**** BEGIN LOGGING AT Sat Feb 10 02:59:57 2007 Feb 10 06:04:47 03rwhitby 07org.oe.dev * r1af65859... 10/ (3 files in 3 dirs): ixp4xx-kernel: Fix redboot breakage in 2.6.20 which was preventing the SlugOS/LE boot. Feb 10 06:04:51 03rwhitby 07org.oe.dev * rb0d6a22a... 10/ (65 files in 11 dirs): packages/obsolete/nslu2: dropped all obsolete nslu2 kernels Feb 10 06:40:35 03rwhitby 07org.oe.dev * re844d673... 10/ (1 conf/machine/include/ixp4xx.conf): ixp4xx.conf: Removed superfluous --faketime in EXTRA_IMAGECMD_jffs2 Feb 10 06:40:39 03rwhitby 07org.oe.dev * r32147916... 10/ (1 conf/distro/include/angstrom.inc): angstrom.inc: Removed SLUGOS_IMAGESEX from throughout OE. Feb 10 06:40:42 03rwhitby 07org.oe.dev * r0e24d38b... 10/ (12 files in 4 dirs): slugos: Removed SLUGOS_IMAGESEX from throughout OE. Feb 10 07:36:00 koen : morning. Feb 10 07:36:26 good morning all Feb 10 07:36:30 koen : i am having trouble with the rep. it seems my key is not recongized as it asks for password Feb 10 07:38:33 Ifaistos: it's 'angstrom@Linuxtogo.org', not '.angstrom@Linuxtogo.org' Feb 10 07:39:42 koen : ok Feb 10 07:40:50 Feb 10 09:12:10 localhost sshd[11783]: Accepted publickey for angstrom from ::ffff:62.169.211.226 port 48943 ssh2 Feb 10 07:58:59 i am uploading... Feb 10 07:59:08 cool Feb 10 08:03:06 firefox does not compile even with a single machine when using icecc. not sure though if icecc is the problem Feb 10 08:03:24 i'll try a "simple" build Feb 10 08:13:19 koen: SLUGOS_IMAGESEX is finally gone. Any other slugos remnants in ixp4xx.conf or angstrom.conf that you can see? Feb 10 08:14:47 ARM_INSTRUCTION_SET = "arm" and THUMB_INTERWORK = "yes" are policy, so they don't belong in a MACHINE.conf Feb 10 08:15:23 same applies for FULL_OPTIMIZATION Feb 10 08:15:58 koen : there seems to be no short.sh in the unshorted dir Feb 10 08:16:20 but there is a sort.sh in the unsorted dir :) Feb 10 08:16:45 need to get coffee Feb 10 08:16:57 ~slap ifaistos Feb 10 08:17:09 * ibot slaps ifaistos, keep your grubby fingers to yourself! Feb 10 08:17:50 koen: so where should ARM_INSTRUCTION_SET and others be initialised to a default? Feb 10 08:18:13 bitbake.conf or tune-thumb.inc Feb 10 08:18:22 and they should be set with ?= instead of = Feb 10 08:19:34 ok, ARM_INSTRUCTION_SET and THUMB_INTERWORK are already in tune-thumb.conf, so I can remove them completely from ixp4xx.conf Feb 10 08:20:24 (and they are ?= there) Feb 10 08:21:53 koen: So FULL_OPTIMISATION should be set in the distro.conf instead of machine.conf? Feb 10 08:22:07 yes Feb 10 08:22:19 ok, I'll move that to slugos.inc Feb 10 08:22:20 it's a policy thing, not an inherent machine thing Feb 10 08:22:24 nod Feb 10 08:22:40 ok, that's those three handled. Feb 10 08:26:27 will push those after I rebuild some images. Feb 10 08:26:49 koen : 1668 packages... Feb 10 08:33:01 koen : done with the index update Feb 10 08:53:10 hi all Feb 10 08:53:37 is here the right palce for some questions about angstrom or it have a dedicated channel ? Feb 10 08:59:52 #angstrom Feb 10 09:02:53 thanks rob_w Feb 10 09:28:17 firefox fails to build for efika either with or without icecc -> http://www.pastebin.ca/348718 Feb 10 09:38:56 Ifaistos: firefox 1.0.7 has no jsautocfg for powerpc Feb 10 09:39:19 Ifaistos: maybe the one from 2.0 works: http://www.openembedded.org/repo/org.openembedded.dev/packages/mozilla/firefox-2.0/powerpc/jsautocfg.h Feb 10 09:55:45 koen: i'll give it a try Feb 10 09:56:16 or build firefox 2.0 :) Feb 10 09:58:44 koen: this is what i am trying first... Feb 10 09:59:10 koen: you never know... Feb 10 10:00:51 i'll let it "bake"... have to go and get a haircut :) Feb 10 10:21:14 03hrw 07org.oe.dev * ra6a1d873... 