**** BEGIN LOGGING AT Wed Mar 22 10:59:56 2006 Mar 22 11:01:29 RP: do you have an idea what the registers mmGRAPHIC_{H,V}_DISP are for? Mar 22 11:03:55 Mardy: Yes. they give details about the current graphics window size Mar 22 11:04:26 Mardy: note that ACTIVE_{H_V}_DISP get set to the same values Mar 22 11:05:04 If you change one, you can do virtual screens of a smaller size. I can't rememebr which is the physical LCD and which is the virtual window though :) Mar 22 11:07:23 zecke: i don't. tomorrow/friday i have to go to the library, though. Mar 22 11:07:32 zecke: beginning from next month i might be at the campus more regularly. Mar 22 11:08:18 pH5: If you feel like drop by at the Solaris room at my faculty Mar 22 11:09:42 funny!!! Mar 22 11:10:05 I'm changing ACTIVE_V_DISP Mar 22 11:10:09 the display pans Mar 22 11:10:43 I still need to understnd *how* i pans, though :-) Mar 22 11:12:38 Mardy: These numbers govern how it combines the data from the video memory into the LCD data stream Mar 22 11:13:15 mmm... this means the framebuffer memory isn't affected, right? :-( Mar 22 11:13:26 Mardy: correct Mar 22 11:13:40 RP: ZKERNEL_VERSION is already set in openzaurus-3.5.4.1.conf Mar 22 11:14:11 Mardy: But by changing the bit of the framebuffer displayed to the LCD, you can effectively scroll the display Mar 22 11:14:27 I think fbcon can make use of that Mar 22 11:14:34 hrw|work: Ah, ok :) Mar 22 11:14:42 hrw|work: I left 354x building last night, after setting the distro to openzaurus-3.5.4.1, and it bombed building linux 2.4.20 Mar 22 11:14:55 Bernardo: I added it in the morning Mar 22 11:15:02 Bernardo: I also dropped 2.4.20 from .oz354x Mar 22 11:15:27 hrw|work: We need to set PREFERRED_PROVIDER_hostap-conf = "hostap-conf" in there as for some reason it defaults to hostap-modules atm :-/ Mar 22 11:15:35 hrw|work: thanks Mar 22 11:15:38 RP: feel free to add it Mar 22 11:15:51 hrw|work: Will do when I get a minute... Mar 22 11:16:38 zecke: you need python2.4-profiler probably installed (non-free) Mar 22 11:16:52 zecke: usr/lib/python2.4/profile.py is in python/python2.4-profiler [non-free] Mar 22 11:17:01 zecke: http://packages.debian.org/cgi-bin/search_contents.pl?word=profile.py&searchmode=searchword&case=insensitive&version=unstable&arch=i386 Mar 22 11:17:24 hrw|work: oh, why is that non-free :} Mar 22 11:17:37 zecke: dont know Mar 22 11:17:51 koen: install that please Mar 22 11:18:03 +Debian note: The license for the "profile" and the "pstats" Mar 22 11:18:04 +modules doesn't conform to the Debian Free Software Guidelines (DFSG). Mar 22 11:18:24 http://packages.debian.org/changelogs/pool/non-free/p/python-profiler/python-profiler_2-5/python2.4-profiler.copyright Mar 22 11:18:38 zecke: http://ftp.debian.org/debian/pool/non-free/p/python-profiler/python-profiler_2-5.diff.gz Mar 22 11:19:25 zecke: done Mar 22 11:20:00 koen: thanks Mar 22 11:20:05 hrm Mar 22 11:20:18 who deleted armeb.patch from openssl? Mar 22 11:21:21 koen: let me guess, jnc? Mar 22 11:22:28 koen: hey, whered you sneak on -) Mar 22 11:22:30 Author: mickeyl@openembedded.org Mar 22 11:22:36 koen: Revision: 4a54c23224bc11e3cb7370ca8f61143214d0cd3b Mar 22 11:22:39 hey XorA Mar 22 11:26:16 koen: this is proper rev? Mar 22 11:28:17 yes, it looks like mickeyl deleted on file too many Mar 22 11:30:30 03koen 07org.oe.dev * r9ed1d43d... 10/packages/openssl/openssl-0.9.7e/armeb.patch: openssl: reinstate armeb.patch Mar 22 11:34:19 hmm, wierd effects building samba_3.0.14a with angstrom distro Mar 22 11:37:13 XorA: gpe-something depend on samba 3.0.14 because koen set it to be like that Mar 22 11:37:33 gnome-vfs-dbus doesn't like the new samba api Mar 22 11:38:39 I know this, but it wont build under angstrom and puts patches in armv5te-linux/ and its source in armv5te-linux-gnueabi/ and when patching fails to find the source Mar 22 11:38:48 koen: in debian it need libsmbclient >3.0.2 and does not conflict with 3.0.21 Mar 22 11:39:58 XorA: sounds like abug in samba or multimachine Mar 22 11:40:12 koen: I cant find a MACHINE= in samba Mar 22 11:40:56 ERROR: Nothing provides dependency libsdl Mar 22 11:40:58 hm Mar 22 11:41:07 bitbake r414 looks a bit broken Mar 22 11:41:36 koen: hmm Mar 22 11:42:04 koen: yeah, better use something earlier for work stuff :} Mar 22 11:42:54 zecke: I know how to work around it :) Mar 22 11:43:36 koen: how? Mar 22 11:43:44 bitbake libsdl-x11 Mar 22 11:44:03 koen: what is the latest stable one with 'working' RDEPENDS Mar 22 11:44:31 zecke: no idea, propably 1.3.2 Mar 22 11:47:23 zecke: is it possible to get bitbake cia commits in #oe? Mar 22 11:48:02 koen: I'm sending commits already somewhere :} Mar 22 11:48:29 koen: hmm I have not touched the hanbdling of 'virtual' Mar 22 11:48:39 koen: and I don't see how it could have any effect :} Mar 22 11:48:46 koen: small test please Mar 22 11:49:02 koen: if you put world into the command line, does it work? Mar 22 11:50:32 zecke: nope Mar 22 11:50:58 koen: well then... Mar 22 11:51:17 koen: then this is the usual RDEPENDS breakage :} Mar 22 11:55:51 koen: What triggered btbake to print that error? Mar 22 11:56:45 RP: who puts RDEPENDS = "libsdl" in the line? or is that in shlibs? Mar 22 11:59:02 zecke: It looks like a DEPENDS error juding by the message but without more information its near impossible to debug Mar 22 11:59:48 RP: 'bitbake libsdl' Mar 22 12:00:10 koen: And you'd expect that to work? Try bitbake virtual/libsdl Mar 22 12:02:33 hehe Mar 22 12:02:55 XorA, your problems are due to samba using S:="" instead of S="". Sorry, I forgot to push this fix. Mar 22 12:02:59 * koen wonders what caused this laps of thinking Mar 22 12:04:11 03pH5 07org.oe.dev * r8ffe1c51... 10/packages/samba/ (samba_3.0.14a.bb samba_3.0.20.bb): samba: replace := with =, fixes builds with TARGET_VENDOR set Mar 22 12:04:32 pH5: cool, cheers Mar 22 12:04:37 pH5: now thats service Mar 22 12:10:00 RP: ping? Mar 22 12:10:23 RP: You wrote the led class stuff, right? Mar 22 12:10:58 so how many devices have had an angstrom image booted now then? Mar 22 12:12:32 NAiL: pong. yes Mar 22 12:12:46 mickey|zzZZzz: hey Mar 22 12:12:47 for h in _handlers: Mar 22 12:12:47 if handler is h: Mar 22 12:12:47 _handlers.remove(handler) Mar 22 12:13:01 that sounds inefficent Mar 22 12:13:16 RP: any idea how much work it would be to add a trigger to the scsi layer? I see there's an IDE-disk trigger, but things like the slug only has scsi (usb) disks. Mar 22 12:13:48 NAiL: triggers and have 5 lines of code. The tricky bit is working out where to put them Mar 22 12:13:54 RP: The LED class stuff hasn't hit mainline yet, has it? I think I want to use it... Mar 22 12:13:54 s/and/can/ Mar 22 12:14:08 Noodles: I'm working on that Mar 22 12:14:13 Its in -mm Mar 22 12:14:23 Ah, cool. I've seen the posts to l-k about it in general. Mar 22 12:14:29 and in use on a lot of slugs and slug-like devices ;) Mar 22 12:14:40 I want to use it for the Amstrad E3 LEDs. Mar 22 12:15:32 Noodles: Does that device have mainline support? Mar 22 12:16:07 Not yet. I'm waiting for my linux-omap-open-source subscription to be approved so I can submit a base patch. Mar 22 12:16:52 Noodles: I'm certainly keen for the interface to get used. It should be straightforward enough :) Mar 22 12:17:17 Excellent. I have 6 LEDs to play with and I like the idea of a NAND access one. :) Mar 22 12:18:53 The NAND access one scares people on Zaurii :) Mar 22 12:19:02 Oh? Mar 22 12:19:16 they will use to this and will miss it if resign from OZ Mar 22 12:19:26 The way it glows is rather surreal :) Mar 22 12:20:20 hrw|work: You are doing an amazing job with the releases... Mar 22 12:21:42 RP: thx Mar 22 12:25:26 hi! i try to run gpe image first time on my system, and i don't see any messages in syslog, someone and idea why? syslog runs ... Mar 22 12:26:23 grma: tried 'logread'? Mar 22 12:27:29 logread shows only that syslog is started Mar 22 12:27:52 is there any dev-node necessary ? Mar 22 12:29:10 i run a older 2.6.14 kernel which is not customized for gpe image but i only want to see how gpe works and if Xfbserver runs and the performance would be ok on our boards ... Mar 22 12:29:36 I will get some food Mar 22 12:30:21 ~lart -Werror Mar 22 12:30:22 * ibot installs PocketPC on -Werror's PDA Mar 22 12:30:42 hrw|work: ahh now something comes out, i think a nfs-root problem maybe, or ... ? Mar 22 12:32:12 or ... xtscal.bin runs with 100% cpu power ... i think that would be the problem ... Mar 22 12:43:39 ~lart so called users for being dumb Mar 22 12:43:40 * ibot decapitates so called users conan the destroyer style for being dumb Mar 22 