**** BEGIN LOGGING AT Tue Jan 22 02:59:56 2008 Jan 22 04:31:19 03john_lee 07org.oe.dev * r824c66fc... 10/ (3 files in 3 dirs): Jan 22 04:31:19 task-openmoko-toolchain-host: fix duplicated text. Jan 22 04:31:19 * duplicated text due to brain damaged patch usage. Jan 22 07:06:12 good morning ! Jan 22 09:10:44 morning all Jan 22 09:12:25 RP: morning Jan 22 09:12:34 yo RP Jan 22 09:13:06 03freyther 07org.oe.dev * r72bb00c1... 10/ (1 packages/qt4/qt_packaging.inc): Jan 22 09:13:06 packages/qt4/qt_packaging.inc: Fix the splitting of the translation packages Jan 22 09:13:06 The variable inside the package name did not get expanded, let us expand Jan 22 09:13:06 it ourselves. Jan 22 09:13:16 03freyther 07org.oe.dev * rf3b18a52... 10/ (1 packages/qt4/qt_packaging.inc): packages/qt4/qt_packaging.inc: Package the phrasebooks Jan 22 09:13:23 03freyther 07org.oe.dev * rdd2b8f6d... 10/ (1 packages/qt4/qt_packaging.inc): packages/qt4/qt_packaging.inc: Attempt to package all possible plugins properly Jan 22 09:13:31 03freyther 07org.oe.dev * r60d7b5ff... 10/ (1 packages/qt4/qt_packaging.inc): Jan 22 09:13:31 packages/qt4/qt_packaging.inc: Remove the attempt to create the meta packages this way Jan 22 09:13:31 It didn't work because the RDEPENDS were not available in the package_write task Jan 22 09:13:31 anymore. We need to create a meta task where all plugins are referred to in some Jan 22 09:13:32 other way. Jan 22 09:13:49 03freyther 07org.oe.dev * rd31528c9... 10/ (1 packages/qt4/qt_packaging.inc): Jan 22 09:13:49 packages/qt4/qt_packaging.inc: Fix packaging of the -dbg packages Jan 22 09:13:49 split in packages.bbclass gets called after the work in do_populate_packages_prepend() Jan 22 09:13:49 is done. This means that on the first packaging attempt no .debug files exist. So Jan 22 09:13:49 create one -dbg package for every plugin we find and assume that the file is getting Jan 22 09:13:52 split out. Jan 22 09:17:50 RP: I took a peek at the wifi problem and it looks like task-base does not include ieee80211_crypt_wep.ko ieee80211_crypt_tkip.ko ieee80211_crypt_ccmp.ko while task-base-exetneded does Jan 22 09:19:19 hvontres|home: You need to talk to someone related to angstrom to find out which it uses and why... Jan 22 09:22:47 RP: I sent a note to the mailing list asking abuout ${MACHINE_TASK_PROVIDER} and how to use it correctly. Hopefully it won't turn into another quasi-religious flame fest Jan 22 09:23:39 ~lart glibc for being such a pig Jan 22 09:23:39 * ibot installs a bad bootloader on glibc and turns glibc into a brick for being such a pig Jan 22 09:23:54 time to go to sleep Jan 22 09:41:09 <__law__> any ideas why i get this problems http://pastebin.com/m321dcdb9, if i try to build a 2007.1 stable angstrom image? gcc34 is on host system Jan 22 09:57:32 morning Jan 22 09:58:05 * hrw knows why hvontes has problem Jan 22 09:58:29 hrw : hi ! Jan 22 10:00:21 hrw : Last night tried to build angstrom-2007.1-legacy for x86 and it fails on glibc-intermediate Jan 22 10:00:39 have you seen it again ? Jan 22 10:00:58 steliosk: time passed since I built angstrom last time Jan 22 10:01:14 I have few Poky builds to maintain Jan 22 10:02:20 Did koen switch glibc to a newer version for the -legacy version ? Jan 22 10:02:37 legacy was never done iirc Jan 22 10:04:47 hmm thought that the 2007.1-legacy was the 2007.1 after they switch to the 2008.1 Jan 22 10:04:53 but it looks like its not Jan 22 10:09:06 no Jan 22 10:09:16 2007.1 is only in separate branch Jan 22 10:09:44 yep, i am getting it now Jan 22 10:18:54 Hi everyone. I've created a bb recipe for libtorrent. It builds fine (bitbake libtorrent works), but it says it cannot find package libtorrent when trying to build an image with 'libtorrent' in the IMAGE_INSTALL ... Jan 22 10:19:15 Any clue ? http://rafb.net/p/QD8B4915.html is the bb recipe in question ... Jan 22 10:19:37 ljp: ping Jan 22 10:24:03 tnt_: What machine? Jan 22 10:24:33 tnt_: how is recipe named? Jan 22 10:26:37 Laibsch: x86-32-nocona Jan 22 10:26:57 hrw: packages/libtorrent/libtorrent_0.12.bb Jan 22 10:27:37 tnt_: basically we do