**** BEGIN LOGGING AT Thu Jun 23 23:59:56 2005 Jun 24 00:04:58 okay that will teach me to check my loops .... raster ecore/evas is too fast :P Jun 24 00:05:20 was spawning canvases faster than i could ctrl+c the server Jun 24 00:05:38 took 15 min to ssh in and kill it Jun 24 00:19:51 good morning all Jun 24 00:33:58 not able to download openembedded; bitkeeper is giving message connection timed out ? any solutions to this problem ? Jun 24 00:34:30 i could not also download bitblake svn also says connection timed out. Jun 24 00:34:55 is there any way i can download tar files and unzip them and work on them ?. Jun 24 00:35:09 does bitblake require svn or cvs will do ?. Jun 24 00:36:16 ram123: looks like you're behind a firewall Jun 24 00:36:30 yes i am behind a firewall. Jun 24 00:36:47 and my comp does not allow anyone to work without a firewall. Jun 24 00:37:01 is there any solution to work behind a firewall. Jun 24 00:38:23 i want to build X and GPE using oe and bitblake Jun 24 00:38:35 once you get the software you should be able to work behind a firewall Jun 24 00:39:09 morning Jun 24 00:39:20 hey hrw|work Jun 24 00:41:16 but how to get the software in the first place sir Jun 24 00:41:38 u mean each of the dependend libs also need to be downloaded with out a firewall ? Jun 24 00:43:02 http://ewi546.ewi.utwente.nl/tmp/bitbake.tar.bz2 Jun 24 00:43:05 koen: give him archived snapshots of bitbake and OE if you can - then rest of work should work for him Jun 24 00:43:26 and http://ewi546.ewi.utwente.nl/tmp/bk-export.tar.bz2 Jun 24 00:43:55 hrw|work: I had to wait for bk export and tar to finish :) Jun 24 00:44:04 ;) Jun 24 00:44:51 so i need to down load these software and do what ?. Jun 24 00:45:03 it does not work with svn right ? Jun 24 00:45:27 is there any diff. downloading using svn and bitkeeper and downloading just the tar file ? Jun 24 00:45:55 the steps in getting started do they hold good if the downloaded files are tar files. Jun 24 00:46:04 i mean the subsequent steps Jun 24 00:46:37 the subsequent steps should be the same Jun 24 00:46:53 so no need to use svn Jun 24 00:47:07 or should i use some other version control sys ?. Jun 24 00:47:23 there shouldn't be any need Jun 24 00:47:56 but does downloading dependend libs for a package say gpe use svn ? Jun 24 00:48:18 gpe release tarballs Jun 24 00:48:22 not that I know off Jun 24 00:48:29 opie release only in cvs Jun 24 00:48:33 maybe uclibc Jun 24 00:48:47 or svn is only used in my case is for local make, build and release ? Jun 24 00:48:48 but oesources.org should have it mirrored Jun 24 00:48:53 and if you will use CVS_TARBALL_STASH then you need only wget/curl working Jun 24 00:49:10 ram123: svn is only used to get bitbake Jun 24 00:49:21 and bk is only used to get the OE data Jun 24 00:49:34 but what about local version control ? Jun 24 01:11:01 reboot.. after 23days Jun 24 01:11:04 brb Jun 24 01:13:58 re Jun 24 01:29:01 hello Jun 24 01:29:22 but what about the local version control and release control. is svn used for that ? Jun 24 01:34:37 ram123: ? Jun 24 01:36:32 i was told svn is only for downloading bitblake. is it also used for local version control system of builds and releases ? Jun 24 01:37:14 we tag repo when there is release Jun 24 01:37:51 you can create distro configs for releases like it is done with openzaurus and familiar in openembedded/conf/distro/ Jun 24 01:38:15 just downloaded and http://ewi546.ewi.utwente.nl/tmp/bk-export.tar.bz2 Jun 24 01:38:15 what should i do next ? Jun 24 01:38:25 what does this file contain bitblake ? Jun 24 01:38:39 http://ewi546.ewi.utwente.nl/tmp/bk-export.tar.bz2 is OE, http://ewi546.ewi.utwente.nl/tmp/bitbake.tar.bz2 is bitbake Jun 24 01:38:40 how different is it from svn downloaded bitblake ? Jun 24 01:38:49 you need both Jun 24 01:39:06 OE is metadata repo, bitbake is a tool which build stuff from OE Jun 24 01:39:34 do i need to download http://ewi546.ewi.utwente.nl/tmp/bitbake.tar.bz2 ? Jun 24 01:39:47 i have downloaded http://ewi546.ewi.utwente.nl/tmp/bk-export.tar.bz2 Jun 24 01:39:58 24 10:38 < hrw|work> you need both Jun 24 01:40:16 have you read GettingStarted? Jun 24 01:41:26 yes Jun 24 01:41:52 so you should know that you need both Jun 24 01:42:57 this is not mentioned in getting started ? Jun 24 01:44:30 ya got it ! Jun 24 01:44:55 To start using OE, you must first obtain the build tool that it needs (bitbake): Jun 24 01:45:01 bitblake.tar.bz2 is bitblake and bk-export.tar.bz2 is meta data Jun 24 01:45:06 Next, you'll need to obtain the actual OpenEmbedded repository, which has our build classes, configuration files, and packages. Jun 24 01:45:14 u need bitblake to build on meta data. Jun 24 01:45:22 just like make on makefiles Jun 24 01:45:25 am i right ? Jun 24 01:45:25 exactly Jun 24 01:45:41 yes Jun 24 01:46:06 but what is this openembedded rep Jun 24 01:46:19 i though bk-exports.tar.bz2 contains that! Jun 24 01:46:41 good morning ! Jun 24 01:46:59 hey mickeyl Jun 24 01:47:06 how was LinuxDay? Jun 24 01:47:52 LinuxTag was very cool. I met nice people, played with a Nokia 770, had some good talks, and the streets were empty enough to finally see how my new car behaves on the Autobahn ;)) Jun 24 01:48:07 hi mickeyl Jun 24 01:48:18 ram123: bk-export is OE Jun 24 01:48:45 mickeyl: closed speedmeter? Jun 24 01:48:59 s/speed/speedo Jun 24 01:49:30 bk-export only contains meta data ?. Jun 24 01:49:36 ram123: yes Jun 24 01:49:49 but why do we require rep ? Jun 24 01:50:02 ? Jun 24 01:50:41 openembedded repository Jun 24 01:50:53 why is it needed ? Jun 24 01:51:16 because that hold the metadata Jun 24 01:51:33 so bk-exports and repository are the same ? Jun 24 01:51:50 yes Jun 24 01:52:10 bk-export is the result of running 'bk export' in the repo Jun 24 01:52:44 the repo which u downloaded using bitkeeper ? Jun 24 01:52:50 yes Jun 24 01:53:04 ie u download repo using bitkeeper and then run bk exports ? Jun 24 01:53:19 hrw|work: heh, not really. When I'm alone I don't driver faster than 180Km/h. Only when I just risk my own live :) Jun 24 01:53:32 s/alone/not alone/ Jun 24 01:53:33 ;) Jun 24 01:53:56 anyway I never drive faster then 140kmh Jun 24 01:54:12 to be honest, that's my speed of choice, as well Jun 24 01:54:30 mickeyl: but your car probably have a power for it Jun 24 01:54:31 ram123: no, we work directly on the downloaded repo Jun 24 01:54:48 ram123: but you were behind a firewall, so I exported it for you Jun 24 01:54:50 hrw|work: yeah, 190PS :) Jun 24 01:55:05 last time I drive 140kmh it was Peugeot 205 with 1.6l - 90PS Jun 24 01:55:45 in many cars, maximum speed is more a question of audio, then of horsepower :) Jun 24 01:55:51 if it's too loud, i don't feel good Jun 24 01:55:56 I hate battery/power management in c7x0 Jun 24 01:56:14 ah btw., on-topic... 2.6.12-mm1 on the C7x0 survived the LinuxTag stress test Jun 24 01:56:20 not one crash Jun 24 01:56:31 mickeyl: you should drive 120kmh in Polish FIAT 125p then ;) Jun 24 01:56:35 and s/r even plays nice with X and wlan Jun 24 01:56:37 hrw|work: eeks ;) Jun 24 01:57:00 mickeyl: at 80 I was shutdown radio - engine was louder Jun 24 01:58:41 try doing 50 on your bike Jun 24 01:58:53 koen: my bike record was 45 Jun 24 01:58:56 heh Jun 24 01:59:48 but no radio.. Jun 24 01:59:51 hrw|work: I cheated by getting pushed by a motorcycle Jun 24 02:01:37 ok i downloaded bitbake and bk-exports untarred them. what should i do now ? Jun 24 02:02:43 bk-exports created a dir called exports and where should i put them if i want the directory structure as mentioned in getting started ? Jun 24 02:03:04 mv exports openembedded Jun 24 02:04:16 i have one more file openembedded-exported-2005-03-01-0700.tar.bz2 what is it the repo ? Jun 24 02:04:18 mickeyl: I got some monotne activity last night, so it seems people are trying it out :) Jun 24 02:05:08 koen: glad to hear. we can assume that it has been decided by now. Jun 24 02:05:11 ram123: openembedded-exported-2005-03-01-0700.tar.bz2 is old one Jun 24 02:05:22 the other guys' had time to complain, but nothing was heard :) Jun 24 02:05:32 so time to install monotone here Jun 24 02:05:48 ya Jun 24 02:05:57 mickeyl: this is in my .bashrc: http://pastebin.ca/15746 Jun 24 02:05:57 ok Jun 24 02:06:01 what should i do now ? Jun 24 02:06:25 i have 2 dirs: bitbake and openembedded ? Jun 24 02:06:45 ram123: and thats good Jun 24 02:07:05 ram123: now read GettingStarted and skip parts which say about obtaining bitbake and OE Jun 24 02:07:16 mornin' Jun 24 02:09:10 hmm.. my debian has /bin/mt which is not included in any package Jun 24 02:09:44 ~lart monotone manual Jun 24 02:09:44 * ibot beats the living hamstercrap out of monotone manual Jun 24 02:09:58 koen: handy Jun 24 02:11:40 you generate one key for all locations or one per each copy of OE? Jun 24 02:11:56 I mean one for work, one for home etc Jun 24 02:13:41 i did not get BBFILES ?. what should i do with it in local.conf ? Jun 24 02:13:57 ram123: its list of recipes Jun 24 02:14:07 BBFILES contains the path to the .bb files Jun 24 02:14:19 ram123: set it to place_with_openembedded/packages/*/*.bb (like in conf/local.conf.sample) Jun 24 02:14:31 koen: should I send pubkey to you? Jun 24 02:14:55 hrw|work: yes please Jun 24 02:15:10 ok Jun 24 02:15:33 Revision e4e062c0df6112a06956f32eaf622fe950c03008: ring us up to date with bk: cset 1.3598.1.1 Jun 24 02:17:08 mickeyl: could you have a look