**** BEGIN LOGGING AT Thu Feb 09 10:59:56 2006 Feb 09 11:00:24 <[lala]> yep Feb 09 11:00:49 <[lala]> RP: btw no one showed any reaction to my hciattach bug report yet :-( Feb 09 11:01:09 <[lala]> that one i reported on kernel.org bugzilla Feb 09 11:01:19 [lala]: Actually, I did see akpm mention it as an issue that needed resolving Feb 09 11:01:39 [lala]: Its just the bluetooth devs themselves don't seem to have noticed/care :-( Feb 09 11:12:24 can I get other recipe PV somehow? Feb 09 11:17:00 ~lart zap_root_password Feb 09 11:17:01 * ibot explains, ever so gently, that if zap_root_password doesn't give the channel more information, they can't help Feb 09 11:23:53 hmmmm Feb 09 11:24:03 is busybox able to not use udhcp? Feb 09 11:25:15 Cobelius: busybox 0wnz udhcp Feb 09 11:25:29 that is what i mean Feb 09 11:26:10 and ifup/ifdown are also implemted by busybox Feb 09 11:26:20 I don't really understand what you mean by "not use". Udhcp is part of busybox, but you aren't obliged to invoke it. Feb 09 11:26:48 Cobelius: You need to reconfigure busybox not to use udhcp Feb 09 11:26:48 but it's used when i use ifup to start an interface Feb 09 11:31:12 Cobelius: Go to /home/stelios/oe/x86/org.openembedded.dev/packages/orig-busybox/busybox-1.01/ Feb 09 11:31:38 Cobelius: go the arch that you are using and change that defconfig Feb 09 11:31:59 Cobelius: so that busybox does not include ifup and udhcp Feb 09 11:32:24 and how can i say that dhcp3-client should be used instead of udhcp? Feb 09 11:32:57 Cobelius: put the ifupdown package Feb 09 11:34:53 ahh thats seems to make sense, thanks a lot! Feb 09 11:42:06 RP: what do you think about making g_ether autoload on c7x0/tosa/poodle? Feb 09 11:55:02 hrw|work: it should autoload a configurable module, not g-ether Feb 09 11:55:25 hrw|work: probably one set with gpe-conf or opie equiv Feb 09 11:55:53 XorA: so it will stay as it is Feb 09 11:57:53 hrw|work: it "should" but without someone putting in the extra scripting/gpe-conf work, it might as well always load g-ether as per your suggestion Feb 09 11:58:52 on 2.4-crapix we configured images to go into usbnet mode by default iirc Feb 09 11:59:06 atleast on 2.4.18 (non usbhost devices) Feb 09 11:59:45 * XorA wishes pxa had enough USB stuff to do storage and usbnet simultaneously Feb 09 12:04:02 dreams... Feb 09 12:04:37 although *doze doesnt like g-file-storage, my mates XP machine got very upset Feb 09 12:07:11 someone know how to get PV of other package in recipe? kind of: kernel_pv = get_package_pv("virtual/kernel") Feb 09 12:07:39 looks like you found a use case for having PN in overrides Feb 09 12:08:08 koen: its not that Feb 09 12:08:20 koen: its for openzaurus checkversion stuff Feb 09 12:09:09 koen: I want kernels like 2.6.14-rc1-3.5.4-r5 but the only check which I can get is 2.6.14-rc1-3.5.4 as I do not know how to get 'r5' which is kernel PR Feb 09 12:30:25 hrw|work: I would much prefer the loading of g_ether was handled by some gui app Feb 09 12:30:53 hrw|work: I've leave it as it is for now as I think something might get done about that soon Feb 09 12:32:40 ok Feb 09 12:34:39 mickeyl: are you really sleeping? Feb 09 12:36:31 I got the X/Qt version of wellenreiter to compile Feb 09 12:37:08 Do you want the modified source code? Feb 09 12:40:38 03rwhitby 07org.oe.dev * r956c0a45... 10/packages/linux/ (27 files in 2 dirs): ixp4xx-kernel: Added patchset and metadata for 2.6.16-rc2 kernel. Feb 09 12:55:46 RP: borzoi/spitz does not have (.oz354fam083) uDrive partitions in fstab - is there a reason for it? Feb 09 13:03:38 hrw|work: It was fixes in .dev after the branch happened and I guess nobody has updated it Feb 09 13:04:01 hrw|work: note that /dev/hda3 is vfat, not ext3 as those files say - needs fixing in .dev as well Feb 09 13:05:04 ok. Feb 09 13:05:44 hda1, hda2 are ext3 by default? Feb 09 13:06:14 hrw|work: yes Feb 09 13:06:19 ok Feb 09 13:11:48 hmm.. Feb 09 13:11:54 25 /dev/hda1 /media/hdd ext3 noauto,defaults,noatime 0 0 Feb 09 13:11:57 27 /dev/hda3 /media/hdd vfat defaults 0 0 Feb 09 13:12:34 how this should be on borzoi finally? Feb 09 13:15:42 afternoon all Feb 09 13:16:44 hrw|work: This is tricky as its different 2.4 to 2.6 Feb 09 13:17:02 RP: I want 2.4 infos Feb 09 13:17:32 hrw|work: /dev/hda3 needs to go somewhere else then Feb 09 13:20:20 the question is.. how this was on sharp? Feb 09 13:24:43 /dev/hdc1 9.5M 1.0M 8.0M 11% /hdd1 Feb 09 13:24:43 /dev/hdc2 9.5M 1.1M 7.9M 12% /hdd2 Feb 09 13:24:44 /dev/hdc3 3.8G 1.5G 2.3G 39% /hdd3 Feb 09 13:24:52 weird.. Feb 09 13:25:06 hrw|work: mount it on /media/hdd1 Feb 09 13:25:35 2.6 uses them as /, /home and /media/hdd Feb 09 13:26:29 I was thinking about /media/hdd3 Feb 09 13:27:06 Fine with me. You know what I think about 2.4 though ;-) Feb 09 13:27:39 You'd better make sure 2.4 uses /dev/hdc rather than /dev/hda actually! Feb 09 13:27:56 ~lart Sharp for broken kernels Feb 09 13:27:58 * ibot gives Sharp a good seeing to for broken kernels Feb 09 13:29:08 so /dev/hda is cf, /dev/hdc is uDrive always? or only if cf is in slot during boot? Feb 09 13:31:54 bbl lunch Feb 09 13:42:21 hrw|work: Good question - I can't remember :-/ Feb 09 14:02:57 heh ;) Feb 09 14:41:49 when i want to make the package i get: http://pastebin.de/4305 Feb 09 14:42:27 th .bb is http://pastebin.de/4306 Feb 09 14:43:08 Cobelius: you need a FILES_${PN} = "/opt/" or something similar Feb 09 14:44:10 Cobelius: use 'install -d' instead of mkdir and 'install' instead of 'cp' Feb 09 14:47:17 koen: what you suggest to edit svg? Feb 09 14:47:36 hrw|work: inkscape or scribus Feb 09 14:47:45 although scribus is more for DTP Feb 09 14:49:20 I'll try sodipodi Feb 09 14:52:00 is that still alive? Feb 09 14:52:12 I thought after the fork sodipodi died Feb 09 14:52:19 (fork = inkscape) Feb 09 14:52:41 ok Feb 09 14:53:58 koen: gfx for gpe-bootsplash has those texts as text or as curves only? Feb 09 14:54:07 curves only Feb 09 14:54:15 or 'paths' iirc Feb 09 14:54:31 ;( Feb 09 14:57:23 I'm still not used to inkscape jargon Feb 09 14:57:30 * koen a coreldraw person Feb 09 15:00:53 * XorA hasnt seen coreldraw for years Feb 09 15:06:49 http://ewi546.ewi.utwente.nl/tmp/hrw/t/openzaurus/ - what do you think? Feb 09 15:07:03 * hrw|work never used vector graphic programs Feb 09 15:12:06 looks ok Feb 09 15:12:19 booting c760 to check how it look on device Feb 09 15:12:42 its black ;( Feb 09 15:14:10 koen: I found problem with my test images (gpe-dm) - S98configure was absent Feb 09 15:14:25 note: do not tweak IMAGE_ROOTFS Feb 09 15:14:39 thank koen it works now Feb 09 15:33:25 koen: build-oz.sh script is nice thing - will add it to contrib Feb 09 15:33:36 hi Feb 09 15:34:04 hi CoreDump|home Feb 09 15:34:47 CoreDump|home: looks like users are thinking about getting poodle for OZ team Feb 09 15:35:08 I saw that on oesf.org Feb 09 15:40:59 opie-image-3.5.4-test5-initrd.bin - one name which tell all Feb 09 15:44:01 If one of us had one, they'd probably end up with a 2.6 kernel and a lot better quality images... Feb 09 15:45:39 * CoreDump|home agrees Feb 09 15:45:52 atleast developer tested images Feb 09 15:45:58 One has to debug the damn thing into a known working state Feb 09 15:46:37 ~poodle Feb 09 15:46:38 hmm... poodle is sharp sl-b500/5600, or a dog Feb 09 15:46:41 the finer details (like working dests in ipkg.conf which were missing in 3.5.1 and 3.5.2 w/o anyone noticing) can be added later Feb 09 15:46:58 b500? it had another name? Feb 09 15:47:05 JustinP: japanese name Feb 09 15:47:10 ah Feb 09 15:47:21 I didn't realize they even released the vertical zaurii in japan... Feb 09 15:47:37 a300, collie, b500 (5600), c700, c750, c760, c860 Feb 09 15:48:19 6000 should be d900 instead ;) Feb 09 15:48:39 * JustinP shakes a fist at the sun....which is in his eyes Feb 09 15:48:49 CoreDump|home: ok, change pushed Feb 09 15:49:05 CoreDump|home: entrance will be broken until you get your fix in Feb 09 15:49:08 JustinP: thanks a lot Feb 09 15:49:32 CoreDump|home: heh...don't know if you'll thank me when you see what a little change this was ;-) Feb 09 15:49:32 JustinP: well, it was never unb0rked for me since the damn /bin/sh thing =( Feb 09 15:49:38 lol Feb 09 15:49:51 it really really should be using bash for everything... Feb 09 15:49:59 are you sure you didn't get an old CVS build or something? Feb 09 15:50:10 JustinP: image from yesterday had the same problem. Feb 09 15:50:12 perhaps you had the CVS build in your "feed"? Feb 09 15:50:23 I built yesterday as well and mine's fine Feb 09 15:50:28 hmm Feb 09 15:50:35 I'll check that later Feb 09 15:50:35 what version of entrance is installed? Feb 09 15:50:39 ok Feb 09 15:50:55 you could also just ls tmp/deploy/ipk/entrance* Feb 09 15:51:19 +mhentges@oe-akita-head:..edded/tmp.dev/deploy/ipk >lgrep entr Feb 09 15:51:19 entrance_20060128-r10_armv5te.ipk Feb 09 15:51:19 libesmart-text-entry0_20060128-r2_armv5te.ipk Feb 09 15:51:23 03justinp 07org.oe.dev * rec946905... 