**** BEGIN LOGGING AT Thu Nov 15 02:59:57 2007 Nov 15 07:50:25 morning Nov 15 07:58:04 * * OE Bug 3301 has been created by jeremy.laine(AT)bolloretelecom.eu Nov 15 07:58:06 * * sqlite3_3.4.1 : source tarball is no longer available Nov 15 07:58:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3301 Nov 15 08:02:13 hrm Nov 15 08:03:24 I can build expat just fine via `bitbake expat`, but when I add it to my image then do_rootfs fails: "Cannot find package expat" Nov 15 08:03:43 yet it is in the deploy directory Nov 15 08:03:54 does anyone have any idea what's going on there? Nov 15 08:22:40 anyone here generating an image that includes expat? Nov 15 08:38:26 gilligan_: expat is a library Nov 15 08:38:35 gilligan_: if you need it, include a program that links to it Nov 15 08:40:36 koen: oh.. I had no idea about that.. well I did know that its a library, but I didn't know that you'd have to handle libraries like that.. why is that the case? Nov 15 08:40:55 because including libs without programs makes no sense Nov 15 08:41:29 and it's not like you are going to LD_PRELOAD libexpat :) Nov 15 08:43:58 koen: can't quite agree on that.. one might want to prepare an image providing certain libraries knowing that they will be needed later on. Which is exactly the case for me right now. In the end a package is a package is a package.. so i find it odd that you can't just install a library like you install any other package to an image Nov 15 08:44:33 "if they are needed later on" implies you are going to install a package Nov 15 08:44:42 which would pull in the libs automagically Nov 15 08:45:28 no.. you don't quite get the scenario here Nov 15 08:45:41 morning Nov 15 08:45:48 hey hrw Nov 15 08:46:24 i am preparing images or rather dev boards for some other developers that don't have the slightest clue about OE.. so what they need/want ideally is a system with everything installed that's required for their stuff Nov 15 08:46:57 now i have to install some random package requiring expat just for the sake of having expat installed Nov 15 08:47:36 I suspect you aren't going to ship the feed OE generates with the BSP? Nov 15 08:50:02 koen: i haven't thought about that.. I was hoping i could just put together an image with all requirements i've been given that just-works(TM) ;] Nov 15 08:51:01 koen: and as my company has very strict firewall rules fetching via angstrom feed won't work Nov 15 08:51:21 just put deploy/ipk on CD or something Nov 15 08:52:12 is it a necessity that libraries can't be installed on their own or has that just been introduced as a rule? Nov 15 08:53:03 they can be installed on their own in a bit of an awkward way Nov 15 08:53:24 since there isn't a use-case for it that makes sense Nov 15 08:53:41 hello! Nov 15 08:54:29 are there any devs more involved in python packaging than others? Nov 15 08:55:06 only mickeyl is involved in python packaging :) Nov 15 08:55:14 the rest of us just bitch about it Nov 15 08:55:57 ah.. ok :) Nov 15 08:56:09 koen: any linux distribution allows for installation of libraries on their own.. simply because you might want to do something that's not part of the packaging system hence won't lead to automatic installation of the library Nov 15 08:56:41 koen: not saying "any linux distrib" == OE.. just trying to make my point clear somehow Nov 15 08:56:44 brb Nov 15 09:22:26 koen: okay.. so i didn't know about that library handling which explains about expat.. but now i'm having probs with net-snmp .. building the package works fine, but including it in my image bitbake reports "No buildable providers available" Nov 15 