at ciabot_monotone.py? Jun 24 02:17:22 sent Jun 24 02:18:05 fine. Jun 24 02:18:20 koen: ya will take a look Jun 24 02:18:26 i want to build a tool chain and not understanding what to modify in local.conf ? Jun 24 02:18:38 for arm processor ? omap 5912 Jun 24 02:19:11 MACHINE="omap5192osk" Jun 24 02:19:13 ram123: MACHINE=omap5912osk Jun 24 02:19:30 ram123: and DISTRO="generic" (for start) Jun 24 02:20:57 what should i do with preferred providers ? Jun 24 02:21:41 hrw|work: key added Jun 24 02:21:45 ok Jun 24 02:22:23 I wonder why it keeps saying "read failed" when you try to pull Jun 24 02:22:48 I pressed ^c in wrong window Jun 24 02:23:30 anyway it do anonymous pull Jun 24 02:23:53 hmm Jun 24 02:24:06 where the hell has fb_sysmove_buf_aligned gone Jun 24 02:24:08 * mickeyl hunts Jun 24 02:24:53 ~lart bedic Jun 24 02:24:53 * ibot judo chops bedic Jun 24 02:26:15 mickeyl: how to add "-fexceptions" into libbedic? it's "inherit palmtop" recipe Jun 24 02:26:38 what should i do with PREFERRED_PROVIDERS in local.conf. should i modify it ? Jun 24 02:27:42 hrw|work: try EXTRA_QMAKEVARS_POST = "CXXFLAGS += -fexceptions" Jun 24 02:28:02 iirc the last -f[no]exceptions counts Jun 24 02:31:10 monotone-viz is pretty cute: http://dominion.kabel.utwente.nl/koen/pda/files/monotone-viz.tiff Jun 24 02:32:20 nice Jun 24 02:32:28 morning pb Jun 24 02:33:38 NOTE: package libbedic-0.9.6-r0: task do_build: completed Jun 24 02:33:49 now I need to make it work automagically... Jun 24 02:34:20 do_configure_append() { patch makefile which we just generated} :) Jun 24 02:34:26 aa ha! Jun 24 02:34:28 - move_aligned(info, &info->pixmap, dst, pitch, Jun 24 02:34:28 - src, idx, image.height); Jun 24 02:34:28 + fb_pad_aligned_buffer(dst, pitch, src, idx, image.height); Jun 24 02:34:43 2.6.12-mm1 contains some fb revamp patches Jun 24 02:34:48 * mickeyl adjusts elpp Jun 24 02:38:18 ok what commands i should give from build directory to build toolchain and libs ? Jun 24 02:38:46 * koen merges org.nslu2-linux.openembedded into org.openembedded.dev : http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=1caf39c2bf8190a1e997277c24f8421e335cdf9e Jun 24 02:38:51 i want with out docs ; becuase i could not compile jade ? Jun 24 02:39:13 bitbake jade should build jade + deps + toolchain Jun 24 02:39:33 actually I'm lying Jun 24 02:39:35 ram123: "bitbake gcc-cross" should give you toolchain Jun 24 02:39:39 we don't have jade yet Jun 24 02:44:46 hey zecke. finally arrived home? Jun 24 02:45:23 03mickeyl 07 * r1.3603.1.1 10openembedded/packages/linux/ (2 files in 2 dirs): linux-openzaurus-2.6.12-mm1: repair elpp which has been broken by fb patches in -mm1 Jun 24 02:45:54 fbpatches does not broke elpp - elpp is b0rken Jun 24 02:45:57 ;) Jun 24 02:46:17 koen, mickeyl: can we start pushing nslu2-linux changes into the monotone repository? Jun 24 02:46:47 rwhitby: you mean into the nslu2 branches? Jun 24 02:47:03 mickeyl: libbedic is weird.. adding CXXFLAGS+=-fexceptions to second qmake in do_configure or into EXTRA_QMAKEVARS_POST does not work Jun 24 02:47:19 koen, we're already doing that, I mean propagating into the .dev branch Jun 24 02:48:14 hrw|work: push it, i'll take look Jun 24 02:48:28 rwhitby: go ahead Jun 24 02:48:56 should i get glibc headers (mentioned in required software) ? Jun 24 02:51:16 koen: thx. has the decision been made, or are we still in the eleventh hour? Jun 24 02:51:57 it's pretty offical now, but we'll wait another day or so before making it really official Jun 24 02:52:06 ok, cool. Jun 24 02:52:12 quasi-official ? Jun 24 02:52:19 we're still debating how to handle the pushes back to oe from nslu2-linux. Jun 24 02:53:20 we will be pulling from oe hourly (to keep in sync), and may even propagate from oe to nslu2 on each pull. we will _not_ propagate from nslu2 to oe except by explicit manual core developer command. Jun 24 02:53:20 and we have to setup the official tree @ oe.org Jun 24 02:54:33 the question is: do you ever want to have nslu2 branches in your repo, or would be prefer that you never see any org.nslu2-linux branches at all, and any transfer between oe and nslu2 is done by the nslu2 core team developers? Jun 24 02:54:56 (and then those core team developers push back to your .dev branch) Jun 24 02:55:06 when i do bitbake gcc-cross i get the following error: Jun 24 02:55:13 ram123: use pastebin.ca Jun 24 02:55:23 ~pastebin Jun 24 02:55:23 hmm... pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.ca Jun 24 02:56:03 03hrw 07 * r1.3604.1.1 10openembedded/packages/zbedic/libbedic_0.9.6.bb: added libbedic 0.9.6 Jun 24 02:56:07 http://pastebin.ca/15748 Jun 24 02:56:19 rwhitby: i'd like to have it the first way Jun 24 02:56:46 mickeyl: look at it if you can Jun 24 02:56:49 rwhitby: i'd like to regularly sync with your tree and to give your core devels direct access to the main tree Jun 24 02:56:58 hrw|work: will od Jun 24 02:57:01 do, even Jun 24 02:57:37 ram123: you need "export PYTHONPATH=place_with_bitbake/lib/bb" or sth like that probably Jun 24 02:59:22 mickeyl: can I ask why you want access to the nslu2-linux branches (we're happy to do that, but I'm not sure we're on the same page yet - I may not have explained our proposal sufficiently) ? Jun 24 02:59:36 agreed on the core devs access to main tree Jun 24 03:00:05 i am getting the same error ? Jun 24 03:00:11 we were expecting that the nslu2 core devs would push to the main oe tree at stable points in nslu2 development (but still fairly regularly) Jun 24 03:00:22 export PYTHONPATH=/usr/local/bitbake/lib/python2.3/site-packages/bb Jun 24 03:00:29 for files that are used by other oe distros, we would push those asap. Jun 24 03:00:46 for files that are only used by nslu2, we would push those after a stable build Jun 24 03:01:08 rwhitby: ah right. sorry, i confused that. Jun 24 03:01:14 rwhitby: I sync mickeyl mixed up 'sync' and 'pull' Jun 24 03:01:40 yep Jun 24 03:01:59 I'm ok with that way of working Jun 24 03:02:12 ok, so the oe repo will serve up org.openembedded, and only contain org.openembedded branches. Jun 24 03:02:40 the nslu2-linux repo will serve up org.nslu2-linux branches, but will also contain org.openembedded branches (by pulling them from the oe repo directly). Jun 24 03:03:15 nslu2-linux project members will usually only access the org.nslu2-linux branches. Jun 24 03:03:44 I suspect monotone can 'protect' branches Jun 24 03:03:44 nslu2-linux core developers will access both sets of branches, to be able to propagate nslu2 changes back to the main oe branch in a controlled manner. Jun 24 03:03:53 koen: yes, in the monotonerc Jun 24 03:04:08 our core developers are the only ones allowed to have @nslu2-linux.org id's for the keys. Jun 24 03:04:41 03hrw 07 * r1.3607 10openembedded/packages/zbedic/ (zbedic_0.9.6.bb zbedic-0.9.6/qtopia17.patch): added zbedic 0.9.6 Jun 24 03:04:45 mickeyl: when you fix libbedic 0.9.6 then please remove 0.9.4 Jun 24 03:05:04 the reason why I prefer that org.nslu2-linux branches don't appear in the oe repo is because there is other stuff in there which is unrelated to OpenEmbedded. Jun 24 03:05:32 so what is policy for keys? oedev@openembedded.org? Jun 24 03:05:54 and since your oe repo is only serving up org.openembedded, then when we push back to it, it will only accept the org.openembedded branches, and will ignore the org.nslu2-linux branches. Jun 24 03:06:29 *nod* Jun 24 03:06:43 mickeyl: and of course we will give access for oe core developers to the nslu2-linux repo, in case you ever need to access it for any reason. Jun 24 03:07:00 (that's write access) Jun 24 03:07:42 so if you have a convention for identifying core oe devs, then that will make that access control easier (by regexp, rather than by listing individual keys). Jun 24 03:09:47 right. we'll think about that. Jun 24 03:15:02 rwhitby: Out of curiosity, the makefile you've created.. Does it work yet? I tried to make, and ended up with monotone screaming misuse ;) Jun 24 03:17:17 NAiL: http://www.nslu2-linux.org/Makefile should be working now. Jun 24 03:17:39 mkdir /home/slug ; cd /home/slug ; wget http://www.nslu2-linux.org/Makefile ; make Jun 24 03:17:49 that's our complete build instructions now :-) Jun 24 03:18:14 hrw|work: does telnet work yet in wrt ? Jun 24 03:18:15 it has been updated since you last used it Jun 24 03:18:51 rwhitby: something like this? Jun 24 03:19:18 mickeyl: openwrt has it by default but it can be removed in their config Jun 24 03:19:30 mickeyl: didnt checked OE yet Jun 24 03:19:35 rwhitby: http://dominion.kabel.utwente.nl/koen/pda/files/montone-OE.pdf Jun 24 03:20:16 i am getting the same error ?. what is the solution ? Jun 24 03:21:11 rwhitby: Do I have to clean out the dir if I used the old makefile? I still get the same error Jun 24 03:21:14 monotone: misuse: branch org.nslu2-linux.openembedded is empty Jun 24 03:21:47 hrw|work: zbedic done Jun 24 03:21:53 mickeyl: cool Jun 24 03:22:07 03mickeyl 07 * r1.3608 10openembedded/ (3 files in 2 dirs): fix libbedic/zbedic 0.9.6 and remove 0.9.4 Jun 24 03:22:08 koen, I think you need a "read" line to from org.nslu2-linux.