**** BEGIN LOGGING AT Fri Sep 05 02:59:56 2008 Sep 05 07:15:30 hi Sep 05 07:37:26 bonjour Sep 05 08:57:25 03koen 07org.oe.dev * r5dc1ac3b... 10/ (8 files in 7 dirs): linux-omap: add missing files Sep 05 09:21:00 hi, all! Sep 05 09:21:24 is it possible to generate locales using qemu in XEN environment? Sep 05 09:21:40 it just segfaults for me :( Sep 05 09:59:00 03koen 07org.oe.dev * r4d007885... 10/ (1 contrib/angstrom/build-feeds.sh): angstrom feeds: add mythtv Sep 05 11:01:04 03koen 07org.oe.dev * rd725cc1a... 10/ (1 packages/gnome/gnome-menus_2.22.2.bb): gnome-menus: update to 2.22.2 Sep 05 11:20:50 slapin: Xen shouldn't make any difference. Sep 05 11:21:14 What distribution are you running on your build host? I think qemu does have trouble with some versions of Red Hat. Sep 05 11:25:47 pb_, I run Debian testing inside XEN, and qemu segfaults, which is not case with normal box. Sep 05 11:26:49 pb_, it runs ok with OpenVZ, but I can't change my xen configuration since it is at service provider. Sep 05 11:28:35 slapin: ah, fair enough. I guess you have to debug why it crashes. Sep 05 11:31:43 03koen 07org.oe.dev * rac2e88d3... 10/ (1 packages/gnome/gnome-desktop_2.22.2.bb): gnome-desktop: update to 2.22.2 Sep 05 11:40:26 hi mickeyl Sep 05 11:40:31 yo florian Sep 05 11:40:40 how can we proceed with OE e.v.? Sep 05 11:40:44 Do we need a conference call next week Sep 05 11:40:45 ? Sep 05 11:40:55 this takes all too long, tbh. Sep 05 11:41:10 mickeyl: Yes, sounds like a plan. Should help to speed things up. Sep 05 11:41:50 florian: good morning Sep 05 11:41:52 mickeyl: good morning Sep 05 11:42:01 hi pb_ Sep 05 11:42:11 morning pb_ Sep 05 11:42:15 03john_lee 07org.oe.dev * r4303b4d9... 10/ (1 packages/dbus/dbus.inc packages/dbus/dbus_1.2.1.bb): (log message trimmed) Sep 05 11:42:15 dbus: enable x support and put dbus-launch into new package dbus-x11 Sep 05 11:42:15 * Enable x support only add rdepends to dbus-launch. Make it a Sep 05 11:42:15 seperate package so other packages like dbus does not rdepend on Sep 05 11:42:16 libx11. Sep 05 11:42:18 * This solves the problem that if you restart X for multiple times, Sep 05 11:42:20 there will be many dbus-daemon left in the system. That's because Sep 05 11:42:46 hrm... I have an old email in my inbox that seems to freeze my mua :-/ Sep 05 11:44:04 johncylee: excellent! Sep 05 12:02:39 03koen 07org.oe.dev * rb122e80b... 10/ (1 packages/gtk-engines/gtk-engines_2.15.4.bb): gtk-engines: update to 2.15.4 Sep 05 12:02:43 03koen 07org.oe.dev * r626f4d1d... 10/ (5 files in 4 dirs): gnome-themes: update to 2.22.2 Sep 05 12:02:49 03koen 07org.oe.dev * r9370d059... 10/ (1 packages/gnome/gnome-themes_2.22.2.bb): gnome-themes: no binaries, so set package arch to all Sep 05 12:04:04 moin Sep 05 12:07:29 03koen 07org.oe.dev * r7f3c3e5a... 10/ (3 files in 3 dirs): eel: update to 2.22.2 Sep 05 12:07:59 any opinions on Vala? I'm currently re-writing a small app in it, and wonder what other people think of it. Sep 05 12:09:40 cbrake: mickeyl loves it :) Sep 05 12:10:10 cbrake: I'm sceptical but don't have first hand experience. I ready the blog entries where people have to fix the "translator/compiler" instead of writing code :) Sep 05 12:10:21 Vala is da bomb Sep 05 12:10:34 even C++ started as a translator Sep 05 12:10:56 seriously, Vala is the only sane future of glib/gobject programming IMO Sep 05 12:11:14 :) Sep 05 12:11:16 (if you don't want to use a scripting language, that's even better) Sep 05 12:11:21 03koen 07org.oe.dev * r42f804f1... 