**** BEGIN LOGGING AT Mon May 16 23:59:56 2005 May 17 00:00:03 * emte goes to find all 21 files May 17 00:01:19 hmm i get up in 6 hours ... i better fix it tomorrow before my gf exicutes me May 17 00:01:41 ok May 17 00:01:48 I can push that simple fix May 17 00:01:53 hehe May 17 00:02:02 its not a prob May 17 00:02:26 just trying to rememebr if undo will pull the while changeset or not May 17 00:02:31 whole* May 17 00:02:52 should May 17 00:03:34 you know the flags off hand May 17 00:03:42 i am trying to find my card May 17 00:04:02 aha found it May 17 00:04:36 bk undo -aCHANGESET May 17 00:09:01 hmm May 17 00:09:33 they should all have the same changeset number no? May 17 00:10:08 guess not ... May 17 00:11:28 when i "bitbake virtual/kernel", i got error May 17 00:11:36 | NOTE: make -e MAKEFLAGS= zImage CC=ccache arm-linux-gcc LD=arm-linux-ld May 17 00:11:37 | scripts/split-include include/linux/autoconf.h include/config May 17 00:11:39 | ccache arm-linux-gcc -D__KERNEL__ -I/stuff/openembedded/tmp/work/friendly2410-2.4-r0/friendly2410-2.4/include -Wall -Wstrict-prototypes -Wno-trigraphs -Os -mapcs -fno-strict-aliasing -fno-common -fno-common -pipe -mapcs-32 -march=armv4 -mtune=arm9tdmi -mshort-load-bytes -msoft-float May 17 00:11:40 -DKBUILD_BASENAME=main -c -o init/main.o init/main.c May 17 00:11:42 | cc1: error: invalid option `apcs-32' May 17 00:11:45 | cc1: error: invalid option `short-load-bytes' May 17 00:13:23 that kernel isnt buildable with that toolchain, clearly. May 17 00:13:32 aha May 17 00:13:37 found it May 17 00:16:15 kergoth: i set May 17 00:16:17 PATH=/usr/local/arm/2.95.3/bin:$PATH May 17 00:16:18 ASSUME_PROVIDED = "virtual/arm-linux-gcc" May 17 00:16:30 that isnt correct. May 17 00:16:33 read the instructions again May 17 00:16:41 and the local.conf.sample, for that matter May 17 00:17:17 hrw:hehe .what you told me is wrong May 17 00:17:24 okay hrw its a no can do May 17 00:17:37 bk wont let me May 17 00:18:24 robert: you're trying to build that kernel with a toolchain that cant build it. its that simple. 2.4 kernels and old 2.6 dont compile with current gcc versions without patching. May 17 00:18:32 bk undo -r1.3305 if you would like to see the nice warning May 17 00:18:34 * kergoth wanders off May 17 00:19:30 i'll repush tomorrow since its 12:20 and i get up at 5 May 17 00:19:56 hopefully overwritting will do it May 17 00:26:01 kergoth:when i dont use oe, i can build the kernel with the toolchain.so the point is how to make oe use the toolchain. May 17 00:34:12 morning May 17 00:45:54 HI guys May 17 00:46:09 some package is causing a build of gcc-cross-kernel May 17 00:46:17 how do I find out which one? May 17 00:46:36 grepping openembedded/packages finds no mention of gcc-cross-kernel May 17 00:52:08 morning all May 17 00:52:18 hi RP May 17 00:52:31 Spyro: Have you run a complete build? May 17 00:52:32 do you know how to answer the above ? May 17 00:52:49 RP: depends what you mean May 17 00:52:53 If you have a complete build, see if it built a kernel May 17 00:53:00 RP: not since trying to add udev :) May 17 00:53:28 Spyro: Did you add anything about virtual/kernel? May 17 00:53:29 RP: are you saying its impossible to know which package is causing the dependency without building everything? May 17 00:53:46 RP: no, I specifically removed those references May 17 00:53:53 No, I'm saying I don't personally know a better way to figure it out May 17 00:54:01 RP: ah :) May 17 00:54:53 The only way gcc-cross-kernel will creep in is if you need it for a kernel May 17 00:55:02 So somewhere you've inherited a kernel May 17 00:55:14 re May 17 00:55:17 What else did you change when adding udev? May 17 00:55:24 hi hrw|work May 17 00:55:53 I presuming it didn't compile the kernel gcc before udev? May 17 00:56:03 RP: well, its got to be one of about a handful of packages, but I'd have thought they depended on the ernel headers, not the kernel... May 17 00:57:08 Spyro: which kernel? May 17 00:57:23 More to the point which packages? May 17 00:57:26 no idea May 17 00:57:56 If one of them wants kernel headers, it'll probably compile a kernel whilst its at it.. May 17 00:58:18 packages: pcmcia-cs wireless-tools irda-utils udev-utils console-tools May 17 00:59:13 One of them will have a dependency on virtual/kernel May 17 00:59:59 pcmcia-cs has for a start May 17 01:00:44 and if you want to build for Zaurus then OE cannot find gcc 2.95.3 so try to build other (random) kernel which needs gcc 3.3.x so it start to build gcc-cross-kernel May 17 01:03:13 Spyro: Most of the above packages will have a dependency on virtual kernel for the headers. virtual/kernel will default to one which needs gcc 3.3.x hence the gcc being built May 17 01:03:58 hrw|work: Spyro is a Toshiba eseries guy so thankfully won't be messing with the external 2.95.3 toolchain :) May 17 01:05:53 Ian Morton... how could I forget.. May 17 01:06:02 iirc? May 17 01:06:10 anyway I go back to work May 17 01:06:42 Close, Ian Molton ;-) May 17 01:07:03 back :) May 17 01:16:05 morning all May 17 01:35:58 morning May 17 01:37:32 hi May 17 01:44:04 oi May 17 01:44:19 hi ljp May 17 01:46:08 hi Bernardo ade ljp May 17 01:46:14 BigAl: May 17 01:46:23 and rest May 17 01:47:29 :) May 17 01:48:33 03RP 07 * r1.3311.1.1 10openembedded/packages/linux/linux-openzaurus_2.6.12-rc4-mm2.bb: linux-oz-2.6: Upgrade to 2.6.12-rc4-mm2 May 17 01:48:50 hrw: May 17 01:50:29 robert: May 17 01:50:34 hrw:kergoth said below is not correct. May 17 01:50:35 PATH=/usr/local/arm/2.95.3/bin:$PATH May 17 01:50:37 ASSUME_PROVIDED = "virtual/arm-linux-gcc" May 17 01:50:49 03RP 07 * r1.3322 10openembedded/packages/linux/linux-openzaurus-2.6.12-rc4-mm2/ (9 files): linux-oz-2.6: Update to 2.6.12-rc4-mm2 May 17 01:51:05 robert: ok - he has right May 17 01:51:53 i dont know how to set it now. May 17 01:52:38 ASSUME_PROVIDED += "virtual/arm-linux-gcc-2.95" May 17 01:52:54 XorA: not -2.95 - he use 2.95 as default May 17 01:52:56 robert: or set DISTRO=openzaurus-3.5.4 May 17 01:53:13 XorA: he dont use zaurus May 17 01:53:30 robert: in conf/local.conf add ASSUME_PROVIDED = "virtual/arm-linux-gcc" May 17 01:53:42 robert: or finally in your_distro.conf May 17 01:54:01 hrw: i have done it May 17 01:54:18 robert: you use glibc or uclibc? May 17 01:54:34 hrw:glibc May 17 01:54:55 what error you get? May 17 01:55:27 ~pastebin May 17 01:55:29 well, pastebin is a place to paste your stuff without flooding the channel - try http://pastebin.ca May 17 01:56:01 | NOTE: make -e MAKEFLAGS= zImage CC=ccache arm-linux-gcc LD=arm-linux-ld May 17 01:56:02 | scripts/split-include include/linux/autoconf.h include/config May 17 01:56:04 | ccache arm-linux-gcc -D__KERNEL__ -I/stuff/openembedded/tmp/work/friendly2410-2.4-r0/friendly2410-2.4/include -Wall -Wstrict-prototypes -Wno-trigraphs -Os -mapcs -fno-strict-aliasing -fno-common -fno-common -pipe -mapcs-32 -march=armv4 -mtune=arm9tdmi -mshort-load-bytes -msoft-float May 17 01:56:05 -DKBUILD_BASENAME=main -c -o init/main.o init/main.c May 17 01:56:07 | cc1: error: invalid option `apcs-32' May 17 01:56:08 | cc1: error: invalid option `short-load-bytes' May 17 01:56:32 in my distro.conf , i set May 17 01:56:33 PREFERRED_VERSION_glibc-initial = "2.2.5" May 17 01:56:35 PREFERRED_VERSION_glibc = "2.2.5" May 17 01:56:52 robert: you need also set FULL_OPTIMIZATION to something which old gcc will handle May 17 01:57:03 robert: look into conf/bitbake.conf for default value May 17 01:58:19 hrw: let me see May 17 01:59:08 FULL_OPTIMIZATION = "-fexpensive-optimizations -fomit-frame-pointer -frename-registers -O2" May 17 02:00:37 robert: hmm.. where does those options (apcs-32 and short-load-bytes) comes.. May 17 02:01:21 makefiles from kernel May 17 02:03:20 hrw:yes May 17 02:04:05 you probably have to look into other OE kernels to get it working (I never did that) May 17 02:04:28 hrw:when i dont use oe, i can build the kernel with the toolchain.so the point is how to make oe use the toolchain. May 17 02:06:14 robert_: you have to choose more original nick and register it May 17 02:06:51 hrw:maybe i was kick out May 17 02:08:16 [Nick collision from services.] == someone who registered nick wanted it back May 17 02:08:19 anyway.. May 17 02:08:31 hehe May 17 02:08:58 robert_: in conf/local.conf (or your-distro.conf) you have ASSUME_PROVIDED with virtual/arm-linux-gcc? May 17 02:09:18 hrw:yes May 17 02:10:09 robert_: so OE does not tried to build gcc-cross? May 17 02:11:23 hrw:but oe build gcc-cross-initial-3.4.3+csl-arm-20050416-r4 also May 17 02:11:52 you have to assume_provide binutils and glibc also probably May 17 02:16:08 hrw:i guess what oe use is not my local arm-linux-gcc ,but is build from gcc-cross-initial-3.4.3+csl-arm-20050416-r4 May 17 02:16:26 robert_: gcc-cross-initial is build to build glibc May 17 02:16:42 hrw:oh May 17 02:17:05 robert_: add ASSUME_PROVIDED += bintuils-cross glibc (look for exact format in conf/distro/sharprom-compatible.conf) May 17 02:17:24 hrw:ok May 17 02:24:55 hrw:le me try again. May 17 02:24:56 Handling BitBake Files cost much time May 17 02:25:20 robert_: you can use BBMASK to limit May 17 02:26:38 hrw:yes ,i know May 17 02:26:52 hi mickeyl May 17 02:26:55 morning May 17 02:26:56 http://www.oesf.org/forums/index.php?act=ST&f=72&t=12636&st=0#entry80157 May 17 02:27:02 i hope people will understand now May 17 02:27:10 should we put that on the OpenEmbedded site ß May 17 02:27:10 ? May 17 02:27:23 (my explanation) May 17 02:27:37 I will read it first May 17 02:28:49 so long May 17 02:29:53 03koen 07 * r1.3320.1.1 10openembedded/packages/gimp/gimp_2.2.7.bb: Add the GIMP 2.2.7 May 17 02:30:19 bring out the gimp May 17 02:30:47 OpenEmbedded is NOT an SDK. - that part is nice May 17 02:32:04 i wonder what we can do to hammer this in peoples minds May 17 02:33:01 thats good write May 17 02:33:11 good May 17 02:33:26 thanks May 17 02:35:00 is opie a duplication of qtopia, I thought it was just another nicer layer built on top? May 17 02:35:17 where are the stones? May 17 02:35:25 *g* May 17 02:35:57 re woglinde May 17 02:36:57 XorA: nicer ?? May 17 02:36:58 XorA: you want an honest answer ? May 17 02:36:58 hi roh May 17 02:36:58 mickeyl: yeah May 17 02:36:58 woglinde: it worked.. i now have a jtag interface and a new bootloader May 17 02:36:58 rh gratz May 17 02:36:58 mickeyl: I never looked that deeply into opie May 17 02:37:27 XorA: Opie started out as a clone of Qtopia 1.4. Now during the years we fixed _many_ _many_ bugs and added lots of infrastructure (libraries) and exciting applications, yet we don't get enough attention since we still are perceived as a half-assed clone of Qtopia, a.k.a "not the real thing" May 17 02:37:27 heh May 17 02:37:40 or at least a working one... but it does not do any output on lcd.. the display is initialized just before displaying a bw-penguin in violett May 17 02:38:10 I will never waste my time for an May 17 02:38:10 effort Sharp is flushing down the toilets with its erratic kernel May 17 02:38:13 migration policy. May 17 02:38:15 nice ;) May 17 02:38:27 yes, i agree with him on some points May 17 02:38:27 mickeyl: oh :-( that is unfortuneate as the bits of opie I looked at looked nice, Im just not up with the whole C++ OO thing May 17 02:38:27 heh **** BEGIN LOGGING AT Tue May 17 02:40:15 2005 May 17 02:40:20 ROM = read only memory, not a "software distribution" May 17 02:40:41 lobotomized zaurus users, *sigh* May 17 02:41:15 koen: what happened? May 17 02:41:46 mickey: Opie _does_ get attention. :) May 17 02:41:47 mickeyl: truckload of work May 17 02:41:58 mickeyl: due last month May 17 02:43:08 ljp: not necessarily positive one :) May 17 02:43:19 koen: bummer May 17 02:43:45 guyhem is yet another cloner... and reminds me zautrix May 17 02:44:06 a bit May 17 02:44:10 he is more knowledged though ;) May 17 02:44:17 when OE team will get 2.6 running then I will look into it May 17 02:44:17 no i take that back, i have no idea May 17 02:44:59 for me that sentence means - I wont help them - will wait for ready to use, tested version May 17 02:44:59 if it wasn't playing like Sharp I'd say lets make our kernel to run only OZ :) May 17 02:45:07 hrw|work: that's true May 17 02:49:03 I will edit