**** BEGIN LOGGING AT Mon Nov 27 02:59:57 2006 Nov 27 04:57:09 03lenehan 07org.oe.dev * r84e86495... 10/ (1 conf/machine/wrap.conf): machine/wrap: Add task-base related settings. Nov 27 04:57:20 03lenehan 07org.oe.dev * r5017a9dc... 10/ (1 conf/machine/titan.conf): machine/titan: Add task-base related settings. Nov 27 06:46:20 "The distro is built on top of Gentoo, one of the more popular source based distributions." Nov 27 06:46:24 ~lart techgage.com Nov 27 06:46:25 * ibot_ throws techgage.com's poor little doggy off a cliff Nov 27 06:46:48 * koen wonders what debian is based on, pixie dust? Nov 27 07:33:31 woooo Nov 27 07:33:35 xcb reaced 1.0 Nov 27 07:41:22 Who is "David Anders" ? Nov 27 07:41:31 prpplague Nov 27 07:42:07 good morning all Nov 27 07:42:07 schurig: leoncamel is the another one Nov 27 07:42:10 morning Nov 27 07:42:22 hrw: Moin! Nov 27 07:42:52 hrw: this weekend I've commit the stub if_pcmcia.c file. With this, the driver compiles without errors, but does nothing, all the CS things are still missing Nov 27 07:43:28 schurig: many pcmcia glue is in gumstix patch Nov 27 07:43:30 the libertas driver works very different from normal CS drivers, e.g. in wlan_main, all calls do hardware stuff (including XXX_probe, XXX_config) are done via SBI Nov 27 07:43:31 2.6.17 one Nov 27 07:44:09 Maybe this afternoon or evening I've time again Nov 27 07:44:37 BTW, there's also a mailing list about the libertas driver, I've put that into the README. Unfortunately, yesterday infradead.org was down. Nov 27 07:44:39 I do not know when will get Nov 27 07:44:47 prprplague should be able to work fulltime on it after he finished moving Nov 27 07:45:59 hi Nov 27 07:46:06 hey Genesis Nov 27 07:46:13 yo koen ): Nov 27 07:46:15 :) Nov 27 07:46:29 hrw: some stuff in the wlan_ things needs to re-added Nov 27 07:47:02 ok Nov 27 07:47:22 hrw: for example, in sbi.h HIM_ENABLE/HIM_DISABLE is different for CF vs USB, but this was deleted in libertas-2.6 tree Nov 27 07:47:36 I have to replace pcmcia->cf2 adapter to pcmcia-cf1 one Nov 27 07:48:21 I think about testing card in progear rather then in Zaurus (more power to recompile) Nov 27 07:48:21 via the libertas mailing list we can get in contact with Marvell developers, because Marvell joined the OLPC project. When infradead.org was not yet down, I saw Marcello asking Marvell technicians questions over there Nov 27 07:48:56 schurig: thats cool Nov 27 07:49:02 need to subscribe there Nov 27 07:49:28 hrw: I try to get this running on a laptop first (Pentium M). Once that is working, I'm going to my PXA255 device. If it doesn't run there, then I have to look for the attribute packed stuff and other things, but I have a reference Nov 27 07:49:50 yes, x86 is always easier Nov 27 07:51:11 RP: bitbake trunk has something which I do not like. 'bitbake bootstrap-image gpe-image' does not build them in order but starts building of all deps. I know that this is probably proper way but I liked old behaviour Nov 27 07:51:46 ok. time to start going to work via misty way Nov 27 07:52:03 hrw: and you can get a race with do_rootfs as well Nov 27 07:52:19 koen: 'nice' Nov 27 07:52:26 hrw: btw, I like "KScope" so far the most when it comes to dig my way throught unknown code Nov 27 07:52:39 schurig: will test it Nov 27 07:52:54 schurig: I compared cf<>usb wlan with kdiff3 :D Nov 27 07:53:11 hrw: will test kdiff3 :-) Nov 27 07:53:15 koen: solution exist: tmp/rootfs/$PN/ instead of tmp/rootfs/ Nov 27 07:53:17 (never heard about that) Nov 27 07:53:28 schurig: very nice merge/diff tool Nov 27 07:53:28 hrw: or proper locking Nov 27 07:53:42 hrw: which we need anyway for multithreaded packaged staging Nov 27 07:53:58 KScope is a GUI for the ancient "cscope" program. You can view the file and view in no time references or definitions of things under your cursor Nov 27 07:54:12 nice Nov 27 07:54:32 schurig: will have to check it. or find good gvim script for it Nov 27 07:54:56 kdiff3 is good schurig :) Nov 27 07:54:57 I do not feel comfortable in !vim gui editors ;) Nov 27 07:55:14 cu - be back in ~1h Nov 27 07:55:29 hrw: for emacs, there is also ctags/etags support and also cscope.el, but so far I likes KScope more Nov 27 07:55:54 what's about anjuta ? Nov 27 08:00:25 anyone here got a mac and a moment to help me out? Nov 27 08:15:18 hrw|gone, hey, did you get replay from Marvell ? Nov 27 08:27:41 a. Nov 27 08:27:41 64 bytes from bugs.openembedded.org (64.136.190.225): icmp_seq=1 ttl=48 time=229 ms Nov 27 08:27:41 64 bytes from bugs.openembedded.org (64.136.190.225): icmp_seq=2 ttl=48 time=228 ms Nov 27 08:27:41 64 bytes from bugs.openembedded.org (64.136.190.225): icmp_seq=3 ttl=48 time=231 ms Nov 27 08:27:41 --- bugs.openembedded.org ping statistics --- Nov 27 08:27:41 6 packets transmitted, 3 received, 50% packet loss, time 5019ms Nov 27 08:27:44 oops sorry Nov 27 08:31:19 Genesis: I have similar times. "mtr" shows you the culprit Nov 27 08:32:10 oki , it's inusual Nov 27 08:32:26 i hope it comes back soon =) Nov 27 08:35:48 i think it's a tip to limitate the bugs ;) Nov 27 08:36:00 morning Nov 27 08:36:08 leoncamel: its schurig who mailed them Nov 27 08:37:14 good morning Nov 27 08:37:32 hey florian_kc Nov 27 08:39:30 03pfalcon 07org.oe.dev * r7a3a6ca0... 10/ (3 files in 2 dirs): (log message trimmed) Nov 27 08:39:30 linux-handhelds-2.6: Add 2.6.18-hh1, update _cvs to it. Nov 27 08:39:30 * 2.6.18-hh1 includes following changes since -hh0: Nov 27 08:39:30 1. h2200, h3900, hx4700 revamped for 2.6.18 (h2200 is not tested though; Nov 27 08:39:30 some flash related changes were made in mainline, so until confirmed to work Nov 27 08:39:33 by port maintainers, flash support is disabled in defconfig). Nov 27 08:39:35 2. Asus MyPal 730 (asus730) port is added. Nov 27 08:40:58 * koen reads http://bec-systems.com/web/index.php?option=com_content&task=view&id=56&Itemid=9 Nov 27 08:42:17 morning Nov 27 08:42:19 hey XorA Nov 27 08:44:33 koen: where doe task-sdk blow up now? Nov 27 08:44:57 it doesn't find a crtdi.o (or something like that) Nov 27 08:45:04 maybe I need a clean rebuild Nov 27 08:46:10 i'll give it a try here too Nov 27 08:48:57 CPU idle vs. CPU usage at 100% makes a difference of 40 watts on this machine... Nov 27 08:50:13 leoncamel: no reply yet Nov 27 08:50:34 leoncamel: but that was to be expected :-( Nov 27 08:55:22 morning all Nov 27 08:57:18 ~lart nfs for being a pain to setup Nov 27 08:57:18 * ibot_ pours gasoline all over nfs, ignites the fire, and then enjoys some toasty marshmallows with the glorious blaze for being a pain to setup Nov 27 08:57:37 koen, : seems to be a good product this moxa Nov 27 08:57:45 i searsh the price Nov 27 08:58:50 $400 Nov 27 09:01:36 8/8 IO , good for my project :) Nov 27 09:01:55 schurig, well, the marvell release 8388 as GPL. So, i think they also do the same thing on 8385. :) Nov 27 09:02:09 schurig, I hope they would do that . :) Nov 27 09:02:15 