* to Joe nslu2 Developer. Jun 24 03:22:25 NAiL: yeah, you need to start from scratch, or do the sync with koen's repo (which is in the setup target in the Makefile now) Jun 24 03:22:26 please refer my error if i run bitbake ::http://pastebin.ca/15750 Jun 24 03:22:49 koen: nice diagram. yeah - an extra read line as dyoung said Jun 24 03:22:49 because there is some other stuff in org.nslu2-linux that only is only of interest to nslu2-linux builder. Jun 24 03:23:14 But other than that, thats a nice drawing. :-) Jun 24 03:24:00 koen: BTW, we specifically put the repo on mtn.nslu2-linux.org instead of just nslu2-linux.org so that we could move it to a different machine later if necessary. Jun 24 03:24:07 dyoung: if someone can write then can read also imho Jun 24 03:24:38 hrw|work: person@nslu2-linux.org is a core developer. Joe User doesn't have an @nslu2-linux.org address. Jun 24 03:24:48 ah.. Jun 24 03:25:38 we're restricting access to @nslu2-linux.org addresses, so that if you ever see an email or key with that, then you know you're talking to someone who has the authority to represent the nslu2-linux project. Jun 24 03:26:07 (there are 10 such people) Jun 24 03:26:25 ok Jun 24 03:45:30 what we will do is add other developer keys to our nslu2-linux repo (so that they can write to the org.nslu2-linux.* branches), but will restrict write access to the org.openembedded.* branches to the core team only. Jun 24 03:45:48 updated http://dominion.kabel.utwente.nl/koen/pda/files/montone-OE.pdf Jun 24 03:47:43 that drawing looks perfect to me. Jun 24 03:48:20 (I'm assuming that person@openembedded.org is a oe core dev) Jun 24 03:48:26 yes Jun 24 03:49:14 Yep, thats wonderful! Jun 24 03:49:29 hmm.. non-core-devs on nslu2 can't write to the nslu2 repo? Jun 24 03:51:51 good point. Jun 24 03:51:58 rwhitby: do you have a sample of your monotonerc? Jun 24 04:01:06 koen: http://pastebin.ca/15753 Jun 24 04:01:48 I haven't done the collection-based access control yet. Jun 24 04:02:22 @handhelds.org? not @openembedded.org? Jun 24 04:03:34 hrw|work: those were the keys that koen and mickeyl sent me Jun 24 04:03:59 aah Jun 24 04:04:22 does Joe User get @handhelds.org addresses, or are those trusted? Jun 24 04:05:33 koen: the non-core developers should be able to write to their own project's repo, but not the other one. Jun 24 04:06:49 we don't have a core/noncore split Jun 24 04:07:30 koen: does opie-image builds for h3600? Jun 24 04:07:35 rwhitby: @handhelds.org addresses only signify a minimal level of trust. We don't do much vetting before handing them out. Jun 24 04:07:58 hrw|work: haven;t tested it since last week Jun 24 04:08:11 koen: but it was working then? Jun 24 04:08:16 koen: do you have one too many d's in your new key? Jun 24 04:08:30 koen@openembeddded.org Jun 24 04:08:51 aaargh Jun 24 04:09:18 koen: the only reason why we have a core/non-core split is to ensure that only experienced core developers have direct write access to your main oe repo. Jun 24 04:10:06 we appreciate the bootstrap which openembedded has given us, and never want to get into a situation where an inexperienced nslu2-linux developer stuffs up the main oe repo Jun 24 04:12:46 pb_: thx - do @openembedded.org addresses get vetted? Jun 24 04:13:46 I guess so, but I'm not sure. I don't have an @openembedded.org address myself. Jun 24 04:14:26 rwhitby: @openembedded.org keys get trusted Jun 24 04:16:21 ok, I'm changing the access to @openembedded, now that koen's key is in. Jun 24 04:16:36 mickey|lunch: feel free to send a new @openembedded.org key Jun 24 04:33:49 please refer my error if i run bitbake ::http://pastebin.ca/15750 Jun 24 04:33:56 unable to run bitbake ? Jun 24 04:34:14 http://pastebin.ca/15750 Jun 24 04:34:29 ram123: can you not prepend http with ::? Jun 24 04:34:56 ram123: you instaled bitbake? Jun 24 04:35:01 yes Jun 24 04:35:26 it is not needed to get it working (atleast I work like that) Jun 24 04:36:14 i am getting bitbake error; please refer the url Jun 24 04:36:37 my environment paths are also ok. Jun 24 04:37:26 I saw that Jun 24 04:37:37 anyway never had that Jun 24 04:37:59 tried removing "bb/" from PYTHONPATH? Jun 24 04:38:20 i tried removing it. but the result is the same. Jun 24 04:41:17 i have posted my local.conf file for your reference http://pastebin.ca/15756 Jun 24 04:41:45 what checks should i carry out on my system to analyse the problem Jun 24 04:41:48 do not use ${HOME} but full path Jun 24 04:41:55 # NOTE: Do NOT use $HOME in your paths, BitBake does NOT expand ~ for you. Jun 24 04:42:03 read comments Jun 24 04:43:13 and I propose disabling TARGET_ARCH (it is set by machine conf), KERNEL_VERSION Jun 24 04:43:47 and unset TMPDIR ot set it somewhere other then /tmp/ to get it existing after reboot Jun 24 04:44:29 done Jun 24 04:44:37 but getting the same error ? Jun 24 04:44:52 replaced $HOME with full path? Jun 24 04:44:56 yes Jun 24 04:45:54 what is this:import bb Jun 24 04:45:55 ImportError: No module named bb Jun 24 04:46:12 its python thing (and I'm not familiar with python) Jun 24 04:46:21 it is not related to local.conf file. some other issue i feel. Jun 24 04:46:58 yep Jun 24 04:47:28 what is PREFERRED_PROVIDERS in local.conf i have not modified it from the original local.conf ? Jun 24 04:49:27 if there are multiple providers for on package (e.g. grep) you can select which one you want Jun 24 04:50:11 i did not undertsand this PREFERRED_PROVIDERS ?. what is it for ? Jun 24 04:51:11 we have multiple packages providing a linux kernel, so you can set a PREFERED_PROVIDER_virtual/kernel=linux-somebranch Jun 24 04:54:41 ok if i want kernel 2.6.11 what should i set ? Jun 24 04:55:11 ram123: you have set kernel provider in omap5912osk.conf Jun 24 04:55:14 see conf/machine/omap5192osk.conf Jun 24 04:55:46 off-topic: do you know a channel where I could ask gdb questions? Jun 24 04:56:07 bitbake is working fine. i changed the PATH and PYTHONPATH variables to point to my uninstalled bitbake and it worked. Jun 24 04:56:33 ram123: things like which kernel to use, whch kernel modules, extra apps to install are listed in MACHINE conf - conf/machine/omap5192osk.conf for your build Jun 24 04:59:08 but what about make menuconfig ? Jun 24 04:59:59 does bitbake do everything by itself ? Jun 24 05:00:29 it does Jun 24 05:00:30 i have a kernel and want to populate the file system only ? Jun 24 05:00:32 ram123: in openembedded/packages/linux/linux-omap(or how that is called)-version/defconfig you have config Jun 24 05:00:35 it will use packages/linux/linux-omap1-2.6.12-rc2/omap5912osk/defconfig Jun 24 05:00:44 how do i proceed ? Jun 24 05:01:08 ram123: "bitbake bootstrap-image" for start probably Jun 24 05:02:16 great Jun 24 05:02:36 i ma building it from scratch. toolchain onwards. Jun 24 05:02:51 would like to know what commands i should give. Jun 24 05:03:11 i just gave bitbake gcc-cross Jun 24 05:03:16 ram123: "bitbake bootstrap-image" is enough - will build all what is needed Jun 24 05:03:41 lunch time... Jun 24 05:03:50 what all will it build ? Jun 24 05:04:16 do "bitbake -n WHAT_YOU_NEED_TO_BUILD" and see Jun 24 05:04:27 bbl Jun 24 05:07:39 ping 192.216.230.225 not working. is oesources up ? Jun 24 05:09:14 it is over here Jun 24 05:09:24 Connecting to www.oesources.org[192.216.230.225]:80... failed: Connection timed out. Jun 24 05:09:33 thats why i am asking ? Jun 24 05:10:21 it works over here Jun 24 05:12:06 andit does seem to block pings Jun 24 05:13:46 i thought so. but why am i getting connection timed out ? Jun 24 05:14:49 can i manually download the file ? Jun 24 05:15:01 and place it in which directory ? Jun 24 05:16:47 I can talk to http://www.oesources.org from USA Jun 24 05:17:33 ram123, when you get OE going, I can talk to you about the OSK specific parts Jun 24 05:17:40 I am familiar with that work Jun 24 05:18:17 i know. i met u in omap chat forum #OL Jun 24 05:18:24 yeah Jun 24 05:18:26 and u helped me Jun 24 05:18:27 I know :) Jun 24 05:18:42 I will be better at the OSK stuff in OE Jun 24 05:18:42 early to office today ? Jun 24 05:18:48 still at home Jun 24 05:19:11 I need to get over early, it has not beeen a productive week Jun 24 05:20:34 if i run bitbake again why is it starting all over gain ?. it has already build up the cache. Jun 24 05:20:39 how to disable it. Jun 24 05:21:59 Crofton: trying to get GPE onto osk Jun 24 05:22:13 hey Jun 24 05:23:05 hey zecke_ Jun 24 05:23:39 ~lart Deutsche Bahn... Jun 24 05:23:39 * ibot drops a truckload of VAXen on Deutsche Bahn... Jun 24 05:24:08 ram123: it has to initialze the cache everytime, but it won't rebuild the toolchain Jun 24 05:31:23 03mickeyl 07 * r1.3609 10openembedded/packages/ (3 files in 2 dirs): ship opie-pcmciaapplet in the default image Jun 24 05:31:44 koen: have you considered how to advise CIA of monotone commits? Jun 24 05:32:06 i manually downloaded quilt-native-0.39-r0] and ran bitbake gcc-cross and am getting the following error ? Jun 24 05:32:36 http://pastebin.ca/15757 Jun 24 05:33:05 mickeyl: pcmciaapplet instead of cardmon one? Jun 24 05:33:18 hrw|work: yeah Jun 24 05:33:32 ram123: looks weird Jun 24 05:33:41 mickeyl: so you broke opie-filemanager and libopie2 Jun 24 05:33:56 ~hail mickeyl Jun 24 05:33:56 * ibot bows down to mickeyl and chants, "I'M NOT WORTHY!!" Jun 24 05:34:08 ~lart us for not shipping stock items Jun 24 05:34:08 * ibot beats us severely about the head and shoulders with a rubber chicken for not shipping stock items Jun 24 05:34:21 mickeyl: http://opie-bugs.oszine.de/view.php?id=1514 Jun 24 05:34:41 rwhitby: I sent mickeyl after ciabot_monotone.py Jun 24 05:34:58 ram123: you still have TMPDIR set to /tmp/something/? Jun 24 05:35:37 koen: can you email me the