12:44:58 hrw|work: life would be easier without users Mar 22 12:45:57 yes! Mar 22 12:50:18 koen: ping Mar 22 12:50:44 pong Mar 22 12:51:15 koen: which method (other then wget -r) can be used to mirror OZ 3.5.4 feeds? Mar 22 12:51:48 koen: mirror by Pendalar need update and I want to contact him Mar 22 12:51:58 * france is away: Away Mar 22 12:56:08 hrw|work: a more intelligent wget script Mar 22 13:17:16 koen: no rsync or other methods? Mar 22 13:20:26 hrw|work: not for people without a shell account Mar 22 13:20:34 ok Mar 22 13:26:44 RP: to avoid OH minions doing duplicate work: florian and I are working on getting tapioca and telepathy into OE Mar 22 13:35:06 koen: ok, I'll mention that to other minions :) Mar 22 13:40:13 RP: do we update to 2.6.16 in next alpha? Mar 22 13:40:51 03koen 07org.oe.dev * rd75e6ebb... 10/packages/sofia-sip/sofia-sip_1.11.7.bb: sofia-sip: add main sofia-sip library Mar 22 13:41:00 hrw|work: Yes, I think so Mar 22 13:41:06 cool Mar 22 13:41:13 unless anyone has a pressing reason why not? :) Mar 22 13:42:31 ~hail ewi and 750K/s Mar 22 13:42:36 * ibot bows down to ewi and 750K/s and chants, "I'M NOT WORTHY!!" Mar 22 13:42:40 hrw|work: I have the TUN defconfig change and a patch to ignore unknown SCR structures pending for .dev which can then go into the branch. Just waiting for someone to test teh SCR patch... Mar 22 13:42:45 Note: BitBake has currently a broken cache Mar 22 13:43:21 RP: there was a guy with such card on oesf - send him PM and kernel info Mar 22 13:44:41 hrw|work: Another member of OH has such a card so I'm awaiting their feedback Mar 22 13:44:52 great Mar 22 13:53:04 hrw|work: I think I have a speed up of 40+ seconds Mar 22 13:53:13 (tmp. I have broken the cache) Mar 22 13:53:22 zecke: nice Mar 22 13:53:39 zecke: 40s on run-with-cache or without? Mar 22 13:54:01 without Mar 22 13:54:26 so stay away from r415 for now Mar 22 13:54:40 hrw|work: use it only to say if things are faster or slower now Mar 22 13:54:46 what is ${STAGING_DATADIR_safe} Mar 22 13:55:09 ignore me, see its defined a line earlier Mar 22 13:55:47 zecke: I use r329 currently Mar 22 13:56:44 my favorite rev Mar 22 13:57:11 if someone finds the time, I would like to know the result Mar 22 13:57:23 r329 vs. r415 without cache, without psyco Mar 22 13:58:05 r329 first run, r329 second run, r415 no cache? Mar 22 13:58:20 zecke: "bitbake umbaumba" will be ok? Mar 22 13:58:27 -p -d Mar 22 13:58:34 we lack umbaumba now Mar 22 13:58:40 or better -d umbaumba Mar 22 13:58:49 -p -d umbaumba? Mar 22 13:58:50 r329 twice, r415 twice Mar 22 13:58:56 hrw|work: no not -p Mar 22 13:59:11 ok Mar 22 13:59:28 shit. removed r329 ;( ave to fetch again.. Mar 22 14:03:47 koen: on angstom distrom, libtool.m4 stages into tmp/staging/arm-linux not tmp/staging/arm-none-linux-gnueabi where .bbs expect to find it Mar 22 14:05:31 zecke: you do not need ouput of bitbake? only time results? Mar 22 14:06:36 <_law_> any ideas why i cant fetch NOTE: Update cvs://anonymous@cvs.savannah.gnu.org/cvsroot/config;module=config;method=pserver;date=20050701 Mar 22 14:08:08 hrw|work: only time Mar 22 14:08:57 ok Mar 22 14:09:35 XorA: ekiga is in OE? Mar 22 14:10:16 hrw|work: dont think so Mar 22 14:10:20 ok Mar 22 14:10:28 XorA: user asked in upgrades thread Mar 22 14:10:45 <_law_> or why it doesnt use version from sources dir? Mar 22 14:10:46 hrw|work: without a headset and someone to talk to, no point me looking at it Mar 22 14:11:13 <_law_> hrw|work, no it ekiga doesnt build (pwlib) Mar 22 14:17:35 zecke: http://pastebin.com/616070 - r329 (ewi) Mar 22 14:20:52 hrw|work: without caches please (both) Mar 22 14:21:12 hrw|work: The first ran is only to warm up the caches Mar 22 14:21:15 zecke: ah. CACHE="" for both? Mar 22 14:21:23 yes please :) Mar 22 14:21:25 ok Mar 22 14:23:15 * koen would recommend minisip for VOIP till we get ekigi,tapioca or telepathy in Mar 22 14:23:33 can I quote this? Mar 22 14:23:43 sure Mar 22 14:23:55 only s/ekigi/ekiga/ Mar 22 14:24:35 someone with the OZ-native SDK should build pwlib :) Mar 22 14:24:48 and apache Mar 22 14:25:05 apache sucks, cherokee rules! Mar 22 14:25:22 thttpd uber alles Mar 22 14:27:09 <_law_> could fix gnu-config problem by renaming tar file in sources dir Mar 22 14:28:38 which sources dir? Mar 22 14:29:53 hm. debian got mc with utf8 patches applied. I would need to rebuild it again.. Mar 22 14:43:06 zecke: http://pastebin.com/616117 Mar 22 14:44:07 hrw|work: that is faster right? Mar 22 14:44:41 zecke: ~1 minute Mar 22 14:46:23 I think that is not too bad Mar 22 14:48:34 <_law_> hrw|work, where bitbake places my sources /home/hrw/devel/sources i think :-) Mar 22 14:48:50 <_law_> o i forgot oe dir :-) Mar 22 14:49:28 _law_: a.. DL_DIR you mean Mar 22 14:49:32 zecke: yep Mar 22 14:49:41 <_law_> hrw|work, right Mar 22 14:50:50 hours of hacking for a one second improvement :} Mar 22 14:50:57 one minute even :} Mar 22 14:52:24 32% time reduction Mar 22 14:52:44 03florian 07org.oe.dev * rabc8d6c0... 10/packages/ (ortp/ortp_0.7.1.bb tapioca/tapioca_0.3.0.bb): ortp, tapioca: new packages. Mar 22 14:52:56 BitBake Broken X now with 32% fluff reduction Mar 22 14:53:32 koen: you makes 13% of my website traffic? "Dutch/Netherlands: 13%" Mar 22 14:54:01 heh Mar 22 14:54:09 hrw|work: We need to benchmark with caches now - well I have to fix the cache first Mar 22 14:55:59 ok Mar 22 14:56:38 koen: we should replace fink with OE :} Mar 22 14:56:42 :) Mar 22 14:56:58 * zecke currently build inkscape for a SVG capable tool Mar 22 14:57:26 that's a bit of a waste of time Mar 22 14:57:34 unless fink that the native gtk+ Mar 22 14:57:46 hrw|work: 31 seconds are spend compiling the python () and python __anonfunc() Mar 22 14:57:55 zecke: http://prdownloads.sourceforge.net/inkscape/Inkscape-0.43-2.dmg?download Mar 22 14:58:02 oh Mar 22 14:58:23 koen: that is download is with native gtk+? Mar 22 14:58:53 no Mar 22 14:58:58 I'm still waiting for that Mar 22 14:59:06 gtk+/X looks like ass on osx Mar 22 14:59:58 koen: gtk+ looks like ass on any OS was what your tried to day? Mar 22 15:00:15 s/day/say Mar 22 15:00:22 *tired* Mar 22 15:02:22 03koen 07org.oe.dev * r0049f684... 10/packages/tapioca/tapioca_0.3.0.bb: tapioca: s/glib/glib-2.0/ Mar 22 15:04:15 koen: doesn't start, I have no open-x11 script Mar 22 15:08:35 03koen 07org.oe.dev * rae3c6599... 10/packages/tapioca/tapioca_0.3.0.bb: tapioca: ship dbus service file in the main package Mar 22 15:09:25 ~seen JustinP Mar 22 15:09:32 justinp is currently on #oe (1d 13h 26m 8s) #openzaurus (1d 13h 26m 8s). Has said a total of 37 messages. Is idling for 10h 11m 48s, last said: 'nacer_: packages/binutils'. Mar 22 15:14:00 zecke: tch, if you insult gtk+ then you insult my entire family Mar 22 15:14:20 pb__: koen, did the insulting :) Mar 22 15:14:31 pb__: and look is highly subjective :) Mar 22 15:14:43 heh Mar 22 15:14:45 * koen should install gpe-conf to swith gtk themes on osx Mar 22 15:14:56 hehe Mar 22 15:16:09 koen: install gconf-editor and feel like nearly like ms windows 'hacker' using regedit Mar 22 15:26:06 yeap :) Mar 22 15:26:13 it's usefull. Mar 22 15:28:21 bitbake now also hides most of the messages... it only displays the 'unpacking' and 'applying patch' messages Mar 22 15:29:52 ArjanS: see mail :) Mar 22 15:32:55 ~seen _law_ Mar 22 15:32:57 _law_ was last seen on IRC in channel #oe, 43m 16s ago, saying: 'hrw|work, right'. Mar 22 15:43:50 zecke: the broken cache mail? sorry, I didn't think they were related Mar 22 15:44:33 ArjanS: hehe Mar 22 15:44:43 ArjanS: BitBake has handlers for events Mar 22 15:44:54 ArjanS: the do_* messages are messages from a handler :) Mar 22 15:45:20 lardman|work: PONG! Mar 22 15:45:22 zecke: ah allright, I'm just a user, knowing nothing about internals :) Mar 22 15:45:31 ArjanS: that is fine Mar 22 15:45:49 zecke: thanks for the explanation. is there a way around it right now? or just to be patient for a few days? Mar 22 15:46:01 ArjanS: When writing the mail I just didn't how to describe these messages Mar 22 15:46:11 ArjanS: use r414 Mar 22 15:46:13 lardman|work: the problems you had seem indicative of a missing dependency. I kept finding these kind of problems with GPE progs not pulling in pixloaders and such.... Mar 22 15:46:16 ArjanS: or wait some more minutes Mar 22 15:46:26 lardman|work: could you try the e-image and see if it works? Mar 22 15:48:14 zecke: ok, sorry for my being impatient Mar 22 15:48:38 ArjanS: as you said, you are a 'user' :) Mar 22 15:48:51 ArjanS: these sort of things are just some kind of law of nature Mar 22 15:54:28 JustinP: You are working on the remove control