not add libraries into rootfs/image but applications. libraries are automatically added as dependencies Jan 22 10:28:36 hrw: Well, that makes senses :) I just didn't create a bb yet for the application :) Jan 22 10:31:52 zecke|away: ping Jan 22 10:32:41 good morning Jan 22 10:33:22 <__law__> hi florian Jan 22 10:40:35 yop Jan 22 11:00:19 i remember that there is a way to list licences that are used on the distribution , but i don't find the trick , any idea ? Jan 22 11:00:45 perharps i should add LICENCE meta in package_ipk.bbclass Jan 22 11:01:22 what for you need it? Jan 22 11:02:26 to show my boss which licence are used in the distro i made Jan 22 11:02:47 XorA: pong Jan 22 11:03:09 we need it before produce our product ( my english is so poor :) Jan 22 11:03:33 Genesis: lot of software is will 'unknown' license Jan 22 11:04:02 SRC_DIST_LOCAL ?= "symlink" Jan 22 11:04:03 INHERIT += "src_distribute_local" Jan 22 11:04:11 ljp: looking at your qtopia-phone stuff now, its not laid out the same way as qtopia-core in the tarballs is it Jan 22 11:04:31 ljp: I was hoping to expand on the qtopia-core recipes already in OE, but it all looks different Jan 22 11:04:50 Genesis: add those two lines into local.conf and rebuild - you will get tmp/deploy/sources/ populated with symlinks to source tarballs which will be organized by license Jan 22 11:05:17 i should rebuild all ? Jan 22 11:05:38 hum i keep the trick , i'll do that latter Jan 22 11:05:40 you can try -cdistsrcall but I do not remember does it work ok Jan 22 11:05:53 i've a buildbot to configure, not easy as i thought Jan 22 11:05:56 Qtopia phone uses a patched version of qtopia core Jan 22 11:06:07 thx for help hrw Jan 22 11:06:28 XorA: as well, a qtopia phone build builds desktop tools (x11) as well Jan 22 11:07:30 ljp: so there is no use trying to harmonise the recipes as its really a different product? Jan 22 11:10:37 yes, unfortuently we keep our own qtopia core fork, so to speak Jan 22 11:12:21 ljp: cool, nice to get confirmation Jan 22 11:13:53 as well, qtopia needs a certain version of qtopia core, which is one reason it comes in the same package Jan 22 11:15:03 ljp: I shall work on getting your recipes into OE then, if they are wrong I'm sure zecke will get enraged enough to fix them up :-) Jan 22 11:17:02 RP: you're involved on bitbake development, right? Jan 22 11:17:15 RP: it looks like 1.8.10 broke GIT support for me. Jan 22 11:17:31 RP: reverting to 1.8.8 solves it Jan 22 11:19:05 * florian imagines RP hiding under his desk ;) Jan 22 11:24:52 I think mickeyl did the extended git stuff Jan 22 11:24:57 RP: I've tested it with git 1.5.3.8 installed Jan 22 11:25:28 XorA: yes, he changed git to support branches and I helped him (but I couldn't test it here and he did the tests) Jan 22 11:25:47 ~blame otavio Jan 22 11:25:48 * ibot blames otavio (and Canada) for all the evil in the world Jan 22 11:25:52 XorA: but he looked to have properly tested it. Dunno if something broke it before or after Jan 22 11:26:25 mickey|zzZZzz: ^ Jan 22 11:27:21 XorA: from what I remeber about the change, it shouldn't change anything when called without branch param Jan 22 11:27:27 XorA: so it shouldn't be it Jan 22 11:47:50 * XorA wonders what goes horribly wrong if we forced all linking to be --as-needed Jan 22 11:51:28 hmm, Wolfgang is getting annoyed with the ARM u-boot custodian Jan 22 11:55:01 I wonder if anybody else is running in eabi breakage with MACHINE=collie? Jan 22 12:02:18 Crofton: which Wolfgang? Jan 22 12:07:03 Wolfgang Denx Jan 22 12:07:47 Harald stirred up some trouble on the u-boot list when he sent in his openmoko patches Jan 22 12:08:12 The ARM maintainer has not been putting in much time with u-boot for a while ... Jan 22 12:09:32 heh Jan 22 12:10:09 I can imagine... the last time I played with u-boot the ARM support looked a little bit hacked. Jan 22 12:10:55 yeah, and patches aren't flowing smoothly in Jan 22 12:11:12 Hopefully, someone with some time volunteers ... Jan 22 12:12:35 I could make a patch for reading SD cards on PXA... if that's still not supported. Jan 22 