10/ (3 files in 3 dirs): lighttpd: simplified config (disabled webdav, auth modules) Feb 10 10:32:46 morning all Feb 10 10:37:06 hi Feb 10 10:37:16 hi hrw Feb 10 10:37:16 RP: got mail from Tim Bird? Feb 10 10:37:50 hrw: yes, I saw that last night Feb 10 10:38:15 RP: I'm writing to oe private now Feb 10 10:38:22 hrw: ok Feb 10 10:42:13 sent Feb 10 10:42:17 hi Paul Feb 10 10:43:02 hrw: if you mean me, then hello :) Feb 10 10:46:20 you Feb 10 10:46:49 koen, RP, mickey: can you read and respond? Feb 10 10:51:14 hi another Paul Feb 10 10:51:32 Hi! ;-) Feb 10 11:03:28 hrw: Have read, will repsond shortly (food first) Feb 10 11:04:12 * koen responds Feb 10 11:06:18 ok Feb 10 11:39:47 * RP responds Feb 10 11:44:22 hey people! Feb 10 11:44:29 hey philippe Feb 10 11:44:42 philippe: are you subscribed to the oe list? Feb 10 11:45:04 hi philippe Feb 10 11:45:40 koen: nope. Anything happening I should know about? Feb 10 11:45:43 hi RP! Feb 10 11:45:48 * philippe goes off to subscribe Feb 10 11:46:32 philippe: http://lists.linuxtogo.org/pipermail/openembedded-devel/2007-February/001412.html Feb 10 11:47:38 ok, will do that immediately. Feb 10 11:47:53 you should update the documentation then, that is where I got it yesterday. Feb 10 11:48:36 http://www.openembedded.org/wiki/MaintainerPolicy Feb 10 11:48:42 /usr/libexec is ${libexecdir}, btw Feb 10 11:49:11 philippe: ehm Feb 10 11:49:23 philippe: did you actually read that page? Feb 10 11:49:38 Oops. Seems I skipped the first line... Feb 10 11:49:51 :) Feb 10 11:50:24 That really big Maintainer Policy title takes the focus away from it... Feb 10 11:51:12 Apologies. I will fix things. Feb 10 12:19:26 03philippe 07org.oe.dev * rafe61f58... 10/ (11 files in 11 dirs): (log message trimmed) Feb 10 12:19:26 packages/gpe-mini-browser/gpe-mini-browser_svn.bb Feb 10 12:19:26 packages/libgpelaunch/libgpelaunch_svn.bb Feb 10 12:19:26 packages/libgpepimc/libgpepimc_svn.bb Feb 10 12:19:26 packages/libgpeplugin/libgpeplugin_svn.bb Feb 10 12:19:27 packages/libgpevtype/libgpevtype_svn.bb Feb 10 12:19:29 packages/libgtkinput/libgtkinput_svn.bb Feb 10 12:20:31 that should fix it. Feb 10 12:27:56 And fixed up the maintainerpolicy page so that other people will not read past that first line as I did. Feb 10 12:28:08 :) Feb 10 12:29:16 * koen pastes http://dominion.kabel.utwente.nl/koen/cms/integer-pixel-operations-for-gtk here as well Feb 10 12:30:39 Aha! The famous patch exists! Feb 10 12:31:28 koen: it should bitrot less when we have it in OE Feb 10 12:32:19 like http://www.openembedded.org/repo/org.openembedded.dev/packages/gtk+/gtk+-2.10.9/integer-pixops.patch ? Feb 10 12:32:25 * koen added it by accident Feb 10 12:32:36 :-) Feb 10 12:33:09 So does it work? According to the explanation on your blog, it doesn't Feb 10 12:33:23 Or is it just incomplete? Feb 10 12:33:59 * philippe will bug the Nokia GTK+ people Feb 10 12:34:07 recent gtk versions do: Feb 10 12:34:08 bilinear_magnify_make_weights (PixopsFilterDimension *dim, Feb 10 12:34:08 double scale) Feb 10 12:34:23 the patch does: Feb 10 12:34:24 bilinear_magnify_make_weights (PixopsFilter *filter, double x_scale_d, double y_scale_d, double overall_alpha_d) Feb 10 12:34:39 they changed the function arguments Feb 10 12:35:24 one more thing on my todo list... Feb 10 12:35:26 [12:46] xan: koen, I've been reading the patch and it seems to be a pretty straightforward conversion Feb 10 12:35:26 [12:46] xan: koen, I can make it, but won't be able to test anything until monday Feb 10 12:35:36 that's what xan said today Feb 10 12:36:01 great. Somebody with some more knowledge than me