10/ (1 packages/angstrom/angstrom-x11-base-depends.bb): angstrom-x11-base-depends: add dbus-x11 Sep 05 12:13:49 mickeyl: :-) the app I'm re-writing is a small device managagment app that I wrote in C#, but we can't tolerate the 8 second startup time. Sep 05 12:14:00 *nod* Sep 05 12:14:06 no go for python or ruby then Sep 05 12:14:17 (if you don't want quicklaunch technology) Sep 05 12:14:46 my biggest complaint with Vala is lack of examples and documentation -- it seems that figuring out how to use the API is mostly experimentation and groking vapi files at this point. Sep 05 12:15:56 well, there 10 or 20 examples Sep 05 12:16:00 but yes, no real documentation Sep 05 12:16:01 03koen 07org.oe.dev * rdb7aa3c6... 10/ (1 packages/nautilus/nautilus_2.22.5.bb): nautilus: update to 2.22.5 Sep 05 12:16:16 johncylee: I think your update breaks my build.. Sep 05 12:16:19 since they still change the language, i would do the same though Sep 05 12:16:22 ERROR: '['/home/alphaone/oe-dev/org.openembedded.dev/packages/dbus/dbus-native_1.2.1.bb']' RDEPENDS/RRECOMMENDS or otherwise requires the runtime entity 'dbus-native-x11' but it wasn't found in any PACKAGE or RPROVIDES variables Sep 05 12:16:31 mickeyl: nod Sep 05 12:21:35 alphaone: *fixing* Sep 05 12:21:48 mickeyl: I think I found it Sep 05 12:22:05 just switching the include/require lines helped Sep 05 12:22:19 ? Sep 05 12:22:27 At least bitbake doesn't complain any more.. Sep 05 12:22:28 we need to yank the RDEPENDS and RRECOMMENDS for the -native packgae Sep 05 12:22:50 Eh, right Sep 05 12:23:01 That I also changed :-) Sep 05 12:23:04 aah Sep 05 12:23:15 let me push it, i have also a stylechange Sep 05 12:23:25 yeah, vala seems pretty awesome. personally I would like GLib bindings for objective-c, but I guess one can't have everything. Sep 05 12:23:49 mickeyl, alphaone: oops, sorry Sep 05 12:23:54 pb_: *nod* Sep 05 12:24:00 objC support would rock Sep 05 12:24:05 johncylee: no problem, fix underway Sep 05 12:24:15 johncylee: No problem Sep 05 12:24:47 admittedly objective-c is, in some ways, teh suck, but we have been using it in my company for a few years and it works pretty well for the most part. Sep 05 12:25:06 i'm slowly falling in love with it Sep 05 12:25:14 using it on some platform that must not be named Sep 05 12:25:18 heh Sep 05 12:25:19 :D Sep 05 12:25:39 mickeyl: Strange, just putting inherit native on top did fix it Sep 05 12:25:40 my main complaint is that selectors don't carry any type information Sep 05 12:25:53 pb_: oh yes, that's a huge miss Sep 05 12:26:13 alphaone: indeed strange Sep 05 12:26:13 03mickeyl 07org.oe.dev * r9fa03717... 10/ (5 files in 3 dirs): uicmoc4-native: remove all versions that do not build, keep the only working one Sep 05 12:26:16 i.e. if you have SEL foo = @selector(myMethodWith:someParameter), and later you want to call through that SEL, there is nothing to stop you accidentally getting either the args or the return value completely wrong. Sep 05 12:26:18 03mickeyl 07org.oe.dev * r80177ad8... 