12:13:22 florian: it is Jan 22 12:13:30 it was in 2006 Jan 22 12:13:59 is anyone here building collie images successfully? Jan 22 12:16:09 no, I have cleaned out tmp, reverted back to a version from Jan 07th and tried to build minimal-image Jan 22 12:16:11 RP, the git fetcher in 1.8.10 has been reported as broken: Jan 22 12:16:13 http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=462077 Jan 22 12:16:14 I'm getting " Error: selected processor does not support `bx r3'" while compiling glibc-intermediate_2.6.1.bb Jan 22 12:16:19 NOTE: Unpacking /export/embedded/sources/gcc-4.1.2.tar.bz2 to /export/embedded/tmp/sonkei-0.0.1/work/armv5te-sonkei-linux/gcc-cross-4.1.2-r13/ Jan 22 12:16:34 Again, arm5vte Jan 22 12:16:41 I was about to send a mail to the list Jan 22 12:16:59 Strange thing is that it used to work fine for me for some time Jan 22 12:17:04 hrw: Doe it work? I remember there was basic support in it for quite a while but it was not able to read from SD but from MMC. Jan 22 12:18:56 Shoragan: I can't really do much without a log from bitbake of the failure :( Jan 22 12:19:09 florian: all I know that u-boot with pdaX patch can be built for Z with sdmmc Jan 22 12:19:17 RP: otavio haves the failure here Jan 22 12:19:21 RP, is the console output enough? Jan 22 12:19:44 Shoragan: yes, that would help Jan 22 12:20:04 ah, /me scrolls up Jan 22 12:20:15 RP, the bugreport is from otavio :) Jan 22 12:20:26 otavio: please report bitbake/oe bugs in OE bugtracker Jan 22 12:20:37 otavio: otherwise they tend to bitrot Jan 22 12:20:54 otavio: Can you pastebin the console logs of a failure please Jan 22 12:21:35 hrw: ah ok, i do ot know that one Jan 22 12:22:54 someone remember which Debian package provides "java" command? Jan 22 12:23:11 hrw, there are several Jan 22 12:23:29 Hopefully, Wolfgang sorts out the ARM maintainer and we can clean out the u-boot patch collection :) Jan 22 12:23:34 Shoragan: I just need one Jan 22 12:24:06 try sun-java6-jre Jan 22 12:24:17 Shoragan: I was asking about working ones Jan 22 12:24:33 oh, that is the only one working for me Jan 22 12:24:40 sun java was broken when last time I checked it Jan 22 12:24:55 amd64 + sid + x11 + sunjava == XCB break Jan 22 12:25:22 :/ Jan 22 12:33:55 hrw, according to http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=440549 you could try http://people.ubuntu.com/~doko/ubuntu/gutsy/ Jan 22 12:34:49 RP: it fails to locate the latest_revision ... it find it as 1... Jan 22 12:35:02 RP: so it tries to do git read-tree 1 Jan 22 12:35:07 RP: and fails Jan 22 12:35:26 Shoragan: I installed few options now and have 'java' command working for cli tool Jan 22 12:36:23 RP: wants me to reproducce the problem again or this was enough? Jan 22 12:37:14 RP: i'm installing .10 again Jan 22 12:37:15 otavio: This doesn't sounds like a bitbake bug Jan 22 12:37:27 RP: .8 works, .10 fails Jan 22 12:37:48 otavio: Which metadata and recipe? Jan 22 12:38:15 RP: i test it with one mine ... Jan 22 12:38:43 otavio: I'd bet it was falling back to SRCREV="1" and you changed something else to make it work ;-) Jan 22 12:38:44 RP: hold... i'll check another one to you be able to reproduce it Jan 22 12:38:55 RP: no Jan 22 12:39:17 RP: I changed nothing to get it to work. Jan 22 12:40:41 otavio: ok, looking at the source change, the way SRCREV worked changed. Did you just change the git.py file? Jan 22 12:41:51 RP: you mean for branch support? Jan 22 12:42:07 RP: it was commited by mickey ... Jan 22 12:42:15 RP: and yes, was just git.py Jan 22 12:42:46 otavio: ok, the problem is not bitbake, I'm 100% sure of that. In this .bb file, what are you setting PV to? Jan 22 12:43:53 http://paste.debian.net/47410 Jan 22 12:44:25 RP: this is the recipe I used Jan 22 12:45:03 otavio: ok, you need to either set SRCREV="somehash" or set SRCREV="${AUTOREV}" Jan 22 12:45:38 RP: may I bother you with the external toolchain stuff again? I tried messing around with dependencies and so on but it will still select it Jan 22 12:45:41 The automatic use of head was disabled in recent versions of OE since we don't want OE to hit