might fix it! Feb 10 12:36:23 * philippe needs more exercise. Feb 10 12:36:56 robtaylor mentioned it could be made even faster with different float->fixed conversions Feb 10 13:06:37 I think I have a slightly newer version of that patch somewhere than the one in the gnome bugzilla. I'll try to find a copy. Feb 10 13:13:15 cu Feb 10 13:17:06 anybody a clue what I am doing wrong here? http://rafb.net/p/lUgToP30.html Feb 10 13:18:52 philippe: ${GPE_SVN} doesn't look like something you'd want with a += ? Feb 10 13:19:22 RP: of course... Feb 10 13:19:34 * philippe needs more sleep it seems. Making stupid mistakes lately... Feb 10 13:20:30 hmmm, seems oe-stylize makes a mistake... Feb 10 13:20:47 if inherit autotools pkgconfig gpe Feb 10 13:20:51 if inherit autotools pkgconfig gpe Feb 10 13:21:08 is not before the SRC_URI it does not fetch from svn. Feb 10 13:21:42 That would be a bug with its reordering... Feb 10 13:22:07 * RP doesn't use oe-stylize for reasons like this... Feb 10 13:22:45 I was just told to do so... grrr, I can fix all those .bb files again now... Feb 10 13:22:48 * koen uses oe-stylize + common sense Feb 10 13:23:22 * philippe will too from now on... Feb 10 13:24:12 does anyone made a gpe-image for qemuarm ? Feb 10 13:24:26 it always fails on pth Feb 10 13:28:18 pth complains about an "unsupported Linux (g)libc and/or platform" Feb 10 13:30:00 wb mickeyl Feb 10 13:34:38 hey mickeyl Feb 10 13:36:57 03philippe 07org.oe.dev * rb841eb67... 10/ (11 files in 11 dirs): (log message trimmed) Feb 10 13:36:57 packages/gpe-mini-browser/gpe-mini-browser_svn.bb Feb 10 13:36:57 packages/libgpelaunch/libgpelaunch_svn.bb Feb 10 13:36:57 packages/libgpepimc/libgpepimc_svn.bb Feb 10 13:36:57 packages/libgpeplugin/libgpeplugin_svn.bb Feb 10 13:36:58 packages/libgpevtype/libgpevtype_svn.bb Feb 10 13:37:00 packages/libgtkinput/libgtkinput_svn.bb Feb 10 13:38:37 cheers guys Feb 10 13:38:46 * mickeyl responds Feb 10 13:39:46 hi mickeyl Feb 10 13:40:14 hi RP Feb 10 13:40:51 mickeyl: I fixed the files you were concerned about... verify if they are ok for you now. Feb 10 13:42:42 excellent, thanks Feb 10 13:42:53 i'm just trying to be a bit strict on new packages Feb 10 13:43:01 old packages still have problems Feb 10 13:43:08 but we should try making it more consistent piece by piece Feb 10 13:43:08 :) Feb 10 13:43:18 s/packages/recipes/ Feb 10 13:43:43 mickeyl: no problem. I should have documented myself a bit better :-) BTW: oe-stylize or bitbake has an annoying bug Feb 10 13:44:03 philippe: which one? Feb 10 13:44:11 mickeyl: see last commit message... Feb 10 13:44:24 ah right Feb 10 13:44:25 well Feb 10 13:44:29 that's not really a bug Feb 10 13:44:46 but rather a problem of oe-stylize not knowing about some tricky orders Feb 10 13:45:02 i don't see how we can fix this by parsing Feb 10 13:45:15 this can be fixed by making oe-stylize _know_ what it parses Feb 10 13:45:30 Well now I know about it is fine, but maybe we should warn about it. Feb 10 13:45:34 Yay, ASoC hit mainline! :) Feb 10 13:45:43 tadaaa Feb 10 13:45:48 ~praise RP, lrg, XorA Feb 10 13:45:53 All hail RP, lrg, XorA! Feb 10 13:46:06 RP: what's your take on dwmw2's battery class from November 2006? Feb 10 13:46:12 I'd really like to have something like that Feb 10 13:46:16 not necessarily in his form Feb 10 13:46:38 mickeyl: Mixed feelings. It doesn't suit embedded hardware but something like it could be made too Feb 10 13:46:59 I have some battery code around I should dig out and clean up... Feb 10 13:47:05 yeah, please Feb 10 13:47:46 I'm taking on trying to clean up the backlight class/drivers: http://git.o-hand.com/?p=linux-rpurdie-backlight;a=shortlog;h=for-mm