setup once you sort it out? Jun 24 05:36:14 rwhitby: sure Jun 24 05:36:23 thx Jun 24 05:36:34 heh, monotone didn't like me dropping my old private keys Jun 24 05:36:44 * koen makes a note on that Jun 24 05:37:18 hrw|work: is it only pcmcia.png or more ? Jun 24 05:37:28 mickeyl: pcmcia.png Jun 24 05:37:35 hrw|work: excellent. I'll fix that now Jun 24 05:38:32 YES... finaly Jun 24 05:38:51 mickeyl: but PLEASE do not do s/cardmon/pcmcia/ Jun 24 05:39:08 mickeyl: it NEEd to work without pcmciaapplet and without cardmon Jun 24 05:39:26 so pcmcia/pcmcia.png (or how it will be called) shuld be in opie-pics pacakge Jun 24 05:42:18 since when am I called "Mr. easy solution" ? Jun 24 05:42:32 i have always been called "Mr fix-it-the-right-way" Jun 24 05:42:35 heh Jun 24 05:42:53 mickeyl: I just want to bump "the right way" parameter Jun 24 05:44:27 jeeze Jun 24 05:44:31 it's really used all over the place Jun 24 05:44:35 * mickeyl shakes head Jun 24 05:44:42 * mickeyl fixes Jun 24 05:45:09 looks like anotherbug reported by me will be fixed Jun 24 05:46:19 tadaa Jun 24 05:46:20 it's done Jun 24 05:47:00 can we remove some DEPENDS opie-cardmon now? Jun 24 05:49:15 mickeyl: this show a nice ancestry diagram: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=26dcbba401e23cdd807b2c6017ed788ce8ae6690e Jun 24 05:49:26 shows* Jun 24 05:49:59 hmm revision id invalid Jun 24 05:50:39 ok, try http://ewi546.ewi.utwente.nl/tmp/viewmtn/headofbranch.psp?branch=org.openembedded.dev Jun 24 05:51:13 woot Jun 24 05:51:15 pretty cool Jun 24 05:51:27 the site looks lame ;) Jun 24 05:51:33 the boxes are not equally sized Jun 24 05:52:48 koen: can you disable "manifest" part? it will be faster then Jun 24 05:53:18 hmm Jun 24 05:53:19 click on a diff Jun 24 05:53:21 traceback Jun 24 05:54:09 which diff? Jun 24 05:54:31 anyone Jun 24 05:54:46 http://ewi546.ewi.utwente.nl/tmp/viewmtn/getdiff.py?id1=1caf39c2bf8190a1e997277c24f8421e335cdf9e&id2=26dcbba401e23cdd807b2c6017ed788ce8ae6690 WFM Jun 24 05:54:52 Exception: Unable to calculate diff: monotone: misuse: unknown path 'packages/linux/linux-openzaurus_2.6.12-mm1.bb' Jun 24 05:54:52 http://ewi546.ewi.utwente.nl/tmp/viewmtn/getdiff.py?id1=1caf39c2bf8190a1e997277c24f8421e335cdf9e&id2=26dcbba401e23cdd807b2c6017ed788ce8ae6690 Jun 24 05:55:07 http://ewi546.ewi.utwente.nl/tmp/viewmtn/getdiff.py?id1=a18267d52d3d7431e7643dafe74393852f1be261&id2=26dcbba401e23cdd807b2c6017ed788ce8ae6690&fname=packages/linux/linux-openzaurus-2.6.12-mm1/add-elpp-stuff.patch does not Jun 24 05:55:12 odd Jun 24 05:55:14 doesn't work for me Jun 24 05:56:04 dif for files does not work Jun 24 05:56:08 work for revisions Jun 24 05:56:36 * koen blames python Jun 24 05:58:03 * mickeyl wanders home Jun 24 05:58:45 oooh Jun 24 05:58:50 nokia770 on ZDF Jun 24 05:58:55 http://www.heute.de/ZDFmediathek/inhalt/22/0,4070,2326614-6-wm_dsl,00.html Jun 24 05:58:59 goooo germans! Jun 24 06:01:54 all I get is a ubuntu video ,,, Jun 24 06:02:08 the last 10 seconds show the 770 Jun 24 06:02:12 ah Jun 24 06:03:36 ~hail mickeyl for closing 1514 Jun 24 06:03:36 * ibot bows down to mickeyl for closing 1514 and chants, "I'M NOT WORTHY!!" Jun 24 06:11:52 i am getting the error : quilt_savannah.nongnu.org_VER_0_39_.tar.gz Jun 24 06:12:18 | patch: **** Can't create file /tmp/poX2r0fG : File exists Jun 24 06:12:40 but when i do a ls -al i cannot find that file Jun 24 06:12:48 i dont understand what is going on ? Jun 24 06:12:56 can anyone explain this ? Jun 24 06:22:18 * rwhitby does the happy dance Jun 24 06:22:39 nslu2-linux just raised USD$1800 of donations in 8 days. Jun 24 06:23:25 Now we can buy a real machine to host the monotone repo for nslu2-linux.org on :-) Jun 24 06:23:39 nice donation Jun 24 06:24:25 hey jamey Jun 24 06:24:52 mickeyl: so we have enough cash to buy your slug Jun 24 06:24:52 hi zecke_ Jun 24 06:26:39 hey jamey Jun 24 06:26:59 rwhitby: sweet, 64bit goodness Jun 24 06:27:21 i am always getting the following error Connecting to www.oesources.org[192.216.230.225]:80... failed: Connection timed out. Jun 24 06:27:29 dont know what to do ? Jun 24 06:27:47 rwhitby: congrats Jun 24 06:27:47 ram123: you have to use proxy at work? Jun 24 06:27:51 i can download using web browser but not through bitblake ? Jun 24 06:27:56 i have a proxy ? Jun 24 06:28:20 cant figure out what's happening ?. any solutions ? Jun 24 06:28:44 ram123, do you have a network administrator there you can talk too? Jun 24 06:29:15 The people here can only be so much help at solving your local networking issues Jun 24 06:30:03 what settings does bitbake require ? Jun 24 06:30:17 rwhitby: do you know how well a nslu2 perform with a usb 100mbit nic? Jun 24 06:30:43 koen: dunno, but [g2] might. Jun 24 06:30:57 ram123: it need working wget or curl Jun 24 06:31:18 ram123: when "wget http://www.example.net/" will work then bitbake will be able to fetch Jun 24 06:31:49 1 sec Jun 24 06:31:49 koen: http://www.nslu2-linux.org/wiki/Info/Performance is all our benchmark numbers that are recorded Jun 24 06:32:21 cool, thanks Jun 24 06:32:30 <[g2]> koen, what rates are you looking to do and what protocols Jun 24 06:32:37 wget does not work ?. but browser works ? Jun 24 06:32:53 do u know what settings are needed to make wget work ? Jun 24 06:32:55 [g2]: 2mbit or so as a router Jun 24 06:33:10 ram123, what broswer are you using? Jun 24 06:33:11 ram123: man wget Jun 24 06:33:15 [g2]: I'm considering getting a slug as a sipproxy and firewall Jun 24 06:33:43 <[g2]> koen, assuming you're not to crazy with iptables/tc should be no problem at all Jun 24 06:33:59 export http_proxy=http://localhost:80 Jun 24 06:34:00 ok cool Jun 24 06:34:09 <[g2]> I've run 100Mbs via thttp Jun 24 06:34:33 <[g2]> and tcp has 70+Mbs throughput on the device itself Jun 24 06:35:01 <[g2]> i've had 2 usb-ethernets plugged into to do transparent bridging/filtering but didn't get to play with it Jun 24 06:35:17 <[g2]> then 3rd would be the mgmt intf Jun 24 06:35:55 <[g2]> I think I can do line rate logging on <10Mbs Jun 24 06:36:11 <[g2]> it may require a little tweeaking though Jun 24 06:37:59 it looks like my ipaq makes mincemeat of the slug Jun 24 06:38:42 mozilla browser Jun 24 06:38:47 crofton: mozilla Jun 24 06:39:09 look in preferences and see if you have a proxy configured Jun 24 06:39:18 I don't know exactly where that will be in the menus Jun 24 06:39:18 i am going through .wgetrc to make wget work ? Jun 24 06:39:26 i know Jun 24 06:39:38 1 sec Crofton Jun 24 06:42:24 got it to work finally Jun 24 06:42:37 great! Jun 24 06:42:39 now the proxy works but get an error Jun 24 06:42:59 ram123: what url exactly? Jun 24 06:43:11 mickeyl: any guestimate when we'll have the flagday for switching? Jun 24 06:43:56 http://pastebin.ca/15760 Jun 24 06:44:14 please fine the error at http://pastebin.ca/15760 Jun 24 06:44:38 URL used is http://www.oesources.org/source/current/quilt_savannah.nongnu.org_VER_0_39_.tar.gz Jun 24 06:44:42 ram123: what you have as TMPDIR? Jun 24 06:44:46 it is the very first file Jun 24 06:44:49 /usr/tmp Jun 24 06:44:56 weird Jun 24 06:45:05 why ? Jun 24 06:45:12 remove the file it is complaining about and try again Jun 24 06:45:37 /tmp/poS Jun 24 06:45:40 ... Jun 24 06:46:17 i know that but dont know why it is comming ?. Jun 24 06:46:32 i changed the TMPDIR=/usr/tmp/oetmp Jun 24 06:46:37 no ideas ? Jun 24 06:46:49 no Jun 24 06:47:00 I never seen OE fail that way Jun 24 06:47:13 i dont have a file called /tmp/po*** whatever when i do ls -l Jun 24 06:47:19 ls -l /tmp Jun 24 06:49:10 see http://pastebin.ca/15763 for my conf file Jun 24 06:55:12 ram123, look at the log files in /usr/tmp/oetmp/work/quilt..../temp Jun 24 06:59:35 patch: **** Can't create file /tmp/poX2r0fG : File exists Jun 24 06:59:46 i get above line in the files Jun 24 07:02:30 is tmp writable? Jun 24 07:03:22 what version of bitbake Jun 24 07:04:56 03koen 07 * r1.3604.2.1 10openembedded/packages/gpe-login/gpe-login_0.77.bb: Add gpe-login 0.77, I blindly removed all patches that failed to apply, but it builds and works for me. Jun 24 07:05:00 Cwiiis[away]: could you check if your collie stuff is still needed in gpe-login? Jun 24 07:06:01 ram123, pastebin log.do_patch .... Jun 24 07:13:15 1 sec Jun 24 07:14:49 i deleted the direcory /usr/oetmp and doing all over again. Jun 24 07:15:10 http://pastebin.ca/15767 Jun 24 07:16:51 i will also paste run do patch Jun 24 07:16:59 forget quilt Jun 24 07:17:19 echo 'INHERIT += "patcher"' >> conf/local.conf Jun 24 07:17:23 then be happy Jun 24 07:18:06 heh. you really don't like quilt, do you? Jun 24 07:18:27 not really. especially i hate it for generating lots of oe-related traffic Jun 24 07:18:33 unnecessary traffic imo Jun 24 07:19:02 if those who were for converting us to use quilt by default would maintain the .bb it would be fine Jun 24 07:19:03 admittedly I don't like quilt much either, but I haven't found patcher to be significantly better. Jun 24 07:19:13 well, patcher builds. that counts for me :) Jun 24 07:19:26 ah, right, I guess that would help Jun 24 07:19:37 'building' is a big for for patcher Jun 24 07:19:47 koen: that's why i like it Jun 24 07:19:49 heh Jun 24 07:19:58 and because i like schurig Jun 24 07:19:58 see http://pastebin.ca/15768 Jun 24 07:20:00 but that's another sotry Jun 24 07:20:02 story, even Jun 24 07:20:13 Crofton: what could be wrong ? Jun 24 07:22:12 Koen: can u have a look into my problem