**** BEGIN LOGGING AT Thu Apr 06 09:59:56 2006 Apr 06 10:09:24 NOTE: package libsm-6.0.3-r1: task do_distribute_sources: started Apr 06 10:09:25 NOTE: Licenses (['MIT-X']) are not all listed in SRC_DISTRIBUTE_LICENSES, skipping source distribution Apr 06 10:10:48 hrw|work: If I remember rightly, MIT-X says you cna do anything with it... Apr 06 10:11:01 ipkg is terribly slow with oz 3.5.4 feeds Apr 06 10:12:23 good morning Apr 06 10:12:34 hi Bernardo Apr 06 10:13:43 hi Bernardo Apr 06 10:13:55 NOTE: package libxcursor-1.1.2-r2: task do_distribute_sources: started Apr 06 10:13:56 NOTE: Licenses (['BSD-X']) are not all listed in SRC_DISTRIBUTE_LICENSES, skipping source distribution Apr 06 10:14:02 Bernardo: Did you have an SD card with an unrecognised SCR structure? Apr 06 10:14:39 NOTE: package libgpg-error-1.0-r0: task do_distribute_sources: started Apr 06 10:14:39 NOTE: Licenses (['GPL', 'LGPL', 'FDL']) are not all listed in SRC_DISTRIBUTE_LICENSES, skipping source distribution Apr 06 10:14:42 hmm Apr 06 10:14:48 RP: i d o Apr 06 10:15:04 ade|desk: Fancy trying a recent 2.6.16 kernel and see if it works? Apr 06 10:16:06 will do as soon as i can get a build running Apr 06 10:16:50 * ade|desk kicks vmware Apr 06 10:17:10 ade|desk: The kernel and modules from http://sam.rpsys.net/openzaurus/3.5.4.1-alpha2/ should work Apr 06 10:17:29 ok that mmakes life easier Apr 06 10:18:52 RP: let me check the dmesg from my akita Apr 06 10:19:20 * Bernardo is looking for a 802.11 (cheap) cf card Apr 06 10:19:53 Bernardo: If ade|desk can test, that's fine - I'd just like to know whether a patch I added helps or not... Apr 06 10:19:57 Bernardo: I look for cheap 802.11 usb stick Apr 06 10:21:08 03mickeyl 07org.oe.dev * re544df2d... 10/packages/efl/ (5 files): evas: stage plugins Apr 06 10:21:09 hrw|work: I have one that works if you have a powered usb hub - a cnet CWD 854 Rev. E Apr 06 10:22:04 Bernardo: I will connect it to desktop pc so power is no problem Apr 06 10:22:07 RP: mine seems to have no problem in 2.6.15 - the only output I have in dmesg after inserting it is "mmcblk0: mmc0:0001 SD 988160KiB" Apr 06 10:22:33 hrw|work: then this one or another rt2570 based stick will serve you well Apr 06 10:22:41 Bernardo: ok, I think I must have been thinking of someone else, sorry :) Apr 06 10:22:49 not as much range as a pcmcia card, but works very well with linux Apr 06 10:23:03 RP: NP. Whatever I can do to help... Apr 06 10:23:07 Bernardo: I need <10m ;) Apr 06 10:24:01 I only had problems with this one when I was more than 30m away, and with at least three stone walls between it and the AP. Apr 06 10:24:34 down in the patio, and I live on the first floor... +- 6 meters up, 25 straight line Apr 06 10:24:50 Bernardo: I want to buy usb wifi stick to have wifi net at work - stick in desktop, cf in zaurus Apr 06 10:25:17 then you'll be ok with this one. I've used it in that config Apr 06 10:25:46 it even has too much range for that, I had a few guys here trying to connect to it... :) Apr 06 10:26:11 now, back to ebay to try to find a replacement for this sandisk... :( Apr 06 10:27:12 Pretec CF Wlan 2.5 card is nice Apr 06 10:27:24 linksys wcf12 is nice Apr 06 10:27:32 dlink dcf-660w is bulky Apr 06 10:29:52 seems like there's none to be found right now in ebay Apr 06 10:34:55 hrw|work: if I add INHERIT += "rm_work" to my local.conf I can no longer build anything, is that the right syntax? Apr 06 10:36:22 XorA: rm_work is a bit faulty but workable - I'm building 3.5.4.1 with it Apr 06 10:46:55 RP: sorry for maxing out sam Apr 06 10:47:11 03hrw 07org.oe.oz354x * rbd06749d... 10/conf/licenses.conf: licenses.conf: added 'MIT-X BSD-X FDL Bitstream Vera' into list Apr 06 10:50:28 hrw|work: http://pastebin.com/643810 Apr 06 10:50:39 hrw|work: bit faulty, Id say, very faulty Apr 06 10:51:16 ok - it is not so usefull - patches welcome Apr 06 10:52:10 I wrote it 19 months ago.. Apr 06 10:52:14 http://oe-devel.bkbits.net:8080/openembedded/related/classes/rm_work.bbclass?nav=index.html|src/|src/classes Apr 06 10:52:51 ade|desk: Its ok, I'm not using that connection much atm so I'df not noticed :) Apr 06 10:54:08 morning all Apr 06 10:54:27 hi dirk Apr 06 10:54:35 hi Dirk Apr 06 10:54:40 RP: i think the kernel is too big Apr 06 10:54:43 hi Marcin, Richard Apr 06 10:55:33 ade|desk: It wouldn't have built it if it were too