driver? Mar 22 15:55:35 do13_: ummm...I got it compiling and "working", yes.... Mar 22 15:55:42 do13_: it *does* nothing, though Mar 22 15:55:44 JustinP: a new one? Mar 22 15:56:11 do13_: the max1111 isn't returning unique values for the keys..... Mar 22 15:56:15 lardman|work: new what? Mar 22 15:56:20 JustinP: Ah, e-image, how do I go about getting that on the machine then, not much room Mar 22 15:56:38 JustinP: I found a schematic: http://home.arcor.de/relei2004/zaurusfernb.html Mar 22 15:56:41 lardman|work: you don't have an SD card? I thought you were testing the e-image on a card before... Mar 22 15:56:46 * chouimat is back. Mar 22 15:57:09 do13_: if that might help you work on the driver I can give you the patch I was using... Mar 22 15:57:19 do13_: I doubt that the schematic will help, though Mar 22 15:57:42 JustinP: thx, but the driver is useless on tosa Mar 22 15:57:46 do13_: the driver I have is basically just a translation of the 2.4 driver into 2.6 (mostly done by RP, just fixe dup by me) Mar 22 15:57:56 JustinP: No, it fitted in my rootfs partition Mar 22 15:58:33 ArjanS: try now please Mar 22 15:58:45 lardman|work: are you sure? I wanted it tested on an SD/CF card because I wasn't sure.... Mar 22 15:59:47 zecke: getting a 'module object has no attribute Registered' messages for every .bb Mar 22 16:00:02 lol Mar 22 16:00:04 * zecke too Mar 22 16:00:25 oh the svn revert :} Mar 22 16:00:57 hi Mar 22 16:01:02 gotta go now... good luck! Mar 22 16:01:32 ~lart maemo in OE Mar 22 16:01:33 * ibot beats the living hamstercrap out of maemo in OE Mar 22 16:01:48 another package to rebuild: firefox Mar 22 16:02:22 ArjanS: now :) Mar 22 16:02:33 firefox 1.0.6 rdepend on thunderbird... argh Mar 22 16:02:40 lol Mar 22 16:04:37 do13_: thanks for trying :-/ Mar 22 16:05:19 I really really wish that RP had one of these remotes..... Mar 22 16:06:42 JustinP: RP will get poodle on days Mar 22 16:10:26 i'll donate a remote to RP if he can get suspend/resume working + fix reboot/halt on my c3000, for 2.6.16 EABI kernel Mar 22 16:14:21 hrw|work: poodle? I'm talking about the CE-RH2 for the Cxx00 Mar 22 16:14:34 jnc: :-\ Mar 22 16:17:01 JustinP: I know but wanted to note that he will get one more toy to support Mar 22 16:20:43 JustinP: unless that happens, i must sell off my zaurus kit :( Mar 22 16:20:53 hrw|work: you have to delete the tmp/staging/shlibs/* for any mozilla build when you build any other moszilla build Mar 22 16:20:55 jnc: :-( a sad day Mar 22 16:21:09 hrw|work: otherwise they all end up RDEPENDing on each other Mar 22 16:21:43 i'll get a flybook or something and run windows on it Mar 22 16:22:22 hoping it doesn't have to happen Mar 22 16:22:53 the SharpROM is a comprimise for now Mar 22 16:23:27 jnc: if you will insist on aiming for a target beyond what the devs are currently planning to release, then theres not much that can be done Mar 22 16:23:41 XorA: or have them packaged into more parts? libnspr etc? Mar 22 16:23:59 XorA: ? he's talking about some hardware problems with his Z...looks like kernel issues, but they haven't figured out what Mar 22 16:24:08 hrw|work: yeah, that might work as well, but I run away from that mess that is mozilla build Mar 22 16:24:13 XorA: without working suspend/resume, this hardware is not useful Mar 22 16:24:23 JustinP: EABI is experimental Mar 22 16:24:39 JustinP: building with gcc4.X is only just being worked on Mar 22 16:25:09 EABI works fine btw, and is probably going to be the default soon for supported zaurus targets Mar 22 16:26:01 ipaq does not appear to get that luxury Mar 22 16:26:14 still if you want working hardware, stop insisting on experimental relases, 3.5.4.1 will work on 3X00 devices Mar 22 16:26:45 http://bugs.openembedded.org/show_bug.cgi?id=762 Mar 22 16:26:49 jnc: heh Mar 22 16:26:51 I love that wording Mar 22 16:27:09 jnc: the main EABI testing is happening on ipaqs Mar 22 16:27:10 it could be a language issue, but I find that disrespectful torwards mickeyl Mar 22 16:27:32 koen: we should Debian omit Thumb Interwork Mar 22 16:27:39 koen: and make EABI usable on my Shark Mar 22 16:27:53 zecke: right now it omits Thumb stuff Mar 22 16:28:10 koen: so I could run it on my Shark? *yipee* Mar 22 16:28:59 koen: I just need to turn on EABI in linux-openzaurus and Im ready to test on c7x0 Mar 22 16:29:23 jnc: try to remember that OE is not a tool for Zaurus machines only Mar 22 16:29:52 JustinP: It appeared to fit without any issues, but I can try on a card if that's what you want Mar 22 16:30:01 jnc: and iirc 1st eabi boot was on hx4700 Mar 22 16:30:37 ok. time to go home Mar 22 16:30:38 cu all Mar 22 16:31:38 lardman|bbiab: I'd appreciate it. And a test of e-image would be good too. Mar 22 16:31:59 lardman|bbiab: my guess is that some locale data is missing or something....it seems to make a difference...sometimes...:-| Mar 22 16:32:11 I just don't know how they're getting included in other images... Mar 22 16:32:28 pH5: have a similar problem to the one you fixed earlier, some libtool files are staged into tmp/staging/arm-linux when I expected them to be staged into tmp/staging/arm-none-linux-gnueabi Mar 22 16:35:00 FWIW, your kernel developer likes the halt on reboot effect so its very low on my to fix list - I'm with mickeyl on that :) Mar 22 16:35:09 we need to fix shutdown though. Mar 22 16:35:36 FreeBSD halts as well Mar 22 16:35:39 pH5: I am figuring its something to do with STAGING_DATADIR_safe = "${STAGING_DATADIR}" to store the STAGING_DATADIR before inherit native, but I dont know enough bb magic to know why Mar 22 16:35:58 I really think having to press a power button isn't the end of the world and there are bigger problems to worry about Mar 22 16:36:13 RP: makes it much easier to flash that way :-) Mar 22 16:36:19 RP: on spitz here, it means the device screen goes off and everything locks up Mar 22 16:36:21 XorA: exactly Mar 22 16:36:35 easier to flash? mine reboots fine when it flashes Mar 22 16:36:43 what is that supposed to mean Mar 22 16:36:57 jnc: It shouldn't lock up. It is normal to have to press the power button Mar 22 16:37:25 oh. for reboot, it seems to physically shutdown and power off okay Mar 22 16:37:29 that behavior is fine, i guess Mar 22 16:37:45 i would hope that you COULD reboot if you wanted to from a technical standpoint, but not high on my list Mar 22 16:37:50 i don't mind having to turn it back on Mar 22 16:37:52 03florian 07org.oe.dev * rafe451a2... 10/packages/openssl/ (4 files in 2 dirs): openssl: add 0.9.7g with pkgconfig support Mar 22 16:38:00 03florian 07org.oe.dev * r569bea40... 10/packages/tapioca/libjingle-tapioca_svn.bb: libjingle: Add Tapioca patched version. Mar 22 16:38:04 for the 'halt' behavior, though, it physically locks up Mar 22 16:38:08 03florian 07org.oe.dev * r58fcb2ba... 10/packages/tapioca/ (tapioca-xmpp_0.3.0.bb tapioca_0.3.0.bb): tapioca-xmpp: Add 0.3.0. Mar 22 16:38:16 03florian 07org.oe.dev * r169621f4... 10/packages/tapioca/tapioca-xmpp_0.3.0.bb: tapioca-xmpp: add staging Mar 22 16:39:04 RP: that behavior should be easy enough to change in the inittab though. change both shutdown and reboot to be a kernel reboot, until the bug is fixed Mar 22 16:39:20 err nevermind Mar 22 16:39:41 i just realized that i'm not using ctrl+alt+del to trigger this Mar 22 16:39:42 heh Mar 22 16:40:21 Keep in mind the device can't actually shutdown - it only ever suspends, be it in the user kernel or the bootloader Mar 22 16:40:57 This might look like shutdown/rebooting from the desktop world Mar 22 16:43:26 * chouimat is away: lunch Mar 22 16:47:00 mickey|zzZZzz: ping Mar 22 16:47:24 koen|away: what is tapioca? Mar 22 16:47:28 koen|away: OSSO's gaim? Mar 22 16:49:48 RP: oh Mar 22 16:50:29 RP: how would the device get to a state like it is in... after you remove all power and press reset button? Mar 22 16:50:43 can't do that programatically? Mar 22 16:51:08 XorA: you are right. i don't know how to fix this, though. i think the problem is that we use the _arm override in angstrom.conf for TARGET_OS and TARGET_VENDOR. it seems that the overrides are only evaluated after parsing the package files and this breaks everything that relies on instant expansion of HOST_SYS. Mar 22 16:52:07 koen|away: it doesn't look too crappy here Mar 22 16:58:34 JustinP: So how do I go about placing the e-image on a CF card? Something about a loop-back mounting? Mar 22 17:01:06 jnc: Not easily. As soon as you apply power, the processor will start running Mar 22 17:02:03 lol Mar 22 17:02:04 http://lists.kde.org/?l=kde-www&m=114297200332213&w=2 Mar 22 17:02:07 pH5: indeed, TARGET_OS needs to