leoncamel: me too :-) Nov 27 09:02:32 leoncamel: the next problem for us (if they don't) would be the firmware Nov 27 09:03:11 leoncamel: adding the PCMCIA stuff is not exactly complicated, especially with a template in from the gumstix people. But then the device won't work without a firmware Nov 27 09:03:40 leoncamel: for now, we probably should extract the firmware from the windows driver. It would be nice if someone can do a script / binary / whatever for this. Nov 27 09:03:51 leoncamel: I'd then concentrate on the kernel stuff Nov 27 09:09:25 morning all Nov 27 09:09:44 hey ade|desk Nov 27 09:10:00 ade|desk, morning Nov 27 09:11:24 yo ade|desk Nov 27 09:12:22 how goes life in the world of oe Nov 27 09:17:45 anyone got any tricks for debugging root on nfs problems? Nov 27 09:18:27 XorA: tried to turn it off and on? Nov 27 09:18:40 heh Nov 27 09:18:41 hrw|work: dont get you Nov 27 09:18:47 someone has watched the it crowd Nov 27 09:18:57 "is it plugged in?" Nov 27 09:19:03 "are you sure?" Nov 27 09:19:03 ah :-D Nov 27 09:19:17 sounds like DOgberts technical support line Nov 27 09:19:43 are your eyes open to read whats on the screen ? Nov 27 09:19:44 I count mount the nfs dir on my zaurus and run init, but on the smdk2440 board it refuses to play Nov 27 09:20:16 XorA: pings works in both directions? Nov 27 09:20:16 wrong ip? norootsquash? noexec? Nov 27 09:21:20 hrw|work: dhcp discovery works Nov 27 09:21:44 koen: I can mount and execute from zaurus, just not samsung board Nov 27 09:21:59 koen: and as this is in kernel stuff I dont really get logs Nov 27 09:31:33 hi all Nov 27 09:34:21 hey likewise Nov 27 09:34:35 hey koen Nov 27 09:35:20 likewise: I'll be attending t-dose on saturday, so dinner at saturday evening? Nov 27 09:38:04 koen: I would like that, it's just that I got an invitation for a familiy meeting and a friend's birthday party. Let me first check how things can fit in, and if/how I can cancel stuff :-) Nov 27 09:38:20 I hate it when everything is on the same day :-) Nov 27 09:38:53 and the NS seems to have a sane schedule for a change Nov 27 09:39:13 koen: for the way back? Nov 27 09:39:20 yes Nov 27 09:39:49 what's the last-but-one option for you? Nov 27 09:40:01 koen: next week is bugdays? Nov 27 09:40:09 s/week/weekend Nov 27 09:40:15 it takes you one and a half hour on the way back? Nov 27 09:40:31 22:15 Nov 27 09:40:41 and ~3 hours Nov 27 09:40:52 2x times a stoptrein Nov 27 09:41:02 argh Nov 27 09:41:30 the intercity goes over utrecht which adds ~30 minutes Nov 27 09:41:50 the last option uses that (22:38) Nov 27 09:41:52 koen: you live in enschede, right? Nov 27 09:41:54 yes Nov 27 09:50:07 morning all Nov 27 09:50:31 RP: mornign Nov 27 09:50:45 RP: you got any tips on debugging root over nfs problems? Nov 27 09:50:48 hey RP Nov 27 09:50:48 hi RP Nov 27 09:51:22 XorA: What's the error you're seeing? Nov 27 09:52:12 RP: http://www.rafb.net/paste/results/QCA7Ox79.html Nov 27 09:52:27 RP: but I can mount it and run init.sysvinit from the zaurus Nov 27 09:53:09 XorA: init=/sbin/init.sysvinit or not? Nov 27 09:53:17 XorA: It will looks for /sbin/init, not init.sysvinit ? Nov 27 09:53:28 RP: init is a link to init.sysvinit Nov 27 09:53:36 XorA: tried without link? Nov 27 09:53:41 XorA: links are a bad idea - copy the file Nov 27 09:53:43 hrw|work: yes :-( Nov 27 09:53:50 ;( Nov 27 09:55:04 I am trying with file copied Nov 27 09:55:13 XorA: Does the nfs server allow root mounts? Nov 27 09:55:32 RP: no_root_squash is et Nov 27 09:55:36 set Nov 27 09:56:19 RP: "All rates include taxes and American Breakfast Buffet" Nov 27 09:56:35 koen: For the Novotel? Nov 27 09:56:40 astrid Nov 27 09:57:03 koen: Go with the astrid then - breakfast = 15Euro at Novotel Nov 27 09:57:27 the astrid even has dry crackers for X hackers that drank too much Nov 27 09:58:00 XorA: I know qemuarm works over NFS. Have a look at http://folks.o-hand.com/richard/qemu.html see if anything there gives a hint Nov 27 09:58:44 new url :) Nov 27 09:59:08 Genesis: time to run urlcatcher over #oe? Nov 27 09:59:59 That page is extremely dated. It should say "look at scripts/runqemu in poky svn" now :) Nov 27 10:00:07 but nfs hasn't changed Nov 27 10:00:22 qemu arm also :( Nov 27 10:00:30 very slow Nov 27 10:00:33 I had this running until last week :-( Nov 27 10:00:57 XorA: What did Liam do to it while you were away? :) Nov 27 10:01:11 Genesis: Poky also has qemux86 now Nov 27 10:01:28 I changd from Ubuntu to openSuse as I needed something to run, Edgy is a total disaster Nov 27 10:01:56 that's why I stayed with dapper on my workstation Nov 27 10:02:14 although both ubuntu and suse are unethical Nov 27 10:02:23 hm. I have 20 bitbake files modified and it still works Nov 27 10:02:43 20 files changed, 4 insertions(+), 28 deletions(-) Nov 27 10:02:53 * RP is still running dapper and doesn't dare "upgrade" Nov 27 10:03:04 * Genesis is happy with gentoo :) Nov 27 10:03:12 I run sid and I'm happy with it Nov 27 10:03:21 I also use RHEL4 Nov 27 10:03:35 i also use openembedded on arm , but it's a crap ;-) Nov 27 10:04:00 I'm happy with my Slackware servers :} Nov 27 10:04:19 I also plan to use angstrom on one of my x86 machines Nov 27 10:04:25 RP: really dont upgrade to Edgy, that distro is a spinning disaster Nov 27 10:04:36 ubuntu is a very bad project Nov 27 10:04:56 XorA: their upstart looks nice - I have it at home on my sid box and it works nice as fasterboot way Nov 27 10:04:58 yeah, ubuntu isn't "source based" Nov 27 10:04:58 XorA: mallum 'banned' us from upgrading the laptops ;-) Nov 27 10:04:59 Edgy ended with most of my apps randomly segfaulting Nov 27 10:05:00 try to turn debian / suse ... people on a disto that only employed packager Nov 27 10:05:49 MS is a buisness man , without other goal that make money. Nov 27 10:07:38 * XorA cries Nov 27 10:08:09 ERROR: No providers of runtime build target libc6-unslung (for ['/home/hrw/devel/oe/org.openembedded.dev/packages/images/unslung-image.bb']) Nov 27 10:08:14 another broken stuff.. Nov 27 10:17:53 03rwhitby 07org.oe.dev * rc9ecfa9b... 10/ (1 packages/linux/ixp4xx-kernel_2.6.19-rc6.bb): ixp4xx-kernel: Updated to latest SVN patchset with full nas100d support including IDE kerrnel module loading Nov 27 10:18:02 03rwhitby 07org.oe.dev * r44f9aad5... 10/ (1 conf/distro/debianslug.conf conf/distro/openslug.conf): slugos: updated debianslug and openslug to include IDE kernel modules in the rootfs Nov 27 10:18:10 03rwhitby 07org.oe.dev * r4782f0e3... 