10/packages/e17/entrance_0.9.0.005.bb: entrance: Move startup script to new location so that update-alternatives can be used Feb 09 15:51:29 yep Feb 09 15:51:34 that's the one from .dev Feb 09 15:51:45 I switched e over to the numbered revisions Feb 09 15:51:52 see preferred-e-versions Feb 09 15:52:07 is this an old tmpdir which had the old ones? Feb 09 15:52:19 oops Feb 09 15:52:29 because the numbered ones *should* be building Feb 09 15:52:34 aha, we finally figure it out ;-) Feb 09 15:52:51 sorry, with e it's been back and forth a lot.... Feb 09 15:53:00 hopefully we can stick to numbered revs now Feb 09 15:53:06 heh Feb 09 15:53:17 wait till raster rewrites it again ;) Feb 09 15:53:25 will we able to release e17 image with oz354/fam083? Feb 09 15:53:28 * koen has been using e since 0.13 Feb 09 15:53:38 I'll just wait for the next "release" Feb 09 15:53:48 hrw|work: we should be able to, yes :-) Feb 09 15:53:58 they both work for me right now Feb 09 15:54:19 e-image from .dev is nice w/ gpe apps installed Feb 09 15:54:23 I just need to add my handy gpe => eap script into the e-image do_rootfs and we'll even have progs installed in the menu in e-image ^_^ Feb 09 15:54:25 2 weeks to fosdem.. Feb 09 15:54:57 and for all this to work we need to copy my revs from dev to oz354fam083 Feb 09 15:55:01 Time is just flying by :-/ Feb 09 15:55:32 in saturday we have local PDA meeting here - mostly windowsmobile users and some palmos ones + /me with linux (or maybe mcgregor with his 3870) Feb 09 15:55:42 I'll go find the rev #'s and add them to the stable bug Feb 09 15:55:51 this time its time for gpe+some stuff Feb 09 15:56:10 JustinP: cool Feb 09 15:56:43 cp e17.dev efl.dev .oz Feb 09 15:57:22 http://bugs.openembedded.org/attachment.cgi?id=502 - why tosa has 'noauto' sd/mmc? Feb 09 15:57:22 JustinP: from where do you "include" the pref version file? Feb 09 15:57:51 hrw|work: should be mounted by udev anyways Feb 09 15:57:59 * CoreDump|home shrugs Feb 09 15:58:04 CoreDump|home: under 2.4 it is noauto too Feb 09 15:58:31 hrw|work: The sharp driver takes care of mounting, doesn't it? Feb 09 15:58:35 in 2.4 there's the /etc/sdcontrol (or somethinglike that) script which does the mounting Feb 09 15:58:42 ah. Feb 09 15:58:50 misterious sharp things... Feb 09 15:58:54 =) Feb 09 15:59:15 let me check status of my .oz tree... Feb 09 15:59:20 11 files changed, 22 insertions(+), 22 deletions(-) Feb 09 15:59:44 CoreDump|home: in the familiar-unstable.conf and openzaurus-unstable.conf Feb 09 16:00:21 ah there it is Feb 09 16:00:44 hrw|work: that's a lot of changes... Feb 09 16:01:00 so should I create patches for all of my revs? Feb 09 16:04:10 hmmm....that would be a super pain... Feb 09 16:04:30 it would probably be easier just to drop the efl/e17 dirs in stable and copy over the ones from dev... Feb 09 16:04:51 cp e17.dev efl.dev .oz =) Feb 09 16:05:04 yes Feb 09 16:11:01 JustinP: you were right of course...while your "versioned" ipk's were built, they were always moved into morgue heh Feb 09 16:11:31 yep Feb 09 16:11:34 :-/ Feb 09 16:11:59 we should have figured that out quicker.... Feb 09 16:12:03 or I should have Feb 09 16:12:24 aargh Feb 09 16:12:26 ~lart me Feb 09 16:12:26 * ibot brandishes Excalibur! "With this sword, I vanquish thee, koen!" and lops off koen's head Feb 09 16:12:38 'why can't I wipe /tmp?' Feb 09 16:12:45 * koen was building as root Feb 09 16:12:55 koen: ;) Feb 09 16:14:23 * koen wonders how many arm binaries are running rampant on his system Feb 09 16:14:27 7 revs pushed Feb 09 16:14:54 koen: check /usr/lib/ipkg on your host heh Feb 09 16:15:13 :) Feb 09 16:15:15 I know ipkg tries to write there during ever do_rootfs Feb 09 16:16:14 koen: I used to do builds as root a lot more often than I should and never came to any harm... Feb 09 16:16:28 I don't now mind ;-) Feb 09 16:18:37 * JustinP wishes for clouds Feb 09 16:18:38 ahh crap Feb 09 16:18:47 * CoreDump|home just trashed his tmp Feb 09 16:19:16 I accidentally did that a few days ago Feb 09 16:19:21 =( Feb 09 16:19:40 I actualy fairly often start form a new tmp so that I can make sure that efl builds from a new stage Feb 09 16:19:53 5 files changed, 15 insertions(+), 6 deletions(-) Feb 09 16:20:04 I can live with those changes Feb 09 16:20:06 well, last time it took ~7hrs to rebuild e-image-core,gpe-image and opie-image Feb 09 16:20:19 yep....:-( Feb 09 16:21:55 well, only deploy/ipk is gone Feb 09 16:22:10 so I can still work on single packages Feb 09 16:23:11 just start your rebuild when you go to bed Feb 09 16:23:22 yep Feb 09 16:24:12 mv ${D}/etc/init.d/entrance ${D}/etc/X11/login-managers/entrance lol Feb 09 16:24:20 should have thought of that heh Feb 09 16:24:32 yep Feb 09 16:24:33 ;-) Feb 09 16:25:37 compiling giac + three monotone instances = slooooooowness Feb 09 16:26:59 03hrw 07org.oe.oz354fam083 * rf50bdc56... 10/packages/libusb/libusb_0.1.10a.bb: Feb 09 16:26:59 libusb 0.1.10a: split libusbpp into separate package Feb 09 16:26:59 - now libusb package does not need libstdc++ Feb 09 16:27:03 03hrw 07org.oe.oz354fam083 * r001519e6... 10/.mt-attrs: Feb 09 16:27:04 sharp-binary-only: removed the line from .mt-attrs which referenced the nonexistent file Feb 09 16:27:06 taken from .dev Feb 09 16:27:08 03hrw 07org.oe.oz354fam083 * re58c4067... 10/.mt-attrs: disapproval of revision '001519e60a14a99181453c8cfe62f1da833e8535' Feb 09 16:27:12 03hrw 07org.oe.oz354fam083 * r19cc4c9b... 10/conf/distro/openzaurus.conf: openzaurus: replace date-time in image-name with distro-version Feb 09 16:27:18 03hrw 07org.oe.oz354fam083 * r003ce1df... 10/conf/machine/ (zaurus-clamshell-2.4.conf zaurus-clamshell-2.6.conf): Feb 09 16:27:19 zaurus-clamshell: remove stuff which is already in handheld-common.conf Feb 09 16:27:21 - zaurus-clamshell.conf include handheld-common.conf so we do not need Feb 09 16:27:23 to add same stuff again in 2.4/2.6 configs Feb 09 16:27:25 03hrw 07org.oe.oz354fam083 * rd9c787c4... 10/packages/linux/ (linux-openzaurus.inc linux-openzaurus_2.6.14-rc1.bb): Feb 09 16:27:30 linux-openzaurus: removed PR from kernel version Feb 09 16:27:32 - OpenZaurus have 'checkversion' script in initscripts which warn user Feb 09 16:27:34 if machine is booted with other kernel then the one provided by us. Feb 09 16:27:36 koen: http://ewi546.ewi.utwente.nl/tmp/hrw/t/openzaurus-3.5.4-test5.conf - FEED_BASE_URI could be then used in gpe/opie/etc-image for easier feeds adding Feb 09 16:27:36 As I do not know how to check virtual/kernel PR in initscripts I drop Feb 09 16:27:38 PR from linux-openzaurus Feb 09 16:27:55 koen: what do you think? Feb 09 16:28:23 hrw|work: sounds ok Feb 09 16:29:01 RP: super awesome, thank you! Feb 09 16:33:25 (i see that it does build if i patch the source to move that function up a little further in the source code) Feb 09 16:33:41 also, and this is important, it builds without modifying the defconfig Feb 09 16:34:01 so that tells me the object code being output is back on track with that of a gcc cross being run on ia32 Feb 09 16:34:04 :) Feb 09 16:36:10 I have to find some time to look at splitfeeds code and generate one probably Feb 09 16:36:12 hrw|work: did you add a RREPLACES = "libusb" in libusbpp ? Feb 09 16:36:21 koen: it does not replace it Feb 09 16:36:36 koen: its rather c++ interface Feb 09 16:36:57 it does ship a file that was present in libusb -> ipkg breakage Feb 09 16:37:32 ah.. Feb 09 16:38:10 both should have RREPLACE for older libusb then.. Feb 09 16:39:02 afaik only libusbpp, but you'd have to ask pigi or pb__ for the correct answer Feb 09 16:39:16 ok Feb 09 16:41:21 koen: armv5 devices supported by familiar use tune-strongarm or tune-xscale? Feb 09 16:42:04 in the branch strongarm since we don't use thumb stuff Feb 09 16:42:15 ok Feb 09 16:42:28 I will do same for oz - simplify feeds Feb 09 16:42:32 yeah Feb 09 16:42:35 and build time :) Feb 09 16:43:40 * JustinP build e-wm-native Feb 09 16:43:45 this should be fun Feb 09 16:43:54 hrw|work: mplayer-atty is armv5 only Feb 09 16:44:15 users would kill me - I will provide machine feeds + one big for rest Feb 09 16:44:29 hrw|work: familiar has a trick for that Feb 09 16:44:48 hrw|work: http://familiar.handhelds.org/releases/v0.8.3/feed/split-world.sh Feb 09 16:44:48 koen: another feed splitter? Feb 09 16:44:57 no, grep -v Source :) Feb 09 16:45:14 saves about 50% -> ipkg twice as fast Feb 09 16:45:57 I know Feb 09 16:47:48 :-( Feb 09 16:47:53 e-native failed Feb 09 16:48:25 koen: first they will have to remove lot of feeds which are in images ;) Feb 09 16:50:28 morning Feb 09 16:50:29 * kergoth yawns Feb 09 16:50:34 hey kergoth Feb 09 16:50:52 03koen 07org.oe.dev * r37a520a3... 