**** BEGIN LOGGING AT Mon Jan 23 02:59:58 2006 Jan 23 03:00:21 pb_: git pull http://ewi546.ewi.utwente.nl/OE/other_scm/git/org.openembedded.oz354fam083 Jan 23 03:00:55 pb_: man git-checkout Jan 23 03:01:03 pb_: man git-branch could be interesting Jan 23 03:06:24 zecke: I've not tried cogito yet... Jan 23 03:06:57 zecke: We should really start the monotone -> git phrase book (plus some documentation of what git actually does) Jan 23 03:07:30 RP: and we should find documentation for synchronisation Jan 23 03:07:39 and merging Jan 23 03:09:07 http://kernel.mirror.nedlinux.nl/pub/software/scm/git/docs/merge-strategies.txt Jan 23 03:09:47 RP: yes but what about git-merge-index Jan 23 03:09:56 RP: it seems only git-merge-one-file is working there Jan 23 03:10:16 RP: which means... I have some fonflicts from uz354 to master I can not resolve :} Jan 23 03:11:33 hehe just 3000 more revs to replay to crofton's svn server Jan 23 03:14:37 zecke: You read http://kernel.mirror.nedlinux.nl/pub/software/scm/git/docs/git-merge-index.html ? Jan 23 03:21:17 hi Jan 23 03:21:45 bugs.openembedded.org down ? Jan 23 03:27:54 zecke: oh, right Jan 23 03:27:57 okay, I'll try that soonest Jan 23 03:31:46 Greetings to all ! Jan 23 03:31:54 RP: leave pulling from ewi and merging to phil please ;) Jan 23 03:37:57 zecke: will do :) Jan 23 03:39:51 managed to get the epia going up to a point... Jan 23 03:40:03 task-bootstrap and kernel Jan 23 03:40:27 but now i have a problem with "bootstrap-image" Jan 23 03:40:41 * pb_ installing git now Jan 23 03:40:58 | /home/stelios/oe/build/tmp/work/epia-linux/bootstrap-image-1.0-r0/temp/run.do_rootfs.31465: line 206: bbimage: command not found Jan 23 03:40:58 NOTE: Task failed: /home/stelios/oe/build/tmp/work/epia-linux/bootstrap-image-1.0-r0/temp/log.do_rootfs.31465 Jan 23 03:41:13 bbimage is on the path... Jan 23 03:41:29 that is weird Jan 23 03:41:32 if i run it manual i get something like this... Jan 23 03:42:20 zecke: what does "fatal: Not a git repository: '.git' Jan 23 03:42:20 " mean? Jan 23 03:42:27 stelios@Athena:~/oe/build$ bbimage Jan 23 03:42:27 NOTE: Executing mkfs.jffs2 --root=/home/stelios/oe/build/tmp/rootfs --faketime --output=/home/stelios/oe/build/tmp/deploy/images/rootfs-epia-20060123103400.rootfs.jffs2 --pad --little-endian --eraseblock=0x40000 Jan 23 03:42:47 any ideas ? Jan 23 03:42:48 pb_: you suck? You need to clone first Jan 23 03:42:51 Ifaistos: that is a bit weird. touch local.conf and try again. Jan 23 03:42:52 zecke: oh, right Jan 23 03:43:11 pb_: yeah you need to clone first Jan 23 03:43:16 okay Jan 23 03:43:18 pb_: or do git db-init or such Jan 23 03:43:41 so you want me to clone from rpsys, right? Jan 23 03:43:45 pb_: yes Jan 23 03:44:02 okay Jan 23 03:44:09 is the url on the wiki page? Jan 23 03:44:10 pb_: git clone pb@oegit.rpsys.net:/home/oegit my-lovely-git Jan 23 03:44:11 * pb_ goes to look Jan 23 03:44:18 ah, very good Jan 23 03:44:31 okay, that seems to be doing its thing Jan 23 03:45:25 pb_: git checkout (co does not work...) 