10/ (4 files in 4 dirs): slugos-init: Now loades IDE kernel modules for nas100d Nov 27 10:20:10 rwhitby: does 'loades' is australian english? Nov 27 10:21:41 * RP suspects the d,e and s keys are too close together :) Nov 27 10:21:48 RP: hrw|work koen the lack of /lib/ld-linux.so.2 is bad :-( Nov 27 10:21:57 heh Nov 27 10:22:00 XorA: very bad Nov 27 10:22:00 that'd do it Nov 27 10:22:14 first time Ive ever used ethereal, it is godlike Nov 27 10:22:27 XorA: isn't it called wireshark now? Nov 27 10:23:06 koen: maybe on the latest distros :-) Nov 27 10:27:07 gah, libw100 svn isnt buildable Nov 27 10:29:53 XorA: same here :( Nov 27 10:29:57 XorA: firmware.h? Nov 27 10:34:43 koen: yup Nov 27 10:35:51 * XorA was trying to make his c7x0 useful again Nov 27 10:36:17 dial back the SRCDATE_libw100 Nov 27 10:36:40 ~summon sirfred Nov 27 10:36:43 apt takes out 20 clean, identical-looking phones, some extra hands, and pretends to be a telemarketer for a large corporation, so he gets delivered a phonelist containing sirfred's coordinates Nov 27 10:38:59 ok, why the hell isnt glibc an RDEPEND for anything Nov 27 10:39:28 that old bug... Nov 27 10:40:10 ah wait, that was a missing libgcc Nov 27 10:41:28 yeah, for some reason libc6 doesnt end up in my images using generic distro Nov 27 10:44:54 * XorA wonders if this is one of those cases a rm -rf /tmp will fix Nov 27 10:52:27 koen: SRCDATE_libw100 = "20061115" Nov 27 10:52:44 XorA: could you add that to sane-srcdates.inc? Nov 27 10:53:52 koen: it might be wiser to use the rev_0.0.2 branch Nov 27 10:53:58 koen: thats from 2 weeks ago Nov 27 10:54:00 fine by me Nov 27 10:54:23 koen: Ill try and remeber to do a .bb later, Im a bit pressured at moment Nov 27 10:54:55 XorA: ok, cool Nov 27 11:03:06 koen: NOTE: Task failed: Fetch failed: http://dominion.kabel.utwente.nl/~florian/themes/foxbox-0.1.tar.gz Nov 27 11:05:34 koen, hrw|work: In the removal.txt file, could/should we list the person who is going to make the removal? Nov 27 11:05:46 RP: good idea Nov 27 11:05:50 anyone can make the removal, right? Nov 27 11:06:14 koen: Yes, but having some spokesperson for the entry would be good Nov 27 11:06:44 who to write with 'please,please do not remove it!' Nov 27 11:06:59 hrw|work: right :) Nov 27 11:07:11 Requested by: person which want entry to be removed Nov 27 11:07:28 03koen 07org.oe.dev * r6665f976... 10/ (1 packages/gpe-themes/gpe-theme-foxbox_0.1.bb): gpe-theme-foxbox: fix SRC_URI Nov 27 11:07:36 Proposed by ? Nov 27 11:07:59 ok Nov 27 11:08:38 NOTE: Task failed: Fetch failed: http://www.intel.com/Please-Read-The-BB-File/IPL_ixp400AccessLibrary-2_1.zip Nov 27 11:08:47 I would prefer file:// urls for it rather then poking intel server each time Nov 27 11:08:59 hrw|work: Those ones have to be manually downloaded... Nov 27 11:09:07 I know Nov 27 11:09:18 but bitbake tries to fetch them when you did not downloaded them Nov 27 11:09:46 add links-native and have fetcher fire it up :-) Nov 27 11:10:30 koen: can angstrom get BBMASK="bluez-libs-2.25"? Nov 27 11:12:00 bluez*2.25 Nov 27 11:12:01 if that works Nov 27 11:30:18 03xora 07org.oe.dev * rf2cd362e... 10/ (3 files in 2 dirs): Nov 27 11:30:18 libw100*.bb : add a release version 0.0.2 and make svn version Nov 27 11:30:18 DEFAULT_PREFERENCE = "-1" Nov 27 11:32:41 koen: tada! Nov 27 11:34:13 yay! Nov 27 11:34:31 * XorA wishes his spitz had a w100 Nov 27 11:36:23 terrier? Nov 27 11:36:34 you means you had more time to work on mplayer ;) Nov 27 11:36:40 +wish somewhere Nov 27 11:37:16 koen: accellerated X + accelerated colourspace conversion would be cool, CXX00 zaurii will only have one of the two :-( Nov 27 11:38:09 * XorA hopes the next OpenMoko phone is MX31 or later based Nov 27 11:38:22 yes!!!! Nov 27 11:41:13 and wifi and bt Nov 27 11:41:39 I can live without 2mpx camera in it Nov 27 11:52:22 What is an "American Breakfast Buffet:? Nov 27 11:53:15 Crofton: probably deep fried fat Nov 27 11:53:21 heh Nov 27 11:53:30 pork rinds Nov 27 11:53:35 03hrw 07org.oe.dev * r7b822d19... 10/ (7 files in 3 dirs): acpid: added 1.0.4 Nov 27 11:53:53 *yawn* Nov 27 11:53:57 morning boys and girls Nov 27 11:54:01 morning mickeyl Nov 27 11:54:05 gm Nov 27 11:54:06 hey mickeyl Nov 27 11:54:16 yo mickeyl Nov 27 11:55:30 hmm... let's see what the postman brought me today... two parcels Nov 27 11:55:34 * mickeyl opens the first one Nov 27 11:56:13 * mickeyl dances a jig Nov 27 11:56:16 a 2.8" portrait VGA display to attach to my s2c2410 developer board Nov 27 11:56:27 * mickeyl opens the 2nd one Nov 27 11:56:28 mickeyl: so your phone will finally get display? Nov 27 11:56:33 hrw|work: hehe Nov 27 11:56:41 * mickeyl dances again Nov 27 11:56:46 gsm module? Nov 27 11:56:47 "Gnome 2.0 developer handbook" Nov 27 11:56:54 now i'm well prepared :) Nov 27 11:57:20 hrw|work: the gsm module is already here, but they fscked up the antenna so it only gets GSM900 networks... all my cards are for 1800 :/ Nov 27 11:57:32 mickeyl: book <50 pages? (gnome is done to be simplest possible so...) Nov 27 11:57:37 heh Nov 27 11:57:43 _developer_ handbook, not _user_ manual :) Nov 27 11:57:45 mickeyl: phone interface is so good it doesnt need screen Nov 27 11:57:48 mickeyl: travel to Poland - here each operator has 900/1800 Nov 27 11:58:51 mickeyl: so it is not simple for developers? blame them Nov 27 11:59:26 bitbake trunk fails in ugly way then someone remove dir for build... Nov 27 11:59:44 hrw|work: i find it much more complicated than Qt. In Qt everything follows a couple of basic principles and you can often use a class without looking into the manual because everything is done in a canonical way. In Gtk it's -- unfortunately -- a bit different. Nov 27 12:00:33 then again, i have 6 years practice with Qt and just 1 with Gtk Nov 27 12:00:35 could be it Nov 27 12:00:43 http://pastebin.ca/259377 Nov 27 12:01:16 mickeyl: I do not like those pseudoobjects in gtk code. too much conversions in code for me Nov 27 12:01:45 *nod* Nov 27 12:02:01 * mickeyl goes and reads a bit in the developer handbook Nov 27 12:02:03 l8er Nov 27 12:02:06 cu Nov 27 12:03:47 why no matchbox-panel 0.9.3 in OE? Nov 27 12:06:57 hrw|work: scary summon mallum power there Nov 27 12:07:20 hey XorA Nov 27 12:07:55 hey mallum Nov 27 12:15:37 hey mickey|reading Nov 27 12:15:42 hey pH5 Nov 27 12:15:46 hey koen Nov 27 12:15:56 mallum: why no matchbox-panel 0.9.3 in OE? Nov 27 12:18:38 hrw|work: no idea Im afraid Nov 27 12:19:45 I know why Nov 27 12:20:05 koen: we have 0.9.3.. but not in angstrom. Nov 27 12:20:06 because I only did 'mtn add' not 'mtn commit' Nov 27 12:20:20 DEFAULT_PREFERENCE = "-1" Nov 27 12:20:33 http://www.angstrom-distribution.org/repo/?action=details&pnm=matchbox-panel Nov 27 12:21:10 03koen 07org.oe.dev * rb53b3fbf... 10/ (1 packages/matchbox-panel/matchbox-panel_0.9.3.bb): mb-panel: add 0.9.3 which has been in my local tree for ages, ooops Nov 27 12:21:11 koen: for progear it build 0.9.2 Nov 27 12:21:28 I couldn't build progear, missing kernel Nov 27 12:21:33 I know Nov 27 12:21:39 will push it today Nov 27 12:28:05 koen: pushed kernel Nov 27 12:28:44 03hrw 07org.oe.dev * rdf355863... 