png Mar 03 08:45:02 koen: nice :) Mar 03 08:46:00 I have a kernel module that needs some firmware but the firmware under /lib/firmware/* never get shipped. Mar 03 08:46:28 RP: could you add your hal fixes to 0.5.7? I'm having trouble starting hal due to some dbus error Mar 03 08:47:18 mproctor_home: try adding FILES_${PN} += "/lib/firmware" Mar 03 08:49:38 Should the external atmelwlandriver and hostap-modules packages compile against 2.6.15? Mar 03 08:50:39 pH5: they should, but at least atmelwlandriver doesn't Mar 03 08:51:53 koen: still get message that they are not shipped. Mar 03 08:51:56 pH5: hrw|work posted some updated hostap stuff in bugzilla, maybe it's worth adding that Mar 03 08:54:30 I have removed atmelwlandriver from my conf for quite some time :-/ Mar 03 08:54:38 * pH5 searches bugzilla Mar 03 08:56:57 koen: thanks had to put it in a differant place in the file. Mar 03 08:58:16 can you 'host hrw.one.pl' and tell ip? Mar 03 08:58:44 ~host hrw.one.pl Mar 03 08:58:47 see nslookup hrw.one.pl Mar 03 08:58:53 hrw|work: 195.149.226.213 Mar 03 08:58:57 ~nslookup hrw.one.pl Mar 03 08:59:00 hrw.one.pl is 195.149.226.213 Mar 03 08:59:15 195.149.226.213 Mar 03 08:59:30 thx Mar 03 08:59:42 so it is propagated now Mar 03 09:00:20 morning all Mar 03 09:00:36 hey katossi Mar 03 09:01:00 morning all Mar 03 09:01:40 morning katossi, greentux_alt2 Mar 03 09:01:50 after updating without the prebreakage thing, I get hundreds of error when I compile anything, am I missing anything new? Mar 03 09:03:45 katossi: probably a recent bitbake Mar 03 09:04:48 I have 1.3.2 Mar 03 09:05:42 old one Mar 03 09:06:46 I just upgraded but I still get a lot of : ERROR: anonymous function: access() argument 1 must be string, not None Mar 03 09:08:03 Does OE/bitbake ever use the gcc/cc/c++ (to build the toolchain or something) of the host, or does it always download it according to .bb files and then builds the toolchain with those downloaded ones? Mar 03 09:08:26 meaning, if my host has gcc 4.0.2, might it cause problems when trying to build things. Mar 03 09:09:23 If this information is somewhere, along the "beginners guide to everything", please tell me where it is. Mar 03 09:09:26 :) Mar 03 09:10:05 host gcc is used to build *-native, glibc-initial, gcc-cross* Mar 03 09:10:42 Allright, thanks. Mar 03 09:25:43 koen, I got problem while bitbake bootstrap-image. Mar 03 09:29:11 koen, I got this error "| Downloading file:/stuff/tmp/deploy/ipk/task-bootstrap_1.0-r25.ipk" Mar 03 09:29:24 koen, "| task-bootstrap: unsatisfied recommendation for dropbear" Mar 03 09:29:39 koen, "| task-bootstrap: unsatisfied recommendation for portmap" Mar 03 09:29:43 koen, any idea ? Mar 03 09:30:07 sounds like portmap and dropbear didn't get built Mar 03 09:30:28 you guys use dropbear? Mar 03 09:30:42 yes Mar 03 09:30:46 it's awesome! Mar 03 09:30:57 cool! Mar 03 09:31:08 mostly because it avoids the linking with openssl mess Mar 03 09:31:13 I have only heard of it because the guy who wrote it is one of the main monotone devs :-) Mar 03 09:31:20 I know :) Mar 03 09:32:20 koen, hmm. how can I deal with it ? Mar 03 09:32:32 koen, how can I rebuild it ? Mar 03 09:32:48 bitbake dropbear portmap Mar 03 09:33:47 koen, but, I _do_ find the file "/stuff/tmp/deploy/ipk/dropbear_0.47-r0_iwmmxt.ipk" Mar 03 09:34:19 hmmm Mar 03 09:34:28 hi Mar 03 09:34:43 leoncamel: is iwmmxt in IPKG_ARCHS? Mar 03 09:34:54 koen, yes .. :) Mar 03 09:38:12 koen, is it a bug ? or not ? Mar 03 09:38:20 koen, and how can I deal with it ? Mar 03 09:39:08 <_guillermo> does anyone have a working local.conf for akita? I'm starting to think that's my problem, since I got oe.db in lots of different ways and still getting errors Mar 03 09:39:31 03koen 07org.oe.dev * rf275ef53... 