Feb 10 13:52:22 cool. one more piece to a completely sane hardware subsystems infrastructure Feb 10 13:52:44 RP: will you be at FOSDEM? Feb 10 13:53:01 philippe: yes Feb 10 13:53:24 RP: Good! Try to catch the lunch discussion on sunday in the embedded room. Feb 10 13:53:46 philippe: What's the topic? Feb 10 13:54:20 http://www.fosdem.org/2007/schedule/events/embedded_open_hardware_manager Feb 10 13:55:12 ah, robtaylor-ware Feb 10 13:55:23 * koen is looking forward to seeing that in action Feb 10 13:56:06 Interesting. I keep wanting to write something like that but it just gets too political and I give up again... Feb 10 14:00:14 hey people Feb 10 14:00:38 how has life been treating all of you? Feb 10 14:03:35 mithro: ok thanks, just busy :) you? Feb 10 14:06:37 okay Feb 10 14:07:10 i need to get a social life :P Feb 10 14:08:02 * RP knows the feeling Feb 10 14:09:47 looking forward to finishing uni in the middle of this year Feb 10 14:11:47 been promoting Thousand Parsec quite a bit, we are running an AI comp to try and generate some intrest Feb 10 14:18:30 mithro: What are your plans after uni? Feb 10 14:18:41 RP: don't know yet Feb 10 14:18:48 probably find a job Feb 10 14:18:58 koen: firefox 2 breaks at exactly the same point Feb 10 14:27:29 mithro: Sadly they're generally the best way to earn money :) Feb 10 14:28:15 RP: dunno where I'm going to end up Feb 10 15:06:09 morning Feb 10 16:24:07 Has anyone made b file for things like libnetfilter_queue? Feb 10 16:31:41 koen, philippe: yeah, they did some fast conversions in cairo. you could steal them Feb 10 16:33:52 robtaylor: who did fast conversions? Feb 10 16:36:57 philippe:umm, cairo authors? Feb 10 16:37:15 philippe: talking about fast double->fixed conversions Feb 10 16:37:18 http://gitweb.freedesktop.org/?p=cairo.git;a=blob;h=fe6c2dc3ee545442286aef3e3291a24dde39733c;hb=eec62c297589ca20f4128c889ecad9e96736a07f;f=src/cairo-fixed.c Feb 10 16:37:23 _cairo_fixed_from_double Feb 10 16:38:00 robtaylor: ah okay. I was a bit confused. I did not exactly get what you meant to say. Feb 10 16:38:05 Now I get it :-) Feb 10 16:38:40 heh :) Feb 10 16:38:58 * robtaylor isn;t sure what to make of koen's term 'robtaylor-ware' =) Feb 10 16:39:22 robtaylor: well it seems koen means it in a positive way... Feb 10 16:39:43 heh :) Feb 10 16:40:05 that's probably like using 'cracktastic' in a postive way ;) Feb 10 16:40:56 RP: nice to see someone else has thought of it too Feb 10 16:41:18 RP: lets see if we can manage the politics.. I'm trying to pursuade davidz to come over Feb 10 16:44:52 * philippe hopes politics and ego's will be manageable too... Feb 10 16:45:17 yeah, i always hope good arguments will win out Feb 10 16:54:24 I could bring my fisheye lens to counter davidz's one :) Feb 10 16:55:01 heh, it'd be a photography fight to the death =) Feb 10 16:55:28 koen: did you see the link above for fast double conversion? Feb 10 16:55:39 I already saw it Feb 10 16:55:49 cool Feb 10 16:55:59 I started running cairo perf tests after daniel put his patches onlne Feb 10 16:56:22 rock Feb 10 16:56:57 well, i think i'm gonna push out dbus-glib 0.73 Feb 10 16:57:16 then i might take a look at this gdkpixmap scaling Feb 10 16:59:06 cool Feb 10 17:07:33 koen, you heard of anyone using hte netfilter user space libraries in OE? Feb 10 17:08:13 enter passphrase for key ID [philippe@openembedded.org]: Feb 10 17:08:13 mtn: bytes in | bytes out | certs out | revs out Feb 10 17:08:16 mtn: 1.3 k | 1.6 k | 0/0 | 0/0 Feb 10 17:08:26 hmmm, is this possible? Feb 10 17:08:40 a new push will say I am up to date Feb 10 17:08:55 Crofton|home: maybe harald does Feb 10 17:09:24 philippe: 'mtn sync' ? Feb 10 17:09:46 koen: I just wanted to be sure and not mess up.. Feb 10 17:09:46 http://www.maniacchallenge.org/index.html Feb 10 17:10:07 I'd like to get their software running on the EFIKA? Feb 10 17:10:23 hmmm, mtn sync does exactly the same... Feb 10 17:16:36 http://rafb.net/p/vM37uW11.html is what I get from mtn log Feb 10 17:16:51 However it does not seem to have transmitted the changes... Feb 10 17:17:21 sync and push do nothing, and when I pull & update it tells me I am up to date Feb 10 17:22:34 philippe: please don't make such monster commits Feb 10 17:22:43 philippe: one commit per package please Feb 10 17:22:56 makes reverting and cherrypicking a lot easier Feb 10 17:39:30 koen: ok, but now nothing has been commited :-( Feb 10 17:39:35 I'll keep it in mind though Feb 10 17:58:42 can't find any info on this problem anywhere. Feb 10 18:12:38 grrr, seems I made a typo... and created an org.openembedded,org branch Feb 10 18:12:50 openembedded-comma-org Feb 10 18:17:02 03philippe 07unknown * r711cd6d7... 10/ (13 files in 9 dirs): (log message trimmed) Feb 10 18:17:02 blueprobe_svn: Feb 10 18:17:02 * added bb recipe Feb 10 18:17:02 * added necessary patches Feb 10 18:17:02 detect-stylus_svn: Feb 10 18:17:03 * added bb recipe Feb 10 18:17:05 * added necessary patches Feb 10 18:17:09 03philippe 07unknown * r2b583c51... 10/ (1 MAINTAINERS): Feb 10 18:17:11 MAINTAINERS: Feb 10 18:17:13 * added myself (Philippe De Swert) Feb 10 18:17:15 * added packages maintained by me Feb 10 18:17:17 03philippe 07unknown * r24993d4d... 10/ (1 MAINTAINERS): disapproval of revision '2b583c51398ab09a8a2369e6add94451c09fa796' Feb 10 18:17:20 03philippe 07org.oe.dev * ra5034239... 10/ (1 MAINTAINERS): disapproval of revision '24993d4dbc28b379aa3f3b199a50ac3ecf716342' Feb 10 18:24:27 philippe: you don't need to use -b when committing Feb 10 18:27:05 * koen fixes http://www.openembedded.org/wiki/MonotonePhraseBook Feb 10 18:27:21 I really, *really* hate when people add wrong info to stuff I wrote Feb 10 18:55:48 03shanevolpe 07org.oe.dev * r42f5a725... 10/ (1 packages/directfb/fusionsound_0.9.26+cvs20070207.bb): fusionsound: add package Feb 10 18:55:52 03shanevolpe 07org.oe.dev * rbad4f5a5... 10/ (1 packages/directfb/lite_0.9.26+cvs20070207.bb): directfb lite: add package to OE Feb 10 19:32:09 isthere a cebit presence this year Feb 10 21:27:09 CoreDump|home: ping Feb 10 21:27:10 rwhitby, sure Feb 10 21:27:30 ok, are you familiar with SlugOS, the nslu2 distribution built with OE? Feb 10 21:27:32 but I do have to warn you. I have no OE env in reach, not even the altboot sources Feb 10 21:27:44 heard about it, that's it Feb 10 21:28:23 ok. we currently have some scripts (in slugos-init) which do a pivot-root after booting from jffs2 to external USB disk or nfs. Feb 10 21:28:48 I'm looking at whether it is a good idea to convert those slugos-specific scripts over to using altboot instead. Feb 10 21:28:55 The constraints are these" Feb 10 21:29:32 1) The configuration interface must be text-mode console only - no GUI is available. Feb 10 21:29:48 2) The run-time boot interface must not require a console at all - the NSLU2 has no display or keyboard. Feb 10 21:31:04 3) The configuration interface must allow selection between booting from (a) internal jffs, (b) a ramdisk copy of the internal jffs2, (c) an external USB disk partition, selectable at configuration time by the user, (d) NFS rootfs, details selectable by the user at configuration time Feb 10 21:31:30 4) If NFS or USB disk is not available, it must failsafe back to the internal jffs2. Feb 10 