10/ (5 files in 4 dirs): Nov 27 12:28:44 linux: added 2.6.19-rc6 kernel without any patches Nov 27 12:28:44 - this recipe can be used as base for those builds which do not need Nov 27 12:28:44 any patches other then mainline Nov 27 12:28:44 - supports only progear for now Nov 27 12:28:44 - adding new machines is easy - create linux/MACHINE/defconfig and build Nov 27 12:29:13 this recipe does not have COMP_MACHINE Nov 27 12:29:29 it does have a defconfig check? Nov 27 12:29:45 no - will fail on do_unpack Nov 27 12:29:47 if ![-e defconfig]; then die? Nov 27 12:29:58 Genesis: I got the windows media codecs running now Nov 27 12:30:25 Genesis: although I have one problem and that is that mplayer only plays in fbdev mode :( Nov 27 12:30:47 oki Nov 27 12:30:54 i'm in fbdev so .. Nov 27 12:31:14 Gensis: I cannot get mplayer to do fullscreen Nov 27 12:31:32 Genesis: In general the video is played in the left corner of the screen Nov 27 12:31:34 gerwinin: fbdev does not support fullscreen Nov 27 12:31:52 gerwinin: you need to -fs -zoom -xy WidthOfFb Nov 27 12:32:25 hrw|work: you missed -fbmode "sixe just bigger than video" Nov 27 12:32:30 gerwinin: fbdev and x11 are stupic outputers which does not do any scaling in fullscreen Nov 27 12:32:48 XorA: I do not use fbmode on my x86 when doing fs on fb Nov 27 12:32:56 hrw: I tried to compile with DirectFB but than it doesn't activate that Nov 27 12:33:04 hrw|work: saves some cpu cycles when scaling Nov 27 12:33:05 NOTE: package gnome-desktop-2.10.2-r1: task do_configure: failed Nov 27 12:33:17 XorA: cpu is powerfull enough ;D Nov 27 12:33:24 | configure: error: Couldn't find scrollkeeper-config. Please install the scrollkeeper package Nov 27 12:33:55 hrw|work: and dont insult poor x11 and fbdev, they cant do scaling because the target doesnt support scaling Nov 27 12:34:39 XorA: from developer point of view I know that they do what they can. I was saying from user POV Nov 27 12:34:50 ~lart users Nov 27 12:34:50 * ibot_ crushes users with a full height scsi disk Nov 27 12:35:54 -vo scale? Nov 27 12:36:08 koen: how did you got gpe-image built when there is not scrollkeeper so gnome-desktop fails? Nov 27 12:36:26 I probably have scrollkeeper locally Nov 27 12:36:31 koen: then push it Nov 27 12:36:43 apt-get install scrollkeeper Nov 27 12:36:48 hrw: How can I activate DirectFB ? Nov 27 12:37:10 koen: thats not sollution Nov 27 12:37:17 gerwinin: never used Nov 27 12:37:39 hrw|work: it is a solution Nov 27 12:38:01 koen: especially on remote boxes without root access and with hard to contact admins Nov 27 12:38:39 crappit db3-native fails to compile Nov 27 12:40:23 skipping gpe-image for now. time for xfce-image Nov 27 12:45:50 * gerwinin is going to find out why DirectFB is not working Nov 27 12:46:16 hrw: How do I autostart apps in gpe on boot ? Nov 27 12:46:42 gerwinin: add them to /etc/X11/xinit.d Nov 27 12:46:44 thats rather question to florian_kc, koen, XorA - I'm mostly opie guy Nov 27 12:47:28 Koen: with a symbolic link ? Nov 27 12:48:03 no, with a helper script Nov 27 12:48:15 gerwinin: I expect you actually want them in Xsession.d when you log in Nov 27 12:48:44 No when gpe is started than I want that firefox boots up Nov 27 12:48:55 gerwinin: instead of login screen? Nov 27 12:49:08 No the login screen I killed Nov 27 12:49:34 then in Xinit.d, but thats gonna be a security nightmare Nov 27 12:50:00 XorA: I agree Nov 27 12:50:43 XorA: what I was wondering as well is that I see Xvia, Xvesa etc... how do I know which one I am running ? Nov 27 12:51:59 ps auxw | grep X Nov 27 13:21:58 hi zecke Nov 27 13:22:10 hey... Nov 27 13:22:32 koen|away: booked your train to eindhoven? Nov 27 13:23:07 otherwise I would consider to chicken out :) Nov 27 13:28:31 re Nov 27 13:32:31 zecke: no need to book, it's free for students :) Nov 27 13:32:39 zecke: I'll be at t-dose saturday Nov 27 13:33:23 I will either book my train or chicken out as I have a test on monday Nov 27 13:33:37 it depends on how much you are upset :) Nov 27 13:36:34 koen: ok, cancelled my family meeting. dinner it is. When zecke is eating all frikandellen at the train station, we can sit in a nearby restaurant. Nov 27 13:39:13 zecke: Dinner on saturday? :-) Nov 27 13:52:04 someone want xfce 4.4-rc? Nov 27 14:02:01 morning Nov 27 14:03:31 hey chouimat Nov 27 14:04:47 likewise: hmm, so no chicken-out from me? Nov 27 14:06:57 is eabi target broken? I tried to compile my stuff for eabi but glibc-2.3.5+cvs20050627 bails out with: :2: Error: cannot represent SWI relocation in this object file format Nov 27 14:07:05 anyone had this? Nov 27 14:07:25 Jin^eLD: mixing 2.3.5 and EABI is never going to work Nov 27 14:08:19 well, it always says that 2.5 is not avaialable Nov 27 14:08:19 my prefereed glibc version is 2.5 Nov 27 14:08:19 so I anyway wonder why it is taking 2.3 Nov 27 14:08:19 hi, all! Nov 27 14:08:19 what is format for tag= attribute for git? Nov 27 14:08:19 Jin^eLD: did you also set glibc-intermediate and glibc-initial to 2.5? Nov 27 14:08:25 let me check.. Nov 27 14:08:59 ok, I set the other two but forgot initial :) thanks Nov 27 14:10:12 Jin^eLD: always check angstrom-2007.1.conf for EABI stuff Nov 27 14:10:34 XorA: I did the same but I guess I missed the initial thing somehow Nov 27 14:12:43 Jin^eLD: blame koen :-D Nov 27 14:13:13 =) Nov 27 14:13:41 still says NOTE: preferred version 2.5 of glibc not available Nov 27 14:14:29 anybody knows about GIT support in bitbake? Nov 27 14:15:13 what should I put in .bb file to force it keep repo at some particular time? Nov 27 14:15:30 like SRCDATE for CVS and SVN? Nov 27 14:16:01 srcdate doesn't make sense for a dscm Nov 27 14:16:47 Jin^eLD: it always says that Nov 27 14:16:56 Jin^eLD: even when it works Nov 27 14:17:07 hmm.. but here it tries to build 2.3.5 again Nov 27 14:17:31 well, I now tried setting PREFERRED_PROVIDER_virtual/arm-linux-gnueabi-libc-for-gcc that seemed to be missing, lets see if it helps Nov 27 14:18:42 ok, seems that did the trick Nov 27 14:19:09 thanks Nov 27 14:20:35 :-D Nov 27 14:21:47 :> Nov 27 14:21:53 is the bugtracker down ? Nov 27 14:22:38 Ifaistos: try bugs.treke.net Nov 27 14:24:16 ~lart pkgconfig Nov 27 14:24:16 * ibot_ hereby declares pkgconfig a troll Nov 27 14:24:21 ~lart pkgconfig.bbclass too Nov 27 14:24:22 * ibot_ nabs the moon and broadsides pkgconfig.bbclass too with the sea of tranquility Nov 27 14:24:48 to get xfce working OK I need to include whole pkgconfig.bbclass and do its mangle and own one too Nov 27 14:26:35 hrw|work: xfce.bbclass time? Nov 27 14:27:14 it exist already Nov 27 14:27:26 resolved with fixstage task Nov 27 14:27:55 26 addtask do_fixstage after do_stage Nov 27 14:28:02 lets test now Nov 27 14:32:10 * XorA can see this expand to do_fixfixfixfixstage() Nov 27 14:32:33 do_fixXorA Nov 27 14:33:13 koen: that is beyond even bitbake powers Nov 27 14:33:27 NOTE: package xfce-mcs-plugins-4.3.99.2-r0: task do_rebuild: completed Nov 27 14:37:01 XorA: you're still using OABI on the s3c2410, right? Nov 27 14:37:21 koen: Im just compiling angstrom for it, as generic distro is busted Nov 27 14:37:47 zecke: well, there is not much to chicken