'REV' Jan 23 03:45:37 pb_: you can use any revnumber or tags or branch names Jan 23 03:45:46 righto Jan 23 03:45:52 this is all very exciting Jan 23 03:45:55 pb_: git checkout oz354fam083 Jan 23 03:46:03 will switch to that branch Jan 23 03:46:09 git branch will list all branches Jan 23 03:46:25 git branch -d master will piss us off Jan 23 03:46:34 excellent Jan 23 03:47:04 pb_: git checkout -f master will undo every change you have in your directory (uncomitted) Jan 23 03:48:34 svk has taken over my home directory Jan 23 03:48:46 cool Jan 23 03:50:27 pb_: fetch, pull, commit are the most important commands Jan 23 03:50:40 righto Jan 23 03:50:51 my clone is still "Packing 30495 objects" at the moment Jan 23 03:50:52 pb_: and merging is still some kind of mess with git. But you can plug any merging algorithm there Jan 23 03:51:05 ah, that doesn't sound too good Jan 23 03:51:45 and I know no-one who audited git... Jan 23 03:51:51 pb: That did the trick :) Jan 23 03:51:56 strange though.... Jan 23 03:52:29 jolly good Jan 23 03:52:47 pb_: On the other hand adding new git mergers is almost trivial Jan 23 03:53:17 zecke: yeah, auditing git sounds like a nightmare. presumably it's all written in eleet inline assembler or something. Jan 23 03:54:05 pb_: probably and it is obscured by #defines like A as #define B #define 1 _ a nightmare Jan 23 03:54:19 yeah, those crazy kernel guys Jan 23 03:54:23 so you write 11SB11 to introduce inline assembly Jan 23 03:54:31 AB even Jan 23 03:54:53 right Jan 23 03:54:53 fear Jan 23 03:56:41 okay I'm off to the lecture Jan 23 03:56:53 pb_: you can pull/merge plenty of objects Jan 23 03:57:06 pb_: http://oegit.rpsys.net:8000/cgi-bin/gitweb.cgi Jan 23 04:03:07 commit diffs are borken - text files are taken as blobs ;( Jan 23 04:03:13 http://oegit.rpsys.net:8000/cgi-bin/gitweb.cgi?p=.git;a=commitdiff;h=61c7ca02f13d22055b799d14c8bab53c8b199ce6 Jan 23 04:07:15 hrw|work: The host might not actually have diff on it... Jan 23 04:07:48 ah. Jan 23 04:09:35 hrw|work: fixed :) Jan 23 04:10:11 RP: thx Jan 23 04:11:16 btw.. did we had soem weird changesets or its tailor fault? http://oegit.rpsys.net:8000/cgi-bin/gitweb.cgi?p=.git;a=commitdiff;h=b806c58824654a44f2437f7b304ce384c433b30e is a bit strange to me Jan 23 04:11:40 messed uclibc and osb-nrcore in one Jan 23 04:13:15 hrw|work: Looks like tailor messed up Jan 23 04:13:29 zecke: You'd better take a look at the above Jan 23 04:29:30 ~lastseen pH5 Jan 23 04:35:42 monotone: 2,3 M | 281,7 k | 1490 | 327 | 102 Jan 23 04:36:56 hardcore.. Jan 23 04:44:01 does the "default" kernel x86 provide MTD block emulation device ? Jan 23 04:47:16 Ifaistos: OE one? Jan 23 04:47:31 yes for x86/epia etc Jan 23 04:47:52 dont know Jan 23 05:31:41 !lastseen pH5 Jan 23 05:31:42 Laibsch: Don't know, don't care Jan 23 05:32:04 ~seen ph5 Jan 23 05:32:07 ph5 was last seen on IRC in channel #oe, 15h 51m 52s ago, saying: 'gremlin[it]: this sounds like it is an issue with bitbake, maybe you still have an old version without the explode_deps code?'