10/conf/machine/include/tune-iwmmxt.conf: tine-iwmmxt: add iwmmxt to ipkg_archs Mar 03 09:40:27 mickeyl, hrw|work: I have cleaned up the hostap recipes: http://en.pastebin.ca/44316. Is this ok to commit? Mar 03 09:41:12 koen: I couldn't build glibc yesterday with tune-iwmmxt, it didn't know about the object file suffix. Mar 03 09:41:38 pH5: weird Mar 03 09:41:48 pH5: try building glibc head ;) Mar 03 09:41:51 koen, the last commit by you is for fixing the problem of mine ? Mar 03 09:42:18 leoncamel: it should fix it Mar 03 09:42:38 koen, what shall I do now ? monotone pulling and monotone update ? Mar 03 09:42:57 yeah, try that Mar 03 09:43:06 koen: will you push gcc 4.1 to .dev? Mar 03 09:43:26 pH5: I can Mar 03 09:43:27 pH5: why RPROVIDES_hostap-modules-cs = "hostap-conf not in .inc? Mar 03 09:43:27 koen, but must I "rm /stuff/tmp -rf" and "bitbake bootstrap-image" ? Mar 03 09:43:41 koen, It will be a very long time .. Mar 03 09:43:43 hrw|work: because it is removed in your latest versions in bugzilla Mar 03 09:44:03 leoncamel: that shouldn't be necessary Mar 03 09:44:17 pH5: no - it is lacking in my versions Mar 03 09:44:32 koen, hmm. what shall I do now ? Mar 03 09:44:37 pH5: I use .oz354fam083 branch which does not use hostap-conf package at all Mar 03 09:45:30 leoncamel: 'bitbake -c clean task-bootstrap bootstrap-image ; bitbake bootstrap-image' Mar 03 09:46:00 hrw|work: ok, if 0.4.7 and 0.3.10 should RPROVIDE hostap-conf, too, I'll add it to the .inc Mar 03 09:46:22 <_guillermo> if anyone could take a look at this http://pastebin.com/581625 and clearify my problems would be nice Mar 03 09:46:45 koen, before bitbaking, I must monotone pull and monotone update ? Mar 03 09:49:19 koen, ok, I am testing .. Mar 03 09:50:54 pH5: gcc 4.1 is in, see the commit message for hints how to solve the error during build Mar 03 09:51:04 koen: thanks! Mar 03 09:52:56 03koen 07org.oe.dev * r66f31781... 10/packages/gcc/ (6 files in 2 dirs): Mar 03 09:52:56 gcc: add 4.1 Mar 03 09:52:56 * ldflags.patch needs to be more thorough, you need to insert the patch to staging_incdir manually for now Mar 03 09:52:56 * default_preference is real low because of the above Mar 03 09:52:56 * this one has the necessary eabi bits Mar 03 09:54:12 <_guillermo> sorry about my questions before, all my probs came from a stupid typo in local.conf Mar 03 09:55:38 koen: The "fix" is "mkdir /tmp/hald-local; mkdir /tmp/hald-runner" Mar 03 09:57:13 RP: yay! it works Mar 03 09:58:05 10:01:21.663 [E] hald_dbus.c:3258: dbus_bus_get(): Failed to connect to socket /var/run/dbus/system_bus_socket: No such file or directory Mar 03 09:58:49 * koen restarts dbus Mar 03 10:02:02 koen, ok. bitbake successed.. :) Mar 03 10:02:57 hostap-modules 0.4.7 doesn't compile either: http://en.pastebin.ca/44317 - any ideas? Mar 03 10:02:59 RP: did you get hal-device-manager to work? Mar 03 10:03:03 good morning Mar 03 10:03:07 