away from - I am not really impressed by most topics. Nov 27 14:37:55 I was wondering if #1616 happens on armv4t OABI as well Nov 27 14:38:38 btw, if I use EABI, do I have to enable CONFIG_AEABI in the kernel? Nov 27 14:38:47 Jin^eLD: yes Nov 27 14:38:54 thx Nov 27 14:39:13 see how linux-openzaurus.inc does it Nov 27 14:39:27 XorA: don't forget to build an eabi kernel :) Nov 27 14:39:28 koen: Did you see my dinner ACK? Nov 27 14:39:34 likewise: yes Nov 27 14:39:47 koen: Is that OE/Gerwin(in) on the speakers list? Nov 27 14:40:00 gerwinin should be organizing it Nov 27 14:40:18 koen: I would look at 1616 but bugtracker has gone AWOL Nov 27 14:40:24 koen: ha, I don't even know the locals :-) Nov 27 14:40:39 koen: for me, nothing but new faces. Nov 27 14:41:14 XorA: http://rafb.net/paste/results/rqj16K42.html Nov 27 14:41:50 likewise: yes, 'hacking hardware' is being brought to us by gerwinin Nov 27 14:42:00 koen: nice Nov 27 14:42:34 koen: nice bug :-) Nov 27 14:42:41 how can i prevent oe to build empty dbg package.. i dont need it for my own app Nov 27 14:43:00 rob_w: you cant Nov 27 14:43:52 oh poor me .. Nov 27 14:44:30 ok then i need to live with it Nov 27 14:44:54 NOTE: Couldn't find shared library provider for libc.so.6 Nov 27 14:45:17 what insanity is going on :-( Nov 27 14:46:50 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Nov 27 14:47:34 something seems to be ill Nov 27 14:48:24 XorA: out-of-oe toolchain? Nov 27 14:49:47 XorA: bitbake glibc -c rebuild? Nov 27 14:50:12 hrw|work: no, all oe built, and Im seeing with both generic and angstrom distros Nov 27 14:50:29 weird Nov 27 14:52:43 XorA: eabi armv4t stuff: http://www.angstrom-distribution.org/unstable/images/20061127/ep93xx/ Nov 27 15:00:07 koen: plans for xfce-image? Nov 27 15:00:28 hrw|work: plans yes Nov 27 15:00:38 but xfce needs to gets fixed first in OE Nov 27 15:00:54 #define fixed Nov 27 15:01:10 MX31 we can have a gnome/Xgl image :-) Nov 27 15:01:18 I'm upgrading it to 4.4-rc now Nov 27 15:01:58 XorA, : which board do you have ? Nov 27 15:02:16 the devkit light .? Nov 27 15:02:48 Genesis: lrg has it Nov 27 15:03:41 oki Nov 27 15:03:49 sounds very good this mx31 Nov 27 15:05:07 hrw|work: 4.3+4.4-rc? Nov 27 15:06:45 how to include only one file which is installed to ipkg? Nov 27 15:07:05 4.3.99.2 Nov 27 15:07:34 slapin_nb: FILES_${PN} = "/only/one/file" Nov 27 15:07:42 koen, thanks a lot! Nov 27 15:11:58 I need to configure source twice for package (one with debug and one is not), how it is best to accomplish? Nov 27 15:16:56 koen, any ideas? Nov 27 15:16:56 NOTE: package xfce-image-1.0: completed Nov 27 15:17:38 slapin_nb: nope Nov 27 15:19:14 Crofton: did you receive your efika yet? Nov 27 15:20:20 koen: you? me not. Nov 27 15:20:43 nothing here Nov 27 15:22:12 let's see if that new ubuntu kernel still will let me boot. Nov 27 15:26:11 looks like the kernel didn't work Nov 27 15:26:25 koen: not surprising if its an ubuntu one Nov 27 15:26:34 :) Nov 27 15:28:45 03hrw 07org.oe.dev * r66321e2f... 10/ (3 files in 2 dirs): DejaVu font: updated to 2.12 Nov 27 15:28:52 03hrw 07org.oe.dev * r41d86dd6... 10/ (1 packages/ttf-fonts/ttf-dejavu_2.12.bb): ttf-dejavu: fixed SRC_URI Nov 27 15:28:58 03hrw 07org.oe.dev * r213411c9... 10/ (1 packages/tasks/task-base.bb): task-base: added 'acpi' machine feature Nov 27 15:29:03 03hrw 07org.oe.dev * rce858b81... 10/ (1 conf/machine/progear.conf): progear.conf: added acpi feature, set fstypes to tar.gz Nov 27 15:30:44 ok, my Angstrom image is also missing glibc, this is a pain Nov 27 15:33:25 anyone been making changes to core .bbclass stuff over the weekend? Nov 27 15:38:26 XorA, : do you know if someone is working on mx21 usbotg ? Nov 27 15:38:44 Genesis: no idea, I dont do mx21 stuff Nov 27 15:38:57 XorA: http://www.openembedded.org/bonsai/query/file/bbclass/?scm_id=3&items=80 Nov 27 15:41:57 koen: can you remind OEML people that task-bootstrap is going to be removed? Nov 27 15:44:39 03koen 07org.oe.dev * r5b764c3f... 10/ (1 conf/machine/guinness.conf): guinness: add acpi Nov 27 15:45:34 man, I hate hand-holding people Nov 27 15:45:49 ? Nov 27 15:45:51 hrw|work: dwmd2 just said that lists.infradead.org is still down, but the mailing list archive is at http://www.infradead.org/pipermail/linux-mtd Nov 27 15:46:08 hrw|work: "please look at removal.txt again, it was updated" Nov 27 15:46:22 hrw|work: "removal.txt was updated again, please read it" Nov 27 15:46:36 hrw|work: there's also a #libertas channel where some develoers sometimes hang around (currently just one, besides me) Nov 27 15:46:41 hrw|work: "you need to inhale oxygen to stay alive, please do so" Nov 27 15:46:46 schurig: ok Nov 27 15:49:48 03hrw 07org.oe.dev * r289c4ecd... 10/ (1 packages/phalanx/phalanx_22.bb): phalanx: fix SRC_URI Nov 27 15:52:37 I'm going to update removal.txt in next push Nov 27 15:56:20 Is there any way to rebuild all updated packages? Nov 27 15:56:32 and these which depend on them? Nov 27 15:58:42 03hrw 07org.oe.dev * rf93795c4... 10/ (1 removal.txt): removal.txt: added 'Proposed by' field, added new entries: mono, zeecookbook, qpe-vexed Nov 27 16:02:07 anyone know why I get "cannot stat `etc/hotplug/{*.{agent,rc},hotplug.functions}': No such file or directory" when trying to bitbake gpe-image? Nov 27 16:02:58 TheCompWiz: it works with .dev - which branch/tree you use? Nov 27 16:03:00 * mrz80 is back (gone 01:16:11) Nov 27 16:03:05 the dev branch. Nov 27 16:03:10 just updated this morning. Nov 27 16:04:51 http://pastebin.ca/259628 Nov 27 16:07:44 apluche Nov 27 16:12:29 anyone know of an arm processor with a FPU? Nov 27 16:12:36 mx31 Nov 27 16:12:44 ep93xx Nov 27 16:13:22 if I am compiling, what is a good flag to use to indicate no FPU? Nov 27 16:13:22 so.... any ideas why nothing is being put into etc/hotplug? Nov 27 16:13:46 TARGET_FPU="soft" Nov 27 16:13:56 TheCompWiz: not until I figiure out why my /lib is devoid of libc6 goodness Nov 27 16:13:58 this is passed to the compiler? Nov 27 16:14:05 sounds fun XorA Nov 27 16:14:23 Crofton: it is (on arm) Nov 27 16:15:49 hmm, I do not see it in eabi build Nov 27 16:16:02 hey Crofton Nov 27 16:16:03 Crofton: -msoft-float Nov 27 16:16:33 Crofton: I now finally got an eabi build too Nov 27 16:16:33 :> Nov 27 16:16:42 Jin^eLD, does it work? Nov 27 16:17:04 03hrw 07org.oe.dev * rf537311d... 10/ (4 files in 2 dirs): lmsensors-apps: updated to 2.10.1, dropped older Nov 27 16:17:11 I did not test much yet. but as far as I can tell the rootfs itself seems to be ok and running Nov 27 16:17:56 dropbear runs? Nov 27 16:18:05 let me check Nov 27 16:18:59 hmm no.. but for other reasons Nov 27 16:19:05 it seems I do not have a random device anymore Nov 27 16:19:13 hmm. I thought there was one Nov 27 16:20:27 hmm, but my server still does not run. a version which I built with a different set of tools works Nov 27 16:20:51 but from the moment I switched... does not work anymore, out