09:24:08 03koen 07org.oe.dev * r7bd7c5bb... 10/ (1 contrib/feed-browser/includes/config.inc): feed-browser: start adding a machine -> arch map Nov 15 09:24:49 koen: merging your hacks^Wwork? Nov 15 09:25:12 koen: nevermind.. seems i have to specify net-snmp-client Nov 15 09:25:34 hrw: merging the extra vars Nov 15 09:25:51 hrw: I can't merge the rest since angstrom isn't using the 'type' field yet Nov 15 09:29:23 I wonder will I get time one day and rewrite it totally Nov 15 09:30:37 hi Nov 15 09:32:03 hi Bernardo Nov 15 09:36:04 hello Nov 15 09:36:10 ph5 is not there now ? Nov 15 09:36:20 he committed some changes to wifi-radar Nov 15 09:36:34 but now launching from the console fails Nov 15 09:36:53 import gtk, gobject Nov 15 09:37:01 cannot find module gtkù Nov 15 09:37:06 gtk Nov 15 09:37:09 ??? Nov 15 09:37:35 it's at the end of /sbin/wifi-radar Nov 15 09:37:53 will try later, now have to work ! Nov 15 09:37:56 See ya Nov 15 10:05:42 ~lart koen badly Nov 15 10:05:43 * ibot grabs a large, mis-shapened log, with squirrels, and beats koen badly until only the nuts remain ... which the squirrels run off with Nov 15 10:05:52 11:04 hrw@home:openmoko$ grep kernel-modules packages/images/* Nov 15 10:05:52 packages/images/x11-gpe-image.bb: kernel-modules \ Nov 15 10:05:52 packages/images/x11-office-image.bb: kernel-modules \ Nov 15 10:05:52 packages/images/x11-pimlico-image.bb: kernel-modules \ Nov 15 10:06:00 WHY? Nov 15 10:06:41 because those are the 'big' images Nov 15 10:06:52 really? Nov 15 10:07:00 I wanted one of them on tosa Nov 15 10:07:09 tosa does not have 26M for rootfs Nov 15 10:07:20 so use x11-image Nov 15 10:07:40 or create another image :) Nov 15 10:07:41 I do not want bare x11-image Nov 15 10:09:04 * * OE Bug 3302 has been created by tobias.pflug(AT)gmx.net Nov 15 10:09:06 * * expat fails to install to image Nov 15 10:09:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3302 Nov 15 10:09:10 mtn rename x11-gpe-image.bb x11-gpe-all-kernel-modules-image.bb;sed -e 's/kernel-modules//' x11-gpe-all-kernel-modules-image.bb> x11-gpe-image.bb; ;mtn add x11-gpe-image.bb;mtn commit . -m"x11-gpe-image: unbreak Koen thinking";mtpush Nov 15 10:09:17 * * OE Bug 3302 has been RESOLVED (FIXED) by tobias.pflug(AT)gmx.net Nov 15 10:09:19 * *  expat fails to install to image Nov 15 10:09:21 * * http://bugs.openembedded.org/show_bug.cgi?id=3302 Nov 15 10:09:28 developer handbag fight :-D Nov 15 10:09:29 * * OE Bug 3303 has been created by jeremy.laine(AT)bolloretelecom.eu Nov 15 10:09:30 * * python-crypt 2.5 has an incorrect FILES definition Nov 15 10:09:32 * * http://bugs.openembedded.org/show_bug.cgi?id=3303 Nov 15 10:09:51 I 'love' how those images look.. Nov 15 10:10:05 pimlico image which gets hal (not used) Nov 15 10:10:44 * hrw probably spent too much time with Poky Nov 15 10:12:55 21M sounds better Nov 15 10:19:25 http://www.flickr.com/photos/koenkooi/2031247323/in/pool-565582@N21/ Nov 15 10:26:51 nice Nov 15 10:33:56 you can even see a zaurus in that picture :) Nov 15 10:36:07 koen: That isn't messy, you should see mine ;-) Nov 15 10:36:36 :) Nov 15 10:37:03 RP: want to add a Poky on Android howto on www.android-internals.org ? Nov 15 10:40:08 "DalvikVM is ~28x slower than optimized C (for this particular test), Nov 15 10:40:09 while the JDK interpreter is ~40x slower." Nov 15 10:45:55 koen: Where? :) Nov 15 10:46:11 http://groups.google.com/group/android-developers/msg/69729d17c013452f?dmode=source Nov 15 10:46:22 Thanks Nov 15 10:46:50 speed kills Nov 15 10:57:26 http://dag.wieers.com/blog/content/obex-for-dummies Nov 15 11:00:32 hrw: we have gnome-vfs-obexftp for that in OE Nov 15 11:01:06 only IF application use gnomevfs Nov 15 11:01:35 03hrw 07org.oe.dev * r5fc01752... 