big? Apr 06 10:56:32 hrw|work: ok, I update bitbake, now it builds but doesnt rm the WORKDIR, I will investiage Apr 06 10:56:41 hrw|work: The last line of rm_work.bbclass should read addtask rm_work before do_build after do_package Apr 06 10:57:04 RP: just going by bug #816 Apr 06 10:57:47 RP: good to know Apr 06 10:58:11 RP: does current way is wrong? Apr 06 10:58:31 NOTE: package libpixman-0.1.6-r0: task do_distribute_sources: started Apr 06 10:58:31 NOTE: Licenses (['X11']) are not all listed in SRC_DISTRIBUTE_LICENSES, skipping source distribution Apr 06 10:58:34 heh.. another one Apr 06 11:04:37 hrw|work: I think it might confuse bitbake Apr 06 11:07:29 thx Apr 06 11:08:14 changed and started build Apr 06 11:12:26 RP: can mail led show sd/mmc transfers? Apr 06 11:15:49 Bernardo: ping Apr 06 11:16:00 or even RP: ping Apr 06 11:16:31 RP: http://bugs.openembedded.org/show_bug.cgi?id=777 - reversed Calendar/Adress on cxx00 - is it kernel bug? Apr 06 11:17:28 RP: on c7x0 Calendar is 59, Address is 60 - keymap-2.6.map is same for c7x0/akita/spitz in this place so it looks like kernel reverse them.. we can fix it in keymap-2.6.map anyway Apr 06 11:18:22 hrw|work: It could well be a kernel bug Apr 06 11:18:46 If it is in the kernel, I'd prefer to fix the kernel rather than the keymaps Apr 06 11:19:02 * hrw|work too Apr 06 11:19:38 hrw|work: If we add an mmc activity led trigger, it can show up on the mail light. Nobody's written such a trigger yet Apr 06 11:19:58 RP: is it hard to write? Apr 06 11:20:26 hrw|work: The trigger code is easy. Finding the right place to put it isn't Apr 06 11:20:34 Although I've not looked Apr 06 11:21:55 ok Apr 06 11:23:12 I will move to sd soon so it would be nice to have it Apr 06 11:24:58 hrw|work: I'll keep it in mind Apr 06 11:25:03 We need more kernel hackers Apr 06 11:26:10 NOTE: package qte-2.3.10-r32: task do_patch: completed Apr 06 11:26:12 uf Apr 06 11:35:21 Bernardo: you are running opie? Apr 06 11:35:39 hrw|work: going to lunch now, is it about the reversed keys? Apr 06 11:35:47 yes Apr 06 11:35:50 hrw|work: yes, I'm running opie Apr 06 11:36:20 I'm pretty sure it's kernel related Apr 06 11:36:23 Bernardo: ipkg install http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4x-experimental/libqte/libqte2_2.3.10-r32_akita.ipk Apr 06 11:36:31 Bernardo: its fixed Apr 06 11:36:35 ok, will do that after lunch Apr 06 11:36:38 thx Apr 06 11:37:01 is that in 354x already? Apr 06 11:37:06 not yet Apr 06 11:37:10 have to run now, wife's calling Apr 06 11:37:15 I want it tested Apr 06 11:37:17 cu Apr 06 11:37:22 greet wife ;) Apr 06 11:37:31 I will, thanks Apr 06 11:38:42 RP: you are added into openzaurus project members at berlios Apr 06 11:38:48 mickey|zzZZzz: you are added into openzaurus project members at berlios Apr 06 11:40:01 hrw|work: TYVM :) Apr 06 11:40:06 ;) Apr 06 11:40:15 hrw|work: I can confirm those keys are a kernel bug Apr 06 11:41:11 ok Apr 06 11:41:35 hrw|work: Its an easy fix, if a bit annoying to have to push into mainline Apr 06 11:50:49 hrw|work: apologies, rm_work is slightly different than I though, it is working here now Apr 06 11:51:16 np Apr 06 11:51:25 it fails sometimes anyway Apr 06 11:51:35 hrw|work: why does it called both before do_build and do_package? Apr 06 11:52:17 dont remember now Apr 06 12:00:48 ~timriker Apr 06 12:00:49 somebody said timriker was my owner http://rikers.org/ mailto:Tim@Rikers.org mailto:TimR@Debian.org maintainer of BZFlag, member of a ton of open source projects http://www.advogato.com/person/timriker/ http://sourceforge.net/users/timriker/ the guy who GPL'd SCO's ABI files, giving every Linux user the right to use them ;-), or a very cool guy. Apr 06 12:01:29 do I remember correctly - he host bugs.oe? Apr 06 12:20:00 ~lart zaurusd Apr 06 12:20:01 * ibot stamps zaurusd on the forehead with the official Troll marker Apr 06 12:32:35 what would you tell for new variable: USE_FEED_FOR_IMAGES? Apr 06 12:37:43 hrw|work: treke (aka Greg GIlbert) hosts bugs.oe Apr 06 12:37:56 http://pastebin.com/643930 - my version of koen patch for using feeds as source of packages for images Apr 06 12:38:08 controlled by USE_FEED_FOR_IMAGES = 1/0 Apr 06 12:38:10 thx Apr 06 12:38:12 ~treke Apr 06 12:38:17 treke is probably Greg Gilbert , or a goon, or a goon with no life, and no neck Apr 06 12:39:36 ~seen treke Apr 06 