. Jan 23 05:34:11 thanks hrw|work Jan 23 05:34:52 hrw|work: Do you happen to know anything about pH5 checking in some code either into a package anthy or nunome? Jan 23 05:34:52 np Jan 23 05:35:12 anthy is in bugtracker Jan 23 05:38:05 it was lacking depends last time when I looked at it Jan 23 05:38:22 40revs to go... Jan 23 05:48:53 just wanted to check the bugtracker. what happened to openembedded.org? Jan 23 05:49:01 dont know Jan 23 05:50:01 hrw|work: You say anthy is in bugtracker. I guess that is on openembedded.org? Anywhere else I can check? What depends were missing? Jan 23 05:50:24 BT is on http://bugs.openembedded.org/ Jan 23 05:50:35 anthy lack all deps in recipe Jan 23 05:51:02 but I dont know - maybe it does not need gtk or qt - plain libc? Jan 23 05:56:12 re zecke Jan 23 05:56:19 wb Jan 23 05:56:49 pb_: did you blow up anything yet? Jan 23 05:57:09 hrw|work: bugs.openembedded.org is down, too. Jan 23 05:58:04 I know Jan 23 05:58:21 RP: is gitweb down? Jan 23 05:58:26 Anybody contacted kergoth, yet? Jan 23 05:59:09 zecke: Works here for me... Jan 23 05:59:20 Laibsch: its treke machine not kergoth iirc Jan 23 05:59:48 pohttp://oegit.rpsys.net:8000/cgi-bin/gitweb.cgi Jan 23 05:59:58 is that the right uri? Jan 23 06:00:32 zecke: s/pohttp/http/ and then yes Jan 23 06:01:44 well ;) Jan 23 06:02:06 i had typed po in my chat line before i pasted the uri Jan 23 06:02:17 zecke: I know ;-) Jan 23 06:02:26 Well, I thought you might Jan 23 06:02:38 It could have also been a genuine mistake... Jan 23 06:03:44 hmm oe.org expires on Jan 21 2007 Jan 23 06:05:32 hrw|work: But kergoth is the domain owner that is why I thought that as previously there was a domain expiry problem or DNS problem. Jan 23 06:05:43 I would guess DNS Jan 23 06:06:02 I remember seeing it would expire in Jan, so he has renewed it Jan 23 06:06:10 but if the domain expired on Jan 21, then it sounds like registry prob. Jan 23 06:06:11 I haven't sen him online since he moved Jan 23 06:06:51 * obergix[work] wonders if using sourceforge, savannah, gna or others wouldn't be an idea ;) Jan 23 06:06:59 sourceforge stinks Jan 23 06:07:13 especially for bugzilla Jan 23 06:07:24 he should move DNS to zoneedit :) Jan 23 06:07:27 you cannot send emails from sf.net Jan 23 06:08:17 zecke, svn working ok? Jan 23 06:09:40 Crofton: yes, may other people use the trac to register accounts?! Jan 23 06:09:49 yes Jan 23 06:10:02 just don't muck up the trac stuff :) Jan 23 06:10:54 they will need to let me know what their id is so I can giv them write access Jan 23 06:11:05 hmmm... seen the secu advisory for track on Debian Jan 23 06:11:58 Crofton: ah okay Jan 23 06:12:17 Crofton: I'm merging one ref after another. 1064 revs are already merged Jan 23 06:12:38 I just looked and it is up to 1076 I think Jan 23 06:12:50 no you are correct Jan 23 06:14:34 Crofton: what is your email address? Jan 23 06:16:16 philip at balister dot org Jan 23 06:18:54 Crofton: RP : it would be nice if we could collect a list of people with write access... Jan 23 06:19:09 morning Jan 23 06:21:39 anyone know why orinoco-modules-0.13e-r4 would not compile? Jan 23 06:21:59 zecke: koen posted a list a while back? Jan 23 06:22:16 LoDown: Against which kernel version? Jan 23 06:22:18 RP: to your repository ;) Jan 23 06:22:32 2.66 Jan 23 06:22:36 *2.6 Jan 23 06:22:38 zecke: Login and ls /home ;-) Jan 23 06:23:04 LoDown: Don't compile 0.13e against 2.6 Jan 23 06:23:17 LoDown: use the orinoco included with 2.6.15 Jan 23 06:23:17 I don't even need it...how do I skip it? Jan 23 06:23:32 LoDown: It should just skip it... Jan 23 