evaluate to a stable value: too many other variables are defined in terms of it for it to change partway through the parse. You could define it in terms of a python expansion, though, if you want. Mar 22 17:02:35 zecke: kde p0rn? awesome Mar 22 17:03:03 pb__: "We want to know our user base - nude" Mar 22 17:03:43 let's just pray that they don't start publishing pictures of the developers in such a state of undress Mar 22 17:05:27 I want to build only gpe and opie things (only gui things) without the hole image, what .bb file would be the best to start with ? Mar 22 17:05:46 grma: bitbake task-gpe, task-opie? Mar 22 17:05:53 grma: bitbake meta-opie, meta-opie-all? Mar 22 17:06:06 ok will look into them. Mar 22 17:08:45 pb__: something like this: TARGET_OS = "linux${@['','-gnueabi'][bb.data.getVar('HOST_ARCH', d, 1)=='arm']}" Mar 22 17:10:57 s/HOST/TARGET/ Mar 22 17:13:29 Is there a fix for the zaurusd fetch error? (http://pastebin.ca/46615) Mar 22 17:14:10 afternoon Mar 22 17:15:22 hi reenoo Mar 22 17:16:20 hey pH5 Mar 22 17:18:01 koen|away: overrides for TARGET_OS break things. what about http://en.pastebin.ca/46618 Mar 22 17:25:46 bonne soiree Mar 22 17:42:02 pH5: add some comments in the file why we need more ugly python and commit it :) Mar 22 17:49:15 umm Mar 22 17:49:51 I am getting NOTE: multiple providers are available (hostap-modules, hostap-conf); while build an image. So i set PREFERRED_PROVIDER_hostap-conf = "hostap-conf" right in the image.bb but it doesn't work Mar 22 17:50:27 no matter what, I always get the message and BB chooses the wrong provider Mar 22 17:53:17 Trying to compile busybox, giving me a set of errors. Anyone mind giving a hand http://pastebin.com/616467 Mar 22 17:55:47 CoreDump|home: Set that in local.conf - I need to add it to the 354x branch Mar 22 17:57:27 lardman|work: you can use altboot's new image creation feature or do it manually Mar 22 17:58:20 CoreDump|home: The image.bb file is far too late for that option to have any effect Mar 22 17:58:30 RP: ahh I see Mar 22 17:58:54 * CoreDump|home tries local.conf Mar 22 18:01:58 Anyone got an idea? Mar 22 18:02:13 Ken|JLime: dunno sry Mar 22 18:03:22 RP: also, akita / spitz / borzoi are still set to use a 2.4 kernel by default ;) Mar 22 18:04:09 JustinP: I'll take a look at it when I get home Mar 22 18:04:17 CoreDump|home: the branch doesn't have the 2.4 kernels anymore Mar 22 18:04:42 CoreDump|home: http://ewi546.ewi.utwente.nl/tmp/viewmtn/revision.psp?id=53f9176d604b66cf1ab5287d10398ab0e92fbf85 Mar 22 18:04:57 koen: tell that my "2.4" gpe-image from yesterday hehe Mar 22 18:05:31 03koen 07org.oe.dev * r48c734b5... 10/packages/tapioca/libjingle-tapioca_svn.bb: libjingle: add a PV Mar 22 18:06:21 CoreDump|home: are you perhaps using oz354fam083? Mar 22 18:06:25 nope Mar 22 18:06:46 hmmm...I saw the commit... Mar 22 18:09:37 Ken|JLime: what is the version of your linux-libc-header package? I assume __kernel_fd_set is undefined at this place (this is defined in linux/posix_types.h) - does adding #include to the beginning of asm/posix_types.h fix those errors? Mar 22 18:10:36 I found a bugreport (http://bugs.openembedded.org/show_bug.cgi?id=657) with the same issues as mine. It compiles nicely after that. Mar 22 18:11:02 using header package 2.6.11 methink, probobly the cause of the issue Mar 22 18:11:55 Ken|JLime: even better :) Mar 22 18:12:05 hi guys Mar 22 18:12:09 hi mickeyl Mar 22 18:13:37 03mickeyl 07org.oe.dev * r085d7150... 10/packages/ (3 files in 2 dirs): add python-itools, a huge library supporting various web technologies Mar 22 18:14:00 hey mickeyl Mar 22 18:16:53 *sigh* a lot of stuff bails out because of gcc 4 Mar 22 18:17:04 heh Mar 22 18:17:09 try using 4.1 ;) Mar 22 18:17:22 that will make it just worse Mar 22 18:17:40 distros should have gcc4 patches Mar 22 18:17:53 I noticed 2.6.16 for Zaurus in there - how to do EABI with this kernel? Mar 22 18:18:00 koen: good idea, I should look there Mar 22 18:18:30 mickeyl: turn on EABI in the defconfig via menuconfig (if RP hasn't done already) Mar 22 18:19:41 RP? Mar 22 18:20:46 ../linux/linux-openzaurus-2.6.16/defconfig-tosa:# CONFIG_AEABI is not set Mar 22 18:20:57 lardman|gone: ask CoreDump|home about it if you can't get it to work. Hopefully he'll have a howto/doc soon. ^_^ Mar 22 18:21:06 has anybody got a complete EABI image running yet? Mar 22 18:21:37 it didn't start because of a missing eabi aware kernel Mar 22 18:21:38 i' Mar 22 18:21:42 m cooking one now Mar 22 18:22:40 jnc apperently did last night Mar 22 18:23:04 ERROR: Cannot satisfy the following dependencies for task-bootstrap: fuser Mar 22 18:23:10 What supplies fuser? Mar 22 18:23:19 util-linux? Mar 22 18:23:23 lol Mar 22 18:23:34 or busybox 1.1 Mar 22 18:23:52 already compiled busybox, so betting my money on util-linux Mar 22 18:32:17 03koen 07org.oe.dev * r1dfaf135... 10/packages/farsight/farsight_20060322.bb: farsight: add todays snapshot of the farsight darcs repo Mar 22 18:33:04 hey zecke Mar 22 18:33:59 03coredump 07org.oe.dev * racaca3ac... 10/packages/initscripts/ (initscripts-1.0/tosa/keymap-2.4.map initscripts_1.0.bb): initscripts: tosa: Enable Fn-ombinations for numbers in kernel 2.4. Courtesy of webmind@#openzaurus. Mar 22 18:34:12 CoreDump|home: yay! Mar 22 18:34:28 pH5: something like that, yeah Mar 22 18:34:29 could you apply that to 354x as well, if hrw|gone doesn't object? Mar 22 18:34:35 hey pb_ Mar 22 18:34:41 mickeyl: yo Mar 22 18:35:24 mickeyl: I'll leave him a message so he can decide if he wants that in or not Mar 22 18:35:32 koen: util-linux didnt supply fuser either, how about coreutils? Mar 22 18:35:56 Ken|JLime: could be, I always confuse those packages Mar 22 18:38:19 CoreDump|home: cool, thanks Mar 22 18:44:39 dammit, coreutils didnt supply fuser either Mar 22 18:45:19 Ken|JLime: psmisc Mar 22 18:45:33 :D Mar 22 18:49:22 hey mreimer Mar 22 18:49:31 hey koen! looks like you've been busy Mar 22 18:52:13 mreimer: just clearing out a backlog of patches :) Mar 22 18:52:44 judging from the commits, it would seem that eabi is working in OE now. Is that true? Mar 22 18:53:08 mickeyl: mickeyl I had tons of questions Mar 22 18:53:09 03koen 07org.oe.dev * rc3f6032f... 10/packages/gstreamer/ (3 files): gst-plugins-{bad,good,ugly}: depend on gst-plugins-base Mar 22 18:53:27 03koen 07org.oe.dev * r5542410c... 10/packages/gstreamer/gst-plugins.inc: gst-plugins.inc: DEPEND on liboil Mar 22 18:53:35 mreimer: there are some problems with socket() and general gcc4.x related problems Mar 22 18:53:48 mreimer: but ph5 has booted a EABI bootstrap image on his hx Mar 22 18:54:17 zecke: I'm afraid I won't have much time to answer, but feel free to fire away Mar 22 18:54:17 koen: sounds like a bit of work. did ph5 say if he noticed a speed difference? Mar 22 18:55:03 mickeyl: I don' remember :} Mar 22 18:55:06 mreimer: I didn't notice anything. As socket() always returns ENOSYS, X doesn't even start. Mar 22 18:55:16 zecke: heh :) next time just mail me :) Mar 22 18:55:26 pH5: that would be a problem :-) Mar 22 18:55:46 zecke: i'm going on a conference tomorrow and need^wwant to spend the evening with Sabine Mar 22 18:55:49 mickeyl: bitbake on ewi is 1 minute faster than before Mar 22 18:55:58 zecke: that's cool! Mar 22 18:56:08 * zecke is going to watch battlestart gal... Mar 22 18:56:30 mickeyl: ah right Mar 22 18:56:40 mickeyl: is there a better copy implementation? Mar 22 18:56:42 SAKS 2006 in Kassel Mar 22 18:57:14 zecke: no, copy.copy is the only one. however, you need to check if you really need deepcopy, because that's slowing down a lot Mar 22 18:57:28 mickeyl: e.g. a transparent copy with ref counting would be nice Mar 22 18:57:49 mickeyl: I have code in data.smart to copy on each setVar Mar 22 18:57:54 mickeyl: to show 'bugs' Mar 22 18:58:29 e.g. someone manipulating a string after it is in the data module Mar 22 18:58:32 and cached Mar 22 18:59:57 i see. unfortunately monitoring object access is tough Mar 22 19:00:46 i once played with something.... Mar 22 19:01:25 ... putting a file online Mar 22 19:01:43 mickeyl: I think I make that configurable, I think I have found most issues Mar 22 19:02:15 http://vanille.de/temp/monitor.py might be a bit interesting Mar 22 19:03:13 although it's probaby tied too much into elan to be useful Mar 22 19:03:19 besides not containing any docs Mar 22 19:04:22 *cough* Mar 22 19:04:22 weakrefs are cool though Mar 22 19:04:42 bbl Mar 22 19:04:56 mickeyl: GPL, that would taint my license :} Mar 22 19:04:56 Null() funny Mar 22 19:04:57 mickeyl: tough stuff Mar 22 19:04:57 mickeyl: I like TERROR Mar 22 19:06:47 is there a variable that contains "${IMAGE_BASENAME}-${MACHINE}-${DATE}${TIME}$AND_WHATEVERS_IS_THE_SUFFIX" + the full path to the image? Mar 22 19:08:19 not one Mar 22 19:09:16 ${DEPLOY_DIR_IMAGE}/{IMAGE_NAME}.