10/ (6 files in 3 dirs): sqlite: added 3.4.2 from Poky - close #3301 Nov 15 11:01:49 fsck.. I killed that rev locally Nov 15 11:01:50 and fuse only works as root and your kernel supports fuse Nov 15 11:02:29 looks like I need to merge 3.5.2 now Nov 15 11:14:29 03hrw 07org.oe.dev * rbabd6607... 10/ (3 files in 2 dirs): sqlite3: upgrade to 3.5.2 - now really close #3301 Nov 15 11:17:23 http://spluch.blogspot.com/2007/11/toilet-themed-restaurant.html Nov 15 11:21:00 koen: btw - why not use koen@ltg as email? Nov 15 11:23:01 hrw: thanks for sqlite3 Nov 15 11:23:26 I had it locally Nov 15 11:23:54 I wasn't sure whether we'd want to stay at 3.4.1 to avoid possible breakeage or whether upgrading to 3.5.2 was an option Nov 15 11:25:09 what timezone is mickeyl on? Nov 15 11:25:19 CET Nov 15 11:25:38 OK, will try to get hold of him for some python bugfixes Nov 15 11:26:12 hrw: because that adds half a day to mail delivery due to greylisting storm Nov 15 11:26:45 i know this isnt the right channel to ask, but on linux 2.6 i can link a cpio archive directly into the kernel, does the 2.4 series kernel provide a similar feature? or a patch that adds this feature? Nov 15 11:30:32 ~hail lumag for 2.6.23 Nov 15 11:30:33 * ibot bows down to lumag for 2.6.23 and chants, "I'M NOT WORTHY!!" Nov 15 11:31:17 hrw: http://kernelnewbies.org/Linux_2_6_24#head-e1c064a53027501a877ab807c99ad455cc0cf248 would be nice for people using flash + sd/cf Nov 15 11:33:55 http://pastebin.ca/774732 is sweet Nov 15 11:38:17 koen: indeed Nov 15 11:38:20 hrw: your madwifi rfc sounds like a good idea Nov 15 11:42:12 I got hit by it with alix Nov 15 12:22:59 dpkg-deb: building package `libglib-2.0-locale-de' in `/OE/build/tmp-new/angstrom/deploy/glibc/deb/armv5te/libglib-2.0-locale-de_2.14.0-r2_arm.deb'. Nov 15 12:24:13 where do I define psplash=false? Nov 15 12:24:32 in the kernel CMDLINE Nov 15 12:24:48 03koen 07org.oe.dev * r7a2606dc... 10/ (1 packages/glibc/glibc-intermediate_2.7.bb): glibc-intermediate: first stab at a recipe for 2.7 Nov 15 12:25:04 * * OE Bug 3301 has been RESOLVED (FIXED) by Nov 15 12:25:06 * *  sqlite3_3.4.1 : source tarball is no longer available Nov 15 12:25:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3301 Nov 15 12:25:22 koen: specifically, which file for tosa Nov 15 12:25:35 there are multiple files :) Nov 15 12:25:47 tosa.conf, linux-rp.inc, linux-rp_2.6.23.bb, etc Nov 15 12:26:33 old problem, the tosa x11-image won't boot past the last 2% of the Angstrom splash screen Nov 15 12:27:06 I remember this was solved before by disabling the psplash Nov 15 12:27:21 I doubt that Nov 15 12:27:55 the 'fix' people reported was 'disable psplash ; start X by hand' Nov 15 12:27:59 which is not a fix Nov 15 12:28:29 you are right. disable psplash was suggested by hrw Nov 15 12:28:43 but I don't know how it was fixed later Nov 15 12:28:55 I'd be more interested in the X log attached to a bug report :) Nov 15 12:29:18 I'm 99% sure it's a gpe problem which can be solved quite easily Nov 15 12:32:14 -rw-r--r-- 1 koen koen 8292929 Nov 15 13:28 Angstrom-console-image-glibc-deb-2008.1-test-20071115-a780.rootfs.tar.gz Nov 15 12:32:16 sweet Nov 15 12:32:57 ~hail openened hand for their .deb work in OE Nov 15 12:32:58 * ibot bows down to openened hand for their .deb work in OE and chants, "I'M NOT WORTHY!!" Nov 15 12:33:30 I wonder how hard it would be to get packageKit working in OE for package_deb.bbclass Nov 15 12:34:26 wait... sorry, it boots after reset Nov 15 