06:23:43 LoDown: Did you set KERNEL_VERSION="2.6"? Jan 23 06:23:59 yes Jan 23 06:24:49 LoDown: It doesn Jan 23 06:25:00 LoDown: It doesn't do that here... Jan 23 06:25:43 RP, what should I set my distro as? Jan 23 06:26:00 LoDown: openzaurus-unstable Jan 23 06:26:09 LoDown: You are using .dev? Jan 23 06:26:14 RP, no Jan 23 06:26:34 LoDown: I have no idea about the status of the release branch. Jan 23 06:26:38 i am using familiar-0.8.3, because I have a h5500 Jan 23 06:26:38 LoDown: Which device? Jan 23 06:26:46 RP, thanks, I think that is the problem Jan 23 06:27:17 LoDown: kernel versions for 2.6 on the ipaq are handled differently - not KERNEL_VERSION but something else Jan 23 06:27:28 right...i have that set as well Jan 23 06:27:52 i think it is because of the file familiar-0.8.3..conf is forcing 13e for some reason Jan 23 06:28:01 LoDown: I fixed this for the zaurus models - I don't know how the ipaq kernels handle things Jan 23 06:28:23 RP, you've pointed me in the right direction though, thank you Jan 23 06:28:29 It could be the ipaq kernels are old enough to mostly work with 0.13e Jan 23 06:28:34 is there a url you can browse the metadata for oe? i know there used to be one, but not sure if it survived after the switch over Jan 23 06:29:02 LoDown: Try 0.15rc2 Jan 23 06:29:17 RP: will do Jan 23 06:29:39 prpplague: http://ewi546.ewi.utwente.nl/tmp/viewmtn/branch.psp?branch=org.openembedded.dev Jan 23 06:30:35 prpplague, there is a test subversion repo that is browsable Jan 23 06:30:43 oh yeah and that one :) Jan 23 06:31:42 zecke, how many revs total? Jan 23 06:31:55 thanks Jan 23 06:35:27 4000 roughly Jan 23 06:38:44 zecke: Did you see the possible tailor issue hrw mentioned? Jan 23 06:39:04 [11:16:39] btw.. did we had soem weird changesets or its tailor fault? http://oegit.rpsys.net:8000/cgi-bin/gitweb.cgi?p=.git;a=commitdiff;h=b806c58824654a44f2437f7b304ce384c433b30e is a bit strange to me Jan 23 06:39:05 [11:17:02] messed uclibc and osb-nrcore in one Jan 23 06:41:25 damn I'm firewalled... Jan 23 06:42:43 The link hosting oegit is probably going to be ighly contended for about 30mins due to data transfer :-/ Jan 23 06:43:10 03hrw 07org.oe.oz354fam083 * r089836c0... 10/ (2 files in 2 dirs): Jan 23 06:43:10 openzaurus branding: new package 'openzaurus-version' contain /etc/openzaurus-version Jan 23 06:43:10 - support in gpe-conf/sysinfo is planned in 0.1.29 Jan 23 06:43:10 - support in opie/sysinfo is in CVS, will be backported Jan 23 06:44:05 RP: that is due tailor+monotone... Jan 23 06:44:18 RP: we travel the tree and apply the changes Jan 23 06:44:30 RP: at some point the trees are different Jan 23 06:44:41 RP: so this could be a merge of two tree or such... Jan 23 06:51:57 hrw|work: thx Jan 23 06:57:08 Is it normal during reorg of build system : ERROR: Cannot satisfy the following dependencies for task-bootstrap: fuser alsa-utils-alsactl alsa-utils-alsamixer alsa-conf modutils-depmod ... etc. Jan 23 06:57:11 ? Jan 23 07:07:02 zecke: Ah, so there should be a changeset prior to that with the "correct" information in. I guess that makes sense, even if the output is a bit confusing... Jan 23 07:13:42 RP: we might need a special version of tailor for the actual