hey Bernardo Mar 03 10:03:17 koen: I never got pygnome sorted out... Mar 03 10:03:50 I remember ending up with x86 python modules Mar 03 10:04:22 pH5: I built every againt 2.4.18-collie. does not checked other kernels Mar 03 10:05:01 pH5: it built against akita 2.4 kernel Mar 03 10:07:49 hm, without kernel_updates.patch it builds... Mar 03 10:08:38 pH5: depends on kernel version... Mar 03 10:09:37 hrw|work: looks like some interface changed in 2.6.15, but those changes are not in 2.6.15-hh0? Mar 03 10:10:26 pH5: For 2.6.15 onwards, use the in kernel version of hostap Mar 03 10:11:10 pH5: In later versions of hostap, they probably added the changes kernel_updates makes Mar 03 10:11:23 RP: would it be possible to unleash a pygtk guru from OH to make a hal-device-manager lite? Mar 03 10:12:22 koen: you'd have to ask mallum. I suspect we'd also need a hal-lite and I'm not going there given the flak I already take on the issue Mar 03 10:12:28 RP: that means we have to remove hostap-modules-cs from BOOTSTRAP_EXTRA_RRECOMMENDS in ipaq-common.conf? Mar 03 10:12:30 <_guillermo> whe I try to build quilt-native I get :ERROR: Nothing provides runtime dependency diffstat-native Mar 03 10:13:40 RP: ok, I'll talk to mallum Mar 03 10:14:19 pH5: handheld-common and ipaq-common shouldn't be in conf/machine, but in conf/distro Mar 03 10:14:23 pH5: The zauruii handle that differently iirc Mar 03 10:14:29 <_guillermo> the truth is, I don't see any of the runtime dependencies of quilt-native Mar 03 10:15:08 _guillermo: See if you have an ASSUME_PROVIDED = instead of an ASSUME_PROVIDED += in your local.conf Mar 03 10:16:13 <_guillermo> I do, but I neved had problems with that before Mar 03 10:16:37 _guillermo: OE changed... Mar 03 10:17:45 <_guillermo> as the philosopher said: something has to change in order to keep all unchanged :) Mar 03 10:20:17 "plus ça change, plus ça c'est la meme chôse" :) Mar 03 10:20:28 koen: I think ipaq-common.conf should be dropped altogether. The kernel modules part should go into the machine files (and/or some cf-drivers.inc in distro/) and kbdd and bl should go into the distro files themselves or into some task- package in meta/ Mar 03 10:21:12 pH5: that's what I actually meant Mar 03 10:21:25 pH5: There has already been a proposal about this made by me on the OE mailing list/OEToDo list on the wiki Mar 03 10:21:35 koen: oh :) Mar 03 10:22:35 pH5: Rather than modules in the machine files, some would go into something like a kernel-modules-pxa27x.inc Mar 03 10:23:03 I really must find time to do another wave of cleanup... Mar 03 10:23:17 <_guillermo> RP: I added the + to ASSUME_PROVIDED, but now I get this: ERROR: /home/colbedded/stuff/org.openembedded.dev/packages/quilt/quilt-native_0.42.bb depends on itself (eventually) Mar 03 10:23:50 _guillermo: You have the latest bitbake? Mar 03 10:23:55 (from svn) Mar 03 10:24:07 <_guillermo> BitBake Build Tool Core version 1.3.3.3, bitbake version 1.3.3.2 Mar 03 10:25:43 _guillermo: not good :-/ Mar 03 10:26:32 <_guillermo> too old? Mar 03 10:26:39 RP: jabber isn't working, so I'll have a shot at gnome-python first :) Mar 03 10:26:42 <_guillermo> I just got that version! Mar 03 10:27:31 _guillermo: No, it looks reasonably up to date. quilt-native shouldn't be depending on itself so I'm a bit puzzled... Mar 03 10:29:32 is