10/ (2 files in 2 dirs): Feb 09 16:50:52 ep93xx: updates Feb 09 16:50:52 * put some kernel modules in the rootfs Feb 09 16:50:52 * put kernel in deploydir Feb 09 16:51:31 hi kergoth Feb 09 16:51:59 hi kergoth Feb 09 16:52:26 JustinP: did you read the followup mail on oz-dev regarding login-managers? Feb 09 16:52:52 morning kergoth Feb 09 16:53:47 oh, i got paid. i can take care of the domain/dns situations now Feb 09 16:53:57 kergoth: great Feb 09 16:54:05 kergoth: excellent :) Feb 09 16:54:29 will see about switching them from easydns to godaddy. cheaper that way. heh Feb 09 16:55:37 CoreDump|home: ok, just saw it Feb 09 16:56:19 sounds like an even better plan. even more flexible and the user could choose which one he likes best vir a config file Feb 09 16:56:59 as long as not having a dm/lm is an option, I'd be very happy to see that in OE Feb 09 16:57:04 CoreDump|home: I didn't see anything about something like that... Feb 09 16:58:06 JustinP: what I was trying to do is launching different LM's via update-alternative selection Feb 09 16:58:21 RP proposes using a script which chooses the LM to use Feb 09 16:58:26 CoreDump|home: I realize that...and it made sense... Feb 09 16:58:33 which is really the same thing Feb 09 16:58:40 actually no Feb 09 16:59:02 if you're saying that you use a script with a *conf file* that's something different.... Feb 09 16:59:15 I'm not suggesting update-alternative is used but just a script Feb 09 16:59:21 yep Feb 09 16:59:28 but then a default conf file has to be included for the wm's with that under update-alternatives, no? Feb 09 16:59:38 well I guess not Feb 09 16:59:54 I suppose it could just choose the first option in /etc/X11/lms or whatever.... Feb 09 16:59:55 Having that script read a config file is the logical extension/conclusion for the idea Feb 09 16:59:56 kergoth: rock Feb 09 16:59:56 that would get overwritten on updates etc Feb 09 17:00:43 no config = choose a sane default (like entrance over gpe-dm) Feb 09 17:00:44 CoreDump|home: go for it Feb 09 17:01:03 entrance isn't a sane option over gpe-dm for non-zaurii Feb 09 17:01:05 kergoth: if needed, I can afford to pay for 1 or 2 domains Feb 09 17:01:13 CoreDump|home: Instead of a config file, just use a symlink to set the user preference Feb 09 17:01:43 RP: then the user would have to mess with symlinks again =) Feb 09 17:01:56 symlinks! Feb 09 17:02:04 how sharp loves them..... Feb 09 17:02:19 CoreDump|home: config file, symlink, same difference. Most users won't have two dm's installed anyway I'd hope Feb 09 17:02:40 good point Feb 09 17:03:23 A symlink just makes it look like another window manager to look for and its given the highest preference - makes the script simple Feb 09 17:03:53 s/window/display/ Feb 09 17:03:53 RP meant: A symlink just makes it look like another display manager to look for and its given the highest preference - makes the script simple Feb 09 17:04:15 I wish we could turn that off :) Feb 09 17:04:24 RP: avoid last / Feb 09 17:04:35 RP: we can, just ask TImRiker Feb 09 17:04:36 s/last/that one which I avoid Feb 09 17:04:41 RP: I have already adjusted gpe-dm and (w/ JustinP) entrance to not install their init-scripts into init.d. Makes thing easy enough heh Feb 09 17:04:45 hrw|work: Will do in future... Feb 09 17:04:58 RP: then he will popup and do something like: ibot chanset sed off Feb 09 17:07:00 bye guys Feb 09 17:10:04 CoreDump|home: depending on what those init scripts actually do, they might not be needed... Feb 09 17:11:36 cu Feb 09 17:12:34 RP: true Feb 09 17:15:46 * JustinP gets off the bus Feb 09 17:29:45 the latest cvs of gaim doesn't build for me, the "desktop-name-cvs.patch" doesn't apply and also configure tries to compile a test program Feb 09 17:29:52 RP: i ordered my Sharp serial cable Feb 09 17:30:05 after hacking around both of those it compiles though, and I'll test it in a sec Feb 09 17:30:13 RP: $49.99usd incl shipping and handling charges :) Feb 09 17:44:23 interesting... http://www.elecdesign.com/Articles/Index.cfm?AD=1&ArticleID=9050 Feb 09 17:50:45 re Feb 09 18:00:50 shadows: ok, hopefully once it arrives we can find the problem Feb 09 18:00:55 :) Feb 09 18:01:06 yes, because your image behaves differently, but it still was hanging my Z Feb 09 18:02:02 and in fact, this gcc 4.0.2 cross initial compiled build on my amd64 host machine, it is the bootstrap image, and i have just tested it. the zaurus is hung but the charge light will toggle on and off when i plug and unplug the AC adapter Feb 09 18:02:08 this is different than previous behavior Feb 09 18:02:31 i am convinced that gcc3 as cross initial is not appropriate for amd64 build hosts Feb 09 18:03:22 i'm looking into how to build a JTAG cable for the SerialIO port Feb 09 18:05:43 shadows: I can show it for you Feb 09 18:07:15 oh? Feb 09 18:07:28 jtag for 5500 Feb 09 18:07:35 ah. Feb 09 18:08:00 i'm trying to think, can i make a JTAG cable out of a sharp Serial cable Feb 09 18:08:02 or have both Feb 09 18:08:23 it would be nice to document the concept Feb 09 18:09:11 http://home.hrw.one.pl/tmp/jtag.png Feb 09 18:09:56 thats from sl-5500 service manual Feb 09 18:10:46 nice scan Feb 09 18:10:48 thanks Feb 09 18:11:27 thats screenshot Feb 09 18:11:35 from a PDF? Feb 09 18:12:14 I cannot share that pdf anyway Feb 09 18:12:28 okay Feb 09 18:12:48 does it have a diagram of the Sharp serial IO connector? Feb 09 18:13:12 ~sharp io Feb 09 18:13:14 [sharp io] http://www.sharp.de/php/a_d.php?par=17_no_no_107 sio.zip, or http://www.sharp.de/pdat/sio.zip Feb 09 18:15:10 when I look at amount of *-locale-* pacakges in stuff I wonder how much of them are needed - maybe we could make feed with just locales? Feb 09 18:15:29 thanks Feb 09 18:15:49 hrw: that's what we do with familiar Feb 09 18:16:08 hrw: ipkg-make-index -L Feb 09 18:16:47 that will error out once (mkdir -p the dir it complains about) Feb 09 18:16:56 koen: where you have 0.8.3 feed? Feb 09 18:16:57 hrw: and it will give you something like: http://familiar.handhelds.org/releases/v0.8.3/feed/locale/ Feb 09 18:17:26 hrw: feeds are at http://familiar.handhelds.org/releases/v0.8.3/feed/ Feb 09 18:17:52 I rsync into world.backup and execute split-world.sh Feb 09 18:18:34 (and remove busybox >r30 till ipkg is fixed) Feb 09 18:18:44 koen: how then user manage to get locales? Feb 09 18:19:01 add http://familiar.handhelds.org/releases/v0.8.3/feed/locale/ as a feed Feb 09 18:19:25 from IMAGE_LINGUAS? Feb 09 18:19:41 maybe I'll put in de, fr and en by default Feb 09 18:19:56 * koen has no idea what IMAGE_LINGUAS does Feb 09 18:20:10 install language glibc support Feb 09 18:20:21 mickeyl: morning - slept well? Feb 09 18:20:26 hey mickeyl Feb 09 18:22:05 ~shoot hh.org bugzilla Feb 09 18:22:06 hi Feb 09 18:22:07 * ibot shoots hh.org bugzilla in the eye with a frozen turkey cannon! Feb 09 18:22:17 hrw: heh, i wish i had time to sleep so much :/ Feb 09 18:22:33 mickeyl: me too ;) Feb 09 18:22:47 mickeyl: 3.5.4-test5 released Feb 09 18:22:54 koen: I used ipkg .159 Feb 09 18:23:16 hrw: thanks. i'll have a go at tosa testing Feb 09 18:23:25 SL-C700 bootloader and jtag... hmm... http://lists.arm.linux.org.uk/pipermail/linux-arm/2003-March/005053.html Feb 09 18:23:34 hrw: someone with poodle told me it hangs before launching opie Feb 09 18:24:50 mickeyl: let he test with test5 if can Feb 09 18:25:11 mickeyl: I cannot find poodle bugs (as rest of OZ team