21:31:44 that is quite possible tho not out-of-the-box Feb 10 21:31:45 CoreDump|home: does altboot meet those requirements? Feb 10 21:32:24 it can auto-boot any available option so there is no real need for a boot-time menu Feb 10 21:32:26 nod - I fully expect some work needs to be done, and I'm prepared to do that work if your prepared to support me getting it to work :-) Feb 10 21:33:02 mickey|sports and I have also discussed whether altboot makes sense for the Neo 1973 OpenMoko phone too ... Feb 10 21:33:06 sure Feb 10 21:33:16 rwhitby, well I dunno about that heh Feb 10 21:33:46 yeah, I'm focusing on SlugOS first :-) Feb 10 21:34:14 The idea is that I will leave SlugOS as-is, but get Angstrom on the ixp4xx machine to support using altboot Feb 10 21:34:53 i see. Well, hacking the scripts to fit a slug shouldn't be much of a problem ;) Feb 10 21:35:41 ok - got any pointers on the best way to start, and how to debug changes without having to actually reboot each time? Feb 10 21:35:59 every single boot-option is a script. There are serveral shell function that make your life easier. Have a look at the NFS boot-scrip Feb 10 21:36:48 I used to debug altboot in a shell while fully booted. Only for some specific things you actually have to reboot Feb 10 21:38:32 the job of the boot-scripts is to make the medium (like an NFS driver, an SD card etc) available, then it launches altboots generic functions to select the boot-image etc Feb 10 21:41:21 CoreDump|home: ok, I'll take a look at it in the next few weeks. thanks for the pointers. Feb 10 21:41:46 np, happy hacking Feb 10 21:43:04 CoreDump|home: I presume you'll be happy to accept nslu2 patches to altboot for the svn repo? Feb 10 21:43:30 of course Feb 10 21:46:21 rwhitby: are you one of the nslu2 guys? Feb 10 21:46:24 sorry if this is offtopic, but can you give me a hint on how to get software into one of the nslu2 distributions? Feb 10 21:46:53 I do not own an nslu2, but I do develop some software that I'd like to see in the nslu2 distros :> Feb 10 21:46:53 Jin^eLD: yes, you could say I'm "one of the nslu2 guys" ;-) Feb 10 21:46:57 :) Feb 10 21:48:32 we're deveoping a UPnP mediaserver, there is already a .bb file for it in OE, we will do a new release soon and I had some requests from nslu2 users Feb 10 21:48:32 Jin^eLD: which nslu2 distro? Feb 10 21:48:47 Jin^eLD: which .bb? Feb 10 21:48:58 mediatomb_svn.bb Feb 10 21:49:16 * rwhitby bitbakes mediatomb for slugosbe Feb 10 21:49:17 right now its not yet finished, the release that is, but I think we will have it quite soon Feb 10 21:49:50 Jin^eLD: if it builds and runs, we're happy to put it in the slugos feeds. Feb 10 21:49:53 which distro - no idea :) I think there are several, right? my problem is that I do not own a unit Feb 10 21:50:03 cool, that would be great Feb 10 21:50:23 for SlugOS, just having it in OE is enough (assuming it builds) Feb 10 21:50:39 ah, ok Feb 10 21:50:45 for Unslung, it needs to be added to Optware. Feb 10 21:51:17 so Optware is not OE based? Feb 10 21:51:23 from the .bb file, it looks like adding it to Optware would be easy too. Feb 10 21:51:48 no, it's a buildroot based set of packages. Designed to be compatible with vendor firmware Feb 10 21:51:56 from what distro are the ipk files on http://ipkgfind.nslu2-linux.org/ ? Feb 10 21:52:09 that site has feeds for many many distros Feb 10 21:52:20 I see Feb 10 21:52:52 I found some wl500g stuff there too I think, how is that related to the nslu2? Feb 10 21:54:01 Optware packages run on 10 different platforms, including the wl500g running Oleg's firmware. Feb 10 21:54:27 Jin^eLD: look at template.mk in