10/ (1 packages/dbus/dbus-native_1.2.1.bb packages/dbus/dbus.inc): dbus-native: yank RRECOMMENDS and RDEPENDS Sep 05 12:26:38 the return value in particular has screwed us up on numerous occasions when we accidentally made our methods return void. Sep 05 12:26:44 pb_: uh oh, i can imagine Sep 05 12:26:45 pb_: you like objective c? Sep 05 12:27:10 zecke: apart from the above, yes Sep 05 12:27:16 pb_: cool! Sep 05 12:27:37 I wrote a UPnP stack in objective c a couple of months ago and it was a fairly painless experience. Sep 05 12:27:40 pb_: how is the GNU runtime? how is it compared to apples one? is gc, properties and the hyped stuff of objective-c 2.0 coming? Sep 05 12:27:58 are... damn... it is too early to use proper grammar Sep 05 12:28:24 i for one, love the fact, that you can mix C++ and C in objC(++) Sep 05 12:29:44 zecke: I think it's had GC for some time, though we don't actually use that feature ourselves Sep 05 12:30:05 there's certainly an --enable-objc-gc option to configure which I think includes the boehm gc in the runtime Sep 05 12:31:16 not sure about objective-c 2.0. I don't think much/any of that other stuff is in the gnu runtime yet. Sep 05 12:31:27 personally I have never really used the apple runtime so I don't know what I am missing :-} Sep 05 12:32:23 lunchtime, brb Sep 05 12:32:34 ~bon appetit Sep 05 12:32:35 methinks bon appetit is smacznego. Guten Appetit. Eet Smakelijk. God Appetitt. Buon Appetito. Buen apetito Bom Apetite. buen apetito Smaklig måltid!. Hyvää ruokahalua. Bo Proveito Sep 05 12:34:08 03mickeyl 07org.oe.dev * rff89d579... 10/ (1 packages/dbus/dbus-native_1.2.1.bb): dbus-native 1.2.1 fix it for real (see previous commit) Sep 05 12:36:27 + Dobrou chut Sep 05 12:38:32 stefan_schmidt: could you take a look at the refcounting of dbus messages in illume's batget? Sep 05 12:38:49 mickeyl: I don't know if you can do this with the stock GNU cc :) Sep 05 12:39:24 zecke: Batget does not have dbus msg's Sep 05 12:40:30 zecke: ah well Sep 05 12:40:42 zecke: my sole use for objc is on apple platforms though Sep 05 12:40:49 yet Sep 05 12:41:09 stefan_schmidt: sorry gsmget :) Sep 05 12:41:18 stefan_schmidt: and I know that it moved to illume itself Sep 05 12:41:38 zecke: gsmget is no more. All dbus is handles in the gadget itself now Sep 05 12:41:43 stefan_schmidt: we had a report that gsmget leaks, then it was moved to illume, now e leaks.. Sep 05 12:41:43 zecke: heh Sep 05 12:42:00 zecke: I can take a look later Sep 05 12:42:20 zecke: Raster fixed some stuff yesterday Sep 05 12:42:37 zecke: I'll take a look anyway Sep 05 12:46:43 stefan_schmidt: it would be very welcome. even if you could give me an answer like "fixed in ..." Sep 05 12:46:46 :) Sep 05 12:48:06 zecke: Are you sure that it leaks dbus msg and nothing else? Raster fixed a wrongly set dbus signal handler yesterday not sure if that would lead to the leakage Sep 05 12:56:53 stefan_schmidt: it leaks so badly fork fails :) Sep 05 12:57:04 stefan_schmidt: it leaks so badly that fork fails :) Sep 05 12:57:16 stefan_schmidt: it is from a user report but from a user i trust on this Sep 05 12:58:09 zecke: wow, that's bad. Do you plan to update a newer version without gsmget or backport fixes? (Later could get tricky as a lot get rewritten) Sep 05 12:59:04 stefan_schmidt: upgrade... Sep 05 12:59:18 zecke: good, I let you know my findings later Sep 05 13:01:05 stefan_schmidt: cool :) Sep 05 13:01:22 http://flickr.com/photos/16845425@N00/2830648062/sizes/o/ <- I'm quite happy Sep 05 13:02:03 NOTE: package illume-0.0+svnrSRCREVINACTION""-r13: task do_fetch: started Sep 05 13:02:12 