if u are free ? Jun 24 07:22:21 rwhitby-asleep: http://www.nslu2-linux.org/wiki/OpenEmbedded/MinimisingBitBakeMemoryUtilisation is not really needed anymore :) Jun 24 07:22:51 ibot: convert 80 EUR to USD Jun 24 07:23:42 foo.org/MinimizingMoneyInYourPocket Jun 24 07:23:53 mickeyl: 'change' Jun 24 07:23:56 ah right Jun 24 07:23:59 thanks koen Jun 24 07:24:05 ~change 80 EUR to USD Jun 24 07:24:08 Content: Get me an ibook Jun 24 07:24:16 80.00 Euro (EUR) makes 96.7275 United States Dollar (USD) (from http://www.xe.com/) Jun 24 07:24:22 here we are Jun 24 07:24:24 ~botsnack Jun 24 07:24:24 :), mickeyl Jun 24 07:24:38 ~users Jun 24 07:24:38 so this is what we get for sharing our unpaid volunteer work with you? complaints and accuses? well done, this clearly supports our motivation to continue working on open source projects. Jun 24 07:25:11 I always have to do that after reading the whining on oesf.org Jun 24 07:25:54 ram123, try what mickey said in your local.conf Jun 24 07:27:30 moin Jun 24 07:28:25 trying echo 'INHERIT += "patcher"' >> conf/local.conf Jun 24 07:28:30 1 sec Jun 24 07:36:41 mickeyl: you've been added to the end of the list: http://www.nslu2-linux.org/wiki/Info/DonationList Jun 24 07:37:19 rwhitby-asleep: excellent. BTW., I just found a shop which has 'em in stock. I'll be getting it tomorrow :) Jun 24 07:37:35 * mickeyl OpenSlug user soon Jun 24 07:38:23 Crofton have u ported GPE to osk Jun 24 07:39:19 I haven't built opie, I haven't tried gpe recently Jun 24 07:39:40 did you get OE from bitkeeper or montone? Jun 24 07:40:14 Crofton|laptop: a bk export Jun 24 07:40:18 koen gave me the link from where i could download the tar files as my system Jun 24 07:40:27 is behind the firewall Jun 24 07:40:29 ok, just curious Jun 24 07:40:40 thanks for that koen Jun 24 07:40:41 mickeyl: you'll be part of our "niche" project now ;-) Jun 24 07:41:01 *cough* ;)) Jun 24 07:41:24 * mickeyl denies having said that ever before Jun 24 07:41:31 ~lart ibot logs Jun 24 07:41:31 * ibot blasts ibot logs with a huge firehose then strangles ibot logs with it Jun 24 07:41:34 now we'll get a client/server wellenreiter Jun 24 07:41:50 since the slug doesn't have a screen ;) Jun 24 07:42:18 koen: use the 1st version of wellenreiter II Jun 24 07:42:22 hehe Jun 24 07:42:25 no kidding Jun 24 07:42:31 the version from hell Jun 24 07:42:34 how long does bitbake boot-strap image takes ? Jun 24 07:42:36 ~lart Wellenreiter 1 people Jun 24 07:42:36 * ibot smacks Wellenreiter 1 people up side the head with a clue-by-4 Jun 24 07:42:45 how fast is your machine? Jun 24 07:42:54 night all Jun 24 07:42:55 about 4 hours on my machine Jun 24 07:42:59 g'night rwhitby-asleep Jun 24 07:42:59 'night rwhitby-asleep Jun 24 07:43:25 ram123, I am running bitbake gpe-image on my machine now to get an idea if there are any problems Jun 24 07:43:38 great Crofton Jun 24 07:44:12 Now I need to get ready to go to school Jun 24 07:46:54 before u do that have a look at my new error. please ... Jun 24 07:47:35 http://pastebin.ca/15769 Jun 24 07:48:03 i dont know how many more errors are going to be there before i get my full image ? Jun 24 07:49:12 it looks like it can't fetch gnu-config from the server Jun 24 07:49:16 afternoon Jun 24 07:49:32 so how do i solve this now ? Jun 24 07:49:54 Ciao all Jun 24 07:51:03 mickeyl ping Jun 24 07:52:01 ram123, try bitbake gnu-config-native Jun 24 07:52:22 It may be that the gnu CVS server is not responding, or cvs can't get through your firewall Jun 24 07:52:37 Try retreiving it from cvs by hand Jun 24 07:52:48 to test your network connectivity Jun 24 07:53:07 and talk to your network administrator Jun 24 07:53:26 Crofton|laptop: it's using cvs over ssh, so port 22 needs to be open too Jun 24 07:55:26 oh Jun 24 07:55:50 network admins are very fussy people when it comes to opening ports ? Jun 24 07:55:55 ram123: set CVSDATE=20050623 and retry Jun 24 07:56:05 ram123: and well they should be Jun 24 07:56:21 ram123, cvs over ssh use only 22 Jun 24 07:56:30 port 22, even :) Jun 24 07:56:56 well network admins are pains in the behinds, but they should help you get your job done, not prevent it from getting done Jun 24 07:57:13 * Crofton|laptop hates fussy network admins Jun 24 07:57:22 it depends on what kind of a netwrokd admin u have ? Jun 24 07:58:35 network ( or firewall ) admin doesn't care too much about your work :) Jun 24 07:58:39 pigi: at your service Jun 24 07:58:51 not so much, mickeyl :) Jun 24 07:59:01 have you seen the bug I filed for libsdl-qpe ? Jun 24 07:59:01 mickeyl: Bad news. I'm giving up with this rotation approximation. Jun 24 07:59:28 mickeyl: It's really more complicated than I thought at a first sight. Jun 24 08:00:13 does my environment variable consist of CVSROOT ? Jun 24 08:00:18 i am not sure ? Jun 24 08:01:00 ram123: that way you will grab it from CVS_TARBALL_STASH Jun 24 08:02:54 Pigi: not yet, let me check Jun 24 08:03:22 SirFred: guess we could hack up some kind of cheating, i.e. use the transformed when rotating Jun 24 08:03:39 mickeyl: I'm working in a transformed approximation now. Jun 24 08:03:53 mickeyl: The two great problems with the other approach were: Jun 24 08:04:07 mickeyl, #117 Jun 24 08:04:08 1.-A lot of APIs needed to be changed, and I was not sure to be following the right way. Jun 24 08:04:40 2.-The worst thing is that on a rotation, all the gfx needed to be recalculated, physically rotated, and all the cliprects for any gfx updated the same way. Jun 24 08:04:54 I didn't find an API to perform this. So more changes to the qte APIs. Jun 24 08:05:17 Sure it's the best way, because the rotation for any gfx is only performed once. Jun 24 08:05:21 so what do i do now ? Jun 24 08:05:23 But it seems too complicated. Jun 24 08:05:40 i cant even download it. it says no such file Jun 24 08:05:45 you are still having trouble retrieving CVS? Jun 24 08:05:48 Pigi: hmm... no idea offhand. since mithro added some patches i didn't look into sdl-qpe for quite a while. i'll keep that in mind Jun 24 08:06:00 yes Jun 24 08:06:01 SirFred: bummer. if only we could question sharp how they did it Jun 24 08:06:13 ~lart network admins Jun 24 08:06:13 * ibot cuts off network admins's head with a halberd that could have been a little bit sharper Jun 24 08:06:27 ram123, see above Jun 24 08:06:53 to use OE effectively you will need to solve the CVS problem Jun 24 08:06:57 that will be done once it works ? Jun 24 08:07:08 mickeyl: I think they were using a transformed approach. Jun 24 08:07:19 it should get you to your next problem :) Jun 24 08:07:33 yes i know. Jun 24 08:07:42 mickeyl: The filling was different. When the framebuffer is physically rotated, it's more noisy visually. Jun 24 08:07:51 oe will download several source files via cvs Jun 24 08:08:05 mickeyl: It seems that something really complicated is happening internally. Jun 24 08:08:25 i thought installation will be smooth. Jun 24 08:08:28 mickeyl: But with the transformed approach, it's, how to say it, smooth. Jun 24 08:08:35 it is getting more complicated. Jun 24 08:08:48 a simple how-to is needed. Jun 24 08:09:06 write one on the OE wiki :) Jun 24 08:09:08 which says these should be there and this is how u do it Jun 24 08:09:19 i will do that once i am through. Jun 24 08:09:23 there are plenty of howtos. your problem is your environment. it isnt our job to teach you how to fix your system to be able to download things Jun 24 08:09:27 mickeyl: I don't think they changed qte that hard. Who wrote the qte driver, sharp or trolltech? Jun 24 08:09:31 but yes, by all means document Jun 24 08:09:43 goooood morning kergoth Jun 24 08:09:48 SirFred: Sharp did Jun 24 08:09:49 hey Jun 24 08:09:50 mickeyl, ok, no problem Jun 24 08:10:00 SirFred: or ATI for that matter Jun 24 08:10:06 hi kergoth Jun 24 08:10:19 mickeyl: Take in mind that the applications compiled against the standard qte api work. Jun 24 08:10:30 hi kergoth Jun 24 08:10:42 mickeyl: On the qte sharp. so , the changes coulnd't have involved deep changes in APIs. Jun 24 08:11:13 SirFred: not in the APIs, but god knows what extra calls to libaticore they embedded in their version of qte Jun 24 08:11:20 ~lart the price of oil Jun 24 08:11:20 * ibot gives the price of oil a good seeing to Jun 24 08:11:26 SirFred: how good are you with recompiling? :) Jun 24 08:11:45 mickeyl: That's not the problem. The problem is that the widgets need to know about these changes. Jun 24 08:12:11 mickeyl: To update their cliprects geometry and also rotate all the gfx. Jun 24 08:12:27 mickeyl: What do you mean with that? Jun 24 08:12:30 pb_: ping Jun 24 08:12:46 somebody said my name? Jun 24 08:12:47 SirFred: in theory you could have a look at Sharps accellerated qte driver. Jun 24 08:12:57 or where they after mickeyl like normal :P Jun 24 08:13:06 hah Jun 24 08:13:14 mithro, yes :) Jun 24 08:13:16 mithro: check #117 in the oe bugtracker. maybe you have an idea Jun 24 08:13:42 mickeyl: You mean reverse engineering the qte driver? Jun 24 08:14:02 SirFred: RE is such a big word. Taking a peek, I'd say Jun 24 08:14:07 mickeyl: :) Jun 24 08:14:53 Luke-Jr: hello Jun 24 08:14:54 mickeyl: I'll try to give it a look. ANyway, I'm pretty sure that the transformed approach was used. Jun 24 08:15:06 pb_: hi Jun 24 08:15:11 mickeyl: All qte is prepared to use transformed screens. But not for hot-rotating framebuffers. Jun 24 08:15:25 mickeyl: We have no trace