12:39:43 re Bernardo Apr 06 12:39:44 treke was last seen on IRC in channel #oe, 86d 12h 58m 21s ago, saying: 'http://zaurus.catstar.org/jmqt/index_e.html'. Apr 06 12:40:37 RP: what do you think about that patch? Apr 06 12:43:06 hrw|work: I think we can make a cleaner version where this gets done in a seperate .bbclass Apr 06 12:43:25 I'll post it as an enhancment in bugtracker Apr 06 12:44:10 hrw|work: ok, its certainly worth doing something with, just a question of the exact implementation :) Apr 06 12:44:22 yep Apr 06 12:45:11 re hrw|work Apr 06 12:45:30 hrw|work: do you want me to test that qte now? Apr 06 12:45:44 yes Apr 06 12:45:53 ok Apr 06 12:47:24 thx Apr 06 12:57:19 hrw|work: the keys are still switched Apr 06 12:58:17 when I check the buttons app, everything is set as it should, callendar launching ko/pi, address launching ka/pi, but pressing callendar I get ka/pi Apr 06 12:58:43 and pressing address I get ko/pi Apr 06 12:59:01 Bernardo: that thing is not libqte but kernel Apr 06 12:59:30 Bernardo: does cursors are ok now? (was rotated with r31) Apr 06 13:00:23 ~lart sd for being so slow... Apr 06 13:00:23 * ibot calls sd on the phone ... the lights are on but nobody's home for being so slow... Apr 06 13:00:55 kingston 512M sd in sync mode is like modem 9600 Apr 06 13:01:43 <_law_> hrw|work, modell? Apr 06 13:02:11 normal lowspeed Apr 06 13:02:37 yes, cursors work fine Apr 06 13:02:48 Bernardo: great Apr 06 13:02:52 I hadn't tested r31, I had r27 right now Apr 06 13:05:59 ok - I know that r31 had cursors problem Apr 06 13:07:07 r27 didn't, I think. I didn't do much testing on them Apr 06 13:07:10 morning Apr 06 13:08:39 hi chouimat Apr 06 13:09:27 Today quote:The Bible teaches to love your neighbor, and Kama-Sutra explains how. Apr 06 13:10:30 I have the following problem. The installation of a software is made by the command "make install". The Software is installed in /usr/realtime. How can I say that somthing like a faked root is used while installation to make the package? Apr 06 13:10:43 03hrw 07org.oe.oz354x * r03036e36... 10/packages/qte/ (5 files in 2 dirs): (log message trimmed) Apr 06 13:10:43 Thanks to the ROAD GmbH we now have a bugfixed QtE kernel-keymap.patch Apr 06 13:10:43 Bugfixes: Apr 06 13:10:43 - Handle keys > 127 (7Bit) Apr 06 13:10:43 - Workaround broken keymaps Apr 06 13:10:44 - Fix Autorepeat Apr 06 13:10:46 - plenty of other fixes Apr 06 13:11:25 Cobelius: fix its makefile how to use DESTDIR Apr 06 13:11:54 can I write ${D} in it? Apr 06 13:12:00 Or in the .config File? Apr 06 13:12:11 Is this variable also availabe in make? Apr 06 13:12:28 let makefile use DESTDIR and you will tell that DESTDIR=${D} Apr 06 13:42:31 how was the formar for .bb filename name_version_revision or name_version-revision? Apr 06 13:42:51 NAME_VER.bb Apr 06 13:43:04 revision in PR inside of bb Apr 06 13:43:17 okay Apr 06 13:43:18 thank you Apr 06 13:43:20 np Apr 06 13:50:04 RP: pcmciautils are not enabled for alpha2 yet Apr 06 13:50:21 hrw|work: I thought I pushed that change? Apr 06 13:52:20 pushed but it looks like not working ;( Apr 06 13:54:11 hmm.. will look more into it Apr 06 13:54:57 false alarm.. something gave me pcmcia-cs also building... Apr 06 13:55:32 sory Apr 06 13:57:41 np :) Apr 06 13:57:59 Is there a variable existing which contains the path to the kernel header files? Apr 06 13:59:54 STAGING_KERNEL_DIR iirc Apr 06 14:00:26 hmm... kernel-modules depend on all modules and is installed ;( Apr 06 14:43:51 50K left for kernel size on c7x0 Apr 06 15:01:11 cu Apr 06 15:31:05 I have a question about mtd Apr 06 15:33:20 i have /dev/mtdblock? but no /dev/mtd?, and i have kernel message complaining about ... Apr 06 15:33:23 jffs2_scan_eraseblock(): Magic bitmask 0x1985 not found at 0x00000400: 0x8010 instead Apr 06 15:33:34 hmm Apr 06 15:33:36 not sure Apr 06 15:39:14 tux1800: thats on the travis still? Apr 06 15:52:57 bonne soiree Apr 06 15:56:39 hey everybody. I'm hitting a "ERROR: Nothing provides runtime dependency libperl5" when building openslug. this with a fresh everything/start from scratch this am. Any Ideas? Apr 06 16:02:07 <_law_> hi all Apr 06 16:04:08 HopsNBarley: Current bitbake head looks totally broken to me - try a slightly older version of bitbake Apr 06 16:07:33 thanks RP. Apr 06 17:13:45 yo Apr 06 17:20:57 hey mickeyl Apr 06 17:21:11 how goes the