the network by default Jan 22 12:45:58 RP: http://paste.debian.net/47411 the log Jan 22 12:45:59 can we have some way of telling the system "do not use extrnal toolchain" Jan 22 12:46:00 Jin^eLD: feel free Jan 22 12:46:08 RP: this is a regression Jan 22 12:46:16 RP: it worked fine and works with .8 Jan 22 12:46:43 RP: look the log. This is reproducable there Jan 22 12:46:43 I think it especially does not make sense when I am building meta-toolchain, so I want to produce a toolchain, and then the external-toolchain gets pulled in Jan 22 12:47:00 which I of course do not have because that will be the output of meta-toolchain Jan 22 12:47:06 RP: bitbake -c fetch -b gtk-theme-torturer_git.bb Jan 22 12:50:11 otavio: Did you try setting SRCREV like I suggested? Jan 22 12:51:18 RP: this makes no sense to change something that used to work ... I can try if you want me to do that but it changes nothing since it does has a bug somewhere Jan 22 12:51:34 RP: if you want, I can put it on the recipe and try Jan 22 12:51:45 otavio: just try it Jan 22 12:51:47 RP: have you tried the one I suggested you? Jan 22 12:52:06 otavio: Yes, it shows the problem you describe, it tried to find this tag "1" Jan 22 12:52:20 RP: is that by chance the source of my problem? http://76.74.156.109/show_bug.cgi?id=3509 Jan 22 12:52:22 otavio: That is entirely expected since OE defaults to SRCREV = "1" Jan 22 12:52:38 otavio: It does this do by default OE does not hit the network Jan 22 12:52:48 s/do/so/ Jan 22 12:53:26 otavio: bitbake 1.8.8 had broken SRCREV. The new behaviour is behaving as intented and it is not a regression Jan 22 12:53:39 * RP agreed we need to make the SRCRE="1" problem more obvious though Jan 22 12:54:34 Jin^eLD: yes, its related Jan 22 12:54:56 RP: any hint for a solution? Jan 22 12:55:08 so far I am deleting external toolchain or using bbmasks but thats really ugly Jan 22 12:55:14 RP :) Jan 22 12:55:20 make that +1 Jan 22 12:55:47 Crofton: I agree its a problem, I just don't know how to make the problem more obvious... Jan 22 12:56:04 RP: doesn't solves Jan 22 12:56:10 if you try to pull r1 print a message asking if this is really what you mean? Jan 22 12:56:37 Jin^eLD: bitbake's preferred version and preferred_provider logic needs fixing basically Jan 22 12:56:50 RP: sounds like a tough task... Jan 22 12:57:20 Jin^eLD: I hate that code in bitbake ;-) Jan 22 12:57:32 heh.. Jan 22 12:58:11 can we at least have some temporary hack, maybe a variable in local.conf that would disable external toolchain or something like that? Jan 22 12:58:33 Jin^eLD: Thats exactly what you're doing with BBMASK Jan 22 12:58:48 otavio: What did you try exactly? Jan 22 12:58:49 well, yes.. but then I need to have a fake external-toolchain.bb in there Jan 22 12:58:56 and probably have some higher revision set so it is taken Jan 22 13:00:15 Jin^eLD: Try changing the RPROVIDES to a PACKAGES_DYNAMIC Jan 22 13:00:25 Jin^eLD: and I never suggested anything that evil ;-) Jan 22 13:00:55 ha, from what I see in the latest pull - someone already did that ;) Jan 22 13:01:13 RP: added autorev... Jan 22 13:01:20 otavio: I took the gtk-theme-torture .bb file, renamed it and tried to build it, it failed. I added a SRCREV="${AUTOREV}" line to it, it worked... Jan 22 13:01:24 RP: I'm hacking this code to see if I solves it. Hold. Jan 22 13:01:42 RP: solved... Jan 22 13:01:49 RP: will give you the diff Jan 22 13:02:04 otavio: the diff for what? Jan 22 13:02:45 Jin^eLD: They just disabled the problematic line Jan 22 13:02:58 leaving the underlying bitbake problem hidden :( Jan 22 13:03:12 RP: the fix Jan 22 13:03:22 RP: so who should be annoyed for a clean fix in bitbake? Jan 22 13:03:38 http://paste.debian.net/47413 Jan 22 13:03:44 RP: ^ Jan 22 13:03:47 Shoragan: ^ Jan 22 13:03:50 Jin^eLD: Any memeber of the bitbake core team I guess ;-) Jan 22 13:04:05 =) Jan 22 13:04:32 RP: git.py was change to avoid using branch names for getting the objects but this one has been forgotten Jan 22 13:06:21 otavio: ok, that change is valid :) Jan 22 13:09:21 03rpurdie * r1009 10/ (4 files in 4 