that sharp went that way. Jun 24 08:15:33 pb_: the 'libtool' file in the work dir is different from the one in my KDE CVS dir-- not sure if that means its the right one? Jun 24 08:15:33 i see Jun 24 08:15:42 mickeyl: I only followed it a little and it's a mad way. Jun 24 08:15:46 mickeyl: Diabolic. Jun 24 08:15:48 Luke-Jr: that does sound suspicious Jun 24 08:16:00 libtool + kde = zecke sees black Jun 24 08:16:02 pb_: the difference all seem to be variables, such as CC Jun 24 08:16:03 SirFred: too bad, it would've earned millions of points for "coolness". Jun 24 08:16:11 SirFred: but an approach that works would be even better :) Jun 24 08:16:14 bbl, dinner time Jun 24 08:16:20 mickeyl: We can always give it a try later. Jun 24 08:16:35 *nod* Jun 24 08:16:37 mickey|dinner: I would like to have something working. There will be time for experimenting. Jun 24 08:16:49 Luke-Jr: ah. if it's _only_ variables, that might be ok. Jun 24 08:16:57 if there are code differences, though, that would be a problem Jun 24 08:17:01 zecke: heh Jun 24 08:17:02 pb_: didn't you say it should be an OE-patched version? Jun 24 08:17:31 yes Jun 24 08:17:42 autotools_do_configure will install OE's own version of libtool into your source directory. Jun 24 08:17:49 if you are doing anything to defeat that, you will have a problem. Jun 24 08:18:09 oe_runconf should do it, right? Jun 24 08:18:31 no Jun 24 08:18:47 autotools_do_configure does a whole pile of stuff before it calls oe_runconf. Jun 24 08:19:11 are you providing your own do_configure() method in the .bb file? Jun 24 08:19:27 yes Jun 24 08:19:36 unable to download using cvs ? Jun 24 08:19:52 am giving up... Jun 24 08:19:55 1) autoconf; 2) oe_runconf w/ args Jun 24 08:20:04 okay, that's no good Jun 24 08:20:28 what goes wrong if you just let autotools.bbclass do its thing? Jun 24 08:20:50 lemme try... moving stuff to EXTRA_OECONF var... Jun 24 08:21:36 time to leave and will catch up tommorrow. hopefully i will try to make it work by tommorrow ? Jun 24 08:21:43 one more thing. Jun 24 08:21:52 what should i ask my sysadmin ? Jun 24 08:21:57 to open port 22 Jun 24 08:22:00 thats it. Jun 24 08:22:08 or anything else ? Jun 24 08:22:14 wget is working fine. Jun 24 08:22:33 what else is needed to change my environment to work on oe Jun 24 08:22:50 can anyone list out the reqts please.. Jun 24 08:23:07 we don't know your environmnet... Jun 24 08:23:41 what is the env needed to work on oe. which ports should be open and things like that ? Jun 24 08:23:53 all ports would work Jun 24 08:24:00 c'mon Jun 24 08:24:05 you should be able to access ssh, cvs and svn port on remote hosts Jun 24 08:24:12 and httpd Jun 24 08:24:18 that's all you need Jun 24 08:24:22 * Luke-Jr thinks blocking outbound anything is evil Jun 24 08:24:33 svn what ports it requires i changed my proxy file and it does not work ? Jun 24 08:25:00 ram123: svn usually works over HTTP, IIRC Jun 24 08:25:35 my proxy does not support HTTP Jun 24 08:25:41 it is not HTTP Proxy. Jun 24 08:27:55 bye Jun 24 08:28:20 time to leave. are u guys comming tommorrow to this room ? Jun 24 08:30:34 pb_: that seems to have fixed the problems :) Jun 24 08:30:53 pb_: (or prompting recompiling the object files, in which case it hasn't reached the linking step yet) Jun 24 08:31:08 pb_: thanks Jun 24 08:33:03 seems to have finished the dcop dir, so should have worked :) Jun 24 08:33:11 ok, very good Jun 24 08:35:39 I was wondering why minimo is built during the buildind of gpe-image ( and the not installed, AFAIK ) Jun 24 08:35:58 Pigi: minimo is part of gpe-image... Jun 24 08:36:43 is it ? I haven't seen in the previous images Jun 24 08:37:29 Pigi: I have a fork, gpe-image-ljr, which has FireFox instead on my site http://utopios.org/~luke-jr/programs/oz-ljr/ Jun 24 08:37:37 (under Live Development Sources/ljr.pkgs) Jun 24 08:38:55 no, I suspect it got build but not installed on ipaq machines: GPE_EXTRA_DEPENDS_bigscreen = "minimo figment" Jun 24 08:39:07 s/build/built Jun 24 08:39:15 ? Jun 24 08:39:56 I think that minimo comes only on machine defined as "bigscreen". My ipaq is definitivelly not a bigscreen Jun 24 08:40:31 so it is built, but then the IPKG_INSTALL don't put it in the image. Jun 24 08:40:43 interesting Jun 24 08:40:59 maybe someone can confirm ? Jun 24 08:41:57 there are some class/machine/*.conf that have :GUI_MACHINE_CLASS = "bigscreen" Jun 24 08:42:06 so I think I'm rigt :) Jun 24 08:42:11 *right Jun 24 08:43:46 Pigi: check for h3900 stuff Jun 24 08:44:02 minimo get's installed in h3900 images Jun 24 08:44:07 and c700 Jun 24 08:44:22 and what on 3800 ? Jun 24 08:44:54 bye Jun 24 08:44:56 no, because that uses h3600 as a config file, which assumes 16 of flash Jun 24 08:45:26 I see. that's explain why I haven't seen it on my previous tests Jun 24 08:45:58 wow, a nslu2 is only 13x9 cm Jun 24 08:46:20 * koen searches for stuff to sell on ebay Jun 24 08:48:20 what exactly is an nslu ? Jun 24 08:48:31 Pigi: http://www.nslu2-linux.org Jun 24 08:48:46 yes, already seen Jun 24 08:49:08 it's a small network server with usb port Jun 24 08:49:20 so you can hook up a usb drive and share it Jun 24 08:49:32 ah, that is. Thx koen Jun 24 09:01:55 pb_: ping Jun 24 09:04:56 hello Jun 24 09:13:33 pb_: the OE libtool stuff seems to have broken the Makefiles? O.o Jun 24 09:13:43 everything after ".NOEXPORT:" is gone Jun 24 09:13:55 any ideas there? Jun 24 09:15:17 no idea Jun 24 09:15:27 libtool doesn't get involved in generating the Makefiles, though, so that seems weird. Jun 24 09:16:01 you might want to check with zecke, he has the kde sk1llz. Jun 24 09:22:04 evening Jun 24 09:24:44 evening Jun 24 09:27:27 03XorA 07 * r1.3601.2.1 10openembedded/packages/mplayer/ (mplayer-atty_1.1.4.bb mplayer-atty-1.1.4/Makefile-libs.patch): mplayer-atty: added needed patch I missed when upping rev Jun 24 09:32:01 * Pigi goes home Jun 24 09:32:05 see you later Jun 24 09:32:56 if i want to build a meta package like openslug-image and want to include udev, do i have to use DEPENDS and RDEPENDS? Jun 24 09:33:25 i think i have to use DEPENDS too because udev should be included in the firmware Jun 24 09:33:51 DEPENDS is build time. RDEPENDS is run time. so, the answer is both Jun 24 09:34:27 ok, so all what should be inside of the image need to be in DEPENDS too, thx Jun 24 09:36:17 reenoo__: and the same with kernel-module-ext3? the module will be built by the kernel package so perhaps only RDEPENDS is needed? Jun 24 09:48:45 mr_claus: right. only RDEPENDS in that case (if virtual/kernel is in DEPENDS) Jun 24 09:49:02 reenoo__: ok, thx Jun 24 10:06:40 hrm Jun 24 10:06:54 autotools class doesn't wipe Makefile.in, does it? :| Jun 24 10:08:09 autotools class runs autoreconf Jun 24 10:08:24 * Luke-Jr isn't aware what that one does Jun 24 10:08:26 if the package uses automake, autoreconf calls automake, which in turn generates Makefile.in from Makefile.am if it exists. Jun 24 10:08:33 autoreconf just calls autoconf, automake, aclocal, etc Jun 24 10:08:36 in the correct order Jun 24 10:08:39 ah Jun 24 10:08:40 Luke-Jr: print the aututools manual and smack the kde devs with it Jun 24 10:09:06 zecke tried to do that earlier with no effect Jun 24 10:09:12 -# KDE tags expanded automatically by am_edit - $Revision: 406248 $ Jun 24 10:09:18 * Luke-Jr thinks that might be the missing key Jun 24 10:10:20 koen: does autotools normally support moc? Jun 24 10:10:43 no idea Jun 24 10:11:05 if not, that would explain needing KDE-specific stuff Jun 24 10:18:44 are bb code segments run in BASH or sh? Jun 24 10:18:57 more specifically, are pushd and popd available? Jun 24 10:20:34 sh Jun 24 10:21:06 koen: any idea if pushd/popd/alias/unalias are valid sh? Jun 24 10:21:38 no idea Jun 24 10:22:11 pushd and popd are shell scripts itself IIRC Jun 24 10:22:33 pushd/popd have functionality no script could implement... Jun 24 10:23:22 () can be used in place of pushd/popd, simple variables in place of alias Jun 24 10:23:44 jbowler: I want to noop-ify part of a class I'm using Jun 24 10:23:51 jbowler: so I'm aliasing the command it uses to echo Jun 24 10:25:10 shell functions are supported Jun 24 10:25:18 ? Jun 24 10:25:40 cat(); {; echo "$@"; } Jun 24 10:25:54 jbowler: I need to override a shell function, tho ;) Jun 24 10:26:08 jbowler: IIRC, aliases have preference over functions Jun 24 10:26:15 ? function(); {; } Jun 24 10:26:43 I need to use the original function afterward Jun 24 10:26:56 Its a damn script - alias is nice and all, but you can probably afford a bit of un-readablity Jun 24 10:27:38 CosmicPenguin: I don't see anything else providing the functionality of alias in this case Jun 24 10:27:40 What is it that does actually use a shell function? Jun 24 10:27:48 I.e. what is the bbclass? Jun 24 10:27:58 autotools_do_configure calls oe_runconf Jun 24 10:28:04 autotools bbclass Jun 24 10:28:14 I need to do some processing prior to oe_runconf Jun 24 10:29:00 now, if most of autotools_do_configure would move to oe_preconf or such... that would be nice =p Jun 24 10:29:07 but I don't like modifying other people's files Jun 24 10:29:33 eck... even alias didn't do it :| Jun 24 10:29:43 Luke-Jr: eh? just write a custom do_configure in yoiur .bb that calls autotools_do_configure Jun 24 10:29:56 kergoth: autotools_do_configure calls oe_runconf, tho Jun 24 10:30:03 I can't find a shell function in autotools.bbclass - I thought there was