dissertation? Apr 06 17:22:27 pretty tough atm. can't really focus on continuing the chapter that describes my architecture.... working on the evaluation chapter instead :) Apr 06 17:22:48 what's your topic/title? Apr 06 17:23:01 adaptive middleware for mobile distributed systems Apr 06 17:23:45 i have developed a component based middleware architecture that reconfigures its building blocks at runtime based on which services / nonfunctional services the applications need Apr 06 17:24:34 why not just use CORBA? Apr 06 17:24:40 heh Apr 06 17:24:41 sounds interesting. current. marketable (-; you planning on work or academia? Apr 06 17:25:35 CORBA is bound to the oo-rpc client/server concept. my middleware can do that, but can also can do peer2peer abstractions + can do tuple space, messaged passing, etc. what you need. Apr 06 17:25:48 hmmm Apr 06 17:26:37 send me a copy when you are done? Apr 06 17:26:39 HopsNBarley: i'm not sure yet. i guess i'd like to be a uni professor, but i don't like to continue doing the same research style i did the last 5 years. i want something else for a while Apr 06 17:26:55 Crofton: sure Apr 06 17:29:09 i'd thought about the prof. route, but then in grad school i saw how much time the profs. spend getting ready to do work (grants, reviews, etc) rather than doing work. that's in the US at least. Apr 06 17:30:28 mickeyl, did you design a language to describe this, or operating at a slightly lower level? Apr 06 17:30:29 * Philippe is back (gone 20:01:04) Apr 06 17:31:40 good evening Apr 06 17:37:51 mickeyl, pH5: yo Apr 06 17:37:56 reenoo: yo Apr 06 17:38:41 evening Apr 06 17:38:44 yo pb__ Apr 06 17:40:54 03eFfeM 07org.oe.dev * r5945a66a... 10/ (2 files in 2 dirs): Apr 06 17:40:54 asterisk: updated download path, added PR Apr 06 17:40:54 slugos-packages.conf: added asterisk to package list Apr 06 17:45:01 HopsNBarley: i settled on describing the architecture using uml2 static and dynamic views. it's not really good, but it's the most commonly accepted description language. the middleware itself is written in Python (like bitbake :) - that helped to make all bindings dynamic etc. Apr 06 17:46:22 HopsNBarley: yeah, that's true, but it wouldn't be a major factor for me. in germany the standard route is to submit a Habilitation thesis before you can apply for a university prof. this would mean 3 to 5 more years doing the same research & publish style - and I'm not sure whether i want that Apr 06 17:46:46 hi pb, REdOG Apr 06 17:46:48 hi reenoo Apr 06 17:46:56 moin mickeyl Apr 06 18:04:07 hrm Apr 06 18:04:37 would be nice if module deps were listed as ipk dependencies Apr 06 18:10:29 reenoo: hey do you maintain the java packages? Apr 06 18:12:18 TheMasterMind1: I do at times Apr 06 18:16:40 reenoo|afk: i want to make non-gui versions of classpath and kaffe.. would ${PN}-console as the package name make sense? Apr 06 18:46:43 1.4M ./usr/lib/ipkg/info Apr 06 18:46:43 592K ./usr/lib/ipkg/alternatives Apr 06 18:46:45 sheesh Apr 06 18:48:10 thats too much overhead for a bootstrap image Apr 06 19:04:46 that does sound like a lot Apr 06 19:06:39 Thats the price you pay for ipkg Apr 06 19:07:23 well, apparently, but 592k of alternatives? Apr 06 19:07:52 it's hard to imagine how you could end up with that much data unless every file on the system was an alternative. Apr 06 19:07:53 true - that is a little high Apr 06 19:08:13 Well, isn't everything in busybox an alternative? Apr 06 19:08:37 s/isn't everything/doesn't everything/ Apr 06 19:08:40 more or less, but how many executables does busybox provide? it's not _that_ many. Apr 06 19:08:41 s/an/have an/ Apr 06 19:09:34 and 1.4MB for info also seems a bit much. the control file should only be a few hundred bytes for each package, plus maybe twice as much again in maintainer scripts. Apr 06 19:10:02 oh well Apr 06 19:10:17 I guess this is why we have jffs2 compression :-} Apr 06 19:12:47 If I do du -sh in my rootfs/usr/lib/ipkg/info dir, it says 3.7M. but 'cat * | wc -c' gives about 304k. For alternatives this is 608k vs. 4674 bytes. Apr 06 19:13:45 pH5: yeah, thats what I get Apr 06 19:14:10 I have 146 alternatives in my directory Apr 06 19:14:48 biggest one is 47 bytes, but du shows 4k per file Apr 06 19:15:29 But almost all of them are busybox related Apr 06 19:28:01 emte: hey Apr 06 19:28:23 emte: +10 devices for familiar? Opie