dirs): git.py: Fix git branch tags fetching Jan 22 13:09:31 RP: this solves my problem Jan 22 13:10:13 otavio: and its committed to bitbake now, thanks Jan 22 13:10:46 (The revision "1" fetching is not a regression though, that bit was intentional) Jan 22 13:10:51 RP: yes .. and now, the SRCREV is required Jan 22 13:11:08 RP: it would be better to detect that it wasn't set and go out gracefully Jan 22 13:11:29 RP, for the debian package, should i just use that patch or wait for a new pointrelease from you? Jan 22 13:11:30 otavio: Agreed, this is probably something for sanity.bbclass to detect Jan 22 13:12:01 Shoragan: I'd just patch it for now to be honest Jan 22 13:12:10 ok Jan 22 13:12:18 Shoragan: thanks Jan 22 13:15:02 otavio, i maintain the bitbake package in collab-maint svn Jan 22 13:15:37 if you havn't yet noticed ;) Jan 22 13:16:31 Shoragan: I haven't; but anyway ... I have enough in my plate ;-) Jan 22 13:16:37 ok :) Jan 22 13:16:43 Shoragan: d-i, parted, grub and like ;-) Jan 22 13:17:16 i see Jan 22 13:17:18 Shoragan: I reported it since it would be nice to be spot out as fast as possible Jan 22 13:17:18 :) Jan 22 13:17:29 yes Jan 22 13:17:32 Shoragan: please fix it once you have time... Jan 22 13:17:34 hrw: IYO is 2430 only about ttf fonts or also about other fonts such as qpf? Jan 22 13:20:44 RP: thanks :-) Jan 22 13:29:17 Laibsch: ttf it was but under opie you need to regenerate for qpf and ttf Jan 22 13:34:52 otavio: you could try https://sirius.lasnet.de/~jluebbe/bitbake_1.8.10-2_all.deb Jan 22 13:36:13 Shoragan: getting it Jan 22 13:37:08 Shoragan: perfect! Jan 22 13:37:36 ok, then i'll upload it Jan 22 13:37:42 Shoragan: great. Jan 22 13:39:38 hrw: You want to move update-qtttffontdir from $bindir to $sbindir? Jan 22 13:39:44 done Jan 22 13:39:48 hrw: uhm, the other way round Jan 22 14:02:15 Laibsch: no Jan 22 14:02:23 Laibsch: update-fonts is root util Jan 22 14:04:32 does anyone know if the feed (using deploy as ipkg feed directory) from Paul are already in? Jan 22 14:09:25 I have a directory www and I want to make a bb package that contains everything in that directory including the sub directories.. what is the best way to do this in OE? install does not seem to work since I can't get it to do recursive. Jan 22 14:14:10 Laibsch: for DISTRO=openmoko MACHINE=collie I have arm-oabi Jan 22 14:14:35 hrw: thanks for testing Jan 22 14:14:42 this changes all the time Jan 22 14:15:01 I also had oabi, but then the problem with do_rootfs Jan 22 14:15:14 hrw: Can you build a complete image? Jan 22 14:17:29 minimal-image will be enough? Jan 22 14:26:44 yes Jan 22 14:27:02 I just want to test do_rootfs functionality Jan 22 14:29:21 ok Jan 22 14:35:08 Laibsch: NOTE: Unpacking /home/hrw/devel/sources/dl_dir/binutils-2.18.tar.bz2 to /home/hrw/devel/build/openmoko/tmp/work/arm-oabi-angstrom-linux/binutils-cross-2.18-r1/ Jan 22 14:35:26 good Jan 22 14:35:59 this is for up-to-date .dev? Jan 22 14:38:06 yes Jan 22 14:43:56 !seen psokolovsky Jan 22 14:43:59 psokolovsky (n=psokolov@nat-users.is.ua) was last seen quitting from #oe 9 days, 22 hours, 23 minutes ago stating (Read error: 110 (Connection timed out)). Jan 22 14:44:08 geez Jan 22 15:01:43 morning Jan 22 15:19:46 03rpurdie 07org.oe.dev * r05ddfe24... 10/ (3 files in 3 dirs): linux-rp-2.6.23+2.6.24-rc8: Tweak collie defconfig to reduce the base kernel size Jan 22 15:19:51 03rpurdie 07org.oe.dev * r8819bbf6... 10/ (6 files in 3 dirs): prelink: Update to the last release (20071027) and pull eabi fixes from poky Jan 22 15:40:00 hrw: What program do you use to create the deb package of openembedded-essential? Jan 22 15:42:09 debuild from devscripts Jan 22 15:43:09 OK, thanks Jan 22 16:02:08 03mickeyl 07org.oe.dev * re38c7c56... 