my post in oesf May 17 02:49:27 which one? May 17 02:50:07 all of them? ;) May 17 02:50:37 That thread annoyed me too - I was ready to let loose with a broadside last night May 17 02:50:38 heh May 17 02:51:53 i like to put up a verbal fight once and then (in case you didn't notice yet), but i don't like false claims and people participating who just complain and don't do anything May 17 02:52:13 (in general, that wasn't necessarily to guylhem) May 17 02:52:56 My annoyance was with the mud slinging at OE & OE people on IRC for no real reason May 17 02:53:55 ya, i don't understand that one. looks like we always get beaten for insisting to do things the way we think it's right - which is not always what users want May 17 02:54:48 Hey there, anyone has an idea what might cause this: May 17 02:54:54 | + ipkg-cl -f /home/rob/oe/build/tmp/work/display-image-1.0-r0/temp/ipkg.conf -o /home/rob/oe/build/tmp/rootfs update May 17 02:54:54 | ipkg: Cannot create directory `/usr/lib/ipkg': Permission denied May 17 02:55:05 http://www.oesf.org/forums/index.php?showtopic=12636&st=0&p=80158&#entry80158 May 17 02:55:18 wrobbie: that's not an error. ignore that May 17 02:55:26 it isn't? :) May 17 02:57:02 oh. that threw me off track.. The error is something else then: | Cannot find package display-image May 17 02:57:11 But i have a ipk for it... May 17 02:57:28 wrobbie: tmp/deploy/ipk/Packages has it listed? May 17 02:59:19 oh. ah.. that should be 'display-files'. ho-hum May 17 02:59:28 will try again :) May 17 03:00:27 * XorA args at the nutters waking up gentoo for Z again May 17 03:01:29 XorA: your pda will be compiling for 3 months straight, but it will the 0,00001% faster afterwards May 17 03:01:33 * Bernardo gave up sourcemage for z a long time ago May 17 03:02:07 koen: and it wont be pda as they usually have / on nfs May 17 03:02:25 koen *g* May 17 03:02:44 if people spent half the effort they put into wierd schemes into the gorgeous build system that is OE, the OE meta data would shoot forward May 17 03:03:38 I investigated a packages.openembedded.org this weekend, but that's going to be quite some work to set it up May 17 03:04:43 XorA: do you have a link to those nutters? May 17 03:05:01 koen: http://www.oesf.org/forums/index.php?showtopic=12626 May 17 03:05:53 koen: meant to ask, hows gnumeric? May 17 03:06:06 XorA: still failing on that xpm thing May 17 03:06:14 I did add the gimp :) May 17 03:07:25 koen: I noticed, Ill have to do a build :-) May 17 03:08:10 hmmm May 17 03:08:14 * koen notices a tiny flaw May 17 03:08:43 you'll need a gtk+native or gtk+-dev on your host system May 17 03:09:21 and I ... ehm... May 17 03:09:33 'mv /usr/include/ usr/include.old' May 17 03:10:04 or remove mathinline.h May 17 03:10:19 we have more bugs like that May 17 03:10:25 unnoticed May 17 03:10:37 that's why you should build on a barebone system May 17 03:10:38 I noticed on my 'clean' buildbox May 17 03:10:41 righto May 17 03:11:52 * mickeyl checks DistroNames for new entries May 17 03:12:08 mickeyl: Viagra linux May 17 03:13:10 koen: make small thing larger? May 17 03:13:19 hrw|work: something like that May 17 03:13:56 ~lart koffce May 17 03:13:56 * ibot grabs a large, mis-shapened log, with squirrels, and beats koffce until only the nuts remain ... which the squirrels run off with May 17 03:14:26 time to install gnumeric as excel viewer May 17 03:14:57 crappit, my bk tree has become corrupt :-( May 17 03:15:01 15M packages ;( why its so gnomish.. May 17 03:15:47 emte: are you on oe@handhelds.org ? If not, I strongly recommend to subscribe this list. May 17 03:19:30 ah, this could be difficult, openembedded.bkbits.net is not allowing cloning May 17 03:20:31 I cloned it yesterday May 17 03:20:44 hrw|work: clone: cannot determine remote pathname May 17 03:21:03 hrw|work: bk clone bk://openembedded.bkbits.net/openembedded May 17 03:21:07 I know May 17 03:21:43 hrw|work: ah well, no OE work for me today :-) May 17 03:22:49 XorA: http://ewi546.ewi.utwente.nl/OE/ May 17 03:24:11 koen: how do I clone that? May 17 03:24:21 wget -r May 17 03:24:42 koen: ah ok, the nasty way :-) May 17 03:28:49 i made a few pics: http://phoenix.hyte.de/jtag_simpad/ May 17 03:33:09 Hmm. How can I debug why my 'bitbake display-image' completely seems to ignore the BOOTSTRAP_EXTRA_DEPENDS (and RDEPENDS) that are in display-image.bb? May 17 03:34:20 check the parsing order of .conf files, use bbread on certain files, etc. May 17 03:35:09 bbread? google time May 17 03:35:23 bbread is one of our utilities May 17 03:35:30 shows the value of attributes May 17 03:35:47 bitbake -D also helps for checking the parse order May 17 03:36:16 but with -D i can't find any reference to _EXTTRA_DEPENDS, it seems May 17 03:37:26 wrobbie: display-image.bb is used to generate image? May 17 03:37:41 wrobbie: BOOTSTRAP* are used by task-bootstrap.bb May 17 03:38:00 yeah, copied from bootstrap-image May 17 03:38:16 i'll do some more checking.. May 17 03:38:59 so.. if anyone kills his simpad in south-germany... send it to me and pack something nice by it and i will flash it back and send it back ... now i know what to do May 17 04:02:02 mickey|lunch: you alive? May 17 04:02:16 why does something depend on sharp-compat-libs? May 17 04:06:12 emte: http://bugs.openembedded.org/show_bug.cgi?id=5 May 17 04:07:20 treke|home: can to do s/greg@treke.net/oe@handhelds.org/ as the default emailaddress? May 17 04:07:42 treke|home: for bugzilla May 17 04:09:49 mithro: it used to be because sharp-compat-libs created the tmp/staging/shlibs stuff that said it supplied qtopia files, and shlib auto DEPENDENCY magic takes the first entry it finds, but HRW fixed that ages ago May 17 04:15:56 XorA: okay cool May 17 04:16:37 anyway why is a script which sharp-compat-libs try to run quickexec and it doesn't exists? or is that a fixed bug too? May 17 04:19:00 mithro: Im a GPE man these days so no idea on that one May 17 04:19:10 * mithro ponders May 17 04:19:41 ~lart openlogging May 17 04:19:41 * ibot judo chops openlogging May 17 04:19:43 ~lart bk May 17 04:19:43 * ibot drops a baby grand on bk May 17 04:19:56 ~lart compat-libs May 17 04:19:57 * ibot makes a balloon animal out of compat-libs May 17 04:22:50 perhaps I'm just completely missing something.. if 'bash' is in my BOOTSTRAP_EXTRA_DEPENDS, will it install it in the rootfs or not? Or does it have to be in B_E_RDEPENDS as well? May 17 04:23:40 (it doesn't install, which probably confirms my suspicion that my B_E_RDEPENDS is completely ignored) May 17 04:24:18 If this well documented somewhere, feel free to flame me :)) May 17 04:24:38 DEPENDS are for build, RDEPENDS for packages May 17 04:24:44 bash need to be in both May 17 04:25:00 wrobbie: DEPENDS=glibc RDEPENDS=libc6 May 17 04:25:29 ok.. May 17 04:27:21 if I do bbread display-image.bb, RDEPENDS is good. But is has no effect. You'd think it is overwritten somewhere, but I just can't find it, all my own files are OK it seems May 17 04:28:29 bitbake -D shows it doesn't touch any of my RDEPEND files. (But can't be used to figure out what happens to my RDEPENDS, i guess) May 17 04:29:40 Is there a super-verbose mode that can be used to track what happens to variables? :) May 17 04:29:52 -D -D -D May 17 04:30:21 ah, will try May 17 04:30:26 koen: http://bugs.openembedded.org/show_bug.cgi?id=5 - fix added May 17 04:31:36 anyone know much about USB? i need to figure out which of the pins are which May 17 04:32:00 (on one of those rectangle connectors) May 17 04:32:14 i'm starving.. time to fetch some substantial indian food. May 17 04:32:59 ~google hwb May 17 04:33:05 Google says hwb is at http://www.hardwarebook.net/ or http://www.hardwarebook.net/connector/ or http://www.hwb-accountants.com/ or http://www.hwb.co.za/ or http://invalid.ed.ntnu.no/~kamben/hwb/ca_ESDI.html May 17 04:33:12 mithro: check in hwb May 17 04:33:38 red is +5V, black is GND but I dont remember which one is D+, D- May 17 04:33:52 -D -D -D, still no mention of RDEPENDS/DEPENDS. meh. May 17 04:34:22 mithro: usually vcc-,data-,data+,vcc+ May 17 04:35:41 some manufactures tend to swap data+ and data- to "encourage" buying their own "special" cables May 17 04:35:46 like Holux May 17 04:37:13 this hwb is pretty cool May 17 04:37:32 mithro: yes May 17 04:37:56 someone has a problem with flex-native 2.5.31? I want to close bug #4 with WFM May 17 04:45:34 I haven't noticed a problem with it May 17 04:45:55 * koen bitbake flex-native May 17 04:47:59 hrw|work: builds fine over here May 17 04:49:14 ok will close it May 17 04:53:03 03koen 07 * r1.3324 10openembedded/packages/evas/ (evas-x11_0.9.9.004.bb evas-x11-fb_0.9.9.004.bb): evas-x11_0.9.9.004.bb, evas-x11-fb_0.9.9.004.bb: fix parsing + maintainer field May 17 04:55:39 i'd like to see all those bugged E files go away May 17 04:55:53 * koen too lazy for that May 17 04:57:25 I'd like to see faster kde May 17 04:58:18 03mickeyl 07 * r1.3323.1.1 10openembedded/packages/python/ (4 files in 2 dirs): fix python-pygtk2 May 17 05:07:25 koen: http://familiar.handhelds.org/releases/download.php looks nice May 17 05:08:16 hrw|work: it's my work in progress script May 17 05:08:41 it looks for files matching a pattern, so it's pretty maintenance free May 17 05:10:37 VW Polo - colour Blue "Tossa" metallic May 17 05:13:54 the bootstrap screenshot rocks May 17 05:14:20 koen: please don't say Qtopia-based May 17 05:14:22 this is wrong May 17 05:14:28 please say 'a Qt/Embedded based...' May 17 05:15:05 this is more along the lines of GPE being a GTK based environment May 17 05:16:24 mickeyl: done May 17 05:17:17 Qtopia-based is more like GNOME-based for gpe May 17 05:17:40 I just copied the old page May 17 05:18:06 http://familiar.handhelds.org/releases/v0.8.2/install/download.html May 17 05:18:16 it's been saying qtopia for ages May 17 05:18:23 (since 0.7.0) May 17 05:18:51 0.8.2 stable, 0.8.2 testing.. May 17 05:19:19 that's a stinking old buggy script May 17 05:20:01 de machine selection is nice, but users still get confused when they click h5500 and get a h3900 image May 17 05:21:14 s/de/the/ May 17 05:21:24 but users will get confused anyway May 17 05:22:49 moin zecke May 17 05:23:08 zecke: did you have a sqlite backend for oe? May 17 05:23:21 koen: thanks May 17 05:23:44 zecke: or some code to put .bbs into a db? May 17 05:26:01 hi all May 17 05:26:10 hey May 17 05:27:03 mickeyl: what do you think of a packages.openembedded.org? May 17 05:27:18 like packages.