{jffs2,tar.bz2,ext2} Mar 22 19:09:59 koen: ahh IMAGE_NAME is great Mar 22 19:10:31 but sometimes the .suffix is not ".$FS" but .img or .bin. Whats the logic behind that? Mar 22 19:10:44 openzaurus hacks Mar 22 19:10:51 meh Mar 22 19:14:28 03koen 07org.oe.dev * r0f1e903d... 10/packages/farsight/gst-plugins-farsight_20060322.bb: gst-plugins-farsight: add 20060322 snapshot Mar 22 19:30:45 whats the difference between boostrap_extra_rdepends and bootstrap_extra_depends? Mar 22 19:31:06 RDEPENDS is runtime, DEPENDS is build-time Mar 22 19:32:12 meaning rdepends will go inside the boot-image ? Mar 22 19:33:02 yes Mar 22 19:33:07 Roger, thx Mar 22 19:33:22 there are also runtime dependencies added automatically when thinks are linked Mar 22 19:35:45 But if i would like to add lets say MC to my bootimage (as a program in /usr/bin) i should add it as bootstrap_extra_rdepends += ...mc Mar 22 19:36:47 03koen 07org.oe.dev * r74d7b2a6... 10/packages/tapioca/tapioca_0.3.0.bb: tapioca: stage more headers Mar 22 19:37:40 mhhh what i have to state in DISTRO do build familiar 0.8.3 ? Mar 22 19:39:52 gremlin[it]: you need the org.openembedded.oz354fam083 branch and DISTRO="familiar-0.8.3" Mar 22 19:40:15 hi reenoo Mar 22 19:40:20 hey pb_ Mar 22 19:40:36 mhhh with dev i can build only familiar-unstabe ??? Mar 22 19:40:52 hi pb_, reenoo :) Mar 22 19:40:57 hi gremlin[it] Mar 22 19:41:02 gremlin[it]: I don't think you can build much at all with .dev right now but I may be mistaken Mar 22 19:42:23 that's FUD Mar 22 19:42:43 I've been building images all day from .dev Mar 22 19:42:50 reenoo yes is true i have some issues :) Mar 22 19:44:03 koen: familiar images? I doubt it Mar 22 19:44:20 no Mar 22 19:44:27 familiar horribly broken Mar 22 19:44:38 you said "I don't think you can build much at all with .dev right now but I may be mistaken" Mar 22 19:45:02 and I responded to the "at all" part of that Mar 22 19:45:18 ah, thanks for correcting that then Mar 22 19:45:50 where 'horribly broken" read "missing a preferred_version for binutils" Mar 22 19:46:41 well. your messing around with binutils preferences broke stuff indeed. but that's not a DISTRO.conf candidate Mar 22 19:46:48 koen : i set it and i was able to build a bootstrap-image ... but now i would like to show opie on the device (customer choose to use qt) ... Mar 22 19:47:22 koen: the binutils version built by default should work for the given arch Mar 22 19:47:49 it works for my arm machines Mar 22 19:48:07 for eabi maybe Mar 22 19:48:08 hey reenoo Mar 22 19:48:12 hey zecke Mar 22 19:49:19 03koen 07org.oe.dev * r9f878719... 10/packages/farsight/farsight_20060322.bb: farsight: stage some headers Mar 22 19:49:22 what's the problem with binutils? Mar 22 19:49:25 03koen 07org.oe.dev * rd5059d33... 10/packages/tapioca/tapiocaui_0.3.0.bb: tapiocaui: add 0.3.0 Mar 22 19:50:22 some mis-algned versions between libc andh binutils ... if i understand right ... Mar 22 19:51:42 no Mar 22 19:52:06 the way binutils behaves acyuses an issue with libgcc taht shows up in glibc Mar 22 19:52:19 causes* Mar 22 19:52:52 but its only in the new cvs one taht has been changed for the EABI stuff Mar 22 19:53:58 okay. so why is the new cvs one the default for a non-eabi configuration? Mar 22 19:54:15 because PREFERED was not set Mar 22 19:54:28 and it defaults to newest Mar 22 19:54:32 pb_: that would be a question for koen Mar 22 19:54:41 emte: well, duh, yeah. why is the preferred version not set? Mar 22 19:54:43 http://www.lisperati.com/logo.html Mar 22 19:54:44 lol Mar 22 19:54:49 new cvs? Mar 22 19:55:01 it uses a binutils release tarball Mar 22 19:55:35 oh, really? what release? Mar 22 19:55:47 if a release version is broken for non eabi that would be rather sad. Mar 22 19:57:00 2.16.91.0.6 Mar 22 19:57:21 ah, that's not a release. that's just a cvs snapsjot. Mar 22 19:57:26 er, snapshot Mar 22 19:57:50 it's still a bit sad that it's broken in the non eabi case, but that's more understandable. Mar 22 19:59:17 anyway, irrespective of the provenance of this code, it shouldn't be the default version for arm-linux if it doesn't work in that configuration. Mar 22 20:00:26 and, if non arm architectures don't derive any benefit from it, it probably shouldn't be the default for them either. Mar 22 20:04:49 * reenoo nods Mar 22 20:09:17 pb_: what is the versioning scheme for binutils anyway? Mar 22 20:10:18 zecke: similar to gcc. major releases are numbered 2.10. 2.11, 2.12; minor bug-fix releases are numbered 2.11.1, 2.11.2, 2.11.3 Mar 22 20:10:36 cvs snapshots are numbered 2.xx.91 Mar 22 20:10:44 pb_: ah Mar 22 20:17:29 you weren't kidding with the half a day stuff were you Mar 22 20:25:02 zecke: as the sheriff, maybe you could clean up this binutils versioning business Mar 22 20:35:19 pb_: I could generate a monotone key for you -as a friend Mar 22 20:35:56 pb_: I hope people change DEFAULT PREFERENCE Mar 22 20:37:35 zecke: heh, let's hope so Mar 22 20:38:38 hmm I should set up a Vmware server somewhere Mar 22 20:40:19 hrm Mar 22 20:40:22 $ bitbake -b ../org.openembedded.oz354fam083/packages/coreutils/coreutils-native_5.1.3.bb -c showdata | grep ^PACKAGE_ARCH Mar 22 20:40:25 PACKAGE_ARCH="arm" Mar 22 20:40:30 that doesn't sound right Mar 22 20:40:39 oh Mar 22 20:52:00 03rpurdie 07org.oe.oz354x * r80f275bc... 10/conf/distro/openzaurus-3.5.4.1.conf: Set PREFERRED_PROVIDER_hostap-conf Mar 22 20:54:39 hi Mar 22 20:55:42 hey hrw|husky Mar 22 20:55:43 CoreDump|afk: feel free to add such fixes as keymap stuff into .oz345x Mar 22 20:56:02 my wifi card loose connection from time to time Mar 22 20:58:30 mickey|bbl: Performance: s += 'string' or s = "%s %s" % (s,string) - which is faster Mar 22 20:58:44 the latter Mar 22 21:00:13 koen, just to know ... the have u try recently to build opie-image with dev repo ? Mar 22 21:00:32 gremlin[it]: mickeyl built one Mar 22 21:00:51 reenoo: much faster? Mar 22 21:00:54 gremlin[it]: look into oe m| archive Mar 22 21:01:00 mhh ok ... so i have a chance :) Mar 22 21:01:22 hrw|husky where ? Mar 22 21:01:33 zecke: iirc mickeyl wrote months ago to use that 2nd notation Mar 22 21:01:52 gremlin[it]: openembedded mailinglist archive on handhelds.org Mar 22 21:01:59 ahh ok :) Mar 22 21:02:15 zecke: not sure. += adds runtime type checking/casting overhead. Mar 22 21:02:27 NOTE: package uicmoc-native-2.3.10-r0: task do_compile: failed <<< sigh !!! Mar 22 21:02:59 gremlin[it]: how can that fail? Mar 22 21:03:20 the performance hints I see talk about = and not += :} Mar 22 21:03:35 time to try timeit Mar 22 21:04:09 lot of compile errors ... Mar 22 21:04:11 | kernel/qmemorymanager_qws.h:129: error: cast from ‘char*’ to ‘int’ loses precision Mar 22 21:04:27 gremlin[it]: hmm Mar 22 21:04:36 gremlin[it]: gcc version? host system? Mar 22 21:04:48 host FC4 on x86_64 Mar 22 21:05:13 gcc what is provided with oe, familiar-unsable, ep93xx Mar 22 21:05:22 gremlin[it]: okay, check the bug tracker for 64bit issues :} Mar 22 21:06:26 ... mhhhhhh ... let's go to read ... Mar 22 21:06:44 gremlin[it]: :) Mar 22 21:07:04 france_: got an Alpha too many? Mar 22 21:08:36 zecke is something relates to opie or gcc-c++ (cross) have u idea ? Mar 22 21:08:42 zecke: I have a rack of them at OSU, that I need to setup. Mar 22 21:09:08 gremlin[it]: it is a native package that is failing Mar 22 21:09:08 DHO !!! is the natice package sorry !!! Mar 22 21:09:12 :) Mar 22 21:10:21 sshit.. 