too) Feb 09 18:25:11 i'll send him a mail Feb 09 18:28:58 mreimer: hi Feb 09 18:29:03 hi hrw Feb 09 18:29:19 hrw, I have a fixed version of irda-utils that I'll commit soon Feb 09 18:29:26 great Feb 09 18:33:19 brb Feb 09 18:34:25 hi, I made this change: http://en.pastebin.ca/40878 to oe_libinstall base.bbclass to make .la files in staging stop referencing other .la files in the work tree. Feb 09 18:34:25 does anyone know a better solution to achieve this? Feb 09 18:38:06 hi zecke Feb 09 18:40:33 pH5: looks like lots of fun Feb 09 18:42:01 uhmmm.... i got this: Feb 09 18:42:03 monotone: misuse: database schemas do not match: wanted e372b508bea9b991816d1c74680f7ae10d2a6d94, got 3a1e88c29c74b0dd910a092f3dccd39885da15cc. try migrating database Feb 09 18:42:09 pH5: I think $WORKDIR could use some more shell escaping Feb 09 18:42:13 ~curse my home machine Feb 09 18:42:14 May the fleas of a thousand camels infest your most sensitive regions, my home machine ! Feb 09 18:42:21 CSMan: monotone <0.24? Feb 09 18:42:26 yeah Feb 09 18:42:30 CSMan: update Feb 09 18:42:36 changed the checksum method ? Feb 09 18:42:56 CSMan: database schema Feb 09 18:43:02 ah Feb 09 18:43:19 does anybody have an updated monotone binary for freebsd ? Feb 09 18:45:10 JustinP: I fear that I'm just working around the real issue. Feb 09 18:45:28 CSMan: what version do you want? Feb 09 18:45:42 monotone >=0.24 Feb 09 18:45:53 time to build 2.6.12 on ewi for my machine and test does hangups which I have are from kernel or hardware Feb 09 18:46:00 CSMan: 0.25 is in ports, isn't it? I'll build you a package Feb 09 18:46:06 koen: it seems to work as I intended. Feb 09 18:46:21 mreimer: yeah, but it takes forever to build on my pentium classic Feb 09 18:46:28 CSMan: what version of FreeBSD are you running? Feb 09 18:46:30 pH5: people could have funny stuff (spaces, umlauts) in their path Feb 09 18:46:36 and i think i couldn't even do it because a library port was broken Feb 09 18:46:41 mreimer: 5.4 Feb 09 18:46:42 pH5: but we can probably ignore those people ;) Feb 09 18:46:45 RELEASE Feb 09 18:46:52 CSMan: good. I'm building one for you now Feb 09 18:46:56 thanks man Feb 09 18:46:59 =) Feb 09 18:47:03 koen: at least until someone compains ;) Feb 09 18:47:14 that's what i call distributed computing =P Feb 09 18:47:18 :-) Feb 09 18:47:54 mreimer: are you the guy who was trying to bring up WPA ? Feb 09 18:48:03 CSMan: no Feb 09 18:48:06 k~ Feb 09 18:48:12 i had WPA working for OZ Feb 09 18:48:21 yeah, everything works on OZ =/ Feb 09 18:48:28 it required i update the station and primary firmwares of my wifi hardware Feb 09 18:48:30 pH5: probably....:-| Feb 09 18:49:01 shadows: yeah, but not for an ipaq h55xx , right ? Feb 09 18:50:12 /data/build/oe/monotone/org.openembedded.dev/packages/sharp-binary-only/sharp-sdmmc-support.bb is so close to being moved to non-working/ Feb 09 18:50:23 koen: what now? Feb 09 18:50:29 one more parse error and *poof* away it goes Feb 09 18:50:47 JustinP: definitely. for example libgobject-2.0.la in the glib-2.0-dev package still contains the reference to libglib-2.0.la in the $WORKDIR. Feb 09 18:50:58 hrw: http://pastebin.com/547016 Feb 09 18:51:00 is there an openembedded libtool guru? Feb 09 18:51:14 pH5: do you think UDEV_DEVFS_RULES is still needed? Feb 09 18:51:51 koen: ok. it need some skip pacakge stuff Feb 09 18:51:57 pH5: if there was, i doubt they'd admit it Feb 09 18:52:00 :P Feb 09 18:53:22 mreimer: familiar still uses UDEV_DEVFS_RULE afaik Feb 09 18:53:52 kergoth: :) Feb 09 18:53:53 koen: does it still need to? most of the rules seemed to be duplicates Feb 09 18:54:02 koen: h2200 works fine without it Feb 09 18:54:05 mreimer: you'd have to ask pb__ Feb 09 18:54:18 ok. udev is only used with 2.6 kernels, right? Feb 09 18:54:22 iirc it's for programs that assume 2.4 style names Feb 09 18:54:46 ah Feb 09 18:54:59 programs I'd call 'broken' Feb 09 18:55:38 pb__: you around? Feb 09 18:55:57 mreimer yes and every day there is a new udev version Feb 09 18:56:16 and every new udev requires a kernel upgrade Feb 09 18:56:36 koen sure *g* Feb 09 18:56:58 maybee a script is better which make all the devices Feb 09 18:57:16 monotone: misuse: unknown path 'packages/initscripts/initscripts-1.0/c7x0/keymap-2.6.map' Feb 09 18:57:30 do I need to be concerned about that? Feb 09 18:57:33 you know what? I hate monotone last days Feb 09 18:57:59 mreimer: do you have pacakges/keymaps/files/c7x0/keymap-2.6.map? Feb 09 18:58:11 i hope monotone 0.26 will resolve some of the issues Feb 09 18:59:04 did anyone read the comments about forward vs backward deltas? Feb 09 18:59:19 hrw: no, and monotone revert doesn't give it to me Feb 09 18:59:26 fsck. Feb 09 18:59:37 isn't it a .mt-attrs problem? Feb 09 19:00:00 http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=eee7e0d8ba89227d29beba3bc804f5340553c590 Feb 09 19:00:48 koen: what's that mean? what's a .mt-attrs problem? Feb 09 19:01:03 hrw: 97d200dc7ec222ba39a0dea542d91ccc2ac67ddd should fix the irda-utils problem Feb 09 19:01:14 koen: 97d200dc7ec222ba39a0dea542d91ccc2ac67ddd should go into oz354fam083 too Feb 09 19:01:21 mreimer: .mt-attrs contains a reference to a non-existant file Feb 09 19:02:20 koen: is that a mt bug? do I just delete those lines from .mt-attrs? Feb 09 19:02:56 mreimer: see? it take a while to build its dependencies Feb 09 19:03:38 CSMan: http://bilbo.vpop.net/~mreimer/monotone-0.25.tbz Feb 09 19:03:49 CSMan: do you need the libboost too? Feb 09 19:03:53 done? wow Feb 09 19:04:01 that's the one that brakes i think... Feb 09 19:04:01 2.8GHZ P4 :-) Feb 09 19:04:13 233Mhz PMMX =P Feb 09 19:04:19 :-) Feb 09 19:04:35 I understand somewhat, as my previous box was a PIII 450 with 256M RAM Feb 09 19:04:49 CSMan: building a boost package for you Feb 09 19:04:58 =) Feb 09 19:05:23 233mhz pmmx is what i'm on now, a laptop for IRC Feb 09 19:05:50 shadows: that's my developer workstation (no laughters, please) =P Feb 09 19:06:23 i used to have a development box comparable to that, it was a k6-2 233mhz machine Feb 09 19:06:47 with small costs of fast hardware today, it would be strange to keep using it though Feb 09 19:06:55 my time is more worthwhile to me than my money Feb 09 19:07:16 well, i'm trying to get out of the hole before getting a real workstation Feb 09 19:07:20 :) Feb 09 19:07:37 i was shocked, my current desktop computer cost relatively little Feb 09 19:07:42 hrw: booting test5 on my corgi Feb 09 19:07:54 koen: gpe? Feb 09 19:07:59 hrw: yes Feb 09 19:08:00 koen: I have it on my husky Feb 09 19:08:08 ~lart google Feb 09 19:08:08 * ibot declares google a moron Feb 09 19:08:11 it is spec'd at 3000xp+ athlon64 cpu, 1gb ram, 80gb hard disk, everything very nice. total cost including a laserjet printer was less than 600usd Feb 09 19:08:51 amd does the magic there Feb 09 19:08:56 * koen is employing sub-GHz arm stuff now instead of super-GHz x86 Feb 09 19:09:33 yes Feb 09 19:10:05 the real convincing point is that power consumption is about equal to the old computers Feb 09 19:10:26 yeah, but memory consumption is 20x more ;-) Feb 09 19:10:27 since it is configured to throttle down to 1ghz mode when fast processing is not needed Feb 09 19:10:31 :) Feb 09 19:10:37 anyway, flash memory is here to stay =) Feb 09 19:12:36 * hrw declare gpe-mini-browser as unfuncional & useless - dont respond to clicks on links Feb 09 19:12:49 it doesn't? Feb 09 19:12:54 it does Feb 09 19:12:59 it does on familiar :p Feb 09 19:13:00 it just won't open a new page Feb 09 19:13:01 I'll recalibrate screen Feb 09 19:13:41 ~lart me again Feb 09 19:13:42 * ibot slaps me again upside and over the head with one freakishly huge killer whale named hugh Feb 09 19:13:43 * CSMan imagines gpe-mini-browser with openSSL and tabbed browsing Feb 09 19:14:15 * johnX imagines it showing google maps Feb 09 19:14:25 almost there! Feb 09 19:14:34 it does some ajax Feb 09 19:14:53 * hrw imagine it workin with Polish pages Feb 09 19:15:13 gpe-i18n Feb 09 19:18:53 mreimer: actually, i think boost is the one that takes a while Feb 09 19:19:21 CSMan: oops, looks like it finished but I wasn't paying attention :-) Feb 09 19:19:48 i was pulling your leg =P Feb 09 19:21:24 cu - this machine is too unstable to even hold amarok + 2xssh Feb 09 19:22:19 CSMan: http://bilbo.vpop.net/~mreimer/boost-1.33.1.tbz Feb 09 19:22:39 great Feb 09 19:22:54 huge library Feb 09 19:33:36 03mreimer 07org.oe.dev * r97d200dc... 10/packages/irda-utils/ (files/init irda-utils_0.9.16.bb): Feb 09 19:33:36 irda-utils 0.9.16: write out /etc/sysconfig/irda in the init script, rather Feb 09 19:33:36 than in postinst. Feb 09 19:33:42 03koen 07org.oe.dev * rad982961... 