Any clue what that might be? Sep 05 13:02:23 lol Sep 05 13:02:29 alphaone: that is an internal bitbake flag :) Sep 05 13:02:39 alphaone: go to bed and sleep Sep 05 13:02:51 ? Sep 05 13:03:14 alphaone: it is obviously not your day (dbus breakage, funny bitbake things) Sep 05 13:03:48 zecke: Well, it wont get better I assume Sep 05 13:04:23 This sort of stuff always happens after some time not using bitbake Sep 05 13:04:29 zecke: What is the goal of your current webkit hacking? Sep 05 13:05:12 stefan_schmidt: Implement DIMS (parts of it) Sep 05 13:05:28 zecke: What does DIMS stand for? Sep 05 13:05:36 zecke: ...interactive...? Sep 05 13:06:30 stefan_schmidt: stefan_schmidt http://www.3gpp.org/ftp/specs/html-info/26142.htm Sep 05 13:06:52 zecke: argh, 3gpp again. It feels they are everywhere... ;) Sep 05 13:07:13 stefan_schmidt: they are? but why again? Sep 05 13:07:17 stefan_schmidt: it is better than OMA Sep 05 13:07:58 zecke: I had to much part at alphaone *fun* with PDU mode :) Sep 05 13:08:59 * stefan_schmidt -> food Sep 05 13:09:08 my current favorite is w3c.org... they are as bad as every other spec but have test cases Sep 05 13:20:01 03koen 07org.oe.dev * r4dbaa78e... 10/ (1 packages/dbus/dbus-native_1.2.1.bb): dbus-native: fix up dependencies Sep 05 13:39:46 re Sep 05 13:57:32 03koen 07org.oe.dev * re9758ef5... 10/ (1 packages/dbus/dbus-native_1.2.1.bb): disapproval of revision '4dbaa78e8c16d8981f8ee837e88654194fab70c6' Sep 05 13:57:38 03koen 07org.oe.dev * rda2adc92... 10/ (5 files in 3 dirs): libpam: update to 1.0.2 Sep 05 13:57:43 03crofton 07org.oe.dev * r5a665541... 10/ (1 packages/gnuradio/gnuradio_3.1.3.bb): gnuradio : Add gnuradio-3.1.3. Sep 05 14:30:26 heh, it seems even mickey's car has irc access now Sep 05 14:30:34 hello mickeymobile Sep 05 14:31:58 HopsNBarley: hey! Sep 05 14:32:19 hi likewise - finally back from mucho travel. Sep 05 14:32:32 pb_: yes, it's his in-car navigation system that can ask for directions on #oe Sep 05 14:33:03 ah yes, of course Sep 05 14:33:12 HopsNBarley: mucho far-awayo? Sep 05 14:33:39 6k round trip: florida and back. Sep 05 14:34:09 hi likewise - finally back from mucho travel. Sep 05 14:34:16 whoops! Sep 05 14:34:33 HopsNBarley: np, I can deal with some redundancy Sep 05 14:34:34 back again? that was a short trip Sep 05 14:34:55 HopsNBarley: ok, that's a decent trip indeed. I have to pickup my girlfriend in town centre. bbiab Sep 05 14:35:08 see ya. Sep 05 14:36:10 hmm Sep 05 14:36:18 mobile IRC clients suck Sep 05 14:36:22 softkeyboards as well Sep 05 14:36:24 *shrug* Sep 05 14:40:46 03utx 07org.oe.dev * r2277b452... 10/ (6 files in 4 dirs): viking: Added new package: program to manage GPS data. Sep 05 14:40:57 good morning to all Sep 05 14:41:02 moinall Sep 05 14:41:42 Hi... I'm new to OE, and I'd like to understand this message : ERROR: Cannot satisfy the following dependencies for task-base * angstrom-version * . Anyone could explain me where to look ? Sep 05 14:41:46 Is there a way to edit the shlib dependenies in executables without relinking ? Sep 05 14:43:34 w_erase: what image are you building Sep 05 14:45:02 w_erase: I don't know if you are already doing this, but I find it helpful $ bitbake -DDD blah-blah >output.out 2>>output.out .... The DDD increases logging and then I search the output.out file to see if there is more detail/hints about were things are going wrong. Sep 05 