3-way merge.. no way to do it on 80x25 term now :( Mar 22 21:10:47 gremlin[it]: http://bugs.openembedded.org/show_bug.cgi?id=535 Mar 22 21:11:03 gremlin[it]: needs to be adopted to uicmoc2? Mar 22 21:12:43 gremlin[it]: http://bugs.openembedded.org/attachment.cgi?id=417 Mar 22 21:12:49 time to apply the patches Mar 22 21:13:03 i found 455 attach Mar 22 21:18:07 grr, crazy xilinx and their 850MB shell scripts Mar 22 21:18:37 wow Mar 22 21:18:59 selfextracting ones? Mar 22 21:19:10 yeah Mar 22 21:19:22 very tedious to edit though Mar 22 21:19:30 yes. Mar 22 21:19:49 I kind of wish they could just use a tar archive like everyone else Mar 22 21:19:55 I worked at company where soft was released as such scripts Mar 22 21:21:18 hm. i will probably drop maemo stuff from .oz345x - its simler way of making it unselectable by default Mar 22 21:22:17 hello all Mar 22 21:24:20 mhhh cpile opie (c++) is really thedious ... an di upgrade the memory few days ago ... Mar 22 21:32:20 hm, which debian package is libXm.so.3 in, again? Mar 22 21:32:23 I can never remember that one Mar 22 21:32:50 pb_: dpkg -S libXm.so.3? Mar 22 21:32:53 pb_: or was it -L Mar 22 21:33:01 ~herring zecke Mar 22 21:33:03 * ibot whacks zecke on the side of the head with a large red herring named alfred Mar 22 21:33:10 that only works if you already have the package installed Mar 22 21:33:25 http://packages.debian.org? :-) Mar 22 21:33:27 pb_: oh you don't have it? Mar 22 21:33:32 pb_: is that motif? Mar 22 21:33:34 zecke: right, that's why I was asking Mar 22 21:33:35 pb_: lesstif? Mar 22 21:33:43 I think lesstif only has Xm.so.2 Mar 22 21:33:44 dpkg: /usr/lib/libXm.so.3 nie odnaleziony. Mar 22 21:33:44 open-motif? Mar 22 21:33:46 njs: good thinking Mar 22 21:34:04 I don't have open-motif in my packages list Mar 22 21:34:07 oh, it might be in non-free Mar 22 21:34:56 ah yes, njs wins Mar 22 21:34:58 usr/X11R6/lib/libXm.so.3 libs/libmotif3 [non-free] Mar 22 21:35:10 *jealous* Mar 22 21:35:14 njs: I hate you :) Mar 22 21:35:20 zecke: neener neener Mar 22 21:37:42 excellent, looks like I'm back in business Mar 22 21:38:05 pb_: Xilinx uses motif - nice Mar 22 21:38:37 zecke: yeah, for some parts of the toolkit Mar 22 21:38:56 the main project navigator doesn't use it, but the UCF editor seems to be based on motif Mar 22 21:39:15 wonderful Mar 22 21:40:41 I think the navigator is written in java, so presumably it uses Swing or something Mar 22 21:41:17 pb_: which might link motif statically :} Mar 22 21:41:19 oh drat, now I have this stupid problem with motif wanting to make tcp connections to the xserver Mar 22 21:41:24 pb_: or are you using freeswing? Mar 22 21:41:34 pb_: wow Mar 22 21:41:39 and debian, in some kind of drive for security, turns off the tcp listener Mar 22 21:42:03 RP: can zaurusd be set to not suspend on hinge off? not everyone is used to this Mar 22 21:42:31 time to learn for the next exam I guess Mar 22 21:42:32 zecke: dunno, all the java stuff comes bundled in this famous 850MB shell archive Mar 22 21:42:35 * reenoo waves Mar 22 21:42:42 reenoo|afk: enjoy Mar 22 21:42:51 pb_: hehe Mar 22 21:43:00 pb_: is it at least free of GNUism? Mar 22 21:43:15 man I would have never thought, I would miss metacity Mar 22 21:43:31 that gnome wm? Mar 22 21:43:43 hrw|husky: yes, twm kind of sucks Mar 22 21:44:00 zecke: i had twm on husky... disaster Mar 22 21:46:41 twm is tragic on any system Mar 22 21:46:49 hi, sorry don't know where to ask this, but say I managed to get a device to find a NAND storage area, might that change the addresses of things and cause haret to fail to mount the ramdisk that it used to find a a certain address? Mar 22 21:47:08 zecke: hard to tell. freebsd is not a supported platform. Mar 22 21:47:17 oh, actually, I think it runs on solaris Mar 22 21:47:19 so you are in luck Mar 22 21:47:43 hi Mar 22 21:48:17 aa_: i would not touch flash on device where i do not have working flash driver and/or jtag cable working Mar 22 21:48:21 hi Bernardo Mar 22 21:48:52 that sounds ominous Mar 22 21:49:37 aa_: better try to get sd/mmc working (iirc hx3715 does not have cf slot) Mar 22 21:50:02 but people who are clever have tried and failed at that and I am really dumb Mar 22 21:51:03 aa_: just detecting the NAND chip is unlikely to do any harm, but if you try writing to that area then you do run the risk of problems. Mar 22 21:51:15 hm. screen flashed once - 25% battery left Mar 22 21:51:17 well, say I am not trying to write to it Mar 22 21:51:32 then you are relatively safe. Mar 22 21:51:36 but say it gets detected, might the address of my ramdisk change? Mar 22 21:51:43 shouldn't do, no Mar 22 21:51:49 note I have no idea what these terms actually mean Mar 22 21:52:13 heh Mar 22 21:52:31 so weirdly, after turning on mtd in the kernel, it is unable to find the ramdisk Mar 22 21:52:50 that is odd Mar 22 21:53:03 I just get a basic kernel panic Mar 22 21:53:05 it's possible that the kernel size has increased far enough that it's now colliding with the ramdisk in memory, or something Mar 22 21:53:13 does it say anything about the initrd having been overwritten? Mar 22 21:53:31 not that I rmeember, I am comiling another kernel Mar 22 21:53:47 it is pretty hard to debug it, I am not sure how at all Mar 22 21:53:58 maybe record the screen and play it back in slow motion Mar 22 21:54:17 probably be easier with a serial cable, if you can get hold of one Mar 22 21:54:17 has it really come to this! Mar 22 21:54:54 i've got the "opposite" problem... my kernel boots, but then doesn't mount the NFS root. Mar 22 21:55:15 HopsNBarley: if I could work out how to do that nfs crap I wouldn't be bothering with this Mar 22 21:55:38 i'll let you know. i've got a ppc/embedded expert joining me in about an hour. Mar 22 21:56:17 oh well, I swear I saw instructions, but I can't find them now Mar 22 21:56:37 aa_: are you using uboot? Mar 22 21:57:27 no, a hting called haret Mar 22 21:58:02 HopsNBarley: haret is wince loader for linux Mar 22 22:00:40 wow, turning on NAND debugging in the kernel is hilarious Mar 22 22:00:45 ok I need that serial thing Mar 22 22:01:11 me too - have to resolder one cable... Mar 22 22:01:14 ok time to sleep - cu Mar 22 22:01:27 HopsNBarley: do you have set the network on boot ? Mar 22 22:01:57 RAMDISK: Couldn't fnd a valid RAM disk image starting at 0 Mar 22 22:02:15 only after I turn on this NAND device Mar 22 22:02:35 but also get this JFFS: trying to mount a non-mtd device Mar 22 22:03:02 now before it wasn't involving jffs at all, it is a compressed ext2 initrd Mar 22 22:04:25 hi Mar 22 22:04:49 hi florian Mar 22 22:08:05 hi kergoth Mar 22 22:08:41 so I am going to start guessing random values for the ramaddr parameter to haret, that won't brick anything will it? Mar 22 22:08:57 hey kergoth Mar 22 22:11:59 aa_: no, that's safe enough Mar 22 22:17:56 03dirk 07org.oe.dev * rc98c235b... 10/packages/lzma/lzma-native_4.17.bb: lzma-native: split path in oe_libinstall call Mar 22 22:18:00 03dirk 07org.oe.dev * re737fdf2... 10/packages/libcap/ (3 files in 2 dirs): libcap: add 1.10 Mar 22 22:44:41 03koen 07org.oe.dev * re5aa5020... 10/packages/scw/scw_0.4.2.bb: scw: add scw, a set of chat related widgets for GTK+ Mar 22 23:04:42 What supplies ttf-dejavu-sans ttf-dejavu-sans-mono for "bitbake opie-image"? Mar 22 23:05:58 opie-ttf-support ..lol Mar 22 23:13:55 hmm, didnt work Mar 22 23:14:24 any ideas? Mar 22 23:16:19 Hey does anyone know how to change a bb to where it installs from the original subversion repository? Mar 22 23:16:28 libgphoto2 to be precise Mar 22 23:17:31 Hey does anyone know how to change a bb to where it installs from the original subversion repository? Mar 22 23:22:01 dtx: what do you mean by "where it installs"? Mar 22 23:22:57 03tmbinc 07org.oe.dreambox * rcff595e5... 10/packages/linux/linux-dm7025.bb: linux-dm7025: remount only on target device Mar 22 23:23:01 03tmbinc 07org.oe.dreambox * r494ca1d2... 10/packages/enigma2/enigma2.bb: enigma2: update to 20060323 Mar 22 23:23:05 03tmbinc 07org.oe.dreambox * rad7dd1ef... 10/packages/dreambox/dreambox-dvb-modules.bb: dreambox-dvb-modules: update to 20060323 Mar 22 23:24:15 h... i should try to update my dbox :-) Mar 23 00:44:22 good nite guys Mar 23 01:23:10 mickey|zzZZzz: ping me when you're up Mar 23 01:52:12 oe run on windows? Mar 23 01:53:26 in cygwin perhaps. i used to mess with it, but havent in ages Mar 23 01:53:36 course i havent touched oe in a year Mar 23 01:53:39 so *shrug* Mar 23 01:53:55 TMM1: how ya been? Mar 23 02:14:20 a year? wow Mar 23 02:14:23 what've you been up to Mar 23 02:29:32 Anyone here? Mar 23 02:32:54 monotone is a strange creature Mar 23 02:33:14 yes Mar 23 02:33:42 Do you have any experience making .bb files? Mar 23 02:34:19 last time i did was like 18 months ago Mar 23 02:34:44 is there a web interface to the OE metadata where i can see the list of packages included Mar 23 02:35:36 Well, I tried making a generic .bb for libgphoto2, and got it to compile and do everything except it also displays this: Mar 23 02:35:51 NOTE: the following files were installed but not shipped in any package: Mar 23 02:36:18 and then it lists all the libs in /usr/lib/libgphoto2_port/ Mar 23 02:36:54 they should be installed also, but they don't get included in the ipk Mar 23 02:37:54 got any ideas? Mar 23 02:38:14 web interface? Mar 23 02:38:35 when you check it out you can just go into org.openembedded.