10/conf/machine/ep93xx.conf: ep93xx.conf: inherit linux-modules Feb 09 19:33:46 03koen 07org.oe.dev * r8e086923... 10/packages/linux/ep93xx-kernel/defconfig: ep93xx-kernel: update defconfig to work around some breakages Feb 09 19:33:51 03koen 07org.oe.dev * rf7940d7b... 10/conf/machine/include/tune-ep9312.conf: tune-ep9312.conf: add file containing optimized parameters for Cirrus-Logic ep9312 cpus Feb 09 19:39:11 Is setting RDEPENDS_${PN} some new style I've missed out on? Feb 09 19:40:53 RP: new? isn't this the same as with FILES, DESCRIPTION and whatever else goes into the ipk? Feb 09 19:41:43 pH5: I know, I was just more used to seeing RDEPENDS = ... Feb 09 19:43:09 yeah, the (erroneous) "RDEPENDS = " construct is unfortunately quite common. Feb 09 19:45:16 mreimer: good evening Feb 09 19:45:19 mickeyl: good evening Feb 09 19:45:27 good evening to any others whose nicks start with "m" as well Feb 09 19:45:35 good evening pb_ Feb 09 19:45:42 good evening rp Feb 09 19:45:54 hrw|gone: after removing keylaunch: http://handhelds.org/scap/port.9288.png Feb 09 19:45:57 what purpose does gpe-soundserver-0.4.1/makefile-breakage.patch serve besides making gpe-soundserver ignore LDFLAGS (which in turn makes it uncompileable here) Feb 09 19:46:06 hi pb_ Feb 09 19:46:11 pH5: good evening Feb 09 19:46:21 pH5: try monotone log on that file Feb 09 19:47:27 * RP ponders cleaning up some package DEPENDS in .dev... Feb 09 19:47:29 "ERROR: ("can't read from file '%s' (%s)", ('/home/jnc/opensource/openzaurus/build/oetmp/staging/mtx-1-linux/kernel/kernel-ccsuffix', )) while parsing /home/jnc/opensource/openzaurus/org.openembedded.dev/packages/linux/linux-mtx-1_2.4.24.bb" Feb 09 19:47:34 nice. hmph Feb 09 19:47:35 hi pb_ Feb 09 19:48:02 shadows: that file should be in nonworking now Feb 09 19:48:10 okay Feb 09 19:49:06 pH5: that patch does look pretty bogus. you need both PACKAGE_LDFLAGS and LDFLAGS; neither one is good enough on its own Feb 09 19:49:20 who added it? Feb 09 19:50:10 xserver_common should have PACKAGE_ARCH = "all" as its all just scripts/config files? Feb 09 19:50:10 koen: hm.. "import clean BK tree at cset 1.3670". it's prehistoric Feb 09 19:50:48 RP: yeah Feb 09 19:51:38 pH5: ah, heh, it seems I added it Feb 09 19:51:42 pH5: try http://oe-devel.bkbits.net/ Feb 09 19:51:42 ~lart pb Feb 09 19:51:42 * ibot strangles pb with a doohicky mouse cord Feb 09 19:51:55 pb__: is UDEV_DEVFS_RULE (which controls whether devfs-udev.rules is installed) still necessary? koen suggested asking you Feb 09 19:52:10 mreimer: no idea. that sounds like a question for ph5 Feb 09 19:52:19 It really depends how familar is going to handle 2.6 /dev Feb 09 19:52:50 In theory that creates a set of udev rules that emulates devfs Feb 09 19:52:56 udev integrates with DBUS, HAL, and so on Feb 09 19:53:12 err Feb 09 19:53:13 nevermind me Feb 09 19:53:13 shadows: thanks for the bulletin Feb 09 19:53:14 which would only be needed for programs expecting to see devfs-style devices, right? Feb 09 19:54:01 yeah, devfs-style device naming is what familiar uses. Feb 09 19:54:06 mreimer: correct Feb 09 19:55:15 pb_: i meant that if you're depending on devfs, then it's not going to fit the "udev way' Feb 09 19:55:16 which devices does familiar care about? many are 2.4 only, like the hal stuff, right? what others that would be relevant in a 2.6 world? Feb 09 19:55:26 pb_: like this: http://en.pastebin.ca/40889? Feb 09 19:55:57 mreimer: disks, serial ports, that kind of thing Feb 09 19:56:12 * RP -> afk Feb 09 19:56:13 framebuffers are another common one Feb 09 19:56:30 and apm, iirc Feb 09 19:56:36 pH5: yeah Feb 09 19:56:43 shadows: ah right, yeah Feb 09 19:56:59 ok Feb 09 19:57:35 mreimer: also, following the principle of least user confusion, making capricious changes to the device layout seems like a dim plan Feb 09 19:57:41 pb_: for 2.4.x kernels i would add my 2 bits and say that you're better off eliminating devfs from the 2.4 kernel setup Feb 09 19:58:07 shadows: for 2.4, there isn't any viable alternative Feb 09 19:58:19 does there need to be? Feb 09 19:58:44 well, er, yes. we can't just stop providing /dev altogether. Feb 09 19:59:25 is making all possible /dev device nodes is not an option? Feb 09 19:59:34 s/is// Feb 09 19:59:35 shadows meant: making all possible /dev device nodes is not an option? Feb 09 19:59:50 certainly not an attractive one, if only because it would waste a pile of flash space for most people Feb 09 20:00:21 how about a script that creates a tmpfs like udev does, and simply creates a load of devices there Feb 09 20:00:39 that'd work, but it would be slower than devfs and it's not at all obvious to me what it would buy us Feb 09 20:00:51 it would eliminate devfs Feb 09 20:01:02 IMO devfs is like the ugliest code i've ever read through Feb 09 20:01:03 we've been shipping devfs in 2.4 for about six years now, and there doesn't seem much point in getting rid of it now just for the sake of it Feb 09 20:01:15 hm Feb 09 20:01:52 it might be ugly, but it works fine for our users. I'm not all that thrilled by the prospect of doing a pile of work to get rid of it if the only benefit will be greater code hygiene. Feb 09 20:02:04 ah Feb 09 20:02:11 particularly given that the 2.4 kernel tree is essentially at EOL anyway Feb 09 20:02:15 yes i would agree with you on that Feb 09 20:02:42 is there a reason that gpe-soundserver depends on diet-x11 and not x11? Feb 09 20:03:01 * hrw for moment Feb 09 20:03:52 pH5: no, that sounds like a bug Feb 09 20:05:39 hmm.. cairo/libsvg => "| svg_parser_expat.c:40: error: static declaration of '_svg_parser_sax_end_element' follows non-static declaration" Feb 09 20:05:52 gcc4 cross-initial being used Feb 09 20:06:07 same sort of issue as the thing with RP's changes to the kernel code i guess Feb 09 20:06:28 shadows: look into debian (or gentoo) for gcc4 patches Feb 09 20:06:57 okay Feb 09 20:07:00 i'm going to try a cvs checkout Feb 09 20:07:02 see if that works Feb 09 20:07:14 CSMan: are those monotone packages working for you? Feb 09 20:07:39 nope, no dice for cvs checkout Feb 09 20:13:05 we have nice things like lost patches, unrelated, https://bugs.freedesktop.org/show_bug.cgi?id=3711 Feb 09 20:13:08 :( Feb 09 20:14:07 shadows: tried debian? Feb 09 20:18:44 koen: libtiff is not fetchable so no gimp Feb 09 20:18:46 cu Feb 09 20:24:10 03coredump 07org.oe.dev * rbf4c4f3a... 10/packages/gpe-dm/ (files/multi-lm.patch gpe-dm_0.48.bb): gpe-dm: Do not install /etc/init.d/gpe-dm any longer, but use /etc/init.d/login-manager instead. Feb 09 20:24:20 03coredump 07org.oe.dev * r9556cfd5... 10/packages/login-manager/ (files/login-manager login-manager.bb): Add "login-manager": A script to select the login-manager to be used on boot. Feb 09 20:24:28 03coredump 07org.oe.dev * rd738d315... 