can live quite happily without glibc localisation Apr 06 19:28:52 applying the 'timezone' argument... well it is late Apr 06 20:01:19 zecke: bitbake head is rather broken but i guess you know this? Apr 06 20:27:10 RP: no Apr 06 20:27:34 RP: no, I don't know that. I suspected it though Apr 06 20:29:30 zecke: Earlier on I tried building something so it parsed the cache, tried to build and failed. I reran bitbake on the same package and it told me nothing provided it Apr 06 20:30:02 zecke: In other words the cache is ending up corrupt :-/ Apr 06 20:31:52 mickeyl: thanks, that seemed to work.....of course now I get a segfault.... Apr 06 20:33:09 RP: oh, I think hrw and koen reported that 'years' ago Apr 06 20:33:48 RP: I have a night shift ahead, we might be able to look into it on the weekend Apr 06 20:34:00 zecke: Ah, I thought it was a result of the cach changes :-/ Apr 06 20:34:26 zecke: np. I just wondered if it was known about. If it is, fine :} Apr 06 20:39:20 RP: OT: Do you work on Monahan Linux support? Apr 06 20:40:12 zecke: no, I'm not even sure what that is :) Apr 06 20:40:38 RP: the new Intel PXA Apr 06 20:42:37 What, are they naming their new processors after congressmen now? Apr 06 20:43:11 zecke: Ah, I just know it by its number :). Sadly I have nothing to do with it Apr 06 20:43:32 RP: by serial number? Apr 06 20:43:41 Serial Numbers from x-y are... Apr 06 20:44:12 zecke: pxa25x, pxa26x, pxa27x etc. I'm not mad enough to memorise serial numbers :) Apr 06 20:44:48 * Philippe is away: visual contact - melancholic dreams Apr 06 20:51:41 RP: do you have an idea how bitbake remembers this error? Apr 06 20:52:02 RP: we have an already failed list, but we will never save it Apr 06 20:52:02 RP: do we 'destroy' the cached file in case of error? Apr 06 20:52:44 zecke: I'm afraid I wasn't able to look into it. We must cache something though :-/ Apr 06 20:54:36 "SDK Light is provided in binary format and includes: Apr 06 20:54:37 pre-built binaries for i386 and ARM" Apr 06 20:54:38 lol Apr 06 20:54:46 TT smells like using EABI already Apr 06 21:22:16 hi Apr 06 21:22:24 hi CoreDump|home Apr 06 21:22:24 'lo CoreDump|home Apr 06 21:22:38 evas is staging now....but I get a segfault from edje.... Apr 06 21:22:54 if anyone else would care to try building entrance I'd be grateful Apr 06 21:23:20 JustinP: .dev or .1? Apr 06 21:23:48 1? Apr 06 21:23:50 .dev Apr 06 21:23:58 oz354* still builds fine Apr 06 21:24:11 changes to evas to make it modular are causing issues in dev right now... Apr 06 21:24:17 JustinP: I'll try to build .dev then overnight and let you know Apr 06 21:25:16 e-image-core will do right? Apr 06 21:25:53 yep, that wold be great Apr 06 21:26:04 make sure to update, mickeyl fixed the staging issue with evas for me Apr 06 21:26:21 will do Apr 06 22:48:57 Hi ! Please, can anyone tell me how to set the cross-compiler to OE so that 'bitbake a_package' use it ? Apr 06 22:49:17 cromain: hi, using ASSUME_PROVIDED Apr 06 22:49:24 I'm a bit confused: in the Wiki, there are instruction using kernel 2.4 and gcc2.95 Apr 06 22:49:24 cromain: and putting your compiler into the path? Apr 06 22:49:43 cromain: right, this is special for kernel compile Apr 06 22:49:49 cromain: but it is very similiar Apr 06 22:50:33 zecke: so ASSUME_PROVIDED = "virtual/arm-linux-gcc-4.1" in my case Apr 06 22:50:47 cromain: no ;) Apr 06 22:51:04 cromain: (we lack documentation and would be glad if one could help us to improve it) Apr 06 22:51:15 zecke: ASSUME_PROVIDED = "/the/path/to/my_compilo" ? :-) Apr 06 22:51:24 cromain: base.bbclass adds 'DEPENDS' for libc and gcc Apr 06 22:51:36 cromain: depending on your TARGET_ARCH and OS Apr 06 22:51:41 zecke: yes, I'll do as soon as I'll succeed ! Apr 06 22:51:56 cromain: you can do ASSUME_PROVIDED += "virtual/arm-linux-libc" Apr 06 22:52:10 cromain: and ASSUME_PROVIDED += "virtual/arm-linux-gcc" Apr 06 22:52:22 cromain: then just make sure you have a arm-linux-gcc binary in your path Apr 06 22:53:14 zecke: Ok, I have armv5te-softfloat-linux-gnu-gcc, so does ASSUME_PROVIDED += "virtual/armv5te-softfloat-linux-gnu-gcc" will do the trick ? Apr 06 22:53:30 cromain: no ;) Apr 06 22:53:36 cromain: open base.bbclass Apr 06 22:53:38 ok should I select virtual/a_package_that_exit_some_where ? Apr 06 22:53:45 um ok, thanks Apr 06 22:53:48 :) Apr 06 22:53:59 cromain: 