*/packages Mar 23 02:38:36 to the monotone repository Mar 23 02:38:41 and it shows all the packages Mar 23 02:39:29 why would you need a web interface if you can see all the packages locally? Mar 23 02:45:29 i can't locally. i don't have a checkout on this machine Mar 23 02:47:50 TMM1: left ti, moved to MN, got a job there for 6 months, left there, moved to arizona for a job there doing less kernel crap Mar 23 02:47:54 heh Mar 23 02:48:31 wow Mar 23 02:49:34 heh, kernel might be crap but it still beats tech support ;) Mar 23 02:52:04 hehe. true that Mar 23 02:55:37 i'm trying to convince the gumstix people to use OE. they're using a buildroot like the one we were using for OZ way way back when Mar 23 03:07:54 TMM1: there is one Mar 23 03:08:01 i can't remeber where however Mar 23 03:08:18 RP might be a good person to help you with the conversion task :P Mar 23 03:13:21 is this a known bitbake problem? Mar 23 03:13:25 File "/home/packages/nslu2/mtn/debianslug/bitbake/lib/bb/fetch/__init__.py", line 193, in ? Mar 23 03:13:25 import svk Mar 23 03:13:25 ImportError: No module named svk Mar 23 03:17:47 hi, folks. It seems bitbake opie-image broken here. Please have a look at : http://pastebin.com/617389 Mar 23 03:18:55 Hmm, It seems it can't find "client.h" while build opie-bluetoothmanager. Mar 23 06:33:56 anyone know what the status is of the power management stuff for the iPAQ H1940? battery charging / indication, that stuff Mar 23 07:29:23 Good morning ! Mar 23 07:41:02 good morning Mar 23 07:42:31 RP: I had to set PREFERRED_PROVIDER_virtual/kernel="linux-openzaurus" and PREFERRED_VERSION_orinoco-modules="0.15rc2" to get oz3.5.4.1 to build Mar 23 07:46:59 mmm... even setting PREFERRED_VERSION it still tries to build orinoco-modules_0.13, which bombs... :( Mar 23 07:53:52 later Mar 23 08:05:30 good morning all Mar 23 08:22:15 morning all Mar 23 08:27:57 hi ! Mar 23 08:32:23 hi Mar 23 08:32:53 hrw: I think we can really speed up 'cached' parsing Mar 23 08:33:33 super Mar 23 08:35:28 most of the time is spend in compiling stuff Mar 23 08:35:35 which was already compiled on the last run Mar 23 08:47:36 ~hail x forwarding Mar 23 08:47:39 * ibot bows down to x forwarding and chants, "I'M NOT WORTHY!!" Mar 23 08:47:54 kdiff3@bitbake. Mar 23 08:57:44 hi pb Mar 23 08:58:07 hi hrw Mar 23 08:59:15 03hrw 07org.oe.oz354x * r1d84ad4c... 10/packages/keymaps/ (files/tosa/keymap-2.4.map keymaps_1.0.bb): keymaps: Enable Fn-combinations for numbers for tosa/2.4. Courtesy of webmind@#openzaurus Mar 23 09:00:31 hrw: will tosa have a 2.6 testing image in 3.5.4.1? Mar 23 09:00:43 koen: no Mar 23 09:00:55 not ready yet Mar 23 09:02:34 hi xora Mar 23 09:02:42 morning Mar 23 09:02:55 hi law Mar 23 09:03:09 hrw: you not working today? Mar 23 09:03:12 hey XorA & _law_ Mar 23 09:03:41 koen: before angstrom release someone need to create good gpe-x.y-preffered-versions.inc to NOT build maemo stuff on 'build world' Mar 23 09:03:46 <_law_> hi koen Mar 23 09:03:52 XorA: power lose was at company Mar 23 09:04:02 _law_: did you contacted Will about collie? Mar 23 09:04:30 <_law_> hrw|work, yes thanks Mar 23 09:04:49 <_law_> hrw|work, he will send me the collie in the next few days Mar 23 09:04:55 cool Mar 23 09:05:10 hrw|work: before the angstrom release we'll have updated maemo bits which won't need the osso gconf,dbus,gnome-vfs, etc Mar 23 09:05:35 koen: I'm now fighting with broken x11 feed because of that maemo stuff Mar 23 09:05:51 today I will do massive 'mt drop' on .oz354x Mar 23 09:06:00 atleast plan to do it Mar 23 09:06:50 I just did 'rm maemo/ dbus/*osso* -rf ; mt drop --missing' a while back Mar 23 09:07:11 which branch? Mar 23 09:07:28 .oz354fam083 Mar 23 09:07:37 I didn't commit the drop Mar 23 09:08:07 morning all Mar 23 09:08:14 hey RP Mar 23 09:10:14 hi RP Mar 23 09:15:28 does anyone needed to disable PARALLEL_MAKE in orbit2 (native and non-native)? Mar 23 09:15:51 hrw|work: iirc it is disabled in .dev Mar 23 09:15:53 * XorA takes a wild stab in the dark that jnc has no idea who mickeyl is Mar 23 09:16:05 ouch Mar 23 09:16:13 watch where you stabbing ;) Mar 23 09:16:16 XorA: why? Mar 23 09:16:29 hrw|work: he is very aptronising towards him in bug reports and emails Mar 23 09:16:34 who's mickeyl Mar 23 09:16:43 is he that one python guy Mar 23 09:16:47 hrw|work: see latest one on oe@ which actually contains zero info Mar 23 09:16:53 TheMasterMind1: mickeyl is one of OE founders Mar 23 09:17:05 ~mickeyl Mar 23 09:17:07 somebody said mickeyl was our local python guru, opie/openzaurus/openembedded/bitbake/opensimpad developer, and generally in charge of way too many things Mar 23 09:17:22 like http://oe.handhelds.org/people.php Mar 23 09:17:26 heh Mar 23 09:17:31 ~hrw Mar 23 09:17:33 hmm... hrw is Marcin Juszkiewicz - collie maintainer, OpenEmbedded/OpenZaurus/OPIE developer Mar 23 09:18:40 collie maintainer eh Mar 23 09:18:40 ibot: hrw is also OpenZaurus 3.5.4.x release maintainer Mar 23 09:18:42 hrw|work: okay Mar 23 09:18:44 so how's about that 1.6 kernel Mar 23 09:18:47 ~hrw Mar 23 09:18:49 from memory, hrw is Marcin Juszkiewicz - collie maintainer, OpenEmbedded/OpenZaurus/OPIE developer. OpenZaurus 3.5.4.x release maintainer Mar 23 09:19:17 TheMasterMind1: 2.6 on collie is not usable yet and I do not have time to work on getting it in OE Mar 23 09:19:41 TheMasterMind1: but I support Pavel Machek (collie/2.6 kernel hacker) with all userland tools which he need Mar 23 09:20:13 TheMasterMind1: I also provided him tcx3106 for making serial cable (which he lacked) Mar 23 09:21:03 03hrw 07org.oe.oz354x * rdf871080... 10/packages/gnome/orbit2_2.10.2.bb: orbit2: disable PARALLEL_MAKE Mar 23 09:21:08 03hrw 07org.oe.oz354x * r70cfdc2f... 10/ (174 files in 22 dirs): dropped whole maemo stuff - we do not use it in OZ and removing it is easier way to fix world build Mar 23 09:21:11 morning all Mar 23 09:21:15 how's touchscreen support in the 2.6 collie kernel doing Mar 23 09:21:35 TheMasterMind1: recently he got it working Mar 23 09:21:41 hi Dirk Mar 23 09:21:45 TheMasterMind1: read his blog Mar 23 09:21:59 Hey Marcin Mar 23 09:22:12 sweet, only took a few years. Mar 23 09:23:13 TheMasterMind1: feel free to work on 2.6 kernel for collie Mar 23 09:23:51 TheMasterMind1: I 'like' people which complain without doing nothing in subject Mar 23 09:25:51 hrw|work: i have =) i wrote the patch that display the cute little framebuffer logo on boot, and hacked a few other drivers. work on touchscreen support for a week or two but didn't get too far Mar 23 09:26:11 TheMasterMind1: http://pavelmachek.livejournal.com/ Mar 23 09:26:14 03xora 07org.oe.dev * r31c5baa2... 10/packages/gpdf/ (files/gpdf-2.10.0-gcc4-2.patch gpdf_2.10.0.bb): gpdf_2.10.0.bb : add gcc 4.X build patch Mar 23 09:28:44 koen: are you running angstrom images yet? Mar 23 09:28:53 morning Dirk Mar 23 09:28:59 someone summarized his touchscreen battle in few words recently: his tslib 'issue' would have been solved withins seconds Mar 23 09:29:06 XorA: no, socket() is still giving us problems Mar 23 09:29:22 koen: ah, that could be a bit of a staller :-) Mar 23 09:29:26 Hi RP Mar 23 09:29:40 TheMasterMind1: I lack time to test collie 2.6 and I want to avoid fetching it from git because of fetch size: 1.2G last time Mar 23 09:30:06 koen: as you may have noticed, I have dual dev environments, just waiting until I can boot it Mar 23 09:30:23 hrw|work: The git fetcher is slowly improving. Its still massive though. I must push my git pruned packed fix for it... Mar 23 09:30:35 someone know how to fetch one version of files from git without whole repo? Mar 23 09:31:00 03koen 07org.oe.dev * r58fe56d4... 10/conf/distro/angstrom.conf: angstrom: insist a bit harder on making angstrom-dev the default maintainer for unmaintained recipes Mar 23 09:31:16 hrw|work: Its not possible for it to work like that afaik Mar 23 09:31:54 RP: suxx Mar 23 09:32:18 XorA: you can boot it, but anything networky (like ehm, X) won't work Mar 23 09:32:19 RP: how do you think - can we pester Pavel to provide patches? Mar 23 09:32:47 koen: heh heh, hmmm, I wonder if there is a pdf viewer for aalib :-) Mar 23 09:33:12 hrw|work: I doubt it - he doesn't work like that. I'll have to pull his tree and turn it into a patch Mar 23 09:33:31 * koen lobs of some heads Mar 23 09:33:33 hrw|work: i gotcha. i wasn't being serious, just very tired so my comments weren't very well worded. i was involved with OZ back when kergoth first split up the 2.4 embeddix patch and starting porting to 2.5.x. I remember you, i think you were more involved with opie back then? Mar 23 09:33:50 anyway. pavel hang around here? Mar 23 09:34:00 TheMasterMind1: no Pavel here ever iirc Mar 23 09:34:15 TheMasterMind1: ah.. now I remind from where your nick looks familiar.. Mar 23 09:34:38 which is why stuff line 'module raw input' TSLIB_TSDEVICE=/dev/input/event0 is taking months for pavel Mar 23 09:34:43 s/line/like/ Mar 23 09:34:54 TheMasterMind1: I got OE images booting and this gave me r/w to OE - then I mostly worked on opie integration Mar 23 09:35:00 ah. maybe i'll email him. I had the tslib stuff working but it was way too noisy to be usable. i know kergoth spent atleast a month trying to write tslib filters to get the noise out with not much luck Mar 23 09:36:23 and the embedix touchscreen code looked so straightforward (for embedix code anyway- what a horrible horrible mess) Mar 23 09:37:07 OT: what can be done with 2 AY chips: http://homepage.mac.com/WebObjects/FileSharing.woa/wa/Madonna_HungUP_ZX_Spectrum128.mp3.mp3?a=downloadFile&user=iuzlik&path=.Music/Madonna_HungUP_ZX_Spectrum128.mp3 Mar 23 09:37:51 hm i suppose 4:45am is a good a time as any to head to bed. Mar 23 09:37:51 night Mar 23 09:38:05 'night TheMasterMind1 Mar 23 09:38:12 sweet dreams TheMasterMind1 Mar 23 09:40:01 03dirk 07org.oe.dev * r8ddf5ef3... 