of memory problem Nov 27 16:20:52 hmm Nov 27 16:21:03 hi Nov 27 16:21:17 hi CoreDump|home Nov 27 16:21:17 hi Nov 27 16:31:04 How can I figure what flags my compiler is passing to the program being compiled ... Nov 27 16:33:15 Crofton: you should see that during compilation, if you mean in OE this will probably be in compile.log Nov 27 16:33:32 I suspect there are more passed than are in the log Nov 27 16:34:13 if you really mean flags and not some defines, I believe they must be on the gcc command line... but well. I may be wrong :) Nov 27 16:35:14 defines ... Nov 27 16:36:34 if it uses autoconf, defines will probably land in config.h Nov 27 16:36:54 and the ones that do not should be visible on the command line, the -Dstuff Nov 27 16:59:37 is there any incobatibility with the latest uclibc and the kernel-headers ? Nov 27 16:59:52 uclibc-0.9.28 Nov 27 17:00:28 uclibc-initial fails to compile for x86 Nov 27 17:10:39 anyone have any good libs for doing fixed point math on integer only machine? Nov 27 17:13:28 re Nov 27 17:13:45 Crofton: you mean for softfloat stuff? isnt libm supposed to take care of that? Nov 27 17:14:14 no Nov 27 17:14:20 like what libmad does Nov 27 17:14:45 oh Nov 27 17:16:13 Crofton : There is Gofast float point lib which is supposed to faster than the gcc ones Nov 27 17:16:25 /jo Nov 27 17:16:28 sorry:) Nov 27 17:18:38 Ifaistos, that is over kill for me Nov 27 17:18:50 I just need to extract the fixed point stuff from libmad Nov 27 17:19:01 and am wondering if that is already done somewhere ... Nov 27 17:20:12 Crofton : I was looking for float to fixpoint to convert some codecs but it turned out to be a lot more difficult than i originaly thought Nov 27 17:20:33 have you looked at libmad? Nov 27 17:22:51 koen|away: have you seen the Trolltech Microcross deal? Nov 27 17:26:21 Crofton : Yes. The problem i had was that the dynamic range of variables changes and you need to find out what type of fixpoint (q15, q31) and what type of scaling is need it Nov 27 17:27:09 yeah, I do not need to do anything that complex Nov 27 17:28:15 Crofton : i think i have some macros from int to fixed point and vice versa. let me look tommorow Nov 27 17:30:55 I think I am going to figure out how to get the stuff from libmad to build Nov 27 17:31:28 mainly need to work out the auto foo needed to set the flags Nov 27 17:40:18 zecke: no, haven't seen it yet Nov 27 17:55:07 How could I setup package to build only for one architecture like for x86, or for ARM? Nov 27 17:55:34 Also - is it possible to patch only for one architecture (or several fixed ones)? Nov 27 18:00:24 bugs.treke.net works? Nov 27 18:01:04 well that's surprising :) Nov 27 18:13:32 ggilbert: do you know if kergoth switched over dns yet? Nov 27 18:13:38 slapin: DEFAULT_PREFERENCE = "-1" ; DEFAULT_PREFERENCE_arm = "1" should do the trick (untested) Nov 27 18:14:03 ~lart gphoto2 for usr/include Nov 27 18:14:03 * ibot_ pulls out his louisville slugger and uses gphoto2's head to break the homerun record for usr/include Nov 27 18:14:05 COMPATIBLE_ARCH would be more like it Nov 27 18:14:38 hey hrw Nov 27 18:14:42 slapin_nb: also SRC_URI_arm += "your-patch.pach" should answer the second question Nov 27 18:15:01 koen: ah I didn't know that variable Nov 27 18:15:19 DEFAULT_PREF doesn't do much when you only have one version of a recipe :) Nov 27 18:15:31 koen, I'll poke him again today. I thought he did it, but it doesn't work for me Nov 27 18:15:58 ggilbert: dns propagation is voodoo for me, so that's why I asked Nov 27 18:16:26 will I have to write COMPATIBLE_ARCH for all i586 i686 ... etc? or there is generic name for x86? Nov 27 18:16:45 koen, it shouldn't take this long to propagate Nov 27 18:17:03 "*.86"? Nov 27 18:17:06 regexps should work Nov 27 18:17:25 it depends on how long you set the time out on the records to be. Nov 27 18:18:38 koen: rather '.*86' Nov 27 18:18:44 ggilbert: btw, during european morning theres lot of packet loss, is that due to cronjobs? Nov 27 18:19:36 koen, I got a ton of pages, but i wasn't sure why. Someone today mentioned that there were some problems in the area Nov 27 18:19:51 coincidentally it was the first rainy day we've had this year Nov 27 18:20:22 heh Nov 27 18:24:35 having ia32 libs in system helps to fsck build Nov 27 18:26:52 we have broken popt Nov 27 18:29:04 gphoto2 use popt and get /usr/include Nov 27 18:36:00 that sucks Nov 27 18:37:24 NOTE: package bash-3.2: completed Nov 27 18:42:57 koen: argh, I ran into your fsck bug :\ Nov 27 18:43:13 heh Nov 27 18:43:20 reminds of my debian times :( Nov 27 18:43:35 it happens when errors were corrected, but not generally after fsck is run Nov 27 18:49:07 hey mreimer Nov 27 18:49:12 hey koen Nov 27 18:49:46 koen, thanks for your help perf-testing cairo Nov 27 18:50:15 bash 3.2-005 will get pushed Nov 27 18:50:26 koen: for whatever reason, the returncode is "3" when error were corrected Nov 27 18:50:46 CoreDump|home: all hail sane return codes! Nov 27 18:51:29 the question now would be what the returncode on a FUBAR FS would be Nov 27 18:58:59 CoreDump|home: "2.3" Nov 27 18:59:30 probably =) Nov 27 19:03:33 CoreDump|home : heh... now how to make sure a fs is fubar... or should that be fscked...:) Nov 27 19:03:50 hvontres|poodle: =) Nov 27 19:04:03 I'll rise the returncode to -gt 3.... Nov 27 19:04:20 current behaviour is idiotic Nov 27 19:06:29 anyone using monotone 0.31? Nov 27 19:07:25 CoreDump|home : I am. I pulled both .dev and .oz354 this weekend and it seemed to work Nov 27 19:08:20 hvontres|poodle: excellent, thanks Nov 27 19:21:31 koen|away: did you stop makeing c7x0 Angstrom images boot from SD :-( Nov 27 19:21:55 evening all Nov 27 19:22:48 what is the best way to add a few packages to a task-base image? Is there a variable in local.conf that can be used? Nov 27 19:23:37 hmm, is SVNDIR broken? git and cvs packages land in the correct directories, but my SVNDIR seems to be ignored, stuff goes to DL_DIR instead Nov 27 19:27:57 XorA|gone: if it booted from sd before it was by accident Nov 27 19:28:56 has anyone solved the annyoing "hostap dies on resume" bug yet? Nov 27 19:29:23 (w/ 2.6.17 that is) Nov 27 19:30:26 cbrake: DISTRO_EXTRA_RDEPENDS/MACHINE_EXTRA_RDEPENDS Nov 27 19:30:28 cu Nov 27 19:30:33 hrw: thx Nov 27 19:30:44 Jin^eLD: for sure bitbake trunk version use SVNDIR. But it seems that previous releases not Nov 27 19:32:15 oh OK.. I have 1.6.3, not the trunk Nov 27 19:32:28 thanks Nov 27 19:33:07 CoreDump|home: Ive always had that as Ive got prism3 cards Nov 27 19:33:23 :\ Nov 27 19:33:31 it used to work fine w/ prism 2.5 Nov 27 19:34:04 CoreDump|home: prism3 cards lose firmware when powered down, so I dont think there is much can be done Nov 27 19:34:34 hi all Nov 27 19:34:48 hi likewise Nov 27 19:34:48 hey mr dubblejoe Nov 27 19:35:06 XorA|gone: maybe reload the f/w via apm.d magic= Nov 27 19:35:11 s/=/? Nov 27 19:35:55 doesn't resume trigger an ifup? Nov 27 19:36:04 althouth with my sandisk card in kernel hostap just oopses in 2.6.17 2.6.18 Nov 27 19:36:41 koen: you mentioned nslu2-linux monotone mirror breakage the other day - can you give me more info so I can fix the problem you see? Nov 27 19:36:42 koen: ifup is useless when the driver is on a FUBAR state :\ Nov 27 19:37:01 rwhitby: your sync only runs sporadically Nov 27 19:37:19 koen: may I abuse oe.org for a 300mb private upload? Nov 27 19:37:36 zecke: I don't own oe.org Nov 27 19:38:06 right, do you imagine this could be a problem? Nov 27 19:38:11 03hrw 07org.oe.dev * r027464b2... 