'virtual' does not exist, as the names says Apr 06 22:54:59 cromain: we see a line "deps += "virtual/${TARGET_PREFIX}gcc virtual/libc" Apr 06 22:55:17 cromain: so as you have your toolchain - ASSUME_PROVIDED virtual/libc Apr 06 22:55:24 cromain: and then the right compiler (gcc) Apr 06 22:55:37 cromain: also look at bitbake.conf to see how TARGET_PREFIX is set Apr 06 22:59:39 zecke: um... I don't have TARGET_PREFIX in my bitbake.conf Apr 06 22:59:57 zecke: I use the release 329 of bitbake. Apr 06 23:00:50 cromain: you have one, look into the one in the OE directory Apr 06 23:01:01 zecke: ok, I found Apr 06 23:01:17 zecke: I looked into the bitbake one Apr 06 23:09:21 zecke: ok I set TARGET_ARCH = "armv5te-softfloat" and TARGET_OS = "linux-gnu" in my org.openembedded.oz354fam083/conf/bitbake.conf Apr 06 23:10:30 cromain: well, you risk a lot of patches not being applied Apr 06 23:11:01 zecke: why ? and what sort of patches ? Apr 06 23:11:26 cromain: OE ones for cross compilation Apr 06 23:11:46 cromain: we apply certain fixes when building for arm (not armv5te-softfloat) Apr 06 23:12:33 zecke: I see. Do you already tested with armv5te ? Apr 06 23:12:55 cromain: sure, we just 'tune' for it though Apr 06 23:13:01 zecke: I can make armv5te a try, and simply using arm if it fails Apr 06 23:13:04 cromain: check conf/machine/* for armv5te Apr 06 23:13:16 cromain: it will fail! and it should fail Apr 06 23:13:34 cromain: and we have no armv5te-softfloat in sites Apr 06 23:15:13 cromain: ASSUME_PROVIDED += "virtual/libc virtual/arm-linux-gcc" Apr 06 23:15:29 cromain: maybe set TARGET_FPU = "soft" (shouldn't be necessary) Apr 06 23:15:47 cromain: make sure arm-linux-gcc is in the path Apr 06 23:15:50 zecke: ok, so should I rebuild and use an arm-softfloat-linux-gnu toolchain ? Or can I still use my armv5te-softfloat-linux-gnu toolchain ? Apr 06 23:15:52 cromain: tune for armv5te Apr 06 23:16:12 cromain: you can still use your one, just rename it ;) Apr 06 23:16:29 zecke: :) Apr 06 23:18:43 zecke: should I specify 'softfloat' in the ASSUME_PROVIDED: virtual/arm-softfloat-linux-gnu-gcc ? or simply virtual/arm-linux-gcc ? Apr 06 23:20:02 cromain: well, no magic involved. Every package has virtual/${TARGET_PREFIX}gcc as DEPENDency Apr 06 23:20:14 cromain: ASSUME_PROVIDED not more, not less -> arm-linux-gcc Apr 06 23:20:55 zecke: ok, thanks Apr 06 23:24:20 zecke: i need some sleep.. thank you very much for your help ! I think I'll have more time this week-end to try OE, eventually with armv5te if arm work without problem :) Apr 06 23:24:40 cromain: wait a second Apr 06 23:24:50 zecke: and of course update the Wiki :) Apr 06 23:24:52 cromain: TARGET_ARCH is arm, x86, mips, sh3... Apr 06 23:25:08 cromain: for what you tune, or what floating point you use is a different story Apr 06 23:25:45 zecke: ah Ok Apr 06 23:25:49 cromain: we regulary build for Xscale (all sorts) Apr 06 23:26:02 zecke: where do I should put softfloat then ? Apr 06 23:26:12 TARGET_FPU = "soft" Apr 06 23:26:21 but as you use a different toolchain it doesn't matter Apr 06 23:26:40 zecke :ok, already set apparently (i use "collie") Apr 06 23:27:32 good nite Apr 06 23:27:36 zecke: thanks again for your help :), bye Apr 06 23:57:47 good evening Apr 06 23:58:41 morning Apr 07 01:52:37 Does anyone know what the ASCII key codes are for the home, menu, mail, etc keys on a 5500? Apr 07 01:52:42 or how i can find out Apr 07 02:39:37 whats the easiest way to create more kernel-module meta packages.. i.e. kernel-modules-sound, kernel-modules-bluetooth etc Apr 07 03:10:03 kergoth: hey Apr 07 04:55:12 echo "EXTRACFLAGS = -DUNIONFS_NDEBUG ${CFLAGS}" > ${S}/fistdev.mk Apr 07 04:55:17 why does that not work in do_configure() Apr 07 05:35:06 Evening Apr 07 05:35:12 Ello Apr 07 05:35:30 How are things going in OE land? Apr 07 05:35:50 Pretty slowly Apr 07 06:48:26 good morning all Apr 07 06:54:20 <_law_> hi koen|away Apr 07 06:54:27 <_law_> hi koen|train Apr 07 07:05:48 good morning all Apr 07 07:06:44 i read something interesting, maybe you are already aware of this : uclinux has been proved as working on a sony psp Apr 07 07:06:46 http://www.psphacks.net/content/view/763/2/ Apr 07 07:10:34 even though this uses a psp targetted SDK, do you think there might be the slightest luck OE tools could help the devs of this project ? Apr 07 07:32:42 morning Apr 07 