10/packages/e17/ (e-wm_0.16.999.023.bb entrance_0.9.0.005.bb): entrance: Do no longer install initscript into /etc/init.d but use "login-manager" to be launched. Feb 09 20:34:22 monotone: multiple update candidates: Feb 09 20:34:22 monotone: dc7030746547ed09bdd941871c775b044c145143 coredump@openembedded.org 2006-02-09T19:31:19 Feb 09 20:34:22 monotone: f7940d7b4ac1dca8c38ef92949c3d012d70c9202 koen@openembedded.org 2006-02-09T19:18:33 Feb 09 20:54:30 gremlin[it]: merge :} Feb 09 20:55:39 thank zecke :) Feb 09 20:58:18 gaim_cvs now compiles a binary to check for %z support in strftime during configure. is there a way to fix this for cross compilation? Feb 09 20:58:36 pH5: site packages Feb 09 20:58:37 put it in the site file, I guess Feb 09 20:58:45 site file, yes Feb 09 20:59:22 any tolerably recent linux system should have %z Feb 09 21:00:04 even freebsd probably has it these days Feb 09 21:01:15 zecke, pb_: thanks Feb 09 21:01:45 03pH5 07org.oe.dev * ra2ba6d1a... 10/packages/gpe-soundserver/gpe-soundserver-0.4-1/makefile-breakage.patch: gpe-soundserver: fix makefile fix not to ignore LDFLAGS Feb 09 21:01:50 03pH5 07org.oe.dev * rf5ff77ed... 10/packages/gpe-soundserver/gpe-soundserver_0.4-1.bb: gpe-soundserver: depend on x11 instead of diet-x11 Feb 09 21:03:42 hrm, but the gaim configure.ac just contains an AC_TRY_RUN block that conditionally AC_DEFINEs a HAVE_STRFTIME_Z_FORMAT. I suppose I can't preempt that without patching configure.ac? Feb 09 21:06:19 dunno, you'd have to inspect the generated configure script Feb 09 21:06:26 I don't remember if AC_TRY_RUN checks the cache or not Feb 09 21:06:43 probably not, since I suspect it doesn't have any way to figure out a cache variable name on its own, but you never know Feb 09 21:08:03 hrm Feb 09 21:08:09 ~stab .mt-attrs Feb 09 21:08:10 * ibot runs at .mt-attrs with an origami Swiss Army knife, and inflicts a nasty paper cut. Feb 09 21:10:06 is busybox 1.01 bad? Feb 09 21:10:31 there's 1.1 now Feb 09 21:10:57 just wondering which one I should stay away from. isn't there some symlink or chown problem or something? Feb 09 21:12:14 pb_: if both glibc and uclibc support %z, is there a reason to keep this test in OE at all? Feb 09 21:12:29 mreimer: there's a bug like that in 1.00. it might be fixed in later versions Feb 09 21:12:36 thanks pb_ Feb 09 21:12:50 pH5: some users might be building against other c libraries Feb 09 21:12:54 e.g. on windows, or solaris Feb 09 21:14:28 mreimer: (this was the "chown /dev/dsp" one; I'm not sure if that's the same bug you were thinking of) Feb 09 21:14:48 pb_: yeah, that's what I was thinking of Feb 09 21:33:46 I have a problem with some auto-renaming of libs Feb 09 21:34:42 edje-dev is renamed libedje-dev in the ipk and hence should be libedje-dev in the RDEPENDS of something which relies on it. However, if I pur libedje-dev in an RDEPENDS it fails saying there are no buildable providers. Changing it to edje-dev allows building but not installing.... Feb 09 21:35:37 what would rdepend on libfoo-dev? Feb 09 21:35:50 that should only contain headers, symlinks and pkgconfig files Feb 09 21:36:46 there's also edje_cc in there Feb 09 21:36:55 which I need for a script Feb 09 21:37:00 it's a binary Feb 09 21:37:09 that shouldn't be in there Feb 09 21:37:12 suppose I could move it to an edje_cc package.... Feb 09 21:37:21 that should probably in a seperate package Feb 09 21:37:25 well, it's not needed for normal edje usage Feb 09 21:37:26 ok Feb 09 21:37:50 edje-utils, or edje-cc Feb 09 21:37:55 03koen 07org.oe.oz354fam083 * r3faa674c... 10/.mt-attrs: .mt-attrs: remove c7x0 keymap so people stop being scared by error messages Feb 09 21:38:01 03koen 07org.oe.oz354fam083 * r4aa5b330... 10/.mt-attrs: .mt-attrs: remove another keymap entry Feb 09 21:38:08 03koen 07org.oe.oz354fam083 * rd1ba8d9d... 10/.mt-attrs: .mt-attrs: and another one Feb 09 21:38:15 03koen 07org.oe.oz354fam083 * r970c8a9d... 10/.mt-attrs: .mt-attrs: and the fun never stops.... (sharp sdcontrol) Feb 09 21:38:21 it seems that this new inferred depend stuff breaks a lot when combined with debian.bbclass Feb 09 21:49:20 pb_: about gaim/strftime, is this what you meant: http://en.pastebin.ca/40921 Feb 09 21:51:55 that's not exactly what I had in mind, but it might be the only option. Feb 09 21:52:02 can you not do something with AC_CACHE_CHECK()? Feb 09 21:56:01 pb_: I know next to nothing about autoconf. if I can put the AC_TRY_RUN macro into AC_CACHE_CHECK's COMMANDS-TO-SET-IT parameter, it should work too. Feb 09 21:58:39 koen|away: can you be more specific about this breakage? as far as I know its mainly limited to -dev packages atm Feb 09 22:02:06 hm, "The COMMANDS-TO-SET-IT _must have no side effects_ except for setting the variable CACHE-ID" Feb 09 22:04:45 pH5: right, within that block you set the variable. after the block you do the real work based on the varaible's value. Feb 09 22:13:00 kergoth, pb_: ok, so I don't think AC_CACHE_CHECK is useful here. Feb 09 22:14:38 kergoth: thanks Feb 09 22:29:51 pH5: okay Feb 09 22:30:18 whats that app called that lets you syunc your opie/gpe contacts and stuff with evo? Feb 09 22:30:20 * kergoth cant recall the name Feb 09 22:30:41 kergoth: opensync/multisync Feb 09 22:31:09 pH5: ping Feb 09 22:31:12 ah thats right Feb 09 22:31:17 zecke: yes? Feb 09 22:31:29 pH5: could you give me a update on the stamp issue :} Feb 09 22:31:34 pH5: what was the issue again? Feb 09 22:32:15 zecke: bitbake references tmp/stamps directly Feb 09 22:32:57 zecke: http://en.pastebin.ca/40937 Feb 09 22:34:24 RP: any more issues that prevent a 1.3.3 release? Feb 09 22:34:38 zecke: no, just that one I know of Feb 09 22:34:51 pH5: you know any other issue? Feb 09 22:35:26 That patch looks wrong though Feb 09 22:35:39 RP: really? Feb 09 22:36:14 Where did pn, pv and pr go? Feb 09 22:36:50 RP: look at bitbake.conf Feb 09 22:36:59 Yes, they're in STAMP, ok :) Feb 09 22:37:01 STAMP = "${TMPDIR}/stamps/${PACKAGE_ARCH}-${HOST_OS}/${PF}" Feb 09 22:37:38 I thought STAMP was a directory, not part of a filename. my mistake :) Feb 09 22:37:54 my cats went nuts.. I should send them torwrds cambridge... Feb 09 22:38:03 RP: you made me recheck the patch :) Feb 09 22:38:20 pH5: off hand you do not happen to know the bug number for this patch? Feb 09 22:38:42 I'm not sure there is one Feb 09 22:39:53 zecke: me neither Feb 09 22:45:19 zecke: I also don't have any other issues with bitbake. Feb 09 22:45:35 good I'm releasing a 1.3.3 now Feb 09 22:46:27 zecke: please keep your cats to yourself, we don't need any more around here Feb 09 22:47:36 damn... Feb 09 23:01:09 03rpurdie 07org.oe.dev * r10fca1d2... 10/packages/xserver-common/ (3 files): xserver-common: Remove invalid DEPENDS and set PACKAGE_ARCH = all Feb 09 23:01:14 03rpurdie 07org.oe.dev * rac9084ad... 10/packages/lsof/lsof_4.75.bb: lsof 4.75: Update SRC_URI Feb 09 23:01:21 03rpurdie 07org.oe.dev * rcba2bacf... 10/packages/linux/ (linux-openzaurus.inc linux-openzaurus_2.6.15.bb): linux-oz-2.6: From now on refer to patches within the archive directory. Feb 09 23:04:55 night all Feb 09 23:05:00 night pb_ Feb 09 23:05:15 night pb_ Feb 09 23:06:08 good night Feb 09 23:06:18 'night pb_ Feb 09 23:09:01 well bitbake-1.3.3 is released now Feb 09 23:09:09 will write a small anouncement tomorrow Feb 09 23:09:38 zecke: Excellent. Should stop a few people complaining :) Feb 09 23:09:50 RP: even released on the berlios site Feb 09 23:10:09 an added bonus :) Feb 09 23:10:17 Did you add the other release tarballs? Feb 09 23:10:26 RP: just 1.3.2 Feb 09 23:10:38 which was the only one missing? Feb 09 23:10:55 1.3.0 and 1.3.1 as well are missing Feb 09 23:12:34 * shadows looks at /openzaurus/build/oetmp/work/armv5te-linux/libsvg-0.0+cvs20060209-r2/libsvg/src Feb 09 23:13:03 why do the .c files have the prototype of the static functions, when they are defined in the header that is included? Feb 09 23:13:06 any particular reason? Feb 09 23:33:12 i think i found a way to make cairo/libsvg compile with gcc4, it involves some automake hackery Feb 09 23:42:44 03justinp 07org.oe.dev * r476f0163... 10/classes/efl.bbclass: efl.bbclass: use += for PACKAGES to allow setting of PACKAGES before inheriting Feb 09 23:42:50 03justinp 07org.oe.dev * r7c4e459d... 10/conf/distro/preferred-e-versions.inc: preferred-e-versions: Add PREFERRED_PROVIDERS for evas, ecore, and imlib2 for x11 Feb 09 23:42:57 03justinp 07org.oe.dev * rd8bd0542... 10/packages/efl/edje_0.5.0.023.bb: edje: move binaries to edje-utils package from edje-dev Feb 09 23:43:01 03justinp 07org.oe.dev * rbefe3287... 10/packages/e17/ (2 files in 2 dirs): e17-gpe-menu-convert: add a new package to convert GPE .desktop files into an e17 menu Feb 09 23:43:08 03justinp 07org.oe.dev * rb252db8b... 10/packages/meta/e-image.bb: Feb 09 23:43:08 e-image: add e17-gpe-menu-convert to installed packages Feb 09 23:43:08 - NOTE: This adds a dep on cpp and edje-utils which will make the image larger Feb 09 23:58:44 03justinp 07org.oe.dev * r21cd02fb... 10/packages/efl/edje_0.5.0.023.bb: edje: move template and include directories to edje-utils, move files in -themes to -examples Feb 10 00:48:41 mreimer: i think the cmov instruction won't let me use the monotone you built Feb 10 00:48:44 =( Feb 10 00:51:22 CSMan: sorry about that. must be my CPUTYPE=p4 Feb 10 00:52:16 CSMan: I'll rebuild Feb 10 00:54:09 * mreimer cd boost; make package Feb 10 01:02:20 hopefully Feb 10 01:07:15 CSMan: boost is done (same URL) Feb 10 01:07:26 * mreimer cd monotone; make package Feb 10 01:23:06 CSMan: http://bilbo.vpop.net/~mreimer/monotone-0.25.tbz and http://bilbo.vpop.net/~mreimer/boost-1.33.1.tbz Feb 10 01:23:42 /me grumbles Feb 10 01:23:45 heh Feb 10 01:24:09 a set of PACKAGES and FILES_... isn't working fright for me... Feb 10 01:24:13 the later one is taking precedence somehow... Feb 10 01:33:52 anyone have any ideas? Feb 10 01:54:48 * emte does Feb 10 01:54:56 give them some love Feb 10 01:55:00 :) Feb 10 01:55:18 emte: my problem is in efl.bbclass Feb 10 01:55:34 emte: ${SRCNAME}-config is getting packed into -examples, not -dev as it should Feb 10 01:55:46 I have no idea why Feb 10 01:55:50 hmm Feb 10 01:56:15 why would it go into examples ... Feb 10 01:56:21 I Feb 10 01:56:33 * JustinP lost the terminal.... Feb 10 01:56:38 03justinp 07org.oe.dev * r26cf91c3... 