14:45:57 khem : http://wiki.openmoko.org/wiki/SHR_Development Sep 05 14:46:07 w_erase: i am not looking at it right now, but I think you can have upto 4 D for the logging...can someone confirm Sep 05 14:46:35 bin1010, thx ! Sep 05 14:46:55 np Sep 05 14:48:19 alphaone: I run into the same problem with illume do_fetch. Let me know if you found the problem Sep 05 14:48:51 w_erase: you can also dump the dependency graphs using bitbake -g and then inspect them Sep 05 14:50:50 khem: morning khem Sep 05 14:56:56 so I am trying to build angstrom and openmoko distros, is it ok to use the same "stuff" tree for both? Sep 05 14:56:58 hi khem, how's it going? Sep 05 14:57:28 Anyone who can tell me what openembedded actually is? I can't seem to find an introduction on the oe website anywhere.. Sep 05 14:57:41 I have successfully built angstrom images, but I have now switched over to openmoko, and fakeroot-native-1.7.1 won't build, it dies on configure Sep 05 14:57:58 blathijs, i view it as a linux distribution construction kit. Sep 05 14:57:59 blathijs: oe is a build system for creating embedded linux distributions Sep 05 14:58:46 like angstrom, openmoko, poky, slugos Sep 05 15:01:25 khem, thx for the tip Sep 05 15:03:03 hans: But I guess it only works for the distros that have been built using oe? Sep 05 15:03:39 Does oe also define a package format and/or file system layout, etc? Sep 05 15:17:48 03stefan 07org.oe.dev * rd131b2a9... 10/ (1 conf/distro/include/fso-autorev.inc): fso-autorev.inc: Fix illume autorev Sep 05 15:24:25 stefan_schmidt: No go here... Sep 05 15:24:35 I even wiped tmp, but no luck Sep 05 15:26:57 blathijs: I think OE can use .deb. and .ipk, it seems to be vaguely debian based Sep 05 16:39:54 how can I set a preferred provider for a runtime library? Sep 05 17:11:35 is our mtn server down? Sep 05 17:44:05 hi someone up and working? Sep 05 17:44:41 i need to understand the do_populate_staging, i have a Error in this step doing a git-native Sep 05 17:45:58 * kergoth`work mulls over alternative oe file formats Sep 05 17:50:41 re Sep 05 17:50:50 i get out of connection for a while Sep 05 18:28:04 kergoth`work: good morning Sep 05 18:28:13 hi flo_lap Sep 05 18:28:22 re Sep 05 18:32:26 hey pb_ Sep 05 19:02:17 the git mirror should be online again Sep 05 19:06:26 wohoo! Sep 05 19:06:44 git was not able to open enough files Sep 05 19:06:48 git-repack fixed it for now... Sep 05 19:24:16 man I hate users... now they want to talk to my supervisor. I will search a sane job now Sep 05 19:24:53 heh Sep 05 19:25:03 where are you working again? Sep 05 19:25:31 kergoth`work: mplayer does not work... Please give me the output of opkg list_installed... well you don't even have it installed... Sep 05 19:25:35 kergoth`work: Openmoko Sep 05 19:25:48 zecke hehe Sep 05 19:25:57 zecke seems you were rude again Sep 05 19:26:03 ah Sep 05 19:26:05 I would talk to your manager too Sep 05 19:26:08 *g* Sep 05 19:26:14 woglinde: eat my shorts Sep 05 19:26:25 which one? Sep 05 19:26:41 woglinde: the really dirty ones.. well they are equally dirty Sep 05 19:26:53 bbl Sep 05 19:27:04 :) Sep 05 19:33:25 hmm, lets say an upstream developer had a basic bitbake recipe, but lacking packaging/files information.. say you move packaging decisions into the hands of distro policy.. i think thatd be clean, but the upstream dev knows what he depends on, so you'd run into