switch Jan 23 07:28:50 <[lala]> hi Jan 23 07:29:05 <[lala]> hrw|work: ping Jan 23 07:31:47 <[lala]> RP: did you have time to look into the hciattach problem? if u haven't i would file a bug report for the kernel people today :) Jan 23 07:38:57 [lala]: pong Jan 23 07:39:23 <[lala]> hrw|work: do you know why qt/e is compiled w/o threads support? Jan 23 07:39:47 [lala]: thats question to zecke rather Jan 23 07:40:54 <[lala]> zecke: then i ask you :-) Jan 23 07:44:15 [lala]: to be Qtopia compatible Jan 23 07:44:57 [lala]: and Qt2 thread support is limited anyway Jan 23 07:46:18 I'm offline now Jan 23 07:51:54 [lala]: I haven't had a chance, sorry Jan 23 07:56:45 pb_: did you push anything yet? Jan 23 08:04:04 no, I am still grappling with angry customers Jan 23 08:04:17 I didn't get any further than cloning yet Jan 23 08:04:42 okay customers are more important Jan 23 08:06:04 yeah Jan 23 08:06:08 I hope to have them neutralised soon Jan 23 08:09:14 https://bugzilla.mozilla.org/show_bug.cgi?id=297788 <- lol Jan 23 08:09:23 how to win developers... Jan 23 08:12:22 just seen embedded issue of Linux Journal (Nokia 770, etc...)... will have to look for mentions of OE Jan 23 08:12:52 obergix[work]: I'm told I get a mention :) Jan 23 08:14:06 RP: you deserve it IMHO ;) Jan 23 08:15:12 obergix[work]: Just to be clear, the mention was 770 related, not OE... Jan 23 08:15:20 ok Jan 23 08:23:19 <[lala]> RP: no problem, i have submitted the bug to the kernel people (http://bugzilla.kernel.org/show_bug.cgi?id=5944) Jan 23 08:23:54 * chouimat is away: Gone voting Jan 23 08:26:17 [lala]: they're the best people to act on that one :) Jan 23 08:34:28 * chouimat is back. Jan 23 08:43:47 03hrw 07org.oe.oz354fam083 * rd9cb2796... 10/packages/libopie/ (libopie2-1.2.1/openzaurus-branding.patch libopie2_1.2.1.bb): Jan 23 08:43:47 libopie: openzaurus branding (CVS backport) Jan 23 08:43:47 taken from .dev Jan 23 08:43:52 03hrw 07org.oe.dev * r6fa36fe3... 10/ (2 files in 2 dirs): Jan 23 08:43:52 openzaurus branding: new package 'openzaurus-version' contain /etc/openzaurus-version Jan 23 08:43:52 - support in gpe-conf/sysinfo is planned in 0.1.29 Jan 23 08:43:52 - support in opie/sysinfo is in CVS, will be backported Jan 23 08:43:56 03hrw 07org.oe.dev * r5cce883a... 10/packages/libopie/ (libopie2-1.2.1/openzaurus-branding.patch libopie2_1.2.1.bb): libopie: openzaurus branding (CVS backport) Jan 23 09:33:05 hi Jan 23 09:34:13 hi CoreDump|home Jan 23 09:58:48 cu Jan 23 10:16:27 * chouimat is away: lunch Jan 23 10:19:15 hello, there Jan 23 10:22:21 when baking glib-2.0-native-2.6.5 , its still having trouble finding libiconv Jan 23 10:22:24 error: #error GNU libiconv not in use but included iconv.h is from libiconv Jan 23 10:22:40 I've added a: Jan 23 10:22:42 PREFERRED_PROVIDER_virtual/libiconv=glibc Jan 23 10:22:54 but it still fails Jan 23 10:26:14 raduga, you building linux or uclibc? Jan 23 10:28:25 that error was fixed a long time ago for most systems Jan 23 10:30:25 er, bitbake target is opie-image. the bb it fails on is glib-2.0-native-2.6.5 Jan 23 10:33:52 any idea where it was fixed? (when it was fixed, if you say, a long time ago) Jan 23 10:34:10 trying to build OZ/opie for akita Jan 23 