10/packages/efl/embryo_0.9.1.023.bb: embryo: split out binaries into embryo-utils Feb 10 01:56:42 03justinp 07org.oe.dev * r57af2cc5... 10/packages/efl/edje_0.5.0.023.bb: edje: Remove unneeded files from edje-utils, add dependency on embryo-utils Feb 10 01:56:47 03justinp 07org.oe.dev * r83e07433... 10/packages/efl/edb_1.0.5.005.bb: edb: move utility binaries to edb-utils from edb-dev Feb 10 01:56:51 03justinp 07org.oe.dev * re1a42505... 10/packages/efl/edje_0.5.0.023.bb: edje: use += just in case Feb 10 01:56:55 03justinp 07org.oe.dev * r96b18d58... 10/packages/efl/embryo_0.9.1.023.bb: embryo: remove extra embryo-dev in PACKAGES Feb 10 01:57:51 grrrr.... Feb 10 01:57:56 emte: any ideas? Feb 10 01:58:09 * JustinP was about to send an e-mail to the list Feb 10 01:58:31 the only way it would do that is if it was miss-identifying the files no? Feb 10 01:59:05 I'm not sure Feb 10 01:59:24 everything I've tried says that the file is identified right in the FILES entry Feb 10 01:59:30 and taht would lead to an inclusion statement or regex i'd think Feb 10 01:59:35 even adding a hard-coded entry for the package I'm testing doesn't help.... Feb 10 02:00:10 then perhaps the issue is outside your class ... Feb 10 02:00:29 thing is it happens to all of my packages which inherit efl.... Feb 10 02:01:09 i'll take a look at efl Feb 10 02:03:24 you expect all your *-config to only be in the bindir? Feb 10 02:04:11 seems like -examples is a catchall Feb 10 02:04:17 screen keeps crapping out on me.... Feb 10 02:04:29 emte: it is...but -dev should catch it Feb 10 02:04:33 FILES_${SRCNAME}-dev += "${bindir}/${SRCNAME}-config ${libdir}/pkgconfig ${libdir}/lib*.?a ${datadir}/${SRCNAME}/include" Feb 10 02:04:33 FILES_${SRCNAME}-examples = "${bindir} ${datadir}" Feb 10 02:04:41 is what your looking at , right? Feb 10 02:04:52 -dev is before -examples in PACKAGES, so it should catch the -config Feb 10 02:04:56 yes Feb 10 02:05:45 and you only have ${SRCNAME}-config ? in -dev ? Feb 10 02:05:46 at least I thought that's how it worked a while ago.... Feb 10 02:05:55 no Feb 10 02:06:04 -dev has everything it should *but* -config Feb 10 02:06:20 bah Feb 10 02:06:23 * JustinP removes -examples Feb 10 02:06:23 now i am lost Feb 10 02:06:31 -dev does not have the -config Feb 10 02:06:38 so you want -config in dev or in examples? Feb 10 02:06:40 is has the pkgconfig and .a/la files Feb 10 02:06:42 in dev Feb 10 02:06:51 okay Feb 10 02:06:52 I want -config in dev Feb 10 02:07:13 do you need to escape the - ? Feb 10 02:07:44 * emte is extreamly rusty on the tiny bit of python i did know Feb 10 02:08:41 hmmm...do I? Feb 10 02:08:45 is *that* it? Feb 10 02:08:51 * emte shrugges Feb 10 02:09:00 its the only one that has a "-" in it Feb 10 02:09:05 true... Feb 10 02:09:34 i remember python has issues with some char but i dont remember which ones Feb 10 02:11:06 That could be it....I never would have thought of that... Feb 10 02:13:44 using ? doesn't seem to help Feb 10 02:13:50 so i don't think it's the -.... Feb 10 02:14:35 * JustinP tries without the FILES_..-examples Feb 10 02:15:58 damn.... Feb 10 02:16:13 if I remove -examples the -config doesn't get packaged at all... Feb 10 02:18:47 as i'd expect Feb 10 02:19:13 and your sure its: ${SRCNAME}-config Feb 10 02:20:12 just seems like it does not match that name Feb 10 02:21:09 I tried hard-coding it to embryo-config Feb 10 02:21:13 doesn't help Feb 10 02:21:25 even ${SRCNAME}* didn't work Feb 10 02:21:50 hmm Feb 10 02:22:39 weird Feb 10 02:22:41 and it's even more screwed up... Feb 10 02:22:57 if I put ${bindir} in FILES_${SRCNAME}-dev it still doesn't work Feb 10 02:23:10 even without the -examples FILES entry Feb 10 02:23:50 maybe its not in bindir ... Feb 10 02:24:18 but -config *does* get put into -examples.... Feb 10 02:24:25 (if I leave its entry in...) Feb 10 02:24:35 and ${bindir} works for other files in that same dir Feb 10 02:33:38 * JustinP uses "peek" Feb 10 02:34:29 well.....WT! Feb 10 02:34:42 BB>> peek embryo FILES_embryo-examples Feb 10 02:34:42 /usr/bin /usr/share Feb 10 02:34:42 BB>> peek embryo FILES_embryo-dev Feb 10 02:34:42 /usr/include /usr/lib/lib*.so /usr/lib/*.la /usr/lib/*.a /usr/lib/*.o /usr/lib/pkgconfig /lib/*.a /lib/*.o /usr/share/aclocal Feb 10 02:35:13 that is *not* what I'm putting in -dev Feb 10 02:37:22 hmmm...lib_package.bbclass has something similar...although is *also* includes ${bindir}-config at the end.... Feb 10 02:37:31 and I'm not inheriting lib_package....AFAIK Feb 10 02:43:56 muther...... Feb 10 02:44:01 conf/bitbake.conf Feb 10 02:45:06 that's where this entry is coming from Feb 10 02:45:14 why isn't my bbclass overwriting it? Feb 10 04:31:50 ~lart FILES_${PN}-dev Feb 10 04:31:50 * ibot beats FILES_${PN}-dev severely about the head and shoulders with a rubber chicken Feb 10 04:32:59 ~lart it again Feb 10 04:33:00 * ibot turns it again into a lifesized tux doll Feb 10 04:34:12 heh Feb 10 04:34:47 JustinP: I've just pushed the e-theme for gpe-bootsplash Feb 10 04:35:36 ah, cool Feb 10 04:35:45 what do *you* know about -dev packages? Feb 10 04:35:55 as far as I can tell you *can't* set FILES for -dev packages... Feb 10 04:36:22 * JustinP pulls Feb 10 04:36:24 well, you should be able to set files. If not, that's a bug. Feb 10 04:36:45 I have been setting FILES for -doc packages in the past IIRC. Feb 10 04:36:51 it's not working for efl.bbclass Feb 10 04:37:07 and if I use the shell and peek (I love this feature) it's completely not what I set Feb 10 04:37:09 hmm Feb 10 04:37:13 in fact, it's what's in bitbake.conf.... Feb 10 04:37:24 but if I change bitbake.conf the value doesn't change Feb 10 04:37:26 hmmm.... Feb 10 04:37:49 I don't remember where thy implemented the -dev, -dec etc stuff Feb 10 04:37:50 * JustinP clears his cache Feb 10 04:37:59 s/dec/doc Feb 10 04:38:00 *looks* like it's in bitbake.conf to me... Feb 10 04:38:22 pushing to ewi, eh? Feb 10 04:38:28 yes Feb 10 04:38:37 * JustinP always pushes to vanille Feb 10 04:38:51 vanille was to unstable in the past =) Feb 10 04:39:02 so you try and if it's down you use ewi ;-) Feb 10 04:39:08 meh heh Feb 10 04:39:14 ah, ok, changing it made a difference this time... Feb 10 04:39:24 I set up some aliases for them ;-) Feb 10 04:39:28 mtpull and mtpullewi Feb 10 04:39:45 my mtpull lets me choose the server =D Feb 10 04:39:56 ::shrug:: Feb 10 04:40:09 :-) Feb 10 04:40:18 but push isn't an alias here, so ewi it is =) Feb 10 04:40:59 btw, entrance now works without the /bin/sh hack *cough* =) Feb 10 04:42:57 hehe Feb 10 04:43:09 glad that miraculously started working again Feb 10 04:44:13 damn morgue Feb 10 04:45:10 * CSMan tries Feb 10 04:45:22 tsk tsk...bad CoreDump|afk Feb 10 04:45:24 hydra Feb 10 04:45:31 * JustinP merges Feb 10 04:45:55 the fun of having more than one upstream server Feb 10 04:47:12 my changes should have been on ewi my now... Feb 10 04:47:14 :-| Feb 10 04:47:17 oh well, no problem Feb 10 04:47:22 * JustinP is the merge master Feb 10 04:48:02 CoreDump|afk: how's this work? Feb 10 04:48:16 just install the theme Feb 10 04:48:28 gpe-bootsplash-theme-e Feb 10 04:48:37 so I also need gpe-bootsplash Feb 10 04:48:48 sure, is pulled in by deps Feb 10 04:48:48 don't think that's part of my e images.... Feb 10 04:48:53 is it? Feb 10 04:48:55 which one? Feb 10 04:49:07 nah deps from the theme i meant Feb 10 04:49:23 not yet part of e-image Feb 10 04:49:42 ok, thought so Feb 10 04:50:00 will test later Feb 10 04:50:07 don't want a bootsplash right now Feb 10 04:50:10 need to see output Feb 10 04:50:17 (or is there a key to turn it off?) Feb 10 04:50:22 btw: re entrance using /etc/X11/Sessions: I think that's the best place anyways Feb 10 04:50:40 JustinP: well, Alt-LeftArrow =) Feb 10 04:50:56 it works? I thought I'd tried that.... Feb 10 04:51:06 works fine Feb 10 04:51:40 hmmm Feb 10 04:51:47 the Alt-LeftArrow thing works w/ 2.6 and 2.4 kernel keymaps now Feb 10 04:51:48 well, will test later.... Feb 10 04:51:54 sure, no hurry Feb 10 04:51:55 how come CoreDump is AFK ? Feb 10 04:51:58 =P Feb 10 04:52:00 trying to track down this FILES-dev thing Feb 10 04:52:15 CSMan: because I gotta go to work in a few minutes heh Feb 10 04:53:44 * CoreDump|afk kicks of another series of bitbakes. tmp should be done when I get home Feb 10 04:54:40 thanks for the bootsplash Feb 10 04:54:43 I'll add it soon Feb 10 04:54:52 you're welcome =) Feb 10 04:54:57 * CoreDump|afk likes eye-candy Feb 10 04:58:05 * JustinP likes it too Feb 10 05:01:59 I'm off to work, cya Feb 10 05:02:28 later Feb 10 05:09:37 03coredump 07org.oe.dev * rde353371... 