{gentoo,debian}.org May 17 05:27:25 koen: good idea May 17 05:27:27 what does that do? May 17 05:27:32 <--- mandriva user May 17 05:27:51 mickeyl: http://packages.debian.org/libc6 May 17 05:28:02 oh a web querying interface for the metadata May 17 05:28:04 yes that'd be really nice May 17 05:28:42 I dont like browsing volkswagen.de - that weird language... May 17 05:28:49 hehe May 17 05:28:57 I like the look of http://packages.gentoo.org/, but I also need access to the .bb May 17 05:29:13 gentoo has no way of showing an .ebuild via the interface May 17 05:31:56 koen: let us get a QA Infrastructure first May 17 05:32:14 koen: It is important to actually be able to fetch the sourcecode we want to build ;) May 17 05:32:25 can someone add a little fix to packages/linux/linux-x86_2.6.11.6.bb that it would aslo appear in all future releases? Just 3 lines at the end ;). May 17 05:33:12 do_configure_prepend() { May 17 05:33:22 install -m 0644 ${WORKDIR}/defconfig ${S}/.config May 17 05:33:32 } May 17 05:34:03 dmit: please send a patch, I'll apply it when I'm home May 17 05:34:39 zecke: cool, but I don't know how it should be done ;) May 17 05:34:49 I mean for OE standarts May 17 05:35:08 dmit: just the usual unified diff, mail it to freyther@handhelds.org or oe@handhelds.oth May 17 05:35:29 zecke: OK May 17 05:36:00 03koen 07 * r1.3324.1.1 10openembedded/packages/linux/linux-x86_2.6.11.6.bb: linux-x86_2.6.11.6.bb: as requested on IRC by dmit May 17 05:36:15 Nice to see that on Z development should be possible now - that's something people have been after May 17 05:36:15 hehe May 17 05:36:38 lardman|work: masochistic people, yes May 17 05:36:41 ;) May 17 05:36:57 koen: yeah, each to his own hey? May 17 05:37:06 lardman|work: I'm adding anjuta to OE atm May 17 05:37:15 koen: what's that? May 17 05:37:22 www.anjuta.org May 17 05:37:22 koen: that is sick May 17 05:37:33 zecke: it is :) May 17 05:37:39 openofice.org2 is fscking big... May 17 05:37:40 That sounds like overkill May 17 05:37:40 Need to get 89.4MB/89.9MB of archives. May 17 05:37:40 After unpacking 247MB of additional disk space will be used. May 17 05:38:09 koen: great ;) anjuta under gpe on sl-5000d would be nice May 17 05:40:00 I wonder if I can get gnome 2.10 into OE before guadec May 17 05:41:26 if one of my qt counterparts would add kde 3.4.... May 17 05:43:39 zecke, mickeyl: have you agreed when to land the patches into svn? May 17 05:44:29 hi there I'm examining my cross-toolchain at powerpc-linux-uclibc/include/c++/powerpc-linux-uclibc/bits/stdc++.h.gch . There are two files here O0g and O2g, both are "precompiled c++ headers". Is that possible that they are of 20Megs? May 17 05:45:29 does anyone know if the bytes_per_inode.patch really needed for genext2fs-native-1.3+cvs20050517-r2 ? I get the error that it can not be applyed (Hunks 1,2,3,4,6 failed, hunk #5 succeeded with fuzz 2)) May 17 05:45:31 koen: today ;) May 17 05:45:48 mickeyl: if you've time feel free to commit :} May 17 05:46:00 mickeyl: or we can wait until 8h tonite May 17 05:46:16 zecke: does that mean a memeory light version of bitbake will be available tomorow? May 17 05:48:17 That guylhem is confused - just read his article May 17 05:48:25 XorA: not tomorrow May 17 05:48:40 XorA: we will do a intermediate release to not force people using bitbake trunk May 17 05:49:20 zecke: cool, but its on the horizon, well done May 17 05:51:20 lunch May 17 05:51:33 smach nego May 17 05:51:43 or something similar May 17 05:53:45 koen: I see you're attempting to raise the Zaurus users' collective IQ by educating us about ROMs :) May 17 05:53:55 koen: I see you're attempting to raise the Zaurus users' collective IQ by educating us about ROMs :) May 17 05:54:06 koen: Hmm, dodgy irc client May 17 05:54:08 lardman|work: insulting pdaxROM by proxy :) May 17 05:54:21 koen: Ah, the hidden meaning, good good ;) May 17 05:57:43 Zaurus users are funny May 17 05:57:48 france: hey May 17 05:58:24 zecke: users are funny in general May 17 06:01:35 urghs May 17 06:01:44 hmm May 17 06:01:45 nm May 17 06:03:00 ~lart SIOCGIFCONF May 17 06:03:00 * ibot turns SIOCGIFCONF into a lifesized tux doll May 17 06:03:03 * ade|desk grumbles about win xp and rebooting May 17 06:03:11 ~lart winxp May 17 06:03:11 * ibot drops a humongous exploding nuke on winxp May 17 06:03:21 the ioctl from hell May 17 06:11:05 does anybody want to take a shot at using graphviz in bb to visualize graphs? May 17 06:11:31 koen: hehe I had that idea roughly a year ago May 17 06:11:41 zecke: me too :) May 17 06:11:49 koen: just make dirgraph emit the dot file May 17 06:12:16 NOTE: package graphviz-2.2.1-r0: task do_configure: started May 17 06:12:54 hmmm May 17 06:13:03 DOT looks a bit complicated May 17 06:13:40 koen: freaking easy format May 17 06:19:44 My fusionsound.bb (part of DirectFB) creates tmp/deploy/ipk/libfusionsound-0.9-22_0.9.22-r0_i686.ipk.. What could be a cause of the double version number? May 17 06:20:09 wrobbie: what is a name of recipe? May 17 06:20:30 fusionsound_0.9.22.bb May 17 06:20:43 can you pastebin it? May 17 06:21:04 .. if i knew what pastebin is :) May 17 06:21:26 ~pastebin May 17 06:21:28 pastebin is probably a place to paste your stuff without flooding the channel - try http://pastebin.ca May 17 06:22:12 ah, goes to show I haven't been on irc from 1996 or so :) May 17 06:25:35 afternoon May 17 06:26:30 hrw: http://pastebin.ca/11981 May 17 06:27:21 hrw: it's based on directfb_0.9.22.bb May 17 06:29:16 My RDEPEND problems were due to a bad IPKG_INSTALL -- it works now! thanks all May 17 06:29:55 hmm.. dont know May 17 06:30:10 hey, btw there is a new one for freenode which anounces it in the channel if you choose May 17 06:30:13 http://paste.lisp.org/ May 17 06:31:31 There is a libfusionsound-0.9.so.22, perhaps that confuses something, somehow.. ? May 17 06:32:12 argh: | /usr/lib/libexpat.so: could not read symbols: File in wrong format May 17 06:32:22 why does arm-linux-gcc look there? May 17 06:32:36 it has /home/koen/OE/build/tmp/familiar/staging/arm-linux/lib/libexpat.so in the cmdline May 17 06:32:46 emte: and fails May 17 06:33:00 emte: 502 proxy error ;) May 17 06:34:58 really? May 17 06:35:01 i am there now May 17 06:35:10 cant say the pink is pretty May 17 06:36:07 now I got it May 17 06:37:19 blech 13 to go May 17 06:38:37 wrobbie: I don't know what the exact problem is, but I suppose you want to use FILE_${PN} += "..." rather than FILES_fusionsound_append May 17 06:38:56 wrobbie: err.. FILES_${PN} May 17 06:40:55 reenoo: ok, thanks May 17 06:41:55 much shorter too May 17 06:42:33 time for bus May 17 07:03:59 hi May 17 07:06:20 ~lart trolltech for makeqpf May 17 07:06:20 * ibot puts on a hockey mask and jumps out at trolltech for makeqpf May 17 07:07:18 opie still doesn't use ttf natively? May 17 07:07:26 What's the preferred provider thing I need to set to change what kernel gets compiled? May 17 07:07:36 koen: opie can May 17 07:07:40 BigAl: virtual/kernel May 17 07:07:51 koen: qt/e only 'can' May 17 07:08:03 hrw|work: thanks May 17 07:09:25 koen: and qt/e use own old version of freetype2 May 17 07:10:05 Will the settings in local.conf override the ones set in the machine definition? May 17 07:11:03 no May 17 07:15:14 I think I've found the kernel size limit on the spitz, too. Needed a nand flash after I tried one about 1.3MB May 17 07:19:25 ljp: ping May 17 07:19:57 BigAl: Its easy to do that. If you generate the kernel within oe, it will warn you about that May 17 07:20:37 BigAl: We should make KERNEL_VERSION="2.6" work for spitz/akita May 17 07:21:03 RP: Yeah, it'd be nice to be able to build a bootstrap image with 2.6 already there for testing May 17 07:21:32 RP: I've made a new spitz-base patch too, it should apply cleanly over the tosa ones. May 17 07:21:58 BigAl: In the usual place? May 17 07:22:23 RP: Yup, spitz-base-r2.patch May 17 07:22:32 mickeyl: Should we split spitz into spitz-2.4 and spritz-2.6 as with corgi/poodle? May 17 07:22:49 RP: rather yes May 17 07:22:52 mickeyl: Or is there a neater way to do it? May 17 07:23:02 RP: yes. please follow the way done with corgi, poodle, and tosa May 17 07:23:14 at least for now it is gives predictable results May 17 07:23:18 s/is/ May 17 07:26:02 ~shoot trolltech guys for using so old freetype May 17 07:26:04 * ibot shoots trolltech guys for using so old freetype in the head with a quantum singularity weapon! May 17 07:26:22 qt/e is shit May 17 07:26:40 too bad qt/x11 is a lot slower right now May 17 07:27:07 * koen has faith in RP May 17 07:27:36 * RP notes he should get that patch applied :) May 17 07:27:38 PREFERRED_PROVIDER_virtual/kernel = "c7x0-kernels-2.4-embedix" May 17 07:27:38 PREFERRED_PROVIDER_virtual/kernel_spitz = "openzaurus-pxa27x" May 17 07:27:43 Will that work? May 17 07:28:06 mickeyl: We have a slightly different problem for spitz - will the above work? May 17 07:28:14 * koen ponders May 17 07:28:22 akita or no akita May 17 07:28:34 I'll deal with akita next :) May 17 07:29:17 mickeyl: do you have the power to change oesf sections? May 17 07:29:23 (The above is to go into zaurus-clamshell-2.4.conf) May 17 07:30:23 koen: no. i just can recommend things to offroadgeek May 17 07:30:26 koen: so can you :) May 17 07:30:54 ok May 17 07:31:15 RP: that should work. try it that, if it doesn't - i'll think about something else May 17 07:31:38 mickeyl: Ok. It lets us keep the code very neat May 17 07:32:57 ~lart debian for makeqpf/qt3-x11 May 17 07:32:58 * ibot smacks debian up side the head with a clue-by-4 for makeqpf/qt3-x11 May 17 07:34:58 re May 17 07:35:14 RP: shouldn't the spitz kernel work on the akita? May 17 07:35:29 Bernardo: No May 17 07:36:36 how do people normally get a 48Mhz clock? May 17 07:36:40 isn't the base hardware the same? May 17 07:37:00 03koen 07 * r1.3327 10openembedded/packages/meta/meta-sectest-gpe.bb: meta-sectest-gpe.bb: bump PR May 17 07:37:00 Bernardo: Some bits are, but there's some big differences. May 17 07:37:16 the second cf slot, and the extra gpio? May 17 07:37:30 Pretty much May 17 07:38:17 well, I'll be offline in 1 hour for god knows how long, so I won't be able to help much testing on the akita... :( May 17 07:38:59 Bernardo: There is no akita 2.6 yet - I'm just putting in the groundwork at the same time as spitz May 17 07:39:01 QPF fonts are #$%#$^%#$^@#$@ (use any offensive words) - fucked TT May 17 07:39:38 hrw|work: I thought qt/e could use TTF fonts? May 17 07:39:47 XorA: yupp it can May 17 07:40:00 mickeyl: We have a load of 2.4 BOOTSTRAP_EXTRA_RDEPENDS in c7x0.conf, spitz.conf and akita.conf. Should we move them to zaurus-clamshell-2.4.conf as machine specific appends ? May 17 07:40:05 XorA: it can.. but getting it compiled is other way May 17 07:40:25 03RP 07 * r1.3322.1.1 10openembedded/conf/machine/ (zaurus-clamshell-2.4.conf spitz.conf akita.conf): Enable KERNEL_VERSION setting for akita/spitz May 17 07:40:47 hrw|work: would have thought it was more worth the time fixing qt/e ttf support than continuing to munge qpf fonts May 17 07:40:50 XorA: opie can use ttf fonts by opie-freetype addon. qt/e 'can' use ttf but use so fscking old freetype... May 17 07:41:07 RP: thnaks May 17 07:41:10 thanks May 17 07:41:41 hrw|work: hehe May 17 07:42:04 I will use old qt-2.3.2 makeqpf who-know-what-is-inside release which do the work May 17 07:42:58 tried few times to build makeqpf/arm or makeqpf/x86 and faster I will go to windows on desktop then to makeqpf again May 17 07:43:02 mickeyl: Specifically move things like BOOTSTRAP_EXTRA_RDEPENDS += "${@base_conditional('KERNEL_VERSION', '2.4', 'kernel-module-pxa-bi', '', d)}" May 17 07:43:22 to BOOTSTRAP_EXTRA_RDEPENDS_append_c7x0 in zaurus-clamshell-2.4.conf May 17 07:44:18 zecke: currently "configure -qconfig makeqpf -static" does not work even on x86 May 17 07:47:35 RP: if they're truly common then yes. May 17 07:49:24 mickeyl: The common factor is KERNEL_VERSION=2.4 and if we can lose all the 2.4 junk into that file, I'm all for it :) May 17 07:49:39 is bk://oe-devel.bkbits.net/openembedded still the current parent for OE May 17 07:51:22 RP: yeah. check though if it really doesn't affect the other machines May 17 07:51:31 iirc even tosa includes zaurus-clamshell.conf for some entries May 17 07:51:52 mickeyl: I'm making them machine specific appends May 17 07:52:41 http://www.hrw.one.pl/index.php/2005/05/17/makeqpf/ May 17 07:54:53 * mithro can't find anywhere to buy 48Mhz Crystal :/ May 17 07:55:10 Has anyone seen this one before: May 17 07:55:17 [*display] ~> ipkg files libfusionsound-0.9-22 May 17 07:55:17 Package libfusionsound-0.9-22 (0.9.22-r0) is installed on root and has the following files: May 17 07:55:27 May 17 07:55:37 /usr/lib/libfusionsound-0.9.so.22.0.0 May 17 07:55:37 /./@LongLink May 17 07:56:19 What could @LongLink be... May 17 07:56:48 mithro: start with a 8Mhz and multiply by three May 17 07:58:24 wrobbie: my firefox ipk has a @LongLink and Ive no idea where it comes from, its a PITA May 17 07:59:35 XorA: it actually prevents libfusionsound-0.9.so.22.0.0 from being installed.. i think the whole ipk is broken/incomplete.. May 17 08:00:06 wrobbie: out of interest, how big is the ipk? May 17 08:00:34 -rw-rw-r-- 1 rob rob 45234 May 17 23:01 deploy/ipk/libfusionsound-0.9-22_0.9.22-r0_i686.ipk May 17 08:00:48 nothing special.. May 17 08:01:20 hmm, an akita is E393 at pricejapan May 17 08:02:02 koen: and since they declare half price for customs, you'll only pay half tax if it gets caught... May 17 08:02:28 however, you need to add 3.9% if using paypal, and also the price of the power adapter May 17 08:04:05 03RP 07 * r1.3329 10openembedded/conf/machine/ (zaurus-clamshell-2.4.conf spitz.conf c7x0.conf akita.conf): Tidy up 2.4 kernel modules for c7x0, akita and spitz May 17 08:04:10 XorA: how do you do that? May 17 08:04:10 ok, the ipaq charger works on a z too May 17 08:04:31 * mithro needs 48MHz input to a USB chip May 17 08:04:58 mithro: actually, 6 * 4 is easier, can just use a clock doubler twice May 17 08:05:33 cant remeber the gate name, whatever one switches on transitions May 17 08:05:34 mithro: Farnell have one May 17 08:05:59 BigAl: That should allow you to build 2.6 kernels for spitz by setting KERNEL_VERSION=2.6 in local.conf May 17 08:06:10 mithro: http://au.farnell.com/jsp/endecaSearch/partDetail.jsp?SKU=112550&N=401 May 17 08:06:29 RP: Excellent, thanks! May 17 08:08:42 BigAl: thanks May 17 08:08:45 RP: Now I've just got to figure out why the CF devices aren't there anymore with 2.6.12-rc4 May 17 08:09:57 ouch $10 May 17 08:10:27 BigAl: It appears CF is broken on the c7x0 as well under 2.6.12-rc4 :-/ May 17 08:10:51 RP: That'd explain it. I thought it was weird that everything was coming up in sysfs ok but cardmgr couldn't find a thing. May 17 08:11:27 BigAl: I'm deliberately rolling forward with mainline to spot problems like this... May 17 08:11:52 BigAl: We just need to find out what killed it now :-/ May 17 08:11:57 RP: Yeah May 17 08:12:07 It could be rc3->rc4 or it could be mm1/mm2 May 17 08:12:38 I guess the only way to find out is to try them and see what happens. May 17 08:13:05 I hadn't noticed until you mentioned it. mm2 has only just started booting on arm (I helped them find the init crash we were seeing) May 17 08:13:39 do13: I found the reason for the init crash on the mm tree - patch is in the bb file in oe May 17 08:13:58 do13: CF is now broken :-/ May 17 08:15:34 RP: I'll have a bit of a poke around tomorrow and see if I can find anything May 17 08:17:14 BigAl: Ok, thanks. I might get it looked at before then - You'll see an update to the bb file if I find the problem in the meantime May 17 08:17:23 PLL stands for phase lock loop right? May 17 08:17:24 RP: Good luck! May 17 08:17:27 Yup May 17 08:18:30 so a 96Mhz PLL is what? a Phase Lock loop which can go up to 96Mhz? May 17 08:19:40 mithro: sounds like it May 17 08:20:34 when you "divide" frequency it goes down right? IE 96MHz divided by 2 goes to 48MHz? May 17 08:20:58 that's the theory May 17 08:22:37 ahh i've figure out this now, it appears that if you put a 4MHz signal into the 96MHz PLL you get a 96MHz signal May 17 08:23:39 so i only need a 4Mhz crystal May 17 08:24:02 mithro: if you put a 2Mhz signal in you should get 48 May 17 08:25:02 well it has a div 2 after it May 17 08:25:33 the block diagram looks something like this May 17 08:26:28 Clock Source -> Pre Divider -> 96Mhz PLL (4MHz input only) -> Post Divider May 17 08:27:32 So I can set it up with, 4Mhz Crystal -> / 1 -> 96Mhz PLL -> / 2 -> 48 Mhz May 17 08:31:00 I'm trying to do an initial clone of openembedded for a build.. I tried bk clone bk://openembedded.bkbits.net/openembedded openembedded and bk clone bk://openembedded.bkbits.net/openembedded May 17 08:31:11 can anyone let me know what the right one should be? May 17 08:32:00 neither seem to work for me, btw May 17 08:32:30 TyggerBob: its broken at moment, go moan at bitmover May 17 08:32:48 ahhhh May 17 08:32:59 I'll just stop moaning, now that I knw. May 17 08:33:14 %s/knw/know May 17 08:33:16 Thanks May 17 08:33:44 TyggerBob: its been annoying me all day as my repository suffered disk corruption May 17 08:41:32 Is there a problem with gnu-config-native? It fails because of config-guess-uclibc.patch May 17 08:42:41 VoodooZ_log: bk pull May 17 08:42:55 VoodooZ_log: http://bugs.openembedded.org/show_bug.cgi?id=2 May 17 08:43:18 I just did but I'll try again. May 17 08:46:11 actually, I forgot I was pulling from the nslu2 repo so I'll pull from yours. May 17 08:46:12 Thanks, May 17 08:47:15 I can't hide the fact that your patching May 17 08:47:15 system, when compared to the way the bsd folks do it, seems incredibly May 17 08:47:16 disorganized, or maybe hyper-organized May 17 08:47:37 That was said about the zaurus kernel patches... May 17 08:50:37 RP: aklml? May 17 08:51:12 koen: private email. Some bsd guy wanting to hack linux for spitz May 17 08:52:38 Nobody put a gun to his head and asked him to switch May 17 08:55:03 CosmicPenguin: The problem is he isn't using oe. Trying to use my patch series without oe is going to be rather painful... May 17 08:55:44 time to go, see you guys in a few days, I hope May 17 08:56:01 cu Bernardo May 17 08:56:07 Bernardo: fingers crossed :) May 17 08:57:00 CosmicPenguin: you got right about gnu-config patch.. it was dont to 20050516 not 20050515 (atleast in OE date as change was done in 20050515) May 17 08:58:55 hrw|work: my fault too - for a while, I thought yesterday was the 15th May 17 08:59:18 hrw|work: all good though - the patch works fine May 17 09:00:12 RP: I don't care what the BSD guys say - quilt has been awesome for me May 17 09:00:32 btw - its good that we switch to bugzilla - atleast bugs are easier to find there May 17 09:00:52 CosmicPenguin: My series isn't in quilt. It ends up in quilt when oe has processed the bb file May 17 09:01:36 CosmicPenguin: Although I use quilt to manipulate it, and yes, I like using patcher/quilt May 17 09:06:07 bye May 17 09:15:55 djeez, some people don't know how to read: http://bugs.openembedded.org/show_bug.cgi?id=2 May 17 09:17:30 To be fair, few people know thats up and running May 17 09:17:51 CosmicPenguin: reopening a bug with a different bug.... May 17 09:18:31 "can't find flex.tar.bz2" isn't a problem with the gnu-config patch not applying May 17 09:19:14 indeed its not May 17 09:30:04 hehe May 17 09:30:20 Though I can confirm flex-native is nonfunctional May 17 09:30:30 The pcmcia problem in the 2.6.12-rc4 kernel was new in -mm2 (not in -mm1 or -rc4) May 17 09:32:38 treke: flex-native builds over here May 17 09:33:17 NOTE: package flex-native-2.5.31: completed May 17 09:33:18 I was seeing the exact same cannot stat flex error, didn't investigage it though May 17 09:33:28 I'm guessing it's a distributionism May 17 09:33:58 the reopened bug number 2 was a case of bad sf mirrors May 17 09:34:25 my mailbox is being flooded :p May 17 09:34:45 treke: s/greg@treke.net/oe@handhelds.org/ May 17 09:34:53 yeah May 17 09:35:13 That's what I'm going to do, I just set it up that way while I tested initially May 17 09:36:02 treke: I can't assign a .bbs maintainer's email to a bug, because bugzilla "can't match", any idea on that? May 17 09:37:23 no account would be my guess May 17 09:37:44 sounds plausible May 17 09:41:39 can accounts be added to bugzilla? May 17 09:41:54 grep MAINTAINER | bugzilla-add May 17 09:42:02 something like that? May 17 09:42:30 I'd rather not May 17 09:43:49 If someone is going to get bugzilla mail, it should be because they signed up for an account May 17 09:43:55 not because someone else signed them up May 17 09:44:01 true May 17 09:44:20 but you could argue they are maintainer for the package in question May 17 09:44:56 Except you can't because we've have examples of people being set as maintainer who never touched bb May 17 09:45:10 look at raster and enlightenment May 17 09:45:43 didn't emte remove those? May 17 09:46:13 probably those particular ones May 17 09:47:03 ah yes, 8 .bbs left May 17 09:47:08 key kergoth May 17 09:47:14 s/key/hey/ May 17 09:47:22 hey May 17 09:47:28 I think anyone who will be using bugzilla should sign up for themselves. If they are marked as mainainers and won't, then remove their email address from the bb May 17 09:47:43 i'm really loving this new system of a down album May 17 09:47:54 cool May 17 09:47:57 mezme something? May 17 09:48:05 yeah, mezmerize May 17 09:48:52 hehe, one of the songs is called "This cocaine makes me feel like i'm on this song." May 17 09:48:57 9/me gets food May 17 09:49:00 s/9// May 17 09:55:27 nice, meteora for E8 May 17 09:55:42 * koen adds to cart May 17 10:01:08 wow and I thought kde's build was a a pain May 17 10:01:30 * koen suspects treke is building gnome May 17 10:01:40 no he's building this kdepim/pi crap May 17 10:01:56 I *like* the way you build gnome May 17 10:02:20 treke: not when you want to update gnome in OE May 17 10:06:43 But gnome's problem isn't the build - its the dependancies May 17 10:07:11 You might add 12 new packages, but at least they're all nicely autotooled May 17 10:07:31 they sure are May 17 10:07:48 the gnome people add a library a day it seems May 17 10:08:10 And each new library requires the very latest of all the other libraries May 17 10:08:38 indeed May 17 10:08:42 hey, they believe in code reuse May 17 10:09:11 treke: they are being religious about it May 17 10:09:35 there's a thin border between code reuse and being too lazy to recode one fscking tiny function but instead depending on yet another library May 17 10:09:38 You don't get any benefit from code reuse if you have to distributed 18 libraries just to show an "about" box May 17 10:10:18 ah but it's a well integrated and consistant about box May 17 10:10:47 yeah, and you can beam the contents via irda, and bluetooth, and even mail it via a multitude of protocols May 17 10:10:48 heh May 17 10:13:09 zecke: Yo May 17 10:13:19 hey May 17 10:13:46 mickeyl: hey, got some time? May 17 10:13:53 zecke: Do I want to be building a specific kdepim branch from svn? May 17 10:14:14 treke: no just HEAD May 17 10:14:17 aeh trunk May 17 10:14:20 ok May 17 10:14:22 darn :p May 17 10:14:37 zecke: a bit May 17 10:14:40 I built the 3.4 branch since I figured trunk was being ported to qt4 May 17 10:14:44 lemme try again May 17 10:14:54 got a call from the movers today, one of their drivers is in dallas today and wants to pick up my shit :) May 17 10:15:05 ths kdepimpi shit is nasty May 17 10:15:06 hehe May 17 10:15:11 doesn't actually build May 17 10:16:25 * zecke fixes something for dinner May 17 10:19:28 * treke tries trunk May 17 10:20:11 treke: trunk is still qt3 May 17 10:20:24 chouimat: excellant May 17 10:20:40 I thought I heard the qt4 stuff had started and had figured it would be in trunk May 17 10:20:42 treke: kde4 is on a branch for now May 17 10:21:11 sweet May 17 10:21:48 treke: won't be moved on trunk before qt4 release and maybe the conference ... May 17 10:26:27 building debs now May 17 10:29:46 kergoth: where you moving to these days? May 17 10:32:31 I'm going to push/commit two patches to bitbakes now May 17 10:35:24 * chouimat is away: watching Star Wars Revelations thingy May 17 10:35:58 zecke: cool May 17 10:36:02 ok, good. May 17 10:36:09 then you will make that 1.3 release or 1.2.1 ? May 17 10:36:46 mickeyl: let us test it first May 17 10:37:06 koen: apply the oe patch now May 17 10:37:38 hi everyone ! May 17 10:39:36 i'm looking for someone who still own the ipk file from the Textmaker beta test. I want to know if the registration process works with sharp-compat-libs. May 17 10:40:01 alan: try #openzaurus May 17 10:41:16 mithro: back to minnesota. still have my house there, so can stay there to cut my expenses while i figure out what i want to do May 17 10:41:26 koen: i did, looks like there is noone for this as noone tried it on openzaurus... May 17 10:42:34 alan: well, it's hasn't anything todo with this channel May 17 10:45:18 koen : ok. sorry ^^ May 17 10:45:33 btw, congrats for you hard work, everyone !! May 17 10:45:37 alan: pop into #openzaurus May 17 10:45:43 oki May 17 10:45:46 bye May 17 10:49:21 woot. they're picking up my stuff today May 17 10:49:29 after a month of back & forth and hassles with the moving company May 17 10:49:31 * kergoth sighs in relief May 17 10:51:28 kergoth: :) May 17 10:52:31 The simpsons are going to Minnisota! May 17 10:54:07 the simpsons are banned from minnesota May 17 10:55:39 yeah, but nobody seems to enforce that ban May 17 10:59:25 god damn using a unpatched bitbake is horror May 17 10:59:45 heh May 17 11:00:53 off to bed May 17 11:00:56 have a gnight May 17 11:04:48 03zecke123 * r157 10bitbake/ (5 files in 3 dirs): May 17 11:04:48 bitbake data module abstraction: May 17 11:04:48 -bb.data is now a delegate to hookable Data implementation. May 17 11:04:48 -bb.data.init() is the 'factory' method to create a instance May 17 11:04:48 of a concrete implementation. May 17 11:04:49 -Kill assumptions that bb.data.init() returns a {} (python dict) May 17 11:04:51 -Add the old Dictionary Based Implementation as data_dict.py May 