10/packages/wlan-ng/ (22 files in 6 dirs): wlan-ng: add 0.2.3 and svn version, split usb, cs and pci driver Mar 23 09:42:05 today I felt adventurous and removed (renamed) my tmp directory, trying to do a completely fresh build from .dev, but I found this error in qemu-native: http://pastebin.com/617682 Mar 23 09:42:15 pb_: ping Mar 23 09:42:22 ArjanS: gcc4 on host? Mar 23 09:42:33 pb_: I think bero introduced the binutils.bb, and he tested it on non eabi builds Mar 23 09:42:41 hrw|work: nope, 3.4.5 Mar 23 09:42:44 no, texi2html isn't installed Mar 23 09:43:16 koen: is that needed on the host? if so, I think the error should be more clear about that Mar 23 09:43:43 ArjanS: | make: texi2html: Command not found :) Mar 23 09:44:08 koen: yeah I know, but I was figuring that should be somewhere in OE Mar 23 09:44:35 we updated RequiredSoftware a few days ago because of that Mar 23 09:45:49 allright, thanks Mar 23 09:47:06 uf.. looks like most of x11 feed is fixed Mar 23 09:47:23 fixed as 'cleaned from maemo stuff' Mar 23 09:48:25 03dirk 07org.oe.dev * r4e45961e... 10/conf/machine/include/tosa-2.6.conf: conf/tosa-2.6.conf: add wlan-ng-modules-usb to RDEPENDS Mar 23 09:49:28 * koen wonders why socket() keeps returning ENOSYS Mar 23 09:49:55 do13_: which module is missing? Mar 23 09:50:27 zecke: ??? Mar 23 09:50:44 do13_: when socket returns NOSYS? Mar 23 09:51:32 hrw|work: very cool, been a long time since I messed with AY music Mar 23 09:52:11 zecke: I don't know Mar 23 09:52:42 sorry :} Mar 23 09:53:04 do13_: the 'koen' didn't highlight too good Mar 23 09:53:08 hrw|work: is there a page to go with that link? Mar 23 09:53:19 do13_: I think I'm a blocker for having a working konq-e in OE Mar 23 09:53:30 do13_: what are the consequences of your QtE-MT change? Mar 23 09:53:43 zecke: lack of time :) Mar 23 09:54:01 zecke: It's on my list Mar 23 09:54:02 do13_: will Opie fit in a 16mb ROM after this change? Mar 23 09:54:05 XorA: Polish Atari forum Mar 23 09:54:19 hrw|work: damn I need a language chip :-) Mar 23 09:54:21 do13_: hehe, nowadays I always try to be not the blocker Mar 23 09:54:29 XorA: search for YERZMYEY Mar 23 09:54:56 XorA: http://parishq.net/proposed//sound/yerzmyey-5th_chnl.mp3 this is 5ch on one AY Mar 23 09:55:15 hrw|work: that guy has too much time Mar 23 09:55:32 zecke: Qte-MT is ~700K bigger Mar 23 09:55:49 do13_: including rtti and exceptions? Mar 23 09:56:13 yes Mar 23 09:56:23 on flash +700k? Mar 23 09:57:10 do13_: if we turn on visibility change we might save that again Mar 23 09:57:30 zecke: This sounds good Mar 23 09:57:48 good morning Mar 23 09:57:50 zecke: really 700K? Mar 23 09:57:52 XorA: did you add thunderbird to OE? Mar 23 09:57:56 hey Bernardo Mar 23 09:58:07 hi koen Mar 23 09:58:08 do13_: I think I saved between 300 and 700k Mar 23 09:58:16 do13_: but I think we can save that :) Mar 23 09:58:24 .. Mar 23 10:00:47 koen: not originally, but I updated it Mar 23 10:01:17 hi Bernardo Mar 23 10:01:40 Bernardo: how goes 3.5.4.1 on akita? Mar 23 10:03:39 i've some pb compiling poky's matchbox ( http://pastebin.com/617719 ) , any idea ? Mar 23 10:06:34 hrw|work: I couldn't get it to build last night and this morning before coming to work Mar 23 10:07:03 I tried all I could, btu when doing "bitbake opie-kdepim-image" it would always try to build orinoco-modules 0.13 and fail Mar 23 10:07:23 seems like it is ignoring PREFERRED_VERSION in local.conf Mar 23 10:07:39 XorA: cp: cannot stat `/home/koen/OE/monotone/org.openembedded.dev/packages/mozilla/files/mozconfig': No such file or directory Mar 23 10:07:41 hm. akita conf fscked? Mar 23 10:07:51 XorA: tb 1.0.7 doesn't build because of that Mar 23 10:08:50 koen: I forgot to copy thunderbird-1.0.6 to thunderbird-1.0.7 Mar 23 10:09:29 Genesis: You need to work out why it has -L/usr/X11R6/lib in that gcc command Mar 23 10:09:43 Its trying to link against system libraries which is bad Mar 23 10:10:26 ho Mar 23 10:11:45 i should recompile all the stuff i imagine , because i lauch oh-image yesterday ;) Mar 23 10:12:08 i'll try to understand that , thx . Mar 23 10:12:30 hrw|work: can't check now... I basically copied the conf I was using from 3.5.4, edited it to use openzaurus-3.5.4.1 as the distro, then had to add a PREFERRED_PROVIDER for it to build the linux-openzaurus kernel and not the jlime one Mar 23 10:12:46 Bernardo: fixed - will push Mar 23 10:12:54 hi! can someone explain me what the difference between meta-* and task-* packages is ? Mar 23 10:13:11 Genesis: I'd be interested to know where that path came from. It could be a broken pkgconfig from some other package Mar 23 10:13:15 grma: task-* define tasks which are used by *-image and meta-* Mar 23 10:13:19 hrw|work: thanks! I'll test it tonight Mar 23 10:13:43 RP, : i need to searsh a bit , i'll tell you if i find an issue Mar 23 10:14:35 03hrw 07org.oe.oz354x * rb55dcb62... 10/conf/machine/ (akita.conf borzoi.conf spitz.conf): cxx00 machines: 2.6 kernel by default once more Mar 23 10:15:03 Genesis: ok, thanks. If you do find the problem I'll fix it in poky/oe as appropriate Mar 23 10:18:18 03xora 07org.oe.dev * r1a896e8b... 10/packages/mozilla/thunderbird-1.0.7/ (extensions-hack.patch mozconfig): thunderbird-1.0.7 : add missing files. Mar 23 10:19:49 hrw|work: thanks, and meta-* ? when i want to build for example gpe(only gui,xserver...) without building the hole image i would do meta-gpe ? Mar 23 10:21:09 grma: yes Mar 23 10:21:19 ahh, thanks Mar 23 10:21:31 grma: thats what for meta-* are Mar 23 10:23:10 hrw|work: wonderfull, i am checking and understanding OE step by step :-) Mar 23 10:27:00 grma: good Mar 23 10:35:45 koen, XorA: does EXCLUDE_FROM_SHLIBS=1 in mozilla.bbclass sounds good for you? this should stop firefox depending on thunderbird Mar 23 10:36:49 hrw| work: now i have builded meta-opei, how do i use it then? ore where are then the compiled things which i can to copy to the target ? Mar 23 10:37:03 s/meat-opei/opie/ Mar 23 10:37:44 grma: use deploy/ipk/ as feed Mar 23 10:38:32 hrw|work: if that fixes it, yes Mar 23 10:39:13 would it be hard to make also a tar.gz file out of the compiled things ? Mar 23 10:44:42 hrw|work: that wont work Mar 23 10:45:04 hrw|work: that excludes it from finding RDEPENDS by shlibs as well, I did do that ages ago Mar 23 10:46:25 XorA: I thought that EXCLUDE_FROM_SHLIBS only worked one way Mar 23 10:47:00 koen: it certainly didnt last time I tried it, it meant firefox.ipk ended up with no Depends: Mar 23 10:47:24 koen: that was a while ago though so it might have changed Mar 23 10:47:42 XorA: so I vote for adding BROKEN=1 into them Mar 23 10:47:48 etc Mar 23 10:48:07 grma: tar.gz instead of ipk? Mar 23 10:48:49 hrw|work: whatever, all I do is update them, mozilla build is a nightmare, it needs a good splitting up and autoconfing :-) Mar 23 10:49:21 hrw|work: i am not that familiar with ipk :-) i wanted easy copy the meta-gpe output to my nfs target :-) Mar 23 10:50:23 hrw|work: experiment with the EXCLUDE_SH* stuff and see if it has changed Mar 23 10:50:41 XorA: I use old bitbake so rather not Mar 23 10:50:49 I am busing finding out mozilla is b0rk3d with gcc4 Mar 23 10:51:51 grma: ipk packages contain dependency stuff which is good to have Mar 23 10:53:55 hrw|work: ok i understand. Mar 23 10:56:04 which version of bugzilla do we run at OE? Mar 23 10:56:18 2.16 iirc Mar 23 10:56:20 * koen checks Mar 23 10:58:00 hrm Mar 23 10:58:08 I can't seem to find the version info Mar 23 10:58:14 zecke: better mail treke Mar 23 10:59:58 hmm kbugbuster doesn't work with the bugzilla :( **** ENDING LOGGING AT Thu Mar 23 10:59:58 2006