10/packages/gpe-bootsplash/ (4 files in 2 dirs): Feb 10 05:09:37 gpe-bootsplash: Use update-alternatives to install pictures Feb 10 05:09:37 Added: gpe-bootsplash-theme-e Feb 10 05:09:42 03coredump 07org.oe.dev * r3af6af59... 10/packages/login-manager/ (files/login-manager login-manager.bb): login-manager: Fix a pathing bug which b0rks login-manager if only one LM is installed. Feb 10 05:11:51 mreimer: monotone won't run , i'll just portupgrade mine Feb 10 05:17:23 where do bitbake bugs go? Feb 10 05:19:05 * JustinP files an OE bug Feb 10 05:26:31 hey JustinP isnt it normal your confs override classes? Feb 10 05:26:53 nope. package metadata can always override configuration metadata Feb 10 05:28:07 kergoth: classes, not package Feb 10 05:28:10 http://bugs.treke.net/show_bug.cgi?id=667 Feb 10 05:28:19 very strange bug.... Feb 10 05:28:25 classes are package metadata by definition. Feb 10 05:28:40 just abstracted out Feb 10 05:28:47 well...bitbake.conf is overriding it Feb 10 05:28:51 then its broken. Feb 10 05:28:54 and += in bitbake.conf didn't help Feb 10 05:28:57 that wasnt the way it was designed to behave Feb 10 05:28:59 hey kergoth , hows the exo going? Feb 10 05:29:02 I figured Feb 10 05:29:09 exo? Feb 10 05:29:10 took a while to figure out what it was, though Feb 10 05:29:36 exoskeleton ... isnt that what your doing now? Feb 10 05:30:01 it isnt an exoskeleton Feb 10 05:30:18 google for land warrior Feb 10 05:30:36 but its going well. clueful group of people working thre Feb 10 05:30:42 and shit, i can get used to 75 degree weather in february Feb 10 05:30:48 :) Feb 10 05:30:56 what about 115 in july? Feb 10 05:31:03 hehe. we'll see Feb 10 05:31:08 air conditioning is my frien Feb 10 05:31:10 d Feb 10 05:31:12 lol Feb 10 05:32:12 okay lets see if i can get my evas event to spawn windows Feb 10 06:52:37 morning kergoth Feb 10 06:55:37 hey Feb 10 07:00:08 hi Feb 10 07:12:44 hi gremlin[it] Feb 10 07:16:20 * Geo_KM is away: Away at the moment Feb 10 07:16:52 good morning pb_ ;) Feb 10 07:23:12 * pb_ heads to his 0730 meeting Feb 10 07:23:14 blrh Feb 10 07:23:16 later all Feb 10 07:46:26 <_law_> morning all Feb 10 07:47:21 * france is away: Away Feb 10 07:51:57 hi law Feb 10 07:52:23 117 testers for oz 3.5.4 test images Feb 10 07:55:41 <_law_> hrw, i?ll test test5 collie image this weekend Feb 10 07:55:50 <_law_> should i use opie or gpe? Feb 10 07:56:19 both ;) Feb 10 07:56:21 03hrw 07org.oe.oz354fam083 * rdfff0763... 10/packages/irda-utils/ (files/init irda-utils_0.9.16.bb): Feb 10 07:56:21 irda-utils 0.9.16: write out /etc/sysconfig/irda in the init script, rather Feb 10 07:56:21 than in postinst. Feb 10 07:56:21 taken from .dev Feb 10 07:56:39 afk Feb 10 08:12:03 03hrw 07org.oe.oz354fam083 * ra1beb18b... 10/packages/libtiff/tiff_3.7.2.bb: libtiff: fix SRC_URI (from .dev) Feb 10 09:02:33 morning Feb 10 09:03:00 morning Feb 10 09:03:28 <_chronic> hey all Feb 10 09:07:46 re Feb 10 09:09:18 morning all Feb 10 09:09:27 hey RP Feb 10 09:09:34 morning RP :) Feb 10 09:24:09 morning all Feb 10 09:25:41 morning Feb 10 09:26:48 cannot check setpgrp when cross compiling for apache. any hint Feb 10 09:26:58 hi dirk Feb 10 09:27:34 good monring all Feb 10 09:27:48 hi koen .. where are u flying to ? Feb 10 09:27:59 Hey hrw|work, koen Feb 10 09:28:06 london Feb 10 09:28:28 have a good trip :) Feb 10 09:28:40 thanks Feb 10 09:30:02 hmm Feb 10 09:30:14 -mcpu=iwmmxt will default to softpa, right? Feb 10 09:30:34 ~lart intel Feb 10 09:30:34 * ibot runs at intel with an origami Swiss Army knife, and inflicts a nasty paper cut Feb 10 09:32:46 koen|airport: have a nice flight Feb 10 09:33:17 hrw|work: I had to 'rm /usr/bin/keylaunch' to get test5 to work Feb 10 09:33:24 Bernardo: thanks Feb 10 09:34:08 <[lala]> morning Feb 10 09:35:10 koen|airport: ~lart gpe ;) Feb 10 09:35:32 ~lart matchbox Feb 10 09:35:33 * ibot explains, ever so gently, that if matchbox doesn't give the channel more information, they can't help Feb 10 09:35:41 stupid keymaps Feb 10 09:36:00 both mb and keylaunch caim a certain key -> crash Feb 10 09:36:50 claim* Feb 10 09:43:36 someone remember when machine conf cleanup will happen? Feb 10 09:47:04 hrw|work: I think we also need a COMPATIBILITY_MODE in local.conf that will stop stuff like -mcpu=iwmmxt, armv5te or ep9312 Feb 10 09:47:22 03nail 07org.oe.dev * rf54776e0... 10/packages/zd1211/ (zd1211-r52/transmitkey.index.fix.diff zd1211_r52.bb): zd1211: Add transmitkey index fix Feb 10 09:48:43 koen|airport: FORCE_ARCH="arm" Feb 10 09:49:28 that deosn't reset TARGET_CC_ARCH, right? Feb 10 09:49:53 I dont know does force_arch exist at all Feb 10 09:50:54 I'd like to put all 'include tune-foo' isnide a base-conditional check Feb 10 09:51:38 DISTRO_TYPE="release" -> ARCH="arm" Feb 10 09:52:00 not quite Feb 10 09:52:24 I can imagine people would a release wit optimizations Feb 10 09:52:44 would like* Feb 10 09:53:06 hmm. true Feb 10 09:53:34 but you get the idea Feb 10 09:53:44 ~stab thunderbird Feb 10 09:53:45 * ibot runs at thunderbird with an origami Swiss Army knife, and inflicts a nasty paper cut. Feb 10 09:53:53 thunderbird + imap = hell Feb 10 09:54:05 thunderbird + imap over gprs = even worse Feb 10 09:54:16 Eww.. Feb 10 09:54:28 koen|airport: I've found tbird's not -too- bad with imap.. not the greatest.. Feb 10 09:54:44 it beats apple's Mail.app Feb 10 09:55:01 hehe, yeah Feb 10 09:55:11 It's better than opie's email cli Feb 10 09:55:12 and Entourage Feb 10 09:55:55 I should try to build sylpheed-claws for OSX Feb 10 09:56:09 or stop using email ;) Feb 10 09:56:36 Hmm.. anyone had success building gcc for the target with OE? Feb 10 09:57:32 yes, but I still haven't looked at setting up native dev Feb 10 09:57:42 uf.. cleaned most of my branch copy Feb 10 09:57:48 koen|airport: Any special packages you had to install to get it to work? Feb 10 09:58:10 gcc and libc6-dev Feb 10 09:58:36 and auto*, make, m4 ,etc Feb 10 09:58:41 the usual suspects Feb 10 09:58:46 Mine keeps bitching about MPFI and GMP Feb 10 09:59:37 my giac build failed on that Feb 10 10:00:06 Er, MPFR Feb 10 10:05:27 hrw|work: Which machine cleanup do you refer to? Feb 10 10:09:14 RP: zaurus 2.4/2.6, 'this machine has bt, wifi' 'this one has cf slot and bt' Feb 10 10:09:49 heh Feb 10 10:09:50 easy Feb 10 10:09:58 this zaurus only has scf Feb 10 10:10:02 cf* Feb 10 10:10:59 so in opie-image I could check do I have to add bt stuff or not etc Feb 10 10:18:52 hrw|work: We agreed how to do this so we just need to actually change the files around Feb 10 10:19:30 I fixed zaurus-clamshell.conf Feb 10 10:25:07 03hrw 07org.oe.oz354fam083 * r7766c833... 10/conf/machine/zaurus-clamshell.conf: zaurus-clamshell: s/c7x0/$MACHINE in IPKG_ARCHS - not every clamshell is c7x0 Feb 10 10:25:12 03hrw 07org.oe.oz354fam083 * r7a65ccf9... 10/ (3 files in 2 dirs): Feb 10 10:25:12 poodle 2.4 config: switch virtual/kernel to poodle-kernels-2.4-embedix Feb 10 10:25:12 this gives us kernel for pxa250 (default) and pxa255 poodles Feb 10 10:25:12 taken from .dev Feb 10 10:25:16 03hrw 07org.oe.dev * r924ef15f... 10/conf/machine/include/zaurus-clamshell.conf: Feb 10 10:25:16 zaurus-clamshell: s/c7x0/$MACHINE in IPKG_ARCHS - not every clamshell is c7x0 Feb 10 10:25:16 taken from .oz354fam083 Feb 10 10:59:48 meeting **** ENDING LOGGING AT Fri Feb 10 10:59:56 2006