17 11:07:24 zecke: the createCopy patch is going in after the release? May 17 11:07:35 koen: no May 17 11:07:48 koen: but this is svn and not bitkeeper ;) May 17 11:08:21 when is the creatCopy patch going in? May 17 11:08:39 * zecke starts python hacked_toss_coin.py False to decide whether to read guy* stuff May 17 11:08:49 koen: once parsing is done May 17 11:14:19 who is muesli321? May 17 11:14:25 anynody we know? May 17 11:15:15 ibot: seen muesli321 May 17 11:15:23 CosmicPenguin: i haven't seen 'muesli321' May 17 11:15:34 03zecke123 * r158 10bitbake/ (7 files in 3 dirs): May 17 11:15:34 createCopy: May 17 11:15:34 -Move a createCopy method into data.py May 17 11:15:34 -Make code creating deepcopies of the data instance May 17 11:15:34 use this new createCopy May 17 11:16:05 03zecke123 * r159 10bitbake/MANIFEST: Add data_dict.py to the MANIFEST May 17 11:18:48 koen: nope May 17 11:21:54 mickeyl: any: could you kick off a built? May 17 11:22:08 waiting for the oe patches May 17 11:22:16 they should be applied as well shouldn't they? May 17 11:22:29 mickeyl: yes all but the fetcher changes May 17 11:22:32 bleh May 17 11:22:35 packaging failed May 17 11:22:37 ok, koen? May 17 11:22:49 or who's going to apply 'em? May 17 11:23:04 zecke: if I were to just run make install in the kitchensync directory, would all be ok? May 17 11:23:11 which patches are those? May 17 11:23:18 zecke: or what packages do I really need updated? May 17 11:23:24 koen: I'll do it May 17 11:24:18 treke: libkdepim, libkdepimidentities, libkcal May 17 11:24:36 treke: but if you install libkcal, you need korganizer as well :} May 17 11:24:42 damn May 17 11:24:48 and korganizer is the one that wont package :p May 17 11:25:05 treke: ah and libical for libkcal May 17 11:25:20 anyone knowing something about docbook? May 17 11:25:45 NOTE: Running the following tests: ['patch_checker', 'source_checker'] May 17 11:25:45 NOTE: Collected 2378 BitBake files May 17 11:25:45 Running test for aalib with machine: h3600 and distro: familiar-0.9.0 May 17 11:25:47 cool May 17 11:27:01 koen: hehe, my machine went out of storage yesterday May 17 11:27:11 hmmm May 17 11:27:11 koen: we need to run wget with a fail fast option :} May 17 11:27:21 I should take a look at that May 17 11:27:38 the machine 'only' has 73gb on the data disk May 17 11:27:55 s/on/as/ May 17 11:28:00 koen: that should work May 17 11:28:36 ERROR: No module named data_smart while parsing /local/pkg/oe/packages/packages/pth/pth_2.0.2.bb May 17 11:28:53 öh May 17 11:29:07 mickeyl: is that out of the cache? May 17 11:29:12 hold on May 17 11:29:41 ok, parsing... May 17 11:31:39 * chouimat is back. May 17 11:31:58 hmmm. need to work out how to make dh_helper just do the package step :p May 17 11:32:29 03mickeyl 07 * r1.3330 10openembedded/ (7 files in 2 dirs): apply zecke's patches to prepare for the low memory bitbake. the repository needs bitbake r159 or better now May 17 11:35:19 mickeyl: hehe, so no release? :} May 17 11:35:31 you said you want to test May 17 11:35:39 mickeyl: hehe May 17 11:35:45 mickeyl: ok /me releases bitbake-r159 May 17 11:36:04 let me do a complete opie-image first May 17 11:36:08 from scratch May 17 11:36:11 will take some hours May 17 11:36:29 * zecke starts a build of the 'other' thingie May 17 11:36:44 * koen starts an oz/gpe build from scratch May 17 11:37:56 I really hope this works :) May 17 11:39:16 me, too May 17 11:39:25 then we can proceed to the interesting patches ;) May 17 11:39:27 03mickeyl 07 * r1.3331 10openembedded/packages/meta/task-python-everything_0.9.1.bb: add python-pygtk2 to task-python-everything May 17 11:39:28 I don't May 17 11:39:37 I still need and excuse for new ram May 17 11:41:21 mickey|tv: gnome-python is still missing from the repo :) May 17 11:42:31 Heh, I am using this as an excuse for more ram May 17 11:44:14 I think my Xeon box might manage 3 or 4 bitbake's at once when the patches are in :) May 17 11:45:34 The electric company is going to love you May 17 11:48:05 lol guyhlems GSM section is ridicoulus May 17 11:48:29 For example the SIMPAD guy pointed me to the MC35 Siemens Modems... They're ARM cores running Java May 17 11:48:47 they follow the regulation, not allowing dialing a number too frequently... and such stuff... May 17 11:49:18 zecke: the zeevo 2001 is also an arm core with 4mbit of flash May 17 11:49:28 (found in the ipaq 2210) May 17 11:50:33 koen: AT^Please wait while uncompressing kernel....... May 17 11:51:31 heh May 17 11:51:41 zecke: let's run mosix on it May 17 11:51:52 our own little cluster May 17 11:52:37 * gb2 pokes around with an n-gage May 17 11:53:28 zecke: parsing succeeded, compiling pkgconfig-native right now May 17 11:54:07 haha /me stepped into the fetcher trap May 17 11:55:34 fetching fakeroot failed May 17 11:55:36 zecke, I am looking through the comments you made about my patch May 17 11:55:53 do you have a kernel bb file you like better than the one I copied? May 17 11:56:16 re May 17 11:56:26 hi hrw May 17 11:56:53 Crofton: openzaurus-sa_2.4.18*emb*.bb May 17 11:57:11 Crofton: just look at COMPATIBLE host May 17 11:57:24 mickey|tv: I think we need to apply the fetcher patch as well May 17 11:57:56 ok May 17 11:58:55 bitbake rev. 159 - start of zecke patches? May 17 12:01:01 hrw: yes but only the nondrastic ones May 17 12:02:22 Why would a "bk pull" just return to the shell prompt without doing anything? May 17 12:02:42 RP: I had that as well May 17 12:02:47 RP: just wait some minutes May 17 12:10:32 Hi folks, sorry to bother you, but I couldn't find a opensimpad channel and it seems to me that opesimpad is based on oe. I want to build the couple of packages to opensimpad, but could not find a opensimpad specific bk repository. Is my understanding correct, that it's now just in the standart oe repository ? May 17 12:10:40 zecke: Any idea what the problem is? May 17 12:11:00 lazy_marmot: you're right here May 17 12:11:05 lazy_marmot: Yes, I'd have thought so May 17 12:11:11 lazy_marmot: we have only one openembedded tree to build for them all May 17 12:11:52 RP: no, When I tried using ethereal it worked May 17 12:11:57 RP: or was that strace May 17 12:12:47 Thanks folks, everything else is clear from wiki. If get my packages built I will then ask more questions how to add them to the repository. May 17 12:13:03 heh.. bt950_cs is not for 2.6 May 17 12:14:57 lazy_marmot: just shoot once you've question May 17 12:16:54 zecke: glibc builds ok too May 17 12:16:57 so far so good May 17 12:21:01 03zecke123 * r160 10bitbake/lib/bb/fetch.py: May 17 12:21:01 lib/bb/fetch.py: May 17 12:21:01 * Remove default params from the fetcher methods to avoid problems May 17 12:21:01 with scoping. The meaning of 'd' is quite overloaded and we try May 17 12:21:01 to undo it one by one. This change is a bit more invasive than May 17 12:21:02 I had hoped for but it is needed. May 17 12:21:04 -- Holger Hans Peter Freyther May 17 12:21:27 ~lart sf.net May 17 12:21:28 * ibot hits sf.net with an anvil and laughs with a contralto voice ... Haha Ha HA Ha May 17 12:21:34 * zecke is the only one without a source dir May 17 12:31:07 I still can't bk pull :-( May 17 12:32:05 I can May 17 12:34:03 * RP stabs bitkeeper May 17 12:35:45 ~lart dtl1_cs May 17 12:35:45 * ibot whacks dtl1_cs upside the head May 17 12:36:45 probably I will add script into /etc/apmd/ to eject/insert card and restart bluetooth if that dtl1_cs card is inserted May 17 12:37:32 hrw: you'll have to keep track of the state when doing that May 17 12:37:46 which state? May 17 12:37:50 on/off May 17 12:37:54 resume/suspend? I know May 17 12:38:20 I mean bluetooth on/off May 17 12:38:31 since the battery doesn't last that long on my corgi May 17 12:39:31 http://pastebin.ca/12004 May 17 12:39:57 koen: I have bt on all time when card is in May 17 12:40:11 zecke, there is a new kernel bb file, we can drop the one I sent earlier, well unless I mixed other stuf in it May 17 12:40:20 I think oe could use a bootstrap-udev-image May 17 12:40:36 hrw: I haven't May 17 12:40:44 saves about 30minutes of battery time May 17 12:41:00 whether udev is included is really a distro decision, and as such belongs in the distro's .conf, not a seperate meta package May 17 12:41:07 heh May 17 12:41:14 Spyro: USE_UDEV = "1" May 17 12:41:15 * kergoth gets food May 17 12:41:49 zecke: seriously? May 17 12:41:53 zecke: awww May 17 12:42:13 zecke: is there any *documented* list of such things? May 17 12:42:58 kergoth: Why would "bk pull" just be returning to the shell on me? May 17 12:43:07 what is that, "documentation"? May 17 12:43:08 no idea May 17 12:43:12 koen: /m hrw http://www.awtrey.com/tony/foss/debianisdifferent.php May 17 12:43:14 Spyro: seriously look at some conf/machine/jornada*.conf May 17 12:43:14 ops May 17 12:43:27 hrw: looks interesting :) May 17 12:44:04 zecke: is it documented anywhere? (or are people expected to guess at the existence of such things? May 17 12:44:41 Spyro: The main documentation of oe are the bb files themselves. We know this isn't ideal... May 17 12:44:54 RP: indeed. May 17 12:45:18 ask what good files are to copy :) May 17 12:45:20 Spyro: I know your busy as well, mind creating a wiki site? May 17 12:45:22 Spyro: patches welcome ;-) May 17 12:45:24 nothing wrong with things being self-documenting, but theres a limit :) May 17 12:45:31 Spyro: Oe_Flags May 17 12:45:34 RP: you know, itd be really easy to define a 'doc' attribute on variables to use for inline variable documentation. May 17 12:45:46 USE_UDEV[doc] = "Set this to 1 to use udev" May 17 12:45:55 then write a small convenience task or binary to dump that May 17 12:45:57 kergoth: good idea May 17 12:46:12 kergoth: It would probably work better than a manual May 17 12:46:14 bipolar: hey May 17 12:46:19 zecke: I would but Im not familiar (sic) enough with OE - there appears to be no standard for how these things should be done May 17 12:46:34 ok. time to go sleep.. tomorrow I go for SW episode III premiere so I wont have much sleep tomorrow... (movie starts at 00:01 in Thursday) May 17 12:46:37 kergoth: you could even have it make a manual :) May 17 12:46:44 zecke: Im adding udev to a custom machine file, you suggested USE_UDEV, and Im sure there are other ways May 17 12:46:46 'night hrw & have fun May 17 12:47:39 Spyro: You can probably stick what you want into your local.conf file rather than a custom machine file May 17 12:48:13 RP: indeed. the only real question is how to document an entire file, like have a description of what a given bbclass does. May 17 12:48:27 we have USE_NLS, USE_VT, USE_DEVFS but no USE_UDEV (yet) May 17 12:49:09 and who said we didn't have USE flags? May 17 12:49:18 ~lart koen May 17 12:49:19 * ibot gives koen an extra strength ACME sleeping pill, sending koen to sleep for 150 years, and awakening to seven strange dwarfs and a large apple May 17 12:49:29 cu May 17 12:49:51 evening May 17 12:49:52 kergoth: We'd probably want to add some kind of class header giving a name and description, then document the functions in the class seperately May 17 12:50:05 yeah May 17 12:50:07 sheesh... how long does it take to generate ahost key :) May 17 12:50:11 * Spyro yawns May 17 12:50:21 think something like mickeyl's machine/distro .conf descriptive metadata May 17 12:50:22 Spyro: use the ts May 17 12:50:25 only for classes May 17 12:50:31 Spyro: or press some buttons May 17 12:50:45 kergoth: I was thinking of that as I was typing the above May 17 12:50:46 koen: why? May 17 12:50:51 :) May 17 12:50:51 Spyro: entropy May 17 12:50:56 ahhhh May 17 12:51:05 forgot about that May 17 12:51:35 zecke, hiya! Whats up? May 17 12:51:42 anyway, the 'doc' attribute thing could be implemented right now with almost no effort. a small bbdoc app and start fleshing out the info May 17 12:52:00 bleh, i need to finish cleaning and packing May 17 12:52:17 you would do much better workin gon oe instead :) May 17 12:52:37 they're picking up my stuff in like 3 hours :) May 17 12:52:41 kergoth: agreed. I think we might have a way forward there May 17 12:52:52 kergoth: best pack then ;-) May 17 12:53:09 no time like the present to procrastinate May 17 12:53:18 Have them walk in on you