http://trac.nslu2-linux.org/optware/browser/trunk/make for how to add a package to Optware. Feb 10 21:54:39 olegs firmware worked very well on my WL500g Feb 10 21:55:09 still does on mine :-) Feb 10 21:55:15 I have a user who keeps trying to get mediatomb running on the wl500g with Olegs firmware, but he did not succeed so far, for some reason configure was running in a loop for him Feb 10 21:55:37 and it was difficult for me to help him because well, I do not have a unit and I did not know much about all the firmwares available and so on Feb 10 21:55:39 rwhitby, try dd-wrt. This firmware kicks serious ass Feb 10 21:56:49 CoreDump|home: optware builds for dd-wrt too :-) the wl500g is my production internet gateway, so I can't easily try out new firmware on it until I get my Freecom FSG-3 running a custom distro. Feb 10 21:57:34 thanks for the link.. looking at the template.. Feb 10 21:58:39 Jin^eLD: http://www.nslu2-linux.org/wiki/Optware/AddAPackageToOptware is the doco that goes with it. Feb 10 21:58:55 I'd be happy to give you svn write access so you can maintain the optware package Feb 10 21:59:08 thanks, will take a look at the link Feb 10 21:59:25 well, is that OK if I can't actually test the package on the target hardware? Feb 10 21:59:35 Jin^eLD: testing is for users :-) Feb 10 22:00:04 hehe :)) Feb 10 22:01:12 (that's a direct quote from Linus) Feb 10 22:01:24 so basically what I would need is - get optware, follow the wiki, create a package and be happy if it just builds? Feb 10 22:02:04 yup Feb 10 22:02:37 or talk to eno and he might even just do it for you. he's an optware package adding machine. Feb 10 22:02:51 :) Feb 10 22:03:19 ok, let's see; first we'll make the release, then I'll come back to you on this Feb 10 22:03:28 at least I now have an idea on how to get going Feb 10 22:03:30 the slugosbe mediatomb build is up to id3lib, so that's looking good. Feb 10 22:03:46 n8 all Feb 10 22:04:14 Jin^eLD: with Optware, you'll get the following targets: "nslu2", "wl500g", "ddwrt", "oleg", "ds101", "ds101j", "ds101g", "mss", "nas100d", "fsg3", "ts72xx" and "slugosbe" Feb 10 22:04:34 that's nice Feb 10 22:04:53 hmm - id3lib build failed. Feb 10 22:04:57 doh.. Feb 10 22:05:31 you could try taglib.. but I found that it has some memory leaks and it sometimes crashes on certain mp3 files.. so less stable Feb 10 22:06:22 or course you could disable id3lib completely but that would be kinda bad, because then we can not parse the metadata and create a nice server layout / sort by artist, genre, etc. Feb 10 22:06:25 armeb-linux-g++: /home/slug/slugosbe/tmp/staging/armeb-linux/lib/.libs/libz.so: No such file or directory Feb 10 22:06:32 s/or/of/ Feb 10 22:06:46 oh yeah, id3lib does depend on zlib Feb 10 22:07:18 hmm - failed on building the examples - the library built fine Feb 10 22:07:37 actually its good to have zlib.h for mediatomb too, activates an extra feature Feb 10 22:07:49 zlib is staged fine Feb 10 22:08:53 Jin^eLD: anyway, I'll leave the building in OE up to you :-) Feb 10 22:09:29 well, it does build... koen has it for Angstrom and I did succesfully build it for my arm based distro Feb 10 22:09:35 let me try.. Feb 10 22:10:06 maybe its an armeb/g++ problem. Feb 10 22:10:29 which distro are you using, -bag, bau, btu, lag, lau, ltu whatever that is? Feb 10 22:10:34 I'll try it on SlugOS/LE as soon as my toolchain finishes building. Feb 10 22:10:41 DISTRO=openslug, MACHINE=nslu2 Feb 10 22:11:04 I'll start a testbuild Feb 10 22:11:10 cool Feb 10 22:11:46 all those unused slugos-foo.bb's are going away soon :-) Feb 10 22:11:57 oh ok :) Feb 10 22:11:59 (I'm in the middle of a big SlugOS rationalisation) Feb 