12:34:34 koen: If I disable the manual autogen.sh call in the navit recipe do_configure breaks: http://nopaste.com/p/aTswA77wj Seen before? Nov 15 12:34:45 koen: I'll have a look at the autogen.sh file Nov 15 12:37:08 stefan_schmidt: I've seen it before Nov 15 12:37:16 it's called 'broken autotools usage' :( Nov 15 12:37:24 koen: hmm, it just calls autoreconf Nov 15 12:37:27 koen: http://navit.cvs.sourceforge.net/navit/navit/autogen.sh?view=markup Nov 15 12:39:36 what is inside the intl/Makefile.in when running autoreconf? Nov 15 12:40:54 koen: Good question, let me check Nov 15 12:41:02 stefan_schmidt: EXTRA_AUTORECONF = " -I m4" Nov 15 12:41:04 try that Nov 15 12:41:06 xjqian: The OpenMoko Distribution -- powered by Angstrom tosa ttyS0 Nov 15 12:43:58 hrw: what do you mean? Nov 15 12:46:11 koen: Congrats, that was the trick Nov 15 12:46:26 koen: I clean it up a bit more Nov 15 12:46:40 stefan_schmidt: run it thru oe-stylize.py Nov 15 12:46:54 xjqian: ops - had you nick written during paste Nov 15 12:46:58 sorry Nov 15 12:48:06 hm.. i am trying to package mod_python which currently fails building because `mod_include.h` is missing. I think that file should be part of apache2-dev ? I added apache2 and apache2-native to DEPENDS.. anyone got any hints for me? Nov 15 12:48:12 re Nov 15 12:50:48 gilligan_: you need to add a do_stage() {} method to the apache2 recipe Nov 15 12:51:43 Hey folks, do we have anybody in the South Carolina area of the US? Nov 15 12:53:00 koen: in which I then install the header files? Nov 15 12:54:46 http://blog.haerwu.biz/2007/11/15/2623-on-tosa/ Nov 15 13:00:46 koen: so what is the standard procedure with the installation of headers? is apache2 "missing" header installation or ... Nov 15 13:01:03 drmikecrowe : Crofton is in Virginia Tech if i remember correctly Nov 15 13:01:53 steliosk: Sounds good. We have just released a device we want ported to OE, and I'm looking for (reasonably local) resources Nov 15 13:02:24 03hrw 07org.oe.dev * r49e3b25c... 10/ (8 files in 8 dirs): base-files: disable /dev/mmcda1 for all Zaurus machines - we use 2.6 not 2.4 Nov 15 13:02:33 drmikecrowe: what type of device? Nov 15 13:02:42 hrw: http://www.linuxdevices.com/news/NS5936183381.html Nov 15 13:02:57 drmikecrowe: which one Nov 15 13:03:32 hwr -- doh, the Zipit instant messenger Nov 15 13:03:39 Ok, asking again about that issue with dependancies and things gone missing. Nov 15 13:03:41 I thought that link took you to ours Nov 15 13:04:00 drmikecrowe: this link is top10 gadgets Nov 15 13:04:05 I want to post a bug report, though it seems strange that so many applications would be uncleanly removed. Nov 15 13:04:06 http://www.linuxdevices.com/news/NS5942157201.html Nov 15 13:04:18 * drmikecrowe is sorry for the confusion :) Nov 15 13:04:41 drmikecrowe: whic wifi chipset did you used? Nov 15 13:04:46 Marvel Nov 15 13:04:54 drmikecrowe : i think there is some "preliminary" support for the zipit, as i was looking at it for a project a week ago Nov 15 13:05:04 drmikecrowe: 8385 or other? Nov 15 13:05:13 I only help people in SC if they vote for Stephen Colbert .... Nov 15 13:05:16 hrw -- I need to look at the schematic Nov 15 13:05:18 where in SC? Nov 15 13:05:26 Crofton: Yeah, I'm going to for him just for giggles Nov 15 13:05:31 Greenville Nov 15 13:05:34 awesome Nov 15 13:05:43 Home of George Hincapie Nov 15 13:05:50 I have family in Charleston Nov 15 13:05:52 steliosk: That was the old revision based on the Cirrus Logic arm. Nov 15 13:06:01 Awesome area, I love it down there Nov 15 13:06:10 koen: What is about this one? http://bugs.openembedded.org/show_bug.cgi?id=3299#c2 