07:48:28 bleh Apr 07 07:48:44 it seems my keys are inside the house and I am not Apr 07 07:48:56 * koen calls the landlord Apr 07 07:50:48 zecke: hi Apr 07 08:00:37 moin Apr 07 08:01:02 zecke: http://bugs.openembedded.org/show_bug.cgi?id=763 Apr 07 08:03:32 wiener (whiner..) Apr 07 08:03:56 capslock handling is unchanged Apr 07 08:04:06 so it was broken before... Apr 07 08:05:31 hrw|work: if numlock is not working, this indicates a keymap 'bug' Apr 07 08:05:46 hrw|work: it used to lock up because the keymap was broken Apr 07 08:05:59 hrw|work: now it doesn't lock up any more, this didn't fix the keymap Apr 07 08:06:28 hrw|work: the ROAD code has a 'fix' for CapsLock but I need to verfiy it first Apr 07 08:06:43 ok Apr 07 08:07:02 zecke: so now we need someone with tosa/2.4 knowledge to fix keymap? Apr 07 08:08:19 hrw|work: I could take a look, but I always 'dispatched' this work Apr 07 08:10:08 you are not the only one with it Apr 07 08:11:53 good morning Apr 07 08:12:33 morning all Apr 07 08:15:16 http://ariya.blogspot.com/2006/04/z-and-kioslave.html ;) Apr 07 08:19:22 :) Apr 07 08:19:58 angstrom boots on spitz :) Apr 07 08:21:52 300GB Seagate SATA for 100 USD... too bad that only in US Apr 07 08:25:36 hrw|work: wb Apr 07 08:25:57 hrw|work: CapsLock is broken, and I will merge the ROAD patch to fix that Apr 07 08:27:07 k Apr 07 08:28:26 RP: can you look why every kernel module is in rootfs? Apr 07 08:28:39 its 7M of modules... Apr 07 08:35:42 hrw|work: Probably because kernel-modules is providing virtual/kernel-hostap Apr 07 08:35:45 morning Apr 07 08:35:46 :-/ Apr 07 08:35:48 hi XorA Apr 07 08:36:34 RP: which is bug Apr 07 08:36:42 hrw|work: correct Apr 07 08:38:02 RP: I also think that hostap-conf and orinoco-conf are not needed with pcmciautils Apr 07 08:38:41 hrw|work: I'm not sure about that... Apr 07 08:41:24 RP: mapping for card<>driver is in driver now Apr 07 08:41:51 hrw|work: It doesn't fall back to userspace mappings at any point? Apr 07 08:42:08 RP: not sure Apr 07 08:43:34 pcmciautils 013 released Apr 07 08:44:11 Did anything we might need change? Apr 07 08:44:56 brb Apr 07 08:49:00 * XorA swears as in the rush to get taxi this morning he forgot to turn on is machine Apr 07 08:50:28 RP: does not look as needed to upgrade Apr 07 08:59:58 hi Cwiiis Apr 07 09:01:34 Morning hrw|work Apr 07 09:03:18 Cwiiis: which device (PDA) you are using? Apr 07 09:03:21 hi Bernardo|away Apr 07 09:03:45 good morning Apr 07 09:03:50 hi hrw|work Apr 07 09:04:02 * Bernardo just flashed with a fresh oz354x image Apr 07 09:04:02 hrw|work: c700 (I have a 5500, and recently a 770 too :)) Apr 07 09:05:54 there's a small bug in /etc/network/if-pre-up.d/hostap-fw-load, it is still looking for cardctl instead of pccardctl Apr 07 09:06:20 Bernardo: good to know - which package is it? Apr 07 09:06:53 ipkg search claims it is from prism3-support Apr 07 09:07:42 morning Apr 07 09:07:46 hi CoreDump|home Apr 07 09:08:19 hi CoreDump|home Apr 07 09:08:36 morning all Apr 07 09:08:42 hi HuntrCkr Apr 07 09:10:45 if I want to compile something (an inputmethod) for Opie/OZ, what's the quickest way to setup a development environment? Apr 07 09:10:55 Bernardo: http://ewi546.ewi.utwente.nl/tmp/hrw/3.5.4.1-test/prism3-support_1.0.0-r1_all.ipk - can you test it? Apr 07 09:11:09 are there any devimg solutions, or do I have to setup the entire bitbake/OE env? Apr 07 09:11:31 ashikase: OZ provide cross-sdk (which lack opie headers/libs anyway) Apr 07 09:12:06 hrw|work: ok, in a few minutes, I'm just finishing setting up myu akita Apr 07 09:12:43 Bernardo: ok. this one has cardctl/pccardctl checks Apr 07 09:13:05 ah, i see there's a native-sdk as well? Apr 07 09:13:27 ashikase: it is but again - noone make opie-dev packages ;( Apr 07 09:13:41 for opie, I just need the headers Apr 07 09:13:47 if necessary, I'll just copy them Apr 07 09:13:57 I'd prefer an onboard compile environment Apr 07 09:14:12 oe seems a little... much for what I want to do Apr 07 09:15:07 understood Apr 07 09:16:33 hi Aman Apr 07 09:17:22 hrw|work: Fixed, thanks! Apr 07 09:20:37 hrw: I found the cross-sdk, but where can I download the native? Apr 07 09:21:40 ashikase: it is in feed - split into packages Apr 07 09:21:44 Bernardo: thx for testing Apr 07 09:22:15 ashikase: I have 512M sd card now so soon will try to get native-sdk better working Apr 07 