10:38:26 it was a gettext and glibc issue Jan 23 10:38:49 but if your building uclibc then its different Jan 23 10:42:50 hm. Jan 23 10:43:12 TARGET_OS is not defined. should I set that to "linux"? Jan 23 10:43:17 or is there a reason to use uclibc? Jan 23 10:43:28 smaller libraries Jan 23 10:43:39 shoudl default to linux tho i belive Jan 23 10:52:17 raduga: set DISTRO="openzaurus-unstable" Jan 23 11:00:15 hiho Jan 23 11:04:23 * chouimat is back. Jan 23 11:05:58 w00t, 1st Akita boot from USB storage Jan 23 11:06:04 hey mickeyl Jan 23 11:08:35 CoreDump|home: cool! Jan 23 11:09:57 RP: thanks. I'm using that already. Jan 23 11:24:58 03coredump 07org.oe.dev * re0cf1c32... 10/packages/altboot/ (7 files in 5 dirs): Jan 23 11:24:58 altboot: Jan 23 11:24:58 - Fixed a bug where the /media/ROM mountpoint wasn't moved into the new rootfs Jan 23 11:24:58 - Added support for booting off USB storage devices Jan 23 11:25:40 CoreDump|home: is your stuff limited to 2.4 or does it also work on the 2.6 kernel? Jan 23 11:26:13 currently altboot is 2.4 only Jan 23 11:27:16 with a little bit of luck, all that is needed for 2.6 is a reworked altboot.cfg tho Jan 23 11:32:06 stupid question: Are SD cards working with kernel 2.6? Jan 23 11:32:31 generally yes Jan 23 11:32:37 great Jan 23 11:32:37 for your particular gadget, I don't know Jan 23 11:33:06 we have 'em working everywhere but collie and poodle Jan 23 11:33:15 excellent! Jan 23 11:35:55 has anyone had success with 1.5gb SD cards? (assuming that any exist) Jan 23 11:36:53 1.5g? Jan 23 11:37:03 gb Jan 23 11:37:09 What an odd value Jan 23 11:37:11 I Jan 23 11:37:23 I've seen 1g, 2g and 4g cards though Jan 23 11:37:27 I've seen 1.5gb CF microdrives Jan 23 11:37:38 (not recently, mind you) Jan 23 11:40:01 anyway, 2g and 4g cards will work, sort of Jan 23 11:41:28 sortof? Jan 23 11:41:49 Linux uses the block size reported by the card Jan 23 11:41:55 I've heard (numerous) reports of not working at all. Jan 23 11:42:06 * raduga nods Jan 23 11:42:07 but most other devices (usb readers, windows etc) assumes a 512 byte sector Jan 23 11:42:25 so you have a compatability issue back and forth Jan 23 11:42:52 speaking of microdrives Jan 23 11:42:56 lol Jan 23 11:42:58 http://cgi.ebay.ca/Apple-iPod-MINI-Hitachi-4GB-Microdrive-Hard-Drive-PARTS_W0QQitemZ5856391678QQcategoryZ48680QQrdZ1QQcmdZViewItem Jan 23 11:43:10 you think the tape is good sign :P Jan 23 11:44:05 emte: if it were duct tape, I'd buy one Jan 23 11:45:10 lol Jan 23 11:47:41 raduga, this might be more for you then Jan 23 11:47:43 http://cgi.ebay.ca/HITACHI-4GB-MICRODRIVE-COMPACT-FLASH-CARD-TYPE-II-NEW_W0QQitemZ7583320565QQcategoryZ50519QQrdZ1QQcmdZViewItem Jan 23 11:47:50 if you build a controller for it Jan 23 11:49:03 has anyone here tired to use bitbake on coLinux Jan 23 11:49:10 pb_: hey ;) Jan 23 11:49:16 ummmm, did anyone mess with task-opie-* in .dev? Looks like the tasks don't RDEPENT the packages in them any longer Jan 23 11:50:41 jatreon, its doubtful it will work Jan 23 11:50:52 there are issues with BSD, OSX, etc Jan 23 11:51:08 didn't kergoth get most of it working with colinux a while back? Jan 23 11:52:06 if he did i would have thought that bitbake/oe would work a bit better