one somehwere, but I don't know where. Jun 24 10:30:07 Luke-Jr: and? Jun 24 10:30:11 kergoth: I need to do something before oe_runconf (run a perl script) Jun 24 10:30:15 yes, and? Jun 24 10:30:16 Oh, ok, right. Jun 24 10:30:18 kergoth: but after the autoreconf Jun 24 10:30:23 okay, then say that. Jun 24 10:30:29 heh Jun 24 10:30:37 * Luke-Jr was trying to =p Jun 24 10:31:38 * Luke-Jr wonders how long an oe_preconf patch would take to get committed Jun 24 10:32:05 Luke-Jr: pushd/popd are bashisms, not standard susv3 or posix sh, afaik. alias is fine. Jun 24 10:32:06 oe_runconf_prepend (or does that only work on tasks?) Jun 24 10:32:22 kergoth: alias didn't work either, tho :( Jun 24 10:32:28 ? Jun 24 10:32:39 kergoth: oe_runconf still ran instead of being an echo Jun 24 10:32:59 then you didnt define the alias after the oe_runconf definition. exactly what you'd expect. Jun 24 10:33:08 http://pastebin.ca/15778 Jun 24 10:33:14 or the alias was in shell function scope, not global. once again, exactly what you'd expect Jun 24 10:33:56 kergoth: would oe_runconf_prepend (suggested by jbowler) work? Jun 24 10:34:22 if oe_runconf is shell defined by oe, that'll work. it works on any bitbake variables, not just tasks Jun 24 10:34:36 mornin kergoth Jun 24 10:34:38 (oe_runconf();{;}; autotools_do_configure) Jun 24 10:34:41 hey Jun 24 10:37:00 What would be better then trying to do stupid shell tricks would be to pull out the autoreconf foo and make it a function, thereby reducing the do_configure to one or two lines Jun 24 10:37:03 and easier to overload Jun 24 10:37:41 CosmicPenguin: hence why I asked how long such a patch would take to get committed into BK Jun 24 10:37:50 BK? Jun 24 10:37:57 BitKeeper Jun 24 10:38:01 Thats all a function of how intellegent said patch ended up being Jun 24 10:38:28 03koen 07 * r1.3610.1.1 10openembedded/packages/ (3 files in 3 dirs): Jun 24 10:38:28 Patch libtool to support '-version-info 1000:0:1000' again. See http://mail-index.netbsd.org/tech-pkg/2005/03/12/0011.html Jun 24 10:38:28 Add newer poppler lib Jun 24 10:38:33 oe_runconf_prepend worked, tho, so it might not be needed Jun 24 10:39:23 ~lart either libtool or gnome for usiDng '1000' Jun 24 10:39:24 * ibot forces either libtool or gnome to use Outlook Express for usiD[D[Dng '1000' Jun 24 10:54:08 kdelibs wants to build and run 'dcopidl' :| Jun 24 10:54:46 looks like you need a dcop-native Jun 24 10:55:13 dcop or dcopidl? O.o Jun 24 10:55:24 isn't that the same? Jun 24 10:55:30 doubt it Jun 24 10:55:34 dcopidl is used at buildtime Jun 24 10:55:40 dcop is like dbus Jun 24 10:56:26 I know that Jun 24 10:56:37 it seems you'll need kdelibs-native Jun 24 10:56:52 see http://packages.debian.org/cgi-bin/search_contents.pl?word=dcopidl&searchmode=searchfiles&case=insensitive&version=stable&arch=i386 Jun 24 10:57:08 good luck with making a x-native env Jun 24 10:57:17 * koen shouldn't gloat Jun 24 10:58:27 some KDE guy says to use scons :\ Jun 24 10:58:57 do I ignore him or make a dcop-native? Jun 24 10:59:11 ignore him Jun 24 10:59:14 ignore him Jun 24 10:59:25 he's saying "we suck, but we aren't going to admit it" Jun 24 10:59:30 err... ignore him & make dcop-native || use scons Jun 24 11:00:14 my question is this Jun 24 11:00:21 how is scons going to help in the slightest? Jun 24 11:00:31 since it's just a different build tool Jun 24 11:02:28 no idea Jun 24 11:07:06 no way to get dcopidl to build for Build instead of Target? :\ Jun 24 11:09:02 well Jun 24 11:09:06 you need it for both :p Jun 24 11:09:30 why? Jun 24 11:09:35 it's only used at build-time Jun 24 11:10:10 koen: BTW, I'd much more soon disable X on kdelibs-native =p Jun 24 11:10:17 koen: KDE does not require X :) Jun 24 11:11:06 * Luke-Jr ponders qemu Jun 24 11:11:43 ~lart opengroupware Jun 24 11:11:43 * ibot chops opengroupware in half with a free Solaris 7 CD Jun 24 11:12:09 ~lart sqlite3 Jun 24 11:12:09 * ibot explains, ever so gently, that if sqlite3 doesn't give the channel more information, they can't help Jun 24 11:12:56 dumb thing can't be built on amd64 Jun 24 11:12:57 ~lart qtopia pim structure also Jun 24 11:12:57 * ibot DoSes qtopia pim structure also Jun 24 11:13:56 to do concat() on data in sqlite3 all content need to be exported and written outside as query Jun 24 11:14:01 treke-: would I get ~lart'd if I use qemu to run dcopidl and such? :| Jun 24 11:15:09 yes Jun 24 11:15:30 hmm Jun 24 11:15:40 what if I require ASSUME_PROVIDED=kdelibs-native? Jun 24 11:15:49 yes Jun 24 11:16:01 hmm... has anyone had a look at the sl-5500 suspend/resume problems? I have a 5500 here with a serial console and whatnot, and I'd like to help. Any ideas/info? Jun 24 11:16:26 what if I steal Gentoo's split-KDE code? Jun 24 11:17:09 eh? When I last looked, kdelibs was a monolithic package in gentoo Jun 24 11:17:41 kdelibs is, but the general splitting code should work Jun 24 11:19:50 I dont know how popular importing X hundred bbs would be, but I doubt anyone would have technical complaints Jun 24 11:20:27 well, I only need to native-build dcop Jun 24 11:20:27 I hope Jun 24 11:20:34 so that's 1 Jun 24 11:22:16 if that works, it shoul dbe fine Jun 24 11:22:32 can bbclasses exist outside of openembedded/conf? Jun 24 11:22:39 or is it better to use a .inc? Jun 24 11:23:05 classes can go into the classes dir for anything in BBPATH Jun 24 11:36:40 anyone fell like answering a question about my efforts to use autoconf? Jun 24 11:36:53 configure: error: cannot run /bin/sh ./config.sub Jun 24 11:37:04 is the error I get from my configure script Jun 24 11:42:16 is it possible to either fix the list archives or drop them from the server entirely? there's one message about it to the list every two weeks, and for every message about it there are probably a half-dozen more confused and discouraged people who don't report it. Jun 24 11:42:42 keturn: talk with hh.org admins Jun 24 11:52:12 Is my bbclass missing something? http://pastebin.ca/15785 Jun 24 11:52:48 I get | /home/luke-jr/src/oe2005/build.alt/tmp/work/kdelibs-dcop-native-3.4.1-r0/temp/run.do_unpack.9193: line 429: kde-split_do_unpack: command not found Jun 24 11:54:16 - isnt a valid character in a shell function's name Jun 24 11:58:22 doesn't it need to be named after the bbclass? O.o Jun 24 11:59:41 in order for EXPORT_FUNCTIONS to behave properly for our "inheritence", yes. but if you're going to just be calling it directly from another function.. Jun 24 11:59:43 brb Jun 24 12:08:15 why not just call it kdesplit instead of kde-split? Jun 24 12:13:00 do we have ipkg on oe.handhelds.org? Jun 24 12:13:06 ipkgs* Jun 24 12:15:02 treke-: I renamed it to kde_split Jun 24 12:21:01 ok Jun 24 12:21:17 if that works :) Jun 24 12:23:32 time to give monotone a shot Jun 24 12:24:32 bang Jun 24 12:24:50 ~kill CP|Laptop Jun 24 12:24:50 * ibot shoots a excited anti-photon gun at CP|Laptop Jun 24 12:32:01 ~kill treke- Jun 24 12:32:01 * ibot shoots a magneto-ionized fluxpositrino gun at treke- Jun 24 12:34:10 ~bitchslap Luke-Jr Jun 24 12:34:10 * ibot beats the sh*t out of Luke-Jr Jun 24 12:34:15 * Luke-Jr cries Jun 24 12:34:24 ~lart treke- Jun 24 12:34:24 * ibot DoSes treke- Jun 24 12:35:18 ~kill CP|Laptop Jun 24 12:35:19 * ibot shoots a charged fluxproton gun at CP|Laptop Jun 24 12:35:21 treke-: happy? Jun 24 12:35:29 very Jun 24 12:36:16 what makes configure.in from configure.in.in? Jun 24 12:42:01 Hey guys Jun 24 12:42:27 I'm looking for x86emu, is it available for oe? Jun 24 12:42:35 it's part of xfree86 Jun 24 12:43:24 I'm fairly new to oe, is there a primary resource for looking up packages? Jun 24 12:43:57 re Jun 24 12:46:17 there is a problem downloading some sources from handhelds, as the url has changed for the feed of sources. Where it should be changed ? Jun 24 12:51:44 Ok, found it. Jun 24 12:52:43 lilalinux: um... nobody uses xfree86 anymore Jun 24 12:53:34 Luke-Jr: I don't need xfree86 itself, just x86emu to initialize a VGA on a meshcube Jun 24 12:54:01 because of the missing 16bit call handler (INT10) Jun 24 12:54:06 callback Jun 24 12:54:17 lilalinux: the xorg xserver is in oe; look in the xserver/ directory Jun 24 12:55:42 Actually I use nylon, but this is primarily an automated buildsystem. The openembedded/packages/ directory doesn't list xserver/ :( Jun 24 12:55:46 where can I download it? Jun 24 12:57:43 There is a xserver folder in my packages dir Jun 24 12:58:27 NAiL: are you using nylon? Jun 24 12:58:40 no Jun 24 13:02:28 Is it possible to download packages without having to install bitkeeper? Jun 24 13:02:52 lilalinux: bk-client is open source Jun 24 13:02:53 use monotone Jun 24 13:03:02 monotone is more recently :) Jun 24 13:03:10 koen: are recent changes in monotone? Jun 24 13:03:20 check the webinterface Jun 24 13:03:57 apting monotone right now\ Jun 24 13:03:59 I mean in general Jun 24 13:04:09 check the webinterface Jun 24 13:04:17 koen: url? Jun 24 13:04:24 koen: check for what? Jun 24 13:04:33 The downloads page says, there is no download page :) Jun 24 13:04:55 Luke-Jr: recent changes Jun 24 13:05:23 koen: ... Jun 24 13:05:57 koen: do you have a commandline at hand for using monotone? I need xserver/ Jun 24 13:06:02 lilalinux: fyi, OE doesn't distribute any packages. only description files to build packages. for those "recipes" you can check http://oe-devel.bkbits.net:8080/openembedded/src/packages?nav=index.html|src/ Jun 24 13:06:09 can only check