10/ (5 files in 4 dirs): Jan 22 16:02:08 python 2.5.1 ship sitecustomize.py that enables some convenient features when Jan 22 16:02:08 working with the interactive interpreter: attribute completion and persistent command history Jan 22 16:34:19 does anyone have a hint on how exactly FEED_DEPLOYDIR_BASE_URI is used? is it combined with FEED_URIS ? Jan 22 16:34:26 hi everybody Jan 22 16:36:51 hi thomas Jan 22 16:38:00 I did read documentation.conf but it is not quite clear to me of those two vars are used in combination or if they are more or less independent from each other and get triggered at different stages of the image generation process Jan 22 16:40:08 thesing: wb Jan 22 16:40:34 did you run into problems compiling images for collie lately? Or did you even compile images? Jan 22 16:45:07 Laibsch: I didn't compile from scratch, but I build an opie-image. Jan 22 16:45:22 when? Jan 22 16:45:30 Can you still build an image? Jan 22 16:48:04 last time was 15.01. Jan 22 16:54:52 Can you try and build a new one? Jan 22 16:55:14 It was around that time that building collie images started failing for me (see mail to list) Jan 22 16:56:05 Laibsch: I can. But not from clean temp (would take at least some days). Do you prefer any image? Jan 22 16:57:00 Laibsch: Do you have the collie service manual available? Jan 22 16:57:20 minimal-image is sufficient Jan 22 16:57:39 have to check about the manual Jan 22 16:57:43 Was that on the CD? Jan 22 16:57:58 I remember I got a bunch of PDFs from hrw Jan 22 16:58:38 It wasn't on the CD. Could you give the pdfs from hrw to me? Jan 22 16:59:22 I'll need to dig them up Jan 22 17:05:39 After searching all my computers I found them Jan 22 17:05:59 (still where they were since hrw had given them to me :-/ ) Jan 22 17:08:18 Laibsch: are they to big for email? Jan 22 17:08:26 a bit Jan 22 17:08:40 I'll have them up on a mailserver in about 5 minutes Jan 22 17:10:00 webserver Jan 22 17:13:51 thesing: http://oz.leggewie.org/zaurus_manuals.tgz Jan 22 17:20:57 Laibsch: Thanks. Got it. Jan 22 17:21:08 Good Jan 22 17:21:29 And apparently one more person monitoring this chat closely ;-) Jan 22 17:21:55 thesing: does it have the information you need? Jan 22 17:22:22 I'm still looking. Jan 22 17:23:57 Laibsch: collie manual is there. Jan 22 17:24:47 Laibsch: I'm working on duplex mode for locomo-spi. The merged mmc-spi driver uses it. I could change the mmc-spi driver but it would be better the other way. Jan 22 17:25:04 what is duplex mode? Jan 22 17:25:08 Sounds like printing Jan 22 17:26:32 Laibsch: sending end receiving at the same time. Jan 22 17:26:45 oh, that would be nice Jan 22 17:26:54 sort of like async mounting? Jan 22 17:28:04 * XorA|gone denies all knowledge of URL linkis Jan 22 17:29:11 no like recording on a sound card while playing music. mmc cards won't be faster. Jan 22 17:29:51 XorA|gone: it was you, then? Jan 22 17:30:08 Laibsch: yes, automatic reaction to URL links :_D Jan 22 17:30:15 hehe Jan 22 17:30:20 "gotta have it" Jan 22 17:30:25 "give it to me" Jan 22 17:30:27 * XorA|gone hasnt actually seen those manuals before Jan 22 17:30:28 ;-) Jan 22 17:30:58 As far as the collie goes, the manual seems to be quite long and descriptive Jan 22 17:31:11 maybe you'll come back for some more Z hacking? Jan 22 17:31:21 Laibsch: the 3100 manual is nice. I with I could read japanese. Jan 22 17:31:36 XorA|gone: Write a sound driver for example ;) Jan 22 17:31:44 thesing: I can help you with general translations Jan 22 17:33:01 Laibsch: I need to play with OE and n810 before doing kernel work :-) Jan 22 17:33:02 There are other people working on 3100. I will concentrate on collie. But when I find a new japanese manual I will tell you ;) Jan 22 17:34:45 looks more like Chinese to me Jan 22 17:35:24 Its all the same for me ;) Jan 22 17:36:11 I believe the file has been corrupted Jan 22 17:36:29 The text has been incorrectly encoded and now it is neither Chinese nor Japanese Jan 22 17:36:38 the pix will have to do in that case Jan 22 17:37:42 slc3100sm.pdf is readable. As the file name suggests, it is the service manual Jan 22 17:38:13 collie manual is best Jan 22 17:39:08 hrw: maybe foreign distributors were supposed to actually use it. Jan 22 17:40:30 thesing: all zaurus manuals we got from one of distributors Jan 22 17:42:17 hrw: Are those files freely redistributable? Jan 22 17:43:04 Unfortunately we didn't get the locomo datasheet. Jan 22 17:45:13 Laibsch: no Jan 