OE up to date? Mar 03 10:29:42 <_guillermo> RP: here http://pastebin.com/581665 you can see the whole error Mar 03 10:30:59 <_guillermo> mine its is Mar 03 10:34:24 This is a valid problem - quilt-native needs autoconf, autoconf needs quilt-native to apply patches Mar 03 10:36:02 RP: I've seen that trying to build for sharp-rom Mar 03 10:36:38 *sigh* Mar 03 10:36:40 afair, it was the same exact error Mar 03 10:36:41 Still no luck Mar 03 10:37:25 _guillermo: A quick fix is to add INHERIT += "patcher" to you local.conf file. The proper solution needs more thought Mar 03 10:37:45 apt-get install quilt Mar 03 10:38:00 Once it builds quilt-native, you won't need it... Mar 03 10:38:10 gcc-cross-sdk doesn't compile. Mar 03 10:38:25 03pH5 07org.oe.dev * rf985fb11... 10/packages/hostap/ (29 files in 5 dirs): hostap-modules: cleanup to avoid duplication Mar 03 10:38:30 03pH5 07org.oe.dev * r1b89afb6... 10/packages/hostap/ (hostap-modules_0.3.10.bb hostap-modules_0.4.7.bb): Mar 03 10:38:30 hostap-modules: add 0.3.10 and 0.4.7 Mar 03 10:38:30 - closes #708 Mar 03 10:38:30 - 0.4.7 still doesn't compile against 2.6.15-hh0 as long as Mar 03 10:38:30 kernel_updates.patch is applied. Mar 03 10:39:05 Cockroach-: Make sure its compiling the same version as your "normal" cross compiler Mar 03 10:39:09 <_guillermo> RP: should I report a bug ? Mar 03 10:39:57 _guillermo: I guess so. Not sure there is a good fix for this though :-/ Mar 03 10:40:07 <_guillermo> too bad Mar 03 10:40:22 pH5: what about having a different SRC_URI for hx4700, h2200 and c7x0? Mar 03 10:40:25 03koen 07org.oe.dev * rf07649d5... 10/packages/python/python-pygtk2_2.6.2.bb: python-pygtk2: add rdepend on python-shell Mar 03 10:40:36 RP: Can you clarify a bit, I'm not sure did I understand your suggestion as I should have Mar 03 10:41:02 Cockroach-: Make sure your gcc-cross version matches your gcc-cross-sdk version Mar 03 10:43:44 hmm, gcc-cross-initial-3.4.4-r3 and gcc-cross-sdk-3.3.4-r2 Mar 03 10:44:21 There are too many moving parts for my experience. :P Mar 03 10:44:33 gcc-cross-sdk-3.4.3 might be close enough - I don't think we have a 3.4.4... Mar 03 10:44:53 set PREFERRED_VERSION_gcc-cross-sdk = "3.4.3" in your local.conf Mar 03 10:45:25 Hmm, ok. Mar 03 10:46:21 Is there any way for a new developer to learn these things except by bugging people on #oe?-) Mar 03 10:47:20 Cockroach-: Start some documentation? Mar 03 10:48:30 Cockroach-: I just happen to have faught that particular problem myself recently. I was using gcc-cross 3.4.3 though and pushed my solution back into OE (a gcc-cross-sdk-3.4.3), Whether it works for 3.4.4, I honestly don't know Mar 03 10:49:52 RP: I've had my share of documenting things at work which I don't really have enough information of. :) And since I'm trying to get things to work so I could develop a certain app as a part of my thesis work, so some I will do some kind of documentation as soon as I understand how things work. Mar 03 10:50:04 but for now I'm quite lost. Mar 03 10:50:32 RP: http://dominion.kabel.utwente.nl/koen/pda/files/lshal-c7x0 Mar 03 10:58:03 koen: cool :). Which util was that with? Mar 03 10:58:37 koen: It doesn't seem to actually know much about the hardware other than that the hardware exists though :) (apart from APM and pcmcia) **** ENDING LOGGING AT Fri Mar 03 10:59:57 2006