playing vice city May 17 12:53:32 hehe May 17 12:53:37 or world of warcraft May 17 12:53:49 and get a ups to power the comp while traveling May 17 12:53:52 kergoth: One thing I will add is that maybe we should have the documents as comments so they don't add to out memory problems May 17 12:53:53 in your underwear of cocurse May 17 12:54:12 s/documents/documentation/ May 17 12:54:19 doxygen style? May 17 12:54:55 RP: not a bad way to go, but increases the amount of work, since you'd have to manipulate the parser. probably a good idea though May 17 12:55:20 ok it boots and I have a login via ssh. thanks chaps May 17 12:55:31 now I need to build X :) May 17 12:56:26 kergoth: The parser isn't going to need to be that clever. A few custom lines of perl would probably do :) May 17 12:56:44 whats the package name for X in OE ? May 17 12:56:52 I grepped for X and found zip May 17 12:56:59 grep for x May 17 12:57:04 Spyro: bitbake gpe-image May 17 12:57:22 diet-x11 and xserver-kdrive would be the most obvious targets May 17 12:58:14 RP: I dont want the entire GPE image :) May 17 12:58:33 diet-x11? never come across that one :) May 17 12:58:52 Spyro: smaller xlib May 17 12:59:25 * chouimat is away: stuff May 17 13:00:16 Spyro: Watching it build will give you some hints as to the bits you need though May 17 13:01:17 How can I find what xserver gpe uses from looking at the bb files? May 17 13:01:19 kergoth: the only problem with [doc] is one would have to hack the parser :} May 17 13:01:42 Spyro: check for PREFERRED_PROVIDER and PREFERRED_VERSION May 17 13:01:46 Spyro: in you configuration May 17 13:01:47 Spyro: xserver is mostly a MACHINE thing May 17 13:01:56 oh May 17 13:01:59 figures May 17 13:02:26 Spyro: check conf/distro/preferred-gpe-versions.inc packages/meta/meta-gpe* for some pointers May 17 13:03:31 zecke: ? May 17 13:05:51 kergoth: USE_UDEV[doc] = "Foo" May 17 13:06:08 kergoth: you need to teach the stuff in lib/bb/parser/* to handle it May 17 13:06:15 kergoth: not hard but no fun either :} May 17 13:06:36 what are you talking about? May 17 13:06:44 thats standard variable flags syntax May 17 13:10:09 kergoth: :} May 17 13:10:20 kergoth: I've seen that for tasks May 17 13:10:20 building kdrive... May 17 13:10:28 but didn't consider it flag syntax May 17 13:22:00 how do you enable soft-float? (presumably a gcc flag ?) May 17 13:22:14 TARGET_FPU="soft" May 17 13:22:19 :) May 17 13:22:35 * chouimat is back. May 17 13:22:51 I think itd be cool to have an 'options class' May 17 13:23:40 so I could run something like 'list_class CPU_ARM' and get every flag that can be tweaked for ARMs in oe (eg. TARGET_FPU) and the paramters allowed May 17 13:24:13 Spyro: we have that: #oe May 17 13:24:28 koen: LOL :) May 17 13:24:56 list_class CPU_ARM_XSCALE May 17 13:25:02 aww :) May 17 13:25:07 *beep* May 17 13:25:16 hehe May 17 13:25:40 somebody feels like donating me a new machine? (I'm at packaging glibc) May 17 13:26:01 zecke: nope :) May 17 13:26:26 zecke: I can't understand how packaging glibc takes so long :-/ May 17 13:26:49 RP: locale generation ? May 17 13:26:51 zecke: Lets get the multitask sorted - that will help :) May 17 13:27:21 Spyro: Its the packaging stage - I know there are a load of locales but that May 17 13:27:24 's no excuse May 17 13:27:41 it's O(n^2) or something May 17 13:28:03 that hurts with 217 locales May 17 13:28:05 koen: How is a packaging stage running an n^2? :-( May 17 13:28:13 yow May 17 13:28:39 RP: a while back pb gave an estimate of it May 17 13:29:12 shit May 17 13:29:19 the movers are gonna be here in like half an hour May 17 13:29:19 hehe May 17 13:29:38 * Spyro is getting a new central heating boiler tomorrow. May 17 13:29:58 my dumb landlord left the current one leaking so long the dumb bastard had to replace it May 17 13:30:15 on top of that new laws require he fit a more expensive (condensing) type ;-) May 17 13:30:30 Im not complaining - it'll almost halve my gas bill May 17 13:30:54 Spyro: no need for OE to heat up your cpu's anymore? May 17 13:31:10 koen: LOL :) May 17 13:31:16 oe is making em want a dualie or quaddie May 17 13:31:34 * Spyro looks up the price of quaddie boards and thinks again May 17 13:31:49 the students in the lab aren't very thankfull for the OE buildbox :) May 17 13:31:56 it's LOUD May 17 13:31:57 :) May 17 13:32:12 and even louder now I plugged in all 3 PSUs May 17 13:32:21 ha ha ha May 17 13:32:27 I have a "dualie" but its also loud and I wouldn't have liked to have paid full price for it... May 17 13:32:42 Its in another room away from me :) May 17 13:33:08 my dual was pretty quiet, till the psu failed May 17 13:33:14 this new psu sucks May 17 13:33:26 wanna know the brand? May 17 13:33:32 "emacs/zippy" May 17 13:33:36 Its the two 10" fans that make that machine noisey. Its server build :-/ May 17 13:34:27 10" fans should be quiet enough. its small fans that do the screamin May 17 13:34:44 Spyro: wanna bet? May 17 13:35:22 unless the 10" are crap, yes May 17 13:36:51 Spyro: They run at the same speed as a small one May 17 13:37:07 They're not crap, just designed to shift serious amounts of air... May 17 13:37:21 So much so I'm not running one of them... May 17 13:37:23 zecke? May 17 13:37:44 no-one needs to shift that much air. if thats the case then the airflow in the machine needs some serious addressing May 17 13:38:51 Spyro: Its just overspeced May 17 13:39:05 Its server grade. Noise doesn't matter :-/ May 17 13:39:18 Crofton: oh I've missed something :} Yes we can start merging your changes May 17 13:39:23 RP: I bet they thought the same thing about wight May 17 13:39:28 weight even May 17 13:39:38 * zecke pets his 20 SMP ARM cores ;) May 17 13:39:50 heh May 17 13:39:57 did you see the pastebin? May 17 13:40:09 http://pastebin.ca/12004 May 17 13:40:25 Don't worry about the rc2 kernel May 17 13:40:47 I will send you a patch based on this one May 17 13:41:02 Crofton: looks ok, sure you don't want to depend on UBOOT anymore? May 17 13:41:09 mickey|tv: python is so fscking great May 17 13:41:11 I need to get uboot building so I have access to mkiamge May 17 13:41:17 mickey|tv: almost as good as haskell May 17 13:41:29 right now I haven't got uboot working, but I put mkiamge in the path May 17 13:41:45 and it is hell to debug since the machine I have now needs more RAM May 17 13:42:06 Crofton: the OMAP or your host? May 17 13:42:15 Will it annoy people I move the kernel image to /tftpboot? May 17 13:42:30 I can build uboot OK outside OE May 17 13:42:45 I just haven't figureed out how to build it for osk inside OE May 17 13:42:53 Crofton: yes ;) May 17 13:43:02 heh May 17 13:43:13 We can comment the line out then :) May 17 13:43:16 Crofton: that parts needs to be kept local, or use OVERRIDES to change the path May 17 13:43:18 I am very lazy May 17 13:43:38 I want a short file name too :) May 17 13:45:32 I would like to get the patches in, since they do build images that are usable on the OSK now May 17 13:45:40 even if they aren't "perfect" May 17 13:46:19 Crofton: I agree with that May 17 13:47:39 The COMPATIBLE_HOST is much nicer than the python May 17 13:47:58 the python came from the kernel file I copied to start with May 17 13:48:13 (see earlier discussion on documentation) May 17 13:48:21 Crofton: consider patching that file as well May 17 13:48:58 Crofton: yes, even if we get bashed from SL6000 users we've gained quite some knowledge in our package database May 17 13:49:07 heh May 17 13:49:20 we could remove the 2.6.9 omap build May 17 13:49:40 it lacks features such as dsp gw support and cf support May 17 13:50:06 the _bk file that may be in my patches came from a need to pull from bk and build against that May 17 13:52:13 ok, I guess kergoth is the omap 2.6.9 maintainer? and I also figure he has not any plan how to continue with it May 17 13:54:53 I think that was the current kernel version when he looked at it May 17 13:55:09 He doesn't work for TI anymore :) May 17 13:55:12 go ahead and kill it. it was current at the time May 17 13:55:24 hmm, movers there yet? May 17 13:55:30 not yet, should be soon May 17 13:55:36 and i'm not fully packed, guess i'm leaving my dishes behind May 17 13:55:37 hehe May 17 13:55:40 heh, well GL May 17 13:55:46 wince they're dirty, i have to clean them before i can pack them May 17 13:55:47 :\ May 17 13:55:49 since, even May 17 13:55:51 I guess they are just moving, not packing? May 17 13:55:57 yeah May 17 13:56:06 cant afford to spend the extra to be really lazy May 17 13:56:10 yeh May 17 13:56:18 kergoth: good luck with everything May 17 13:56:20 i needed to _use_ the dishes. i wish they had given me a couple days notice May 17 13:56:21 :\ May 17 13:56:22 thanks May 17 13:56:40 I sure hope you don't have any mover horror stories May 17 13:57:03 Max sis OK, but the same truck that picked up his stuff,dropped it off May 17 13:58:07 s/did/sis May 17 14:15:08 anyone try getting usbmon going in OE? May 17 14:15:59 Crofton: what is usbmon May 17 14:16:17 apparently something to monitor crap that goes over usb May 17 14:16:26 I am having problems with a usb device May 17 14:16:31 doing research now May 17 14:16:39 if you dont put someting in RDEPENDS does that mean it wont end up in the rootfs/ dir ? May 17 14:16:42 ah the latest stuff May 17 14:16:49 Spyro: no May 17 14:17:06 Spyro: it ends up in the rootfs dir when you create an image May 17 14:17:16 I did create an image May 17 14:17:23 my machine file has May 17 14:17:30 BOOTSTRAP_EXTRA_DEPENDS = "udev sysfsutils xserver-kdrive" May 17 14:17:41 and it built the xserver stuff May 17 14:17:46 but I cant find it in my rootfs May 17 14:18:05 I think you need to add it to RDEPENDS also May 17 14:18:08 Spyro: DEPENDS is a buildtime dependency May 17 14:18:15 RDEPENDS is runtime... May 17 14:18:18 I am a little fuzzy on DEPENDS and RDEPNEDS though May 17 14:18:37 Spyro: Change it to an RDEPENDS May 17 14:19:01 RP: Im still not clear on the difference :) May 17 14:19:05 RP: trying now May 17 14:19:42 Spyro: Runtime is what you actually want on the image. Buildtime is just required to compile it May 17 14:19:44 Spyro: RuntimeDepends May 17 14:20:01 and a DEPENDS differs how? May 17 14:20:20 Spyro: autoconf is a DEPENDS for example May 17 14:20:21 Its *build* time. Only needed when compilng May 17 14:20:39 it's needed for the build, but won't be needed to run the app May 17 14:21:06 that's not the point though May 17 14:21:22 dependecies handled by OE are in DEPENDS May 17 14:21:39 dependencies handled by the package manager (ipkg) are in RDEPENDS May 17 14:21:57 OE doesn't care about RDEPENDS directly May 17 14:22:21 reenoo_: Surely oe builds what it finds in RDEPENDS though? May 17 14:22:25 so what happens if you put something in RDEPENDS and not DEPENDS ? May 17 14:22:30 RP: no May 17 14:22:37 whatever you add to RDEPENDS must have been built as a package beforehand May 17 14:22:47 evil May 17 14:23:26 reenoo_: I can see do_rootfs going a bit made if you miss an RDEPENDS out a DEPENDS then which is a bit silly :-/ May 17 14:23:27 you can RDEPEND on kernel-module-iptables or xine-plugin-w100 May 17 14:23:40 OE doesn't know about those May 17 14:24:02 so you need to depend on virtual/kernel and libxine May 17 14:24:17 s/depend/DEPEND/ May 17 14:26:05 Spyro: there's nothing evil about it actually. it's the only way you can deal with this kind of thing. the majority of packages (.ipk, .tgz) are generated automatically using some regex filter or another. OE cannot know about those beforehand... May 17 14:26:36 progress on your builds? May 17 14:26:56 gpe-image almost done May 17 14:27:18 so far so good May 17 14:27:21 zecke: I have a bitbake from svn and a bitbake with all patches applied running from the same oe data, all going well so far May 17 14:27:49 *ARG* May 17 14:28:20 god damn it, this is dependency hell. it didnt add X to the rootfs for some reason despite me adding it to the RDEPENDS May 17 14:30:24 Spyro: you have re-built task-bootstrap though, haven't you? May 17 14:30:41 reenoo_: doesnt bootstrap_image do that for me ? May 17 14:31:15 no May 17 14:31:30 oh May 17 14:31:32 why not ? May 17 14:33:26 because the stamps of task-bootstrap say