for specific changes Jun 24 13:06:13 reenoo__: thx Jun 24 13:06:44 lilalinux: and fwiw, I don't think there's a .bb for x86emu Jun 24 13:07:26 reenoo__: x86emu is part of xserver Jun 24 13:08:37 not in xorg cvs afaict Jun 24 13:09:42 ic :( Jun 24 13:09:51 xorg <-> xfree86 :( Jun 24 13:09:56 err.. ignore that Jun 24 13:10:01 found it? Jun 24 13:10:14 xorg > xf86 Jun 24 13:11:03 reenoo__: what should I ignore? Jun 24 13:13:05 lilalinux: the "not in xorg cvs" Jun 24 13:14:26 k, so I installed monotone Jun 24 13:14:37 how to I comment out a block in python? Jun 24 13:14:46 how do i get the recipe? Jun 24 13:15:07 wget doesn't sound smart Jun 24 13:15:18 monotone pull? Jun 24 13:15:39 monotone: misuse: no database specified Jun 24 13:15:57 <--- newbie Jun 24 13:16:21 http://venge.net/monotone/monotone.html <--- manual Jun 24 13:16:53 ;) Jun 24 13:21:51 hmmm... does gmp not play well with x86_64? Doesn't look like it does Jun 24 13:23:31 it'd be a bit odd if it didn't. Jun 24 13:23:36 what problem are you having? Jun 24 13:25:10 /tmp/cckJ2i3O.s: Assembler messages: Jun 24 13:25:10 /tmp/cckJ2i3O.s:24: Error: suffix or operands invalid for `bsf' Jun 24 13:25:29 mm, that sounds more like a gcc bug. Jun 24 13:25:45 or, at least, a disagreement between gcc and gas Jun 24 13:25:45 thats not entirely impossible Jun 24 13:26:59 if you know amd64 assembler, you could eyeball the .s file to figure out which one of those is at fault. Jun 24 13:27:15 i.e. whether gcc is generating duff code, or whether gas is turning its nose up at something that ought to be valid Jun 24 13:29:35 There is a lot of processor specific code in that general vicinity, though Jun 24 13:30:00 ah, you think it might be coming from an asm() statement? Jun 24 13:30:04 gmp does have a few of those Jun 24 13:30:46 The file is mpn/dive_l.c Jun 24 13:31:23 and by l, I do of course mean 1 Jun 24 13:36:03 On Opteron/AMD64 systems with GCC, count_trailing_zeros in longlong.h may cause errors from the assembler about incorrect registers used with bsfq. Patch. Jun 24 13:36:06 thar she blows Jun 24 13:39:52 aha. Jun 24 13:41:06 Should I toss the patch into BK, or is that lliable to get lost? Jun 24 13:50:33 not yet Jun 24 13:55:29 CosmicPenguin: I'm doing a periodic bk export and push that into the monotone tree Jun 24 13:56:04 ok Jun 24 13:56:52 if you notice http://ewi546.ewi.utwente.nl/tmp/viewmtn/headofbranch.psp?branch=org.openembedded.dev isn't up to date, ping me or mickeyl Jun 24 14:01:57 * CosmicPenguin kicks CIA Jun 24 14:02:19 Bah, mabye my e-mail is fsked up Jun 24 14:02:32 Anyway, bmp fix is up for any opteron usin' fools Jun 24 14:02:39 s/bmp/gmp/ Jun 24 14:02:45 03jcrouse 07 * r1.3613 10openembedded/packages/gmp/ (gmp_4.1.2.bb gmp-4.1.2/amd64.patch): Add x86_64 fix for bad opcode message Jun 24 14:20:17 CosmicPenguin: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=526c5e54a8723f6b4ff9f8426dd64c69f4ab3db5 Jun 24 14:30:18 koen: nice Jun 24 14:30:19 thansk Jun 24 14:32:56 koen: jbowler just did our first propagate of nslu2-linux changes back to org.openembedded.dev - can you check it to make sure we're all in sync and haven't lost anything? Jun 24 14:33:00 what's our favourite hotplug implementation today? or whatever package it is I need to make-pcmcia-card-automatically-get-a-dhcp-address. Jun 24 14:33:32 I pushed to dominion: Jun 24 14:33:33 oe_site=dominion.kabel.utwente.nl Jun 24 14:33:33 oe_collection=org.openembedded Jun 24 14:33:41 keturn: for most kernels, that'd be pcmcia-cs. Jun 24 14:33:50 hotplug for pcmcia is a fairly recent innovation Jun 24 14:35:09 jbowler-away: could you try the same with ewi546.ewi.utwente.nl as a host? Jun 24 14:36:34 koen: in progress... Jun 24 14:36:39 I don't seem to have any changes Jun 24 14:37:09 Try now Jun 24 14:37:16 still nothing Jun 24 14:37:17 see http://ewi546.ewi.utwente.nl/tmp/viewmtn/index.psp Jun 24 14:37:26 fa34be49ded63214198ca4b776e49650e67ae189 jbowler@nslu2-linux.org 2005-06-24T21:23:07 Jun 24 14:37:48 Seems to be there now. Jun 24 14:37:50 koen: http://ewi546.ewi.utwente.nl/tmp/viewmtn/ <- shows nslu2-linux branches in it. any idea how those go in there when ewi is only serving org.openembedded? Jun 24 14:38:05 I imported them Jun 24 14:38:13 the hh.org repo won't have them Jun 24 14:38:17 oh, ok. Jun 24 14:38:38 for a second there I thought our nslu2-linux proposal was flawed :-) Jun 24 14:38:38 It doesn't matter anyway - it's just a name+tag for stuff which exists anyway. Jun 24 14:38:57 jbowler-away: yeah, but it will be out of date and confusing unless it is pulled regularly. Jun 24 14:39:34 What I meant was that it couldn't be pulled from ewi - but I guess it's confusing in viewmtn Jun 24 14:39:51 koen: there is an "invisible branches" thingo in monotonerc which could be used to hide them Jun 24 14:39:53 Ok, so should I be pushing to ewi, not dominion? Jun 24 14:40:06 jbowler-away: correct Jun 24 14:40:52 koen: I have changed mtn.nslu2-linux.org to sync from ewi Jun 24 14:41:24 jbowler-away: still don't see your push at http://ewi546.ewi.utwente.nl/tmp/viewmtn/ Jun 24 14:43:52 http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=d91164277f525d52f5d93b713de003dfb955b82a Jun 24 14:45:31 I'm not sure that's the correct one though I don't see how it got there otherwise. The date is weird and the id doesn't match. Jun 24 14:46:04 that's the one which koen pulled a while back Jun 24 14:46:16 http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=526c5e54a8723f6b4ff9f8426dd64c69f4ab3db5 doesn't show any of your latest changes Jun 24 14:47:29 Yes, it's not there - i.e. the revision fa34be49ded63214198ca4b776e49650e67ae189 isn't there. Jun 24 14:49:31 koen: when/where will the monotone testing [ pubkey .... ] for openembedded be made avail, so I can init, read and pull my --db="????.db" Jun 24 14:50:41 jbowler-away: The samba push you did, wouldn't that show up somewhere in there? I can't find it, and the initscript doesn't show up. Am I missing something? :) Jun 24 14:50:59 I got the initscript when I pulled, so it's there Jun 24 14:52:19 jbowler-away: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=fa34be49ded63214198ca4b776e49650e67ae189 <- revision not found in database Jun 24 14:53:38 All my latest changes seem to be missing. Jun 24 14:54:18 rwhitby: yes, and I checked on nslu2-linux too. Something weird is happening. Jun 24 14:54:29 nigth all Jun 24 14:55:34 Ok, I synced to the nslu2-linux database, that revision is there now, so that works. Jun 24 14:56:22 Dang, there are two heads. Jun 24 14:57:33 hufnus: anynomous pull is enabled Jun 24 14:59:25 The nslu2-linux org.openembedded.dev has the same head (now) as mine. Jun 24 15:01:51 koen: I think my push may have caused your db to have two heads on the openembedded.dev branch. Jun 24 15:02:20 (And I don't understand this - my model of the branching structure is flawed somewhere.) Jun 24 15:02:30 ckoen@dominion:/data/build/oe/monotone/org.openembedded$ monotone merge Jun 24 15:02:30 monotone: misuse: branch 'org.openembedded.dev' is merged Jun 24 15:02:50 What is the head of that branch? Jun 24 15:03:11 koen@dominion:/data/build/oe/monotone/org.openembedded$ monotone heads Jun 24 15:03:11 monotone: branch 'org.openembedded.dev' is currently merged: Jun 24 15:03:11 526c5e54a8723f6b4ff9f8426dd64c69f4ab3db5 koen@openembedded.org 2005-06-24T21:19:31 Jun 24 15:03:58 Unchanged from before my push. Yet I have a different head in my database. Jun 24 15:04:10 try merging it Jun 24 15:04:36 I get the 'misuse' message. Jun 24 15:05:02 koen: I checked out but there is no xserver/ Jun 24 15:05:09 I did: monotone --db=meshcube pull ewi546.ewi.utwente.nl org.openembedded Jun 24 15:05:19 and: monotone --db=meshcube co --branch org.openembedded.dev Jun 24 15:06:09 lilalinux: according to http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=526c5e54a8723f6b4ff9f8426dd64c69f4ab3db5 there is Jun 24 15:07:52 Did I do something wrong? See my commandlines above Jun 24 15:08:04 oh, my bad Jun 24 15:08:09 I looked in the classes dir Jun 24 15:08:11 not the packages Jun 24 15:08:14 there it is Jun 24 15:08:42 jbowler-away: you could take a look at the db with monotone-viz Jun 24 15:12:12 jbowler-away: ewi has 93 certs, when that changes, you'll know it succeeded Jun 24 15:12:34 Just googling monotone-viz Jun 24 15:12:50 dyoung, I restarted my build, but it's rebuilding samba (I think it changed) Jun 24 15:12:59 jbowler-away: http://venge.net/monotone/tools.html Jun 24 15:13:00 oops wrong channel Jun 24 15:13:58 hmmmm Jun 24 15:14:13 monotone docs are quiet on the dropping of branches Jun 24 15:16:54 'night all Jun 24 15:32:20 how do I select whether a package is targetting native or cross? Jun 24 15:32:36 putting -native in the name doesn't seem to do it ^^; Jun 24 15:32:55 "inherit native" Jun 24 15:51:08 what is xserver-kdrive? Jun 24 15:51:20 small xserver Jun 24 15:51:29 I think knoppix uses it Jun 24 15:51:38 thx Jun 24 15:51:44 np Jun 24 16:24:30 jbowler-away: trying to do a monotone pull but what branch and url??? Jun 24 17:10:47 pb_: hey Jun 24 17:11:00 pb_: I'm getting that old "Host.c:53: error: `FamilyServerInterpreted' undeclared" error a log says you got-- what's the solution? :) Jun 24 17:18:03 or rather, reenoo got in xhost Jun 24 18:24:18 morning Jun 24 21:13:02 anyone here made a serial cable for a Zaurus before? **** ENDING LOGGING AT Fri Jun 24 23:59:56 2005