discrepencies in package naming/splitting when expressing the runtime dependencies of the package.. Sep 05 19:34:02 * kergoth`work wonders about the possibility of file level dependency, not just cross package/system like rpm has done (depend on /bin/sh), but also dependency among files in a package Sep 05 19:36:35 i'd think 99% of it could be automated, just associations between files and documentation would be hard to represent, and auxilliary files would have to be manual Sep 05 19:50:19 re zecke Sep 05 19:53:13 wow, I'm a bit pissed off Sep 05 19:53:45 * mwester says it wasn't he -- this time! Sep 05 19:57:29 zecke hm Sep 05 20:44:23 hi polyonymous Sep 05 20:47:28 hi bluelightning Sep 05 20:47:42 hi woglinde Sep 05 20:55:05 ah nice libxcomp did compile Sep 05 20:55:10 next step nxproxy Sep 05 21:01:33 morning all Sep 05 21:03:06 hi thesing Sep 05 21:03:09 hi ant Sep 05 21:03:21 hello woglinde Sep 05 21:03:41 thesing: he he kopfwasch... Sep 05 21:03:50 wie immer von K. Sep 05 21:04:49 ant__: he has a valid point Sep 05 21:04:56 hm... I don't see anything different from linux-rp-2.6.2x Sep 05 21:05:16 (defconfigs) Sep 05 21:05:43 linux-rp is not the best kernel recipe in oe. Sep 05 21:06:06 well, all for zaurus is ackward Sep 05 21:06:20 because of the proprietary origin :-) Sep 05 21:06:27 and mostl other recipes use machine dirs. Sep 05 21:06:37 I'd just ask RP Sep 05 21:06:49 I think RP just didn't want to mess with existing kernel recipe. Sep 05 21:07:58 I think the kernel build process in oe needs some love. Sep 05 21:08:11 +standardization Sep 05 21:08:22 = kernel in /boot Sep 05 21:11:53 woglinde: btw linux-kexecboot would be a trivial port on simpad Sep 05 21:13:04 just dunno if you need it..if you have no space limits you could just use the uclibc initramfs-bootmenu-image Sep 05 21:16:04 ant I only wanted a easy way for nfsboot Sep 05 21:17:13 or start from my cf card Sep 05 21:17:37 so i can safe cycles on the internal flash Sep 05 21:18:10 and time (several partitions supported on CF/SD) Sep 05 21:18:56 4gb flash costs around 20 -30 eur Sep 05 21:19:54 thesing: I have one strange mini-issue... after choosing the boot source and pressing OK the screen is somehow corrupted for a moment, before kexec. This did not happen with the first versions Sep 05 21:20:21 really a no issue Sep 05 21:21:00 ant yeah but with cf I have no network Sep 05 21:21:11 because of the one pcmcia slot Sep 05 21:21:28 ok, I use CF for Bluetooth Sep 05 21:21:43 or ethernet at home Sep 05 21:21:58 do you have usb on simpad ? Sep 05 21:22:41 there is usb-client on simpad. But I don't think that the current kernel has the necessary patch. Sep 05 21:22:56 it seems zaurus have even been reported to boot from usb Sep 05 21:23:13 same usb-client on c7x0 Sep 05 21:23:16 ant yes but it is crappy Sep 05 21:23:22 iirc Sep 05 21:23:27 because of armv4 architecture and sa1100 Sep 05 21:23:46 he, it seems 2.6.27 would be better Sep 05 21:23:56 ant? Sep 05 21:24:07 1 sec Sep 05 21:24:29 http://lists.linuxtogo.org/pipermail/angstrom-distro-devel/2008-September/thread.html Sep 05 21:24:30 thesing the last stuff I see was 2 years ago in the balloon cvs Sep 05 21:24:32 bottom Sep 05 21:24:40 that wasnt working well Sep 05 21:25:22 stanislav and dmitry are our beloved kernel hackers.... Sep 05 21:25:27 There is a patch from balloon that I got working on collie. Sep 05 