task-bootstrap has been built successfully May 17 14:34:33 I rebuilt task-bootstrap May 17 14:34:42 now doing bootstrap-image May 17 14:35:12 that's a nice one May 17 14:35:19 'night all May 17 14:35:33 'night koen|sleep May 17 14:35:37 hm. task-bootstrap didnt build anything May 17 14:36:01 and (perhaps unsurprisingly) bootstrap-image also failed to do anything different (still no X) May 17 14:39:10 you have 'bitbake -c clean'ed it and it's still the same? May 17 14:39:26 that'd surprise me a bit :) May 17 14:39:29 reenoo_: I have to rebuild all of task-bootstrap? May 17 14:39:41 thats gonna take forever :( May 17 14:40:14 task-bootstrap is just a meta-package. "a" as in "one", "a single one" May 17 14:41:20 bitbake -c clean task-bootstrap will only clean task-bootstrap May 17 14:41:29 phew May 17 14:41:32 not any of its dependencies May 17 14:43:54 either we do a release 'now' or we need to wait until tomorrow May 17 14:44:41 zecke: bitbake with your patches? May 17 14:45:36 reenoo_: no May 17 14:45:42 reenoo_: just the non intrusive part May 17 14:45:48 to start going mad with trunk May 17 14:45:54 zecke: I say go for a release May 17 14:46:09 zecke: Its working for me and seems ok for koen :) May 17 14:46:22 zecke: ah, I see May 17 14:46:36 zecke: I want to try four builds at once :) May 17 14:47:03 ARGH. now its moaning about missing deps... May 17 14:47:17 zecke: opie-image went through May 17 14:47:24 zecke: I'm for releaseing May 17 14:50:54 it fails with: May 17 14:51:12 | ERROR: Cannot satisfy the following dependencies for task-bootstrap: May 17 14:51:15 | xserver-kdrive May 17 14:51:15 NOTE: Task failed: May 17 14:51:15 NOTE: package bootstrap-image-1.0-r0: task do_rootfs: failed May 17 14:51:15 ERROR: TaskFailed event exception, aborting May 17 14:51:15 NOTE: package bootstrap-image-1.0: failed May 17 14:51:15 ERROR: Build of bootstrap-image failed May 17 14:51:19 And yet: May 17 14:51:30 bash-2.05b$ bitbake xserver-kdrive May 17 14:51:32 NOTE: Psyco JIT Compiler (http://psyco.sf.net) not available. Install it to increase performance. May 17 14:51:32 NOTE: Using cache in '/home/ian/projects/openembedded/stuff/tmp/cache' May 17 14:51:32 NOTE: Parsing finished. 2270 cached, 0 parsed, 47 skipped, 0 masked. May 17 14:51:32 NOTE: Building provider hash: [####################] (100%) May 17 14:51:32 NOTE: build 200505172249: started May 17 14:51:33 OE Build Configuration: May 17 14:51:35 TARGET_ARCH = "arm" May 17 14:51:37 TARGET_OS = "linux" May 17 14:51:39 MACHINE = "eseries" May 17 14:51:41 DISTRO = "" May 17 14:51:43 TARGET_FPU = "" May 17 14:51:45 NOTE: build 200505172249: completed May 17 14:51:47 Build statistics: May 17 14:51:49 Attempted builds: 0 May 17 14:51:51 xserver-kdrive is clearly built May 17 14:51:52 Spyro: pastebin May 17 14:52:13 :) May 17 14:52:26 are you sure the ipk has been built? May 17 14:52:35 ls -l tmp/deploy/ipk/*kdrive* May 17 14:52:51 Spyro: That doesn't say xserver-kdrive is built - it tried to build nothing... May 17 14:53:12 RP: it built earlier - I was watching it :) May 17 14:53:18 heh - I didn't know genext2fs had an active development c ommunity May 17 14:54:22 mickey|tv: its in tmp/deploy May 17 14:54:42 ati, epson, fake, xfbdev... all there May 17 14:55:02 hmm. should the RDEPEND be for xserver-kdrive-fbdev ? May 17 14:55:12 yeah May 17 14:55:17 bah :) May 17 14:55:20 hah May 17 14:55:21 :) May 17 14:56:28 * Spyro will one day write the evilmastr-packagemanager-wot-does-everything-and-actually-behaves-itself (after teatime when Ive conquered the world) May 17 14:57:21 mickey|tv: next question can I do a release? :} May 17 14:57:36 zecke: no idea. i never did one with the berlios interface May 17 14:57:41 I'm going to call it 1.2.1? May 17 14:57:51 yes, that makes sense May 17 14:58:01 I'm wondering if you should call it a pre-release? May 17 14:58:05 ahh, doesn't matter I guess May 17 14:58:22 yeah, if it's broken, we do 1.2.2 :) May 17 15:02:06 03zecke123 * r161 10bitbake/ChangeLog: Add a ChangeLog to lay out the 'API' changes of bitbake May 17 15:02:50 do we have an AUTHORS file? May 17 15:02:55 if not, could you add one while you're there? :) May 17 15:03:58 can I skip the tedios 'handling bitbake files' bit every single time ? May 17 15:04:07 bitbake -h May 17 15:04:12 there it will tell you about -b May 17 15:04:24 you will disarm the dependency handling in that case though May 17 15:04:26 so be warned May 17 15:04:37 I tried -b but it bitched. it seems you need to specify the path to the .bb file if you use -b May 17 15:04:45 right May 17 15:04:48 -b May 17 15:04:58 mickey|tv: from what I saw so far it wouldnt be much of a loss ;-) May 17 15:05:18 uhm May 17 15:05:21 it works very good for us May 17 15:05:35 mickey|tv: note winky May 17 15:05:52 sure May 17 15:07:41 ~lart our broken router May 17 15:07:41 * ibot takes out a seltzer bottle and sprays our broken router in the face. You know, one of those old-school seltzer bottles clowns have? Yeah those. Anyway, consider yourself spritzed May 17 15:07:45 wtf svn copy complains that bitbake is readonly May 17 15:08:11 ibot: yeah, that'll really help ;-) May 17 15:08:55 Yay! I have Xness. May 17 15:09:38 congrats May 17 15:11:03 03zecke123 07bitbake-1.2.1 * r162 10/: May 17 15:11:03 Now it is official, bitbake 1.2.1 is tagged. May 17 15:11:03 Bitbake1.2.1 features a revamped data module structure. We will build on that May 17 15:11:03 for future improvements May 17 15:11:12 hmm May 17 15:11:20 well May 17 15:11:22 yeah May 17 15:14:46 mickey|tv: do you want to commit your scoping fix to bin/bitbake after my commits? May 17 15:14:46 cool. Xfbdev stiffed it May 17 15:14:52 now I have aminimal debug case May 17 15:14:57 zecke: yes May 17 15:15:01 zecke: now? May 17 15:15:23 * chouimat is away: food May 17 15:16:19 03zecke123 * r163 10bitbake/lib/bb/build.py: May 17 15:16:19 lib/bb/build.py: May 17 15:16:19 Do not assume we hold a reference to the variable May 17 15:16:19 we got with a bb.data.getVar*. Always set changes May 17 15:16:19 back to the variable. May 17 15:17:59 I added an AUTHORS file May 17 15:18:51 hmm May 17 15:18:52 svn: Connection is read-only May 17 15:20:06 03zecke123 * r164 10bitbake/lib/bb/ (data.py data_dict.py make.py): (log message trimmed) May 17 15:20:06 lib/bb/data.py: May 17 15:20:06 ?\194?\183 -Add a method to get the modify time of named bb.data instance May 17 15:20:06 ?\194?\183 -Add a method to create a named bb.data instance. These instances May 17 15:20:06 ?\194?\183 inherit the global configuration. May 17 15:20:07 lib/bb/data_dict.py: May 17 15:20:09 ?\194?\183 -Add persistents to the data_dict implementation, on the commit May 17 15:20:20 oops May 17 15:20:40 mickey|tv: svn co svn+ssh://svn.berlios.de/svnroot/repos/bitbake/trunk/bitbake bitbake May 17 15:20:45 ah May 17 15:20:46 thanks May 17 15:20:48 anyone seen this when building strace ? May 17 15:20:59 | NOTE: make DESTDIR=/home/ian/projects/openembedded/stuff/tmp/work/strace-4.5.11-r1/image install May 17 15:20:59 | make: *** No rule to make target `install'. Stop. May 17 15:21:06 nopie May 17 15:21:10 never May 17 15:22:40 03mickeyl * r165 10bitbake/ (AUTHORS MANIFEST): add authors file May 17 15:22:51 t'works May 17 15:24:31 'night all May 17 15:25:09 03zecke123 * r166 10bitbake/lib/bb/ (data.py make.py): (log message trimmed) May 17 15:25:09 lib/bb/data.py: May 17 15:25:09 -Provide a special 'collection/dict' for the collected May 17 15:25:09 bb-files. If we use a persistent storage (CACHE) we will May 17 15:25:09 not hold any data instance in memory but load it from disk May 17 15:25:10 on demand. May 17 15:25:12 This reduces bitbakes memory usage to roughly 40MB. The initial May 17 15:25:35 mickey|tv: good May 17 15:25:45 where does bitbake put the packages its compiling? May 17 15:25:58 mickey|tv: I'm comitting to bin/bitbake now May 17 15:26:08 ok May 17 15:26:09 Spyro: tmp/work/* May 17 15:26:26 mickey|tv: but not the 'def main():' change May 17 15:26:40 ok, this will be done by my OO refactoring anyway May 17 15:27:00 oops I've broken something May 17 15:27:03 theres no way thats where the source is... strace is definately more than 40 files in size May 17 15:27:33 Spyro: I see that as well May 17 15:28:14 zecke: I take it its broken then? May 17 15:31:15 03zecke123 * r167 10bitbake/bin/bitbake: May 17 15:31:15 bin/bitbake: May 17 15:31:15 -Access pkgdata[] less. This means unpickling the data less often. May 17 15:31:15 This results in a faster initial parsing. We still access each May 17 15:31:15 file three times though May 17 15:31:26 Spyro: yupp. I'm currently into QA and soon such problems will be noticed earlier May 17 15:31:37 zecke: how do I find the scripts or whatever that do the actually do the building? May 17 15:31:39 mickey|tv: I need to fix a bug in data_dict.py now I guess May 17 15:31:43 the .bb files seem to not be those May 17 15:32:00 Spyro: you've tmp/work/PackageNM-PackageVersion-PR/ May 17 15:32:02 zecke: ok. i will commit the scoping fix tomorrow May 17 15:32:06 eyes are falling May 17 15:32:08 good work May 17 15:32:09 g'night May 17 15:32:15 zecke: yes May 17 15:32:19 Spyro: there you've temp/ with the scripts May 17 15:32:28 Spyro: and logs (if you asked OE/bitbake to emit them) May 17 15:32:31 Spyro: and the source tree May 17 15:33:18 scripts are run.xxx ? May 17 15:35:23 whats a 'do_stage' ? May 17 15:38:33 Spyro: installing header to tmp/staging/ May 17 15:38:41 Spyro: yes May 17 15:38:47 oops I've broken something May 17 15:38:48 what header ? May 17 15:39:11 Spyro: header files or other resources (*.h) May 17 15:39:21 Spyro: when you build a library or such May 17 15:39:34 Spyro: for strace it can be empty though May 17 15:39:36 from that package then, not the ones needed to build that package May 17 15:39:59 the run_* files dont differ much... are they auto generated ? May 17 15:40:12 Spyro: yupp May 17 15:40:39 what generates them ? May 17 15:40:44 Spyro: bitbake May 17 15:40:53 Spyro: depending on your package definition May 17 15:40:54 from what ? May 17 15:41:00 Spyro: from the .bb files May 17 15:41:03 the .bb ? May 17 15:41:07 hm. May 17 15:41:12 Spyro: a .bb file inherits a bbclass May 17 15:41:28 Spyro: a bbclass defines tasks and their order (a depend on b but b b depends on c) May 17 15:41:36 Spyro: and then bitbake just executes the tasks May 17 15:41:41 inherit autotools May 17 15:46:48 zecke: ok so I can see an autotools.bbclass May 17 15:46:58 zecke: how does that tell bb to unpack the strace source? May 17 15:47:07 zecke: more to the point, why didnt it? May 17 15:47:25 03zecke123 * r168 10bitbake/lib/bb/ (data.py parse/ConfHandler.py): May 17 15:47:25 lib/bb/data.py: May 17 15:47:25 -Remove default arguments from the data methods. They May 17 15:47:25 cast errors. May 17 15:47:25 (we need to update the test case though) May 17 15:47:26 lib/bb/parse/ConfHandler.py: May 17 15:47:28 -Use the right dictionary May 17 15:54:17 zecke: looks like the problem is a failure to fetch straces source May 17 15:54:25 but do_fetch didnt fail May 17 15:55:17 or rather it failed silently May 17 15:55:20 Spyro: does it come from sourceforge? May 17 15:55:25 03zecke123 * r169 10bitbake/lib/bb/ (data.py data_smart.py): (log message trimmed) May 17 15:55:25 lib/bb/data.py: May 17 15:55:25 Use the new data_smart dictionary May 17 15:55:25 lib/bb/data_smart.py: May 17 15:55:25 The old COW Implementation, made persistent and pass May 17 15:55:26 the test suite. May 17 15:55:28 I commit it because the persistent mode in data_dict.py seems May 17 15:55:34 Spyro: they love to append ?use-other-mirror May 17 15:55:39 zecke: checking May 17 15:55:57 ah yes. May 17 15:56:01 bastards May 17 15:56:23 Spyro: I plan to fetcher bitbakes fetcher in that case May 17 15:56:30 how come do_fetch() doesnt actually fail ? perhaps do_fetch should test to see if it *actually* downloaded what it thought it did ? May 17 15:56:43 Spyro: yupp, I plan to do it May 17 15:56:55 Spyro: look at lib/bb/fetch.py if you want to fix it May 17 15:57:13 zecke: if I get the time and learn python I will :) May 17 15:58:39 Spyro: no hurry, you do other cool stuff May 17 15:59:13 zecke: it appears to be