22 17:45:41 thesing: that one is NDA only and no one from guys which signed it in the past decided to share it Jan 22 17:46:26 hrw: who are those guys? Jan 22 17:46:43 kergoth and few others - would have to dig in mails Jan 22 17:47:10 oh, hrw Jan 22 17:47:24 I hope you are not mad Jan 22 17:47:28 but? Jan 22 17:47:29 that is nothing to take lightly Jan 22 17:47:53 Laibsch: OpenMoko-minimalist-image-glibc-ipk-P1-Snapshot-20080122-collie.rootfs.jffs2 Jan 22 17:48:01 hm Jan 22 17:48:13 * Laibsch wonders if that is good or bad Jan 22 17:48:16 Laibsch: 22 18:48 < Laibsch> I hope you are not mad" Jan 22 17:48:22 what do you mean by that? Jan 22 17:48:50 well, I shared the files which apparently were not meant to be distributed Jan 22 17:49:45 Laibsch: they cannot be put officially to distribute Jan 22 17:49:53 Laibsch: I already had them but they are on the hds of my broke pc. (sata) The only pc I have available ist only pata. Jan 22 17:50:01 Laibsch: officially I do not even know they exist Jan 22 17:50:08 thesing: Did you get them officially? Jan 22 17:50:18 Well, then I don't either ;-) Jan 22 17:50:24 hrw: Well, then I don't either ;-) Jan 22 17:50:26 Laibsch: no. I got them from hrw. Jan 22 17:50:33 ;) Jan 22 17:50:38 Well, happy sharing ;-) Jan 22 17:51:32 hrw: was the talk about the nda on OZ mailing lists? Jan 22 17:51:40 moment Jan 22 17:52:19 * XorA|gone got those manuals of some random link on IRC :-) Jan 22 17:52:44 XorA|gone: lol Jan 22 17:54:11 I'm afraid that my reading of the NDA indicates that I can't give out Jan 22 17:54:11 copies of the documents. I believe I can answer questions by refering Jan 22 17:54:12 to my notes though. So feel free to ask. Jan 22 17:54:19 thesing: that was mail Jan 22 17:54:30 thesing: I can forward it to you probably Jan 22 17:54:54 thesing: it was discussion between group of zaurus kernel hackers/maintainers Jan 22 17:55:54 hrw: would be nice. Jan 22 17:56:10 hrw: so you read the locomo datasheet? Jan 22 17:58:02 thesing: forwarded Jan 22 17:58:08 thesing: never saw on Jan 22 17:58:10 e Jan 22 18:04:43 hrw: from the guys in the mail, can you tell me who doesn't have the datasheet? Jan 22 18:06:35 thesing: Brian and Chris are the ones who have Jan 22 18:06:43 thesing: so once you know who doesn't have it you can ask them not to show it to you? Jan 22 18:07:20 hvontres|home: to ask them questions and for comments. Jan 22 18:08:31 hrw: do you think they might have a poodle service manual? Jan 22 18:09:36 hvontres|home: none of our zaurus hackers have it Jan 22 18:09:51 hvontres|home: our friendly distributor do not have it too Jan 22 18:10:27 hvontres|home: you can use c7x0 service manual for many things in poodle as c700 was poodle based Jan 22 18:11:24 hrw: where could I not find that one? Jan 22 18:12:47 ops. c7x0 is only parts list ;( Jan 22 18:13:03 hrw: np. Jan 22 18:13:41 service manuals for 5500 (EN) and 3100 (JP) exists. for 6000/750/860/3000 we have parts lists Jan 22 18:14:48 have a nice evening guys Jan 22 18:43:51 morning mickeyl Jan 22 18:44:06 hi hvontres|home Jan 22 18:44:10 how are things? Jan 22 18:48:53 thesing: Did your image build finish? Jan 22 18:49:07 let me look Jan 22 18:49:14 I am again at the problem of allegedly unsatisfiable dependencies Jan 22 18:49:20 mickeyl: not too bad. Staying home to watch the kids becase grandma is sick.. Jan 22 18:49:40 Laibsch: NOTE: package minimal-image-1.0: completed Jan 22 18:50:19 so no problems here. But that was no clean build. Jan 22 18:50:27 Hm, strange Jan 22 18:55:32 a+ Jan 22 19:09:37 RP: are you there? Jan 22 19:10:22 RP: i have problems with the branch support for git. It only works, when I remove the complete git tree. On subsequent updates I get something like "fatal: failed to unpack tree object d298a16f00ed32570c6d039777e80f353f0f2cda' Jan 22 19:13:22 mickeyl: you using svn, RP checked in a fix for something this afternoon Jan 22 19:14:00 oh Jan 22 19:14:04 * mickeyl tries svn update Jan 22 19:15:04 mickeyl: I think it was this http://paste.debian.net/47413 Jan 22 19:17:27 oh Jan 22 19:17:28 interesting Jan 22 19:17:35 