21:25:32 thesing too ;-) Sep 05 21:25:49 I could ping through usb. Sep 05 21:25:53 ant__: ;) Sep 05 21:26:34 thesing yes ping Sep 05 21:26:40 but ssh is pain Sep 05 21:27:12 and data transfer is the ko Sep 05 21:27:12 Well its a point where one can start from. Sep 05 21:27:14 what does it mean: For more recent kernels you'll need to fix one line - change Sep 05 21:27:14 #include to #include Sep 05 21:27:25 fixed ? Sep 05 21:27:36 thesing I really looked into it Sep 05 21:27:45 and dont understood much Sep 05 21:27:56 how the usb stuff works Sep 05 21:28:07 and there were many workarounds Sep 05 21:28:10 for silicon bugs Sep 05 21:28:29 I think one needs to read part of the usb spec and the usb part of SA Manual to do serious work on that driver. Sep 05 21:28:38 yes Sep 05 21:28:40 but not me Sep 05 21:28:43 well, I know usb can work, by luck, but usually does not Sep 05 21:29:07 and there will be no other Sep 05 21:29:12 bescause armv4 is dead Sep 05 21:30:18 Well there are still a bunch of people that work on debian for 68K processors. Sep 05 21:30:43 thesing yes I know Sep 05 21:30:44 collie and simpad are quite cheap now and they are far from useless. Sep 05 21:31:22 ah nice Sep 05 21:31:28 now libxcomp is in shape Sep 05 21:31:51 next step nxproxy nad nxcl tommorrow which will be easier Sep 05 21:31:55 and then qtnx Sep 05 21:31:59 and oe has nx supporr Sep 05 21:32:00 t Sep 05 21:32:14 I know two persons who said that they wan't to look at that usb driver. Sep 05 21:32:55 thesing who? *g* Sep 05 21:33:42 gremlin and Christian Varga Sep 05 21:35:05 want? or won't? Sep 05 21:35:14 want Sep 05 21:35:31 hm Sep 05 21:35:38 thesing: the actual c7x0 kexecboot kernel w/out net is 1164604 Sep 05 21:35:55 go crappy mtn Sep 05 21:36:00 I want to get in my bed Sep 05 21:36:07 ah done Sep 05 21:36:08 nite Sep 05 21:36:16 nite Sep 05 21:36:31 ant__: is this small enough? Sep 05 21:36:50 thesing: yes, >250kb left, but I'd restart copying the ones in linux-rp-2.6.26 Sep 05 21:37:03 hrw did some fix to sound on c7x0 Sep 05 21:37:18 and dmitry / stanisalv did edit the configs Sep 05 21:37:25 in the last weeks Sep 05 21:37:34 03woglinde2 07org.oe.dev * r664c9037... 10/ (5 files in 4 dirs): Sep 05 21:37:34 libxcomp: inital release Sep 05 21:37:34 * sa_restorer is taken from latest debian unstable package Sep 05 21:37:34 * first step for nx-client support in openembedded Sep 05 21:40:23 03 07master * rd2e90f61be 10OE.dev/packages/libxcomp/ (4 files in 2 dirs): Sep 05 21:40:23 libxcomp: inital release Sep 05 21:40:23 * sa_restorer is taken from latest debian unstable package Sep 05 21:40:23 * first step for nx-client support in openembedded Sep 05 21:40:25 03 07org.openembedded.dev * rd2e90f61be 10OE.dev/packages/libxcomp/ (4 files in 2 dirs): Sep 05 21:40:25 libxcomp: inital release Sep 05 21:40:29 * sa_restorer is taken from latest debian unstable package Sep 05 21:40:31 * first step for nx-client support in openembedded Sep 05 21:40:36 03koen 07org.oe.dev * rbda06e19... 10/ (4 files in 2 dirs): gdb: add 6.7.1 with avr32 support Sep 05 21:40:41 03koen 07org.oe.dev * r7f8cad96... 