making no attempt whatsoever to get the file... May 17 15:59:26 mickey|zzZZzz: if you feel like there is one bug in data_dict.py May 17 16:00:00 mickey|zzZZzz: damn copying Cow stuff to it wasn't the smartest move :} May 17 16:02:06 zecke: theres no {run,log}.do_fetch May 17 16:06:30 if I do bitbake -b openembedded/packages/strace/strace_4.5.11.bb it should still download the package right ? May 17 16:08:26 I skipped the download stage and did it by hand May 17 16:08:43 if anyone works out why it wouldnt download, do let me know :) May 17 16:14:11 Spyro: If it tries a download and it fails you can be left with a 0 sized md5 file in the sources directory May 17 16:14:24 wiping that would make it retry the download May 17 16:14:37 Spyro: This is one of the things that needs fixing May 17 16:16:54 RP: aha May 17 16:19:03 I'm sending a bunch of patches to the oe@handhelds.org list May 17 16:19:16 do they look alright, or am i just spamming :) May 17 16:20:00 Spyro: some are python methods somre are shell functions May 17 16:21:11 Spyro: RP: please use the bug tracker May 17 16:22:42 <_ggilbert> RP even has his own section May 17 16:22:57 Does he? That's news to him :) May 17 16:24:22 _ggilbert: Do I need to create an account or do I already have one? May 17 16:24:26 python drives me insane, there is something wrong with the data_dict.py implementation May 17 16:24:39 <_ggilbert> you need to create one May 17 16:24:52 <_ggilbert> I never bothered to mention it since kernel developers seem to hate tools like bugtrackers May 17 16:25:02 the new bugzilla? May 17 16:25:21 ok cya later May 17 16:25:34 data_smart.py should work. data_dict.py doesn't... whatever May 17 16:25:45 kergoth: thumbs up, good luck May 17 16:26:44 Hmm, the files i included in my emails got scrubbed May 17 16:26:45 dammit May 17 16:28:35 _ggilbert: This would be the Zaurus kernel Request area? I can see that giving me headaches :) May 17 16:29:03 Ken|JLime: There's a size limit on the list - providing links will probably be best May 17 16:29:21 RP, roger i'll do that May 17 16:33:46 Btw who usually applies/adds files and patches? May 17 16:40:51 what's wrong if "bitbake bootstrap-image" fails on the first do_fetch on quilt (without printing any error), whileas "bitbake -cfetch -b ../openembedded/packages/quilt/quilt-native_0.39.bb" doesn't fail? (building on ppc64 for mipsel) . no files except the directory "tmp/work/quilt-native-0.39-r0/" are generated in tmp/. no error is displayed. BBDEBUG="yes" is set. May 17 16:41:10 Ken|JLime: Anyone who has access and knows something about what they apply against May 17 16:41:12 i can manually (bitbake -b ..) build the package - the next package then fails May 17 16:42:07 same setup on x86 as host works fine.. except that i'm using psyco there which isn't available for ppc64 here May 17 16:42:15 tmb_arg: check your sources directory for 0 byte sized files May 17 16:42:53 It sounds like quilt isn't getting installed properly :-/ May 17 16:43:22 as said, if i "bitbake -b .../quilt-native*.bb", quilt builds properly. using bitbake without -b fails on the next package then May 17 16:44:22 and completely emptying sources didn't helped May 17 16:44:55 I really don't know then - sorry, irc client crashed there May 17 16:45:33 I'm off to bed. 'night all May 17 16:46:47 nighty May 17 17:21:16 whee, movers almost done, i have no place to sit other than floor now May 17 17:21:17 heh :) May 17 17:29:40 :D May 17 17:29:43 sounds like fun May 17 17:31:24 i suck at packing.forgot to pack the glass top for the coffee table and stuff May 17 17:31:33 they charge extra for packing, and even more for glass for liability reasons May 17 17:31:35 heh May 17 17:31:38 my first time dealign with movers May 17 17:31:53 movers are shitheads May 17 17:34:25 they have your stuff and they know it. May 17 17:34:48 laters guys May 17 17:34:55 heh. nothing of any real value is going with them. computer equipment, knives, etc goes in the car May 17 17:34:58 later May 17 17:35:05 night :) May 17 17:35:25 Spyro: some movers are good May 17 17:35:43 Spyro: I have even had some that arranged the furniture, the same way in the new place as the old May 17 17:35:52 Spyro: even the silverware drawer was the same. May 17 17:36:03 france: lucky you :) May 17 17:36:20 http://www.smartphonetoday.com/articles/2005/4/2005-4-4-FCC-Spills-iPAQ.html May 17 17:36:34 A friend of my dads had thieves who did that... stole a table and layed out the tablecloth and ornaments onthe floor just like it was on the table May 17 17:36:38 Spyro: even re-hooked up the tivo and vcr. May 17 17:36:48 heh May 17 17:36:53 neatness does count. May 17 17:36:56 :) May 17 18:10:18 * chouimat is away: tv and pop corn May 17 18:25:03 hi May 17 18:57:37 milan vs liverpool ,next week , so expected May 17 19:35:56 hi hrw May 17 19:36:46 hi kergoth May 17 19:37:05 are you here May 17 20:45:13 hi May 17 20:45:18 anybody here May 17 20:45:35 no May 17 20:46:23 ..... May 17 20:46:29 kergoth::) May 17 20:51:08 "It appears that handhelds.org it being hosted with all the proper approvals in May 17 20:51:08 place, but openembedded.org is considered a unauthorized site. On midnight May 17 20:51:08 of May 31st, CRL will no longer be able to host openembedded.org, but we can May 17 20:51:08 host oe.handhelds.org and/or openembedded.handhelds.org." May 17 20:51:10 hmmmm May 17 20:51:26 guess i should look into moving the content to berlios or something May 17 20:52:05 kergoth:i use "bitbake bootstrap-image", it generate rootfs in tmp/. i want to add libs i need to tmp/rootfs/lib/.how to do it? May 17 20:52:43 eh wtf? " unauthorized site" what's that supposed to mean? May 17 20:52:47 the easiest way is to add the package to BOOTSTRAP_EXTRA_DEPENDS and BOOTSTRAP_EXTRA_RDEPENDS, then clean task-bootstrap and rebuild bootstrap-image May 17 20:53:01 CoreDump|home: i think they just want everything they host to be on their domain for some reason May 17 20:53:13 which is somewhat understandable May 17 20:53:16 ah i c May 17 20:53:57 later May 17 20:55:16 kergoth,:i have done like you said except cleaning task-bootstrap.and it is no use May 17 20:55:29 'except' being the key word May 17 20:55:32 i didnt say it for no reason. May 17 20:55:39 those steps are all critical to it working May 17 20:55:44 oh May 17 20:55:57 kergoth: i could provide hosting, but i think you want something with more then 1 server? May 17 20:55:59 bitbake -c clean task-bootstrap May 17 20:56:25 kergoth:thanks .let me try May 17 20:56:50 mithro: dunno, actually. i have no idea what our traffic is like May 17 21:00:12 kergoth: if it's under 100g per month I can do hosting if people want (can provide shell accounts etc) I'm sure there are most probably better offers however (ie more then just 1 server :P) May 17 21:01:17 if you want complete root access i can provide a UML May 17 21:07:51 uptime has generally been pretty good (only 2 down times in the last 8 months) May 17 21:12:18 I can provide some traffic info on a oz.pdai.org but that's soley used for feed purposes May 17 21:12:54 can probably get info from hh about it May 17 21:14:38 =-) sup kergoth May 17 21:14:44 nada May 17 21:14:51 movers finally came by and got my shit, so i'm finally moving May 17 21:15:36 Yet Again (tm) May 17 21:15:44 out of Texas this time or? May 17 21:15:47 yep, back to MN May 17 21:18:26 I've been out of the loop talking to you.. what happened at this last job? May 17 21:18:31 I knew about the TI mess May 17 21:18:50 kergoth: run into the screw where the movers won't touch half your stuff? May 17 21:19:03 one of my coworkers told a horror story about that recently May 17 21:19:20 long list of things the movers won't touch because of liability concerns May 17 21:19:21 Pendalar: just wasnt the right job for me, so after a year and a half of hating it, i quit it and decided to move back to mn, to my house, and take some time to figure out what i want to do May 17 21:19:30 didnt hit that problem May 17 21:19:36 course i dont really own much of value, so.. May 17 21:19:40 :) May 17 21:19:54 heh May 17 21:20:24 what were you doing for a career at MN before you left? May 17 21:21:00 it included stuff like any glass, so they basically wouldn't touch any of the dishware, stuff like that May 17 21:21:01 unix tech support for Digi International May 17 21:21:40 gb2: jeeze, thats terrible. i had them pack my glass, like the seperate top of my coffee table and stuff. because they packed it rather than me, itll cost quite a bit for the extra insurance for the liability concerns May 17 21:21:45 so its just pricey, they still do it May 17 21:22:12 i think my coworkers just picked movers poorly May 17 21:22:26 another coworker is moving to california this weekend May 17 21:22:39 he and his girlfriend are doing the pod thing May 17 21:22:54 their "pod" was delivered last weekend, and they've been packing it full of stuff May 17 21:23:14 once they find an apartment in california, the pod will appear again May 17 21:23:31 seems like a really convenient way of doing it May 17 21:23:51 storage for upto 3 months or something is basically free May 17 21:23:55 nice May 17 21:24:06 the place i went with was really cheap, quite a lot cheaper than pods, they specialize in small moves May 17 21:24:13 people with one room, or just a small apartmetn or whatever May 17 21:24:27 was cheaper to pay them to move me than it wouldve been to rent a one way uhaul and bring the shit myself May 17 21:24:30 especially useful for moving somewhere new without a place in advance May 17 21:24:36 yeah, thats cool May 17 21:24:45 "Take care of my computer, it's all I have left" ~ Kergoth, with a tear in his eye May 17 21:24:48 the place i went through has storage too for doing that sort of thing, but its definately not free May 17 21:24:50 Pendalar: hehe May 17 21:24:59 computer, knives, clothes, and bedding comes with me in the car May 17 21:25:07 which is basically the stuff i actually give a shit about May 17 21:25:22 the pod thing when moving between cities, most of the cost is the transportation May 17 21:25:34 * kergoth nods May 17 21:25:45 yeah, my friends are moving all their computers with them, never trust harddrives to movers May 17 21:26:41 I hope I can get something better than my lab job soon May 17 21:27:00 getting tired of the bitchy women who are over me.. and haven't a clue how to do their own jobs, let alone mine May 17 21:27:09 what sort of job are you looking for? May 17 21:27:54 I assume that was directed at kergoth..? May 17 21:28:15 i'd assume it was you, since you were talkinga bout finding something better than your lab job May 17 21:28:18 i'm not even job hunting yet May 17 21:28:43 Pendalar: that was directed at you looking for something better than your lab job May 17 21:29:09 ah May 17 21:29:21 well.. something more computer related hopefully.. less chemical May 17 21:29:36 in the end it's almost a glorified QA lab position May 17 21:30:01 but I haven't finished my degree yet and my local area is hicksville =-/ May 17 21:30:15 so what little I'd care for within 30 miles, I don't have the degree for May 17 21:30:40 ah May 17 21:32:25 heh I'd love to be into game developing ultimately May 17 21:32:36 which even with a degree isn't a good market in the US May 17 21:32:44 unless I want to translate =-e May 17 21:34:05 i remember people considering that when they were considering a graphic arts focus in their CS degrees May 17 21:35:00 one of my friends went on to work for ILM though, which is pretty cool, she commented on the game industry at the time being rediculously competetive May 17 21:36:30 well and too you have to think 5-10 yrs back every highschool and college was plugging CIS/CS as the end all next gen degree May 17 21:36:47 then you get knee deep in it and realize you've learned nothing so far .. and the market's floating to India May 17 21:36:52 so what's the bloody point? May 17 21:38:47 to that end, you start thinking of mixing EET and CIS.. hoping to do both sides of the 'ware May 17 21:48:15 haha, I read that as CSI May 17 21:57:15 that would be more fun assurdly May 17 21:57:17 night guys **** ENDING LOGGING AT Tue May 17 23:59:56 2005