Nov 15 13:06:32 drmikecrowe : Crofton is your man Nov 15 13:07:01 what are we talking Nov 15 13:07:07 Crofton: Mind if I pm you? Nov 15 13:07:11 sure Nov 15 13:08:32 drmikecrowe: mini-sd instead of microsd? Nov 15 13:08:42 hrw: yeah Nov 15 13:08:44 drmikecrowe: any discounts for developers? :D Nov 15 13:09:02 hrw: For the right ones, you bet Nov 15 13:09:18 zipit, Aeronix? Nov 15 13:09:34 koen: I also got the confirmation that -fPIC can just be removed. Nov 15 13:10:31 openzipit.org ha(d)(s) some OE files, but the dude was extremely uninterested in getting it into OE Nov 15 13:10:54 koen: it was for Z1 Nov 15 13:13:24 right Nov 15 13:15:24 koen: If you are fine with this version of the recipe I would push it. http://nopaste.com/p/a6BCKZdos Nov 15 13:15:40 stefan_schmidt: fix the PV first Nov 15 13:15:53 0.0.2 is the latest release Nov 15 13:15:56 not 0.0.3 Nov 15 13:16:12 koen: grr, right. Overseen that Nov 15 13:16:18 http://www.openembedded.org/versioningpolicy Nov 15 13:18:07 koen,crofton,florian,mickeyl : I talked to zecke yesterday and as he is swamped with work, we decided to take over the matter of refunds for the travel expenses, so i would appreciate if you could send me all the invoices you have for refund at skoroneos at digital-opsis dot com Nov 15 13:19:15 steliosk: oh, ok - yes, i'll do this - do you want a scan/pdf or whatever from the original one? Nov 15 13:19:38 koen: NOTE: package navit-0.0.2+cvs20071115: completed Nov 15 13:19:46 florian : a pdf would be fine Nov 15 13:21:22 * florian digs in a pile of travelling stuff Nov 15 13:21:23 ~change 150 usd to eur Nov 15 13:21:27 hallo Nov 15 13:21:28 i got some error trying that Nov 15 13:21:50 ~30 :p Nov 15 13:21:59 hi antvr Nov 15 13:22:09 100 eur + 22% vat + posting Nov 15 13:22:18 150 eur probably Nov 15 13:22:35 hi florian Nov 15 13:22:56 hi stefan_schmidt Nov 15 13:23:08 BTW did you compile speechd Nov 15 13:23:11 ? Nov 15 13:23:15 for navit Nov 15 13:23:31 wifi-radar also uses it Nov 15 13:23:36 antvr: Not yet Nov 15 13:23:41 we already have flite Nov 15 13:23:51 is it enough or do we need festival Nov 15 13:23:55 bloat Nov 15 13:23:56 antvr: I like to start with a working supset Nov 15 13:24:04 i agree 100% Nov 15 13:24:11 antvr: I not even tried it on my laptop with specchd Nov 15 13:24:19 speechd Nov 15 13:24:25 hrw : the way the dollar falls even the bot has problems to do the calculations :) Nov 15 13:24:52 any amount in dollars is approx 0.10GBP Nov 15 13:25:01 we established that at ELC2007 Nov 15 13:25:25 stefan, about GPS, did you note they released Gpsdrive 2.10pre4 ? Nov 15 13:26:05 antvr: No, but I never was happy with GPSdrive. I like real routing, not only showing maps. Nov 15 13:26:10 on my Zaurus c7x0 the actual 2.10pre2 shows somes gtk glitches... Nov 15 13:26:26 If you happen to have time...thanks in advance! Nov 15 13:27:05 antvr: ATM my GPS interest is navit only. Nov 15 13:27:16 I agree it's the only REAL app Nov 15 13:27:56 but 2 days ago I was driving from Italy trough CH till Frankfurt and Gpsdrive was better than nothing! Nov 15 13:28:25 on the Alps it shows aprecision of 1,6 meters Nov 15 13:28:29 antvr: There are still roadster and roadnav. But navit plays with some bought maps which gives you a nice experience if your country is not covered by OSM/TIGER Nov 15 13:28:31 maemo-mapper rocks Nov 15 13:28:53 koen: But also only map showing. Or did they changed that? Nov 15 13:29:05 it does routing thru google maps Nov 15 13:29:12 and it did not build on c7x0 on my last try Nov 15 13:29:24 koen: With the fixed version fine to push? Nov 15 