on the otehr full platforms before a hybrid ... Jan 23 11:52:12 other* Jan 23 11:52:18 especially BSD Jan 23 11:54:12 emte: well that glibc and want GNU binutils is not kergoth's issue ;) Jan 23 11:55:16 zecke, yeah i know Jan 23 11:55:58 but i would expect that for it to work in winodws that an "enviroment" would have to be used Jan 23 11:56:23 as opposed to an open hybrid system Jan 23 11:57:01 which reminds me( since i dont use windows) is that what cygwin provides? Jan 23 12:02:06 emte: it uses GNU utils and is POSIX compilant :} Jan 23 12:03:26 CoreDump|home: You're probably looking for meta-opie Jan 23 12:03:56 lol Jan 23 12:04:06 zecke, was that a yes to my question then? Jan 23 12:04:27 well, I'm not. bitbake opie-image failed because the tasks don't build all he apps anymore Jan 23 12:04:41 what is the name of the ipaq sleeves that look like a cross? not the compaq/hp type Jan 23 12:05:25 CoreDump|home: Did you update to the latest bitbake? Jan 23 12:05:37 yep, 2 minutes ago Jan 23 12:06:16 emte: if I didn't get the term hybrid system ;) Jan 23 12:06:17 CoreDump|home: opie-image RDEPENDS on the tasks in task-opie so that should work... Jan 23 12:06:52 the tasks are build fine. the RDEPENDS of these tasks however werent for some strange reason Jan 23 12:07:21 * CoreDump|home trashes cache Jan 23 12:08:08 mickey|dinner: poodle should have working SD btw Jan 23 12:08:35 wait, I've mixed up RDEPENDS and DEPENDS Jan 23 12:08:51 task-opie* doesen't depend on anything Jan 23 12:09:09 no wonder tha apps inside task* aren't compiled Jan 23 12:09:09 CoreDump|home: Which is correct after the recent changes to bitbake Jan 23 12:10:00 tell that opie-image please, having to build meta-opie just for a simple opie-image is a joke Jan 23 12:10:09 (IMO) Jan 23 12:10:15 CoreDump|home: I never said that Jan 23 12:10:22 RP: did pb_ pull yet? Jan 23 12:10:24 ok, how is it done now? Jan 23 12:10:31 CoreDump|home: If you build opie-image it will follow all the RDEPENDS Jan 23 12:10:36 RP: I wonder if I should provoke a merge conflict for him... Jan 23 12:10:43 right Jan 23 12:11:10 CoreDump|home: opie-image inherits a variable BUILD_ALL_DEPS which makes it do that Jan 23 12:11:18 huh Jan 23 12:11:18 (all images do) Jan 23 12:11:42 so I need to export BUILD_ALL_DEPS=1? Jan 23 12:12:01 CoreDump|home: no, image_ipk.bbclass already does that for you Jan 23 12:12:29 aha found it Jan 23 12:13:27 zecke: not yet Jan 23 12:13:39 bbl Jan 23 12:13:45 however, my customers are neutralised and my phone system is (mostly) working, so life is good Jan 23 12:13:48 * pb_ going home now Jan 23 12:13:54 I'll pull rp's tree after dinner Jan 23 12:14:15 pb_: take your time Jan 23 12:14:34 zecke: Creating a merge conflict would be evil :) Jan 23 12:15:02 RP: thanks, svn up + rebuilding cache seems to fix opie-image here Jan 23 12:16:40 CoreDump|home: That's a relief. I keep worrying people will find genuine problems with the RDEPENDS code Jan 23 12:17:05 may I ask what the motivation was behind this change? Jan 23 12:17:37 CoreDump|home: Have a look at the old meta-opie and compare it to the new one (and the added task-opie) Jan 23 12:17:51 will do Jan 23 12:18:59 CoreDump|home: Also look for (the