09:22:15 my feeds must not be set properly Apr 07 09:22:21 ah Apr 07 09:22:30 i guess I'll try the cross instead Apr 07 09:25:06 hrw|work: Did you see the conversation in #openzaurus yesterday about touchscreen buttons? Apr 07 09:25:24 yes Apr 07 09:25:34 sounds reallly nice Apr 07 09:25:45 good evening Apr 07 09:26:08 hrw|work: is there any reason we use matchbox-wm_0.9.5cvs$something instead of -wm_1.0? Apr 07 09:26:17 hi TheMasterMind1 Apr 07 09:26:17 btw, the updated qte libs are working nicely, thanks... nice to be able to use Fn+Shift again Apr 07 09:26:42 CoreDump|home: dont know Apr 07 09:26:52 I'll install and see Apr 07 09:27:00 CoreDump|home: ask koen or #gpe Apr 07 09:28:12 heh Apr 07 09:28:32 my front door looks a bit 'odd' Apr 07 09:28:47 koen: hacked to get it? Apr 07 09:28:56 s/hacked/whacked Apr 07 09:29:07 the locksmith used brute force on the lock Apr 07 09:29:51 next time search for a burglar instead of locksmith Apr 07 09:30:32 the next time I won't forget my keys :) Apr 07 09:31:34 CIA-4: ping Apr 07 09:32:00 03hrw 07org.oe.oz354x * rb4e30279... 10/packages/meta/e-image-core.bb: e-image-core: add X11 feed Apr 07 09:32:04 03hrw 07org.oe.oz354x * r88381ea2... 10/packages/prism3-support/ (files/hostap-fw-load prism3-support_1.0.0.bb): prism3-support: fixed to work with pcmciautils, set PACKAGE_ARCH to 'all' Apr 07 09:32:10 03hrw 07org.oe.oz354x * rda91f395... 10/packages/prism3-support/prism3-support_1.0.0.bb: bumped PR to same as in .dev Apr 07 09:32:13 03hrw 07org.oe.dev * r81b55e46... 10/packages/prism3-support/ (files/hostap-fw-load prism3-support_1.0.0.bb): prism3-support: fixed to work with pcmciautils Apr 07 09:35:56 03rpurdie 07org.oe.oz354x * r06e16908... 10/packages/linux/ (4 files in 2 dirs): linux-oz-2.6: Fix kernel-hostap RPROVIDES problem, enable KEXEC in defconfigs Apr 07 09:36:01 03rpurdie 07org.oe.oz354x * r4bd1dd8b... 10/conf/distro/openzaurus-3.5.4.1.conf: zaurusd: Update to 20060407 svn to get softbutton updates. Apr 07 09:36:04 03rpurdie 07org.oe.oz354x * r973cded6... 10/.mt-attrs: Fix .mt-attrs Apr 07 09:36:57 RP: can you add a post on oz.org about tskeys working? Apr 07 09:39:42 hrw|work: softbuttons now work on the clamshells? Apr 07 09:40:24 Bernardo: almost, we need to adjust some keymaps first, but yes Apr 07 09:40:51 even the space _between_ two softkeys is mapped =) Apr 07 09:41:22 cool Apr 07 09:41:31 way cool indeed Apr 07 09:41:46 * Bernardo will once again try to learn graffiti so he can use his akita in landscape mode... :) Apr 07 09:41:52 I mean portrait Apr 07 09:42:16 * CoreDump|home hates to use the stylus to enter text Apr 07 09:42:31 keyboard or nothing.. Apr 07 09:42:36 at least I need to edit the virtual keyboard for pt... Apr 07 09:43:32 well, anyone has a updated zaurusd for the akita, or I'll have to wait until tonight? ;) Apr 07 09:44:57 it is of no use currently as a) the softkeys have not been mapped to the hardware keycodes yet, or b) the softkeys are not configured to do anything right now Apr 07 09:45:19 it i on my todo for today however Apr 07 09:45:36 ~hail CoreDump|home Apr 07 09:45:40 * ibot bows down to CoreDump|home and chants, "I'M NOT WORTHY!!" Apr 07 09:45:43 nahh Apr 07 09:45:48 ~praise RP Apr 07 09:45:50 All hail RP! Apr 07 09:48:27 hrw|work: I'll post on oz.org once the keymap is in place and working Apr 07 09:48:35 oka Apr 07 09:54:20 hrw|work: I'll re-enable the 200DPI bug work-around. The bug has not been fixed yet... Apr 07 09:54:53 ok Apr 07 09:55:06 at least the panel menu is fixed Apr 07 09:55:37 CoreDump|home: what part of the bug still exists? Apr 07 09:55:39 03rpurdie 07org.oe.dev * r4e688eca... 10/packages/linux/ (4 files in 2 dirs): linux-oz-2.6: Fix kernel-hostap RPROVIDES problem, enable KEXEC in defconfigs, make 2.6.16 the default kernel. Apr 07 09:56:22 XorA: everything. The unreadable window-titles you experience in most themes? That _is_ the bug. W/o bug they are perfectly readable Apr 07 09:56:47 CoreDump|home: they can be patched in themes really easilly Apr 07 09:57:16 CoreDump|home: remove the font line from the theme and it defaults to Gtk/Font like everything else Apr 07 09:57:36 i see Apr 07 09:58:37 which is about the same as running with 100DPI heh Apr 07 09:59:05 IMO the themes should hard code a pixel size Apr 07 09:59:09 shouldnt **** ENDING LOGGING AT Fri Apr 07 09:59:56 2006