no idea 'bout git in general Jan 22 19:17:45 that calls for a new release soon Jan 22 19:44:31 re Jan 22 19:58:25 Is there a reason the autotools_stage_all isn't done automatically ? Jan 22 20:03:11 which package do I need to clean before rebuilding the rootfs, so that changes in FEED_URIS are processed? Jan 22 20:03:35 * Chriva[A] is now away - Reason : Auto-Away (Away from Keyboard) Jan 22 20:06:42 tnt_, not all packages need to stage things, not all packages use autotools, ,etc Jan 22 20:11:08 anyone? Jan 22 20:16:22 RP: Are you there? Jan 22 20:18:36 You might recall that I recently committed apm-wifi-suspendfix and made (hostap|orinoco)-modules RRECOMMENDS it Jan 22 20:20:00 Are those really the right packages? Jan 22 20:20:25 On my spitz, I don't have either of those packages, but I do have kernel-module-(orinoco|hostap) Jan 22 20:20:41 Do I need to fiddle with linux-rp.inc? Jan 22 20:24:23 Laibsch: AFAIK hostap and orinoco modules are in the main kernel. Jan 22 20:38:08 steliosk, I have powered up my ml403 and run the linux demo .... Jan 22 20:41:24 Laibsch: you could try RDEPENDS_modulename in linux-rp or in kernel.bbclass Jan 22 20:41:43 That is what I was considering Jan 22 20:41:54 But I wanted to hear what RP has to say Jan 22 20:42:06 * Laibsch did not know about kernel.bbclass Jan 22 20:42:19 that might even be better than linux-rp.inc Jan 22 20:45:55 Laibsch: in linux-rp.inc there ist "inherit kernel" that says that there is an kernel.bbclass and the file will use it. Jan 22 21:03:29 bye Jan 22 21:31:38 Laibsch: I am now... Jan 22 21:33:58 Laibsch: I'd say there was a need for that RDEPENDS in the kernel class. The only issue is it should only really apply for kernel versions from about 2.6.17... Jan 22 21:54:47 RP: I guess that could be taken care of by some python magic Jan 22 21:55:04 I'll see if I can come up with something Jan 22 22:11:10 * flo_lap thinks we need to fix the monotone server Jan 22 22:20:27 flo_lap: isn't that YOUR server ? Jan 22 22:21:04 ~lart glibc binary package generation for taking so @#$%ing long Jan 22 22:21:04 * ibot pries glibc binary package generation's back open with a screwdriver and flashes a new bootldr to glibc binary package generation for taking so @#$%ing long Jan 22 22:21:30 hvontres|home: no, it gets forwarded to koens machine Jan 22 22:22:07 flo_lap: ohh, the ewi machine that was suppsed to go offline sometime last year????? Jan 22 22:22:32 hvontres|home: no idea, i think koen is the one who owns the tunnel Jan 22 22:23:24 flo_lap: I was thinking that since the machine is doing much better with the new kernel now that we could get mtn back to ltg Jan 22 22:23:38 I always thought the machine was capable of handling it Jan 22 22:24:03 It was just I/O that was wasting 95-99% of CPU cycles in wait state Jan 22 22:24:44 hvontres|home: ewi546.ewi.utwente.nl yess... that one seems to be down Jan 22 22:26:23 Laibsch: That would be worth a try, the only really bad thing with mtn is that it eats up memory Jan 22 22:28:35 I don't think 1G is too tight Jan 22 23:16:06 ~seen zecke Jan 22 23:16:10 zecke was last seen on IRC in channel #oe, 1d 7h 12m 38s ago, saying: 'rschuster: see the topic'. Jan 22 23:17:01 has anybody tried the qtopia-core-gui-image yet? Jan 22 23:30:45 hvontres|home: I have not, but zecke told me there are no programs for it and it is rather for the gumstix only Jan 23 00:05:39 Laibsch: ahh...I see. That would explain the lack of usefull stuff in the qemu-build :) Jan 23 01:05:15 hi, how do I update the Packages* files in deploy/ipk ? Jan 23 01:07:03 redguy: "bitbake package-index" Jan 23 01:10:54 Laibsch: thanks Jan 23 01:11:32 Laibsch: I googled for an hour and I couldn't find it. It it anywhere on the wiki? Jan 23 01:12:14 RP: How about http:///oz.leggewie.org/wip/kernel.bbclass.patch Jan 23 01:12:21 redguy: Not sure Jan 23 01:12:37 Some see this as out of scope for OE Jan 23 01:13:03 And even official stuff is not always well documente although I find the user-manual to be excellent Jan 23 01:14:23 RP: RFC necessary? **** ENDING LOGGING AT Wed Jan 23 02:59:57 2008