lack of) BOOTSTRAP_EXTRA_DEPENDS Jan 23 12:19:30 is there an ipkg command to upgrade a single package (and its dependencies), rather than all packages? Jan 23 12:20:14 trust me I will since I have to rewrite my own bootstrap and -image files when I go to dev then Jan 23 12:22:09 good evening !!! Jan 23 12:22:23 hi gremlin[it] Jan 23 12:22:31 CoreDump|home: Its honestly not difficult. They'll work as is, or you can remove all references to DEPENDS and add BUILD_ALL_DEPS="1" instead Jan 23 12:24:09 If they inherit the image bbclass, you don't even need to add the BUILD_ALL_DEPS line... Jan 23 12:26:39 hi CoreDump|home Jan 23 12:27:02 RP: I'll see =) Jan 23 12:28:30 mhh doing a 'monotone diff' i notice i had modified a couple of files and so maybe they are old respect to database ... how i can force an overwrite from database to working copy ? Jan 23 12:32:48 gremlin[it]: mt revert blah Jan 23 12:34:10 CoreDump|home are u sure ... i mean there isn't 'revert' in the man of monotone :( Jan 23 12:34:31 works here =) Jan 23 12:34:38 very useful command Jan 23 12:35:57 CoreDump|home yes it work ... and 'info monotone' report it ... .... ticket to monotone guys ... update man pages other than info :) :) :) Jan 23 12:42:52 Anyone know what's the endianness of a PXA270? Jan 23 12:46:03 little endian Jan 23 12:46:10 by default Jan 23 12:49:32 evening Jan 23 12:57:08 reenoo_: hi Jan 23 12:57:41 reenoo_: we have a git tree and svn up and running Jan 23 12:57:56 reenoo_: I'm currently svk smerging the to the server Jan 23 12:59:10 mreimer: ipkg upgrade is the closest to what you want, but I don't think it upgrades dependencies Jan 23 12:59:20 thanks pb_ Jan 23 13:01:55 hey zecke Jan 23 13:03:16 zecke: excellent. wished I had the time to help with the testing. instead I'm struggling with non-sensical assignments :/ Jan 23 13:03:30 is there a way to tell ipkg to keep its hands of certain packages, like 'kernel*'? Jan 23 13:03:41 reenoo_: hehe I need to program something as well... Jan 23 13:06:01 zecke: well. the problem with this course is that you're constantly fighting broken software (a piece of crap *shareware* prolog interpreter written in java this time). Jan 23 13:06:15 reenoo_: well you should at least get an account so you may cast your vote... Jan 23 13:06:18 NOTE: package libtool-native-1.5.10-r3: task do_compile: failed Jan 23 13:06:30 reenoo_: hehe Jan 23 13:06:32 | configure: error: source directory already configured; run "make distclean" there first. Jan 23 13:06:57 zecke: good call. I'll install git tomorrow. who do I need to contact for git/svn accounts? Jan 23 13:07:21 reenoo_: http://oe.handhelds.org/cgi-bin/moin.cgi/SvnScmTrial Jan 23 13:07:36 http://oe.handhelds.org/cgi-bin/moin.cgi/GitScmTrial Jan 23 13:08:01 http://oe.handhelds.org/cgi-bin/moin.cgi/ScmVotingProcedures?action=show Jan 23 13:08:08 feel free to fix any errors you see Jan 23 13:08:58 once we have darcs, mercurial up for testing I will mail oe again Jan 23 13:09:43 ~lart t-offline for fscking up dns queries again Jan 23 13:09:43 * ibot executes killall -KILL t-offline for fscking up dns queries again Jan 23 13:10:44 ah, finally Jan 23 13:10:48 * reenoo_ reads Jan 23 13:11:28 mreimer: ipkg flag hold