10/ (1 packages/gnash/gnash-minimal.inc): gnash-minimal: fix packaging Sep 05 21:40:45 thesing: is in zaurus-2.6.inc Sep 05 21:40:47 KERNEL_IMAGE_MAXSIZE = "1294336" Sep 05 21:40:47 KERNEL_IMAGE_MAXSIZE_collie = "1048576" Sep 05 21:41:07 1408k is uboot :-) Sep 05 21:49:17 bitbake -c menuconfig linux-rp-2.6.26 Sep 05 21:49:24 oops wrong window Sep 05 21:52:17 anyone know why creating a test-base-image.bb that simply does a 'DISTRO_INIT_MANAGER = "" require packages/images/base-image.bb' does 'not' keep sysvinit from appearing in the image? It would appear that DISTRO_INIT_MANAGER is not seen by task-boot Sep 05 21:53:35 tharvey: I think its because the task is parsed before the image Sep 05 21:54:15 03 07org.openembedded.dev * ra432077c37 10OE.dev/packages/gnash/gnash-minimal.inc: gnash-minimal: fix packaging Sep 05 21:54:15 03 07org.openembedded.dev * r0e150de965 10OE.dev/packages/gdb/ (gdb-cross_6.7.1.bb gdb.inc gdb_6.7.1.bb): gdb: add 6.7.1 with avr32 support Sep 05 21:54:16 03 07org.openembedded.dev * r2489897f13 10OE.dev/: Sep 05 21:54:16 merge of '664c903716e2388f94905f48588afd5542b78fc6' Sep 05 21:54:17 and '7f8cad96c8dde93c85aa34ce420a3a852cbdf6ad' Sep 05 21:54:20 03 07master * ra432077c37 10OE.dev/packages/gnash/gnash-minimal.inc: gnash-minimal: fix packaging Sep 05 21:54:22 03 07master * r0e150de965 10OE.dev/packages/gdb/ (gdb-cross_6.7.1.bb gdb.inc gdb_6.7.1.bb): gdb: add 6.7.1 with avr32 support Sep 05 21:54:25 03 07master * r2489897f13 10OE.dev/: Sep 05 21:54:27 merge of '664c903716e2388f94905f48588afd5542b78fc6' Sep 05 21:54:31 and '7f8cad96c8dde93c85aa34ce420a3a852cbdf6ad' Sep 05 21:55:09 thesing, ah... because tasks are not included/required they are always in namespace? Sep 05 21:55:36 Crofton: when is a good time to ping you? Sep 05 21:57:01 anybody knows about libssp? Sep 05 21:57:30 tharvey: hm. looking at this in more detail I think the task and the image get parse separately (they don't include each other) Sep 05 21:57:52 right... I noticed no inclusion of task anywhere other than by name Sep 05 21:58:53 so are all tasks 'evaluated' prior to parsing a recipe then? Meaning all vars defined are considered defaults Sep 05 21:58:53 i think the image set a DEPEND on IMAGE_INSTALL and image.bbclass installs everything to some dir an packages it. Sep 05 21:59:28 tharvey: the tasks are evaluated when they are build. Which is before the image is build. Sep 05 22:01:00 tharvey: if you don't want initscripts in your image don't use a task that include them I guess. Sep 05 22:02:59 well cearly the task was designed so that init is a variable (DISTRO_INIT_MANAGER) - I'm just trying to understand where that variable can be set Sep 05 22:03:20 it would appear that you can 'not' set it in a recipe so that means it must be set in a conf Sep 05 22:03:44 which describes the behavior I'm seeing, but certainly is not intuitive Sep 05 22:04:20 andrea@mizar /oe/org.openembedded.dev $ grep -R DISTRO_INIT_MANAGER . Sep 05 22:04:20 ./conf/distro/openprotium.conf:DISTRO_INIT_MANAGER = "sysvinit" Sep 05 22:04:20 ./packages/tasks/task-boot.bb:DISTRO_INIT_MANAGER ?= "sysvinit sysvinit-pidof" Sep 05 22:04:20 ./packages/tasks/task-boot.bb: ${DISTRO_INIT_MANAGER} \ Sep 05 22:04:20 ./packages/tasks/task-openprotium.bb: ${DISTRO_INIT_MANAGER} \ Sep 05 22:04:25 is a task considered a 'recipe' (because its a bb file)? If so, then my statement above isn't quite correct... Sep 05 22:04:55 ant__, right... notice how its only set in conf or tasks Sep 05 22:05:14 yes, no ptrace in packages Sep 05 22:05:32 ant__, no 'ptrace'? Sep 05 22:05:44 s/ptrace/trace/ Sep 05 22:05:54 tharvey: you can set it in your local.conf Sep 05 22:06:13