10/ (1 packages/libexif/libexif_0.6.13.bb): libexif: added 0.6.13 Nov 27 19:38:14 this will be downloaded by one person and only once :) Nov 27 19:38:15 koen: I changed that about three or four days ago so that it runs reliably every 5 minutes Nov 27 19:38:26 03hrw 07org.oe.dev * r9bac02fe... 10/ (4 files in 3 dirs): bash: added 3.2 with upstream patches: 001 -> 005 Nov 27 19:38:27 zecke: not with my limited imagination Nov 27 19:39:03 koen: we now have a separate monotone database which syncs against monotone.nslu2-linux.org and against monotone.openembedded.org every five minutes. Nov 27 19:39:30 making it a separate database instance stopped it locking the main server, so it runs reliably now Nov 27 19:39:33 it seemed you had a hickup yesterday Nov 27 19:40:12 suddenly ~5 commits by you and mwester appeared Nov 27 19:40:45 it could be that both you and mwester sync to nslu2 at the same time, though Nov 27 19:41:10 it created multiple heads, which is why I as mumbling at broken syncing Nov 27 19:41:14 actually, those commits were close in time Nov 27 19:41:43 I'll check with mwester to make sure he's using the Master Makefile targets to push (which do a pull and merge first) Nov 27 19:42:55 ah well, multiple heads are unavoidable Nov 27 19:43:46 ok, let me know if you see anything else untoward. From our end, it seems 100% more reliable now, but it hasn't been running that way for more than a few days yet. Nov 27 19:45:44 hmm.. the OE wiki says that sed 4.x should be fine; for some reason I get an error when using oe_libinstall. log says: sed: -e expression #1, char 47: unknown option to `s' and then some exlanation on that follows Nov 27 19:45:55 I have sed 4.1.5 Nov 27 19:46:12 any ideas? Nov 27 19:46:32 Jin^eLD: GNU sed? Nov 27 19:46:49 yep, FC6 default, GNU sed version 4.1.5 Nov 27 19:51:00 thats what it says http://pastebin.ca/259843 Nov 27 19:51:12 Jin^eLD: this sounds familiar to me, but I cannot remember the details. It had nothing to do with sed, but with an empty variable somewhere. Nov 27 19:51:27 (only now looking at your pastebin - hold on) Nov 27 19:51:41 oe_libinstall -so build/libBriansAPIfbdev_1bpp ${STAGING_LIBDIR} Nov 27 19:51:45 thats what triggered it Nov 27 19:52:21 later Nov 27 19:52:39 Jin^eLD: seems like the '/' in path is not escaped, hence the bad expression Nov 27 19:53:02 aah, so I screwed up.. as always Nov 27 19:53:26 it would be saffer to use # or : delimiter rather than '/', especaily if path are involved in expressions Nov 27 19:54:30 you mean in my bb file I can just use oe_libinstall -so build:libBriansAPIfbdev_1bpp ${STAGING_LIBDIR} ? Nov 27 19:54:35 hmm.. this does not seem to work though Nov 27 19:54:37 | is valid as well and is nice and readable Nov 27 19:55:18 Jin^eLD: no, I mean something like s:build/libBriansAPIfbdev_1bpp::g Nov 27 19:55:34 or s/build\/libBriansAPIfbdev_1bpp//g Nov 27 19:56:03 in the .bb file?? Nov 27 19:56:15 'cuz it's not me who is sedding, its somewhere below Nov 27 19:57:01 Jin^eLD: hmm, you could try 'build\/libBriansAPIfbdev_1bpp' in your .bb files... i'm looking for a better solution though Nov 27 19:58:19 btw, when I have a library, do I really need to manually write a do_stage() for it? Nov 27 19:59:26 Jin^eLD: ah maybe you need to use the '-C' option of oe_libinstall Nov 27 20:00:22 Jin^eLD: other .bb files use non escaped '/', so '-C' might be what you need Nov 27 20:00:31 thanks, I will try that Nov 27 20:00:40 in the meantime I figured out that I have to search for the libs in the install directory :> Nov 27 20:00:54 is there a variable for it? Nov 27 20:01:09 doh.. each package does it in a different way Nov 27 20:02:33 Jin^eLD: yes, look at the conf/bitbake.conf :) Nov 27 20:03:51 hmm, interesting that none of the library packages I checked so far uses it.. Nov 27 20:04:13 Jin^eLD: you probably need ${D}${libdir} or ${D}${libexecdir} Nov 27 20:08:21 hmm Nov 27 20:08:24 ${D} is image Nov 27 20:08:32 but thats empty for some reason Nov 27 20:09:44 oh what the hell, I'll just do ${WORKDIR}/install Nov 27 20:11:26 ok, that was not the way :> Nov 27 20:12:00 03coredump 07org.oe.oz354x * rdf0ba4e0... 10/ (3 files in 3 dirs): Nov 27 20:12:00 initscripts: Nov 27 20:12:00 - Reformat some echo lines to be more readable on QVGA Nov 27 20:12:00 - echo RETURNCODE after fsck run Nov 27 20:12:00 - raise RETURNCODE for a FUBAR FS to -gt3 Nov 27 20:12:06 03coredump 07org.oe.oz354x * r07da5cac... 10/ (3 files in 3 dirs): zaurusd: Update Poodles asound.state to account for asoc changes Nov 27 20:12:14 03coredump 07org.oe.oz354x * r53a14a7a... 10/ (3 files in 3 dirs): linux: Update 2.6.17 to latest asoc Nov 27 20:12:27 03coredump 07org.oe.dev * r82f9c471... 10/ (3 files in 3 dirs): Nov 27 20:12:27 initscripts: Nov 27 20:12:27 - Reformat some echo lines to be more readable on QVGA Nov 27 20:12:27 - echo RETURNCODE after fsck run Nov 27 20:12:27 - raise RETURNCODE for a FUBAR FS to -gt3 Nov 27 20:12:39 03coredump 07org.oe.dev * r0f87f8be... 10/ (3 files in 3 dirs): linux: Update 2.6.17 to latest asoc * untested in .dev * Nov 27 20:24:18 03crofton 07org.oe.dev * r7dcd3559... 10/ (6 files in 2 dirs): ossie : Add more components. Nov 27 20:24:27 03crofton 07org.oe.dev * rf364b08c... 10/ (3 files in 3 dirs): Nov 27 20:24:27 ossie : Adjust RDEPENDS so that waveforms require the component they need. Nov 27 20:24:27 task-ossie need only depend on the required waveforms to pull in Nov 27 20:24:27 the waveform components. Nov 27 20:36:41 cu Nov 27 20:38:10 * cbrake watches opie build, even though I only asked for xfbdev and gtk+ ???? Nov 27 20:38:49 03coredump 07org.oe.oz354x * r165d86fa... 10/ (3 files in 3 dirs): zaurusd: Unb0rk last cset Nov 27 20:39:05 so then I re-run bitbake bootstrap-image, and the image builds fine, even though libsdl-qpe-1.2.7: failed Nov 27 20:39:20 must be some stray dependencies somewhere. Nov 27 20:39:56 well... you're one step farther than me. I can't even get the dang thing to build. Nov 27 20:44:38 TheCompWiz: x86/generic, and compulab-pxa270/generic seem to be working pretty well. Nov 27 20:46:25 I can't get that far on mine... and I can't figure out why either. Nov 27 20:56:57 * cbrake suspects things will work much better once I set: PREFERRED_PROVIDERS += " virtual/libsdl:libsdl-x11" Nov 27 21:11:19 isn't this line: "install -d /stuff/build/tmp/work/armv5te-linux/linux-hotplug-20040920-r8/image/etc/hotplug/{usb,pci}" supposed to make 2 directories? usb & pci? instead of 1 {usb,pci}? Nov 27 21:13:49 no, globbing doesn't work like that. Nov 27 21:13:57 didn't think so. Nov 27 21:14:09 03mreimer 07org.oe.dev * r2b33aa7e... 10/ (3 files in 3 dirs): Nov 27 21:14:09 mplayer_1.0pre8.bb: Add a patch to hook up the imageon driver (making it Nov 27 21:14:09 actually usable), and build the imageon driver for hx4700 as well. Nov 27 21:14:13 so why would you create a directory "{usb,pci}? Nov 27 21:14:14 that expression would expand to two directories if they already existed, but of course you wouldn't be using "install -d" in that case Nov 27 21:14:32 um.... isn't that what's specified? Nov 27 21:14:56 I'm just trying to deb0rk the bitbake... Nov 27 21:15:24 mreimer: how does mplayer perform on the hx? Nov 27 21:15:29 yeah, I think the install command is just wrong. Nov 27 21:15:37 mreimer: haven been playing too much with the h2200 lately :) Nov 27 21:15:42 so.... should I report it? Nov 27 21:15:43 s/haven/I have/ Nov 27 21:15:55 yup, I guess so Nov 27 21:16:03 grumble... grumble... Nov 27 21:16:04 it should be easy to fix: just change it to specify both directories separately Nov 27 21:16:24 koen: I don't have time to test it right now. A quick test just scrambled the screen, so probably need to talk to libw100 people Nov 27 21:16:46 koen: you seem to like your h2200 :-) Nov 27 21:16:48 hi mreimer Nov 27 21:16:55 hey pb_! Nov 27 21:17:07 mreimer: it's so nice, small, fast and frugal Nov 27 21:17:13 koen: yep Nov 27 21:17:22 guess I should start yelling about how bugs.openembedded.org is down.... Nov 27 21:17:53 I guess I should start yelling about the topic of this channel Nov 27 21:20:22 * TheCompWiz glares at "bugs.treke.net"'s non-response. Nov 27 21:21:09 argos-wlan:~ koen$ ping bugs.treke.net Nov 27 21:21:09 PING treke.net (64.136.190.225): 56 data bytes Nov 27 21:21:10 64 bytes from 64.136.190.225: icmp_seq=0 ttl=53 time=178.073 ms Nov 27 21:21:15 um... bugs.treke.net has an alias to itself. Nov 27 21:21:23 it durn't work that way. Nov 27 21:21:27 http://nwtools.com/default.asp?prog=lookup&host=bugs.treke.net Nov 27 21:21:45 * ggilbert isn't sure how it was set up Nov 27 21:23:52 Looking at it, bugs.treke.net is an alias to treke.net Nov 27 21:24:33 which in turn is not responding. Nov 27 21:24:51 and it looks like the alias is pointing to itself. treke.net is the dns server answering the query. Nov 27 21:25:08 grumble grumble. Nov 27 21:25:31 no it isn't Nov 27 21:25:44 reguardless... the web server isn't accepting connections. Nov 27 21:25:50 or is... and is b0rken. Nov 27 21:26:10 looks like mod_perl is fucked or something Nov 27 21:26:18 since appache seems to be answering Nov 27 21:26:23 yep. Nov 27 21:26:32 as is mod_php Nov 27 21:26:35 * TheCompWiz goes back to grumbling. Nov 27 21:27:44 it shouldn't be responding to pings though Nov 27 21:27:46 that's strange Nov 27 21:28:00 I'm not the admin... so I couldn't tell ya. Nov 27 21:32:14 ah. Part of it is the new t3 that was installed for the server Nov 27 21:32:24 * ggilbert is always the last person to hear about this sort of thing :) Nov 27 21:32:31 well not specificallyf or that server :) Nov 27 21:34:43 koen|away: that would probably explain the packet loss you were seeing this morning Nov 27 21:34:55 * TheCompWiz is content to manage his own fiber connections... Nov 27 22:11:04 pb_: salve! do you think you could make it to FOSDEM'07? I really would like to drink 1,2,3... beer with you again Nov 27 22:18:01 * CosmicPenguin is laying the groundwork to get approval to go to FOSDEM Nov 27 22:20:20 CosmicPenguin: excellent! Nov 27 22:20:26 mickeyl: hello! yes, I think I should be able to manage that. Nov 27 22:20:46 pb_: ah pretty good news. can't wait Nov 27 22:23:47 yeah, should be fun Nov 27 22:26:33 should xserver-kdrive-fbdev work on a x86 system if the fb driver is built for the video card? Or should I be building full X? Nov 27 22:27:52 kdrive-fbdev would be fine - assuming you could deal without having accelerations Nov 27 22:28:35 CosmicPenguin: I don't need the performance, so I'm hoping fbdev will work -- just for simplicity. Nov 27 22:28:59 if its a standard BIOS, use vesafb - that way you get the least amount of churn Nov 27 22:29:07 unless you need a non vesa resolution Nov 27 22:30:25 CosmicPenguin: its a pretty standard pentium-III system. Is vesafb built as part of kdrive? I don't see an ipkg for it. Nov 27 22:30:43 vesafb would be the kernel driver - it would provide a standard /dev/fb interface to kdrive Nov 27 22:30:51 I'm not sure if kdrive has a built in vesa driver Nov 27 22:31:36 CosmicPenguin: I see a Xvesa -- maybe that is it. Nov 27 22:33:17 CosmicPenguin: I'll also try fbvesa in the kernel config. thx for the help. Nov 27 22:40:46 g'night Nov 27 22:42:41 'night mickey|zzZZzz Nov 27 22:56:26 hi florian Nov 27 22:56:33 re Nov 27 22:56:38 hey pb_ Nov 27 22:57:12 pb_: you seem to have a time consuming job now Nov 27 22:57:39 yeah, it seems Nov 27 23:02:36 CosmicPenguin: kernel fbvesa driver + Xvesa (kdrive) seems to work :-) Nov 27 23:02:52 excellent Nov 27 23:02:59 seems to work is always good Nov 27 23:03:38 CosmicPenguin: and without too much pain is good too :-) Nov 27 23:03:54 .me is back Nov 27 23:04:08 * v8jlene needs coffee obviously Nov 27 23:04:14 morning Nov 27 23:04:57 * cbrake needs a simple gtk2 app for testing Nov 27 23:06:55 * cbrake tries dillo2 Nov 28 01:07:15 Anybody knows which mini-ITX cpu would be faster: Eden or C3? Nov 28 01:08:43 I'm looking into a new board for my next robot project which has plenty of processing power but light on the power. Nov 28 01:09:00 a pentium M is probably much better ($$$) Nov 28 01:10:14 how much processing power do you think you need? Nov 28 01:13:27 I read on the oe list about a multithreaded bitbake. I would like to try it, any pointers? Nov 28 01:14:18 emte: well to experiement with stuff like face recognition and lots of vision stuff so a good amount. Nov 28 01:14:34 but not too much as to sacrifice the power demand as it's for a mobile robot. Nov 28 01:20:31 jacques: try bitbake trunk Nov 28 01:21:30 JustinP, cool. is there something I need to do to enable multithreaded > Nov 28 01:21:31 ? Nov 28 01:22:50 jacques: no idea, I don't touch trunk currently. I have no need to test broken bitbake on top of broken OE :-/ Nov 28 01:22:58 i.e. I don't have time to fix bitbake bugs too Nov 28 01:23:21 JustinP, heh :-) Nov 28 01:24:08 I saw on the list that multithreaded was working well enough for developers to use it, just that the error messages were very hard to decipher Nov 28 01:25:17 * JustinP shrugs Nov 28 01:25:24 once it's in a release branch I'll use it ;-) Nov 28 01:25:42 sounds like that won't be til there's a ncurses UI for bitbake **** ENDING LOGGING AT Tue Nov 28 02:59:59 2006