10 22:12:33 just testing LE before I push the changes. DISTRO=openslug will still work. Feb 10 22:12:50 lets see how far I get Feb 10 22:22:58 ...caching complete, build started :> Feb 10 22:53:12 03rwhitby 07org.oe.dev * r5e4ea63e... 10/ (3 files in 3 dirs): Feb 10 22:53:12 ixp4xx.conf,slugos.inc: Removed ARM_INSTRUCTION_SET and THUMB_INTERWORK, cause Feb 10 22:53:12 they are already set to defaults in tune-thumb.conf. Moved FULL_OPTIMIZATION to Feb 10 22:53:12 slugos.inc cause it's distro policy not inherent to the machine. Feb 10 22:53:15 03rwhitby 07org.oe.dev * r856725f8... 10/ (12 files in 3 dirs): slugos: Removed the obsolete freeze-packages.inc stuff (but left freeze.inc itself there in case any other distro is still using it). Feb 10 22:53:19 03rwhitby 07org.oe.dev * raec96775... 10/ (10 files in 5 dirs): slugos: Replaced SLUGOS_IMAGENAME with SLUGOS_DISTRO, and localised the usage of SLUGOS_DISTRO to slugos-*.* files only. Removed SLUGOS_FLASH_IMAGE completely (made it always enabled) and cleaned up slugos-image.bb Feb 10 22:53:23 03rwhitby 07org.oe.dev * r9bfe666a... 10/ (7 files in 3 dirs): slugos: Subsumed slugos-btu.conf into ucslugc.conf, and dropped slugos-{b,l}{a,t}u.conf Feb 10 22:53:27 03rwhitby 07org.oe.dev * r1dd9be1f... 10/ (9 files in 3 dirs): Feb 10 22:53:27 slugos: Subsumed slugos-lag.conf into slugosle.conf, slugos-bag.conf into Feb 10 22:53:28 slugosbe.conf, and dropped slugos-{b,l}ag.conf. Deprecated openslug.conf in Feb 10 22:53:30 favour of slugosbe.conf. Deprecated debianslug.conf in favour of slugosle.conf. Feb 10 22:53:32 Eventually this all might be done with multimachine. Feb 10 22:53:34 03rwhitby 07org.oe.dev * rf3aeb26e... 10/ (3 files in 3 dirs): slugos: Added task-slugos to separate out rootfs contents from image creation. Feb 10 22:53:39 03rwhitby 07org.oe.dev * r6072f1d3... 10/ (3 files in 3 dirs): slugos: Update the native distro stuff Feb 10 22:53:41 03rwhitby 07org.oe.dev * rd0ee6e73... 10/ (5 files in 3 dirs): arm-kernel-shim: Upstream bug fixes for little-endian Feb 10 22:53:45 03rwhitby 07org.oe.dev * r70c08428... 10/ (1 packages/tasks/task-slugos.bb): task-slugos: Fixed package generation Feb 10 22:54:30 phew! Feb 10 22:54:40 :) Feb 10 22:55:51 :) Feb 10 22:56:57 one step closer to slugos multimachine. Feb 10 22:57:10 * rwhitby goes to the park with the kids. Feb 10 22:57:46 I guess it's not 00:30 for you then ;) Feb 10 22:58:02 :) Feb 10 23:13:24 yawn Feb 10 23:13:31 re idlib3 problem Feb 10 23:13:39 that's an autotools problem Feb 10 23:13:39 easy fix Feb 10 23:13:43 remove do_configure() entirely Feb 10 23:13:46 and let it regenerate Feb 10 23:13:51 that way it works Feb 10 23:13:59 sorry, didn't have a chance to push that upstream Feb 10 23:14:21 * mickey|sports does now Feb 10 23:14:59 cool... Feb 10 23:15:07 thx Feb 10 23:15:58 rwhitby is in .au Feb 10 23:16:05 the other side of the world :) Feb 10 23:16:14 :) Feb 10 23:16:44 well, I did think that it must be some daytime at his place if he's taking kids to the park :) Feb 10 23:17:25 righto Feb 10 23:40:59 03mickeyl 07org.oe.dev * r5503536a... 10/ (1 packages/id3lib/id3lib_3.8.3.bb): id3lib: fix zlib linking issue by reconfiguring autotools stuff Feb 11 00:04:11 mickey|sports: I can confirm that id3lib is building.. thx Feb 11 00:38:35 rwhitby: ok.. I was able to compile mediatomb for openslug now, had to fix a bug in my configure script, so please make sure to use at least rev 1088 when you try it :) thx Feb 11 00:38:40 rwhitby: nice work. Feb 11 00:55:56 hmm - diffutils configure suddenly fails. wonder when it was last changed. Feb 11 00:56:53 false alarm. Feb 11 01:01:57