13:29:26 it gets a .gpx file with instructions from the web Nov 15 13:29:35 stefan_schmidt: yes, push it Nov 15 13:29:39 koen: That means network connection... Nov 15 13:29:58 stefan_schmidt: yeah, or downloading a bunch of gpx files :) Nov 15 13:30:07 koen: It is cool for now, but I like to see real navigation in the future. Nov 15 13:30:10 but I have flatrate gprs in .nl :) Nov 15 13:30:32 koen: And a country with full OSM data. :) Nov 15 13:30:57 I'm waiting from Frosse Reiseplanner from Amazon.de Nov 15 13:31:08 Grosse natürlich! Nov 15 13:31:10 koen: I also have a UMTS flat rate in .de, but sometimes I'm not in my home country. :) Nov 15 13:31:33 antvr: Yeah, I like to buy the newest one, too Nov 15 13:31:47 I ordered the 2007/2008 with Italy Nov 15 13:32:05 the 2008 only was A DE CH only Nov 15 13:33:10 antvr: I could live with that. ;) Nov 15 13:35:26 costs some Eur less ! Nov 15 13:35:57 BTW I think it's in German only Nov 15 13:36:09 I can live with thath !!! Nov 15 13:37:04 Hope to receive the DVD next week. Already shipped by Deutsche Post Nov 15 13:39:06 koen: thanks for supporting me on e-devel Nov 15 13:39:10 morning guys Nov 15 13:39:25 hey mickeyl Nov 15 13:39:38 mickeyl: I have too much exotic hardware on my desk :) Nov 15 13:41:37 03Alessandro 07org.oe.dev * r0c08620c... 10/ (1 packages/navit/navit_cvs.bb): navit_cvs.bb: CVS recipe. Thanks for review from koen. Fixes #3299 Nov 15 13:44:22 03stefan 07org.oe.dev * r7f5c53f2... 10/ (1 packages/navit/navit_cvs.bb): navit_cvs.bb: Remove unused patch from SRC_URI Nov 15 13:48:08 koen: why did you bump openmoko-dialer2 to a broken version? Nov 15 13:48:19 XorA: I did? Nov 15 13:48:33 koen: 3361, its no longer a singleton :-( Nov 15 13:48:36 I could have sworn I tested it on athe neo Nov 15 13:49:11 koen: sorry but I don't quite understand what I am supposed to do with staging the include file etc (mod_install.h / packaging mod_python etc) .. the include file is part of the apache2-dev package which however doesn't help during the build of mod_python Nov 15 13:49:32 koen: and after OM announced 3306 was last working version until Thos finished his work :-D Nov 15 13:49:45 gilligan_: you must implement a do_stage() {} method for apache2 Nov 15 13:49:50 XorA: announced where? Nov 15 13:50:03 openmoko-apps Nov 15 13:50:28 it was also announced that right-click was being fixed Nov 15 13:51:18 heh Nov 15 13:51:25 koen: which is exactly what you told me before and wich isn't exactly of much use for me because i am not sure what the function is supposed to do Nov 15 13:52:07 speaking about that Nov 15 13:52:10 | Collected errors: Nov 15 13:52:10 | ERROR: The following packages conflict with openmoko-session2: Nov 15 13:52:10 | matchbox-common Nov 15 13:52:26 mickeyl: you need to mtn pull Nov 15 13:52:31 koen: just install the header(s) i need to ${STAGING_DIR} ? Nov 15 13:52:31 again? Nov 15 13:52:33 and read the angstrom mailinglists Nov 15 13:53:03 gilligan_: add do_stage() { autotools_stage_all } and test Nov 15 13:53:28 mickeyl: we have a huge problem with RREPLACE, its pretty much a useless operation if more than one DISTRO builds from OE Nov 15 13:54:05 hrw: that was the first thing I did and it didn't change anything unfortunately Nov 15 13:54:25 * XorA goes off to epoch openmoko-dialer2 Nov 15 13:54:33 guys: native libs should have 'installed=yes' or 'no' in their .la files? Nov 15 13:55:13 XorA: ah ok... ~lart ipkg Nov 15 13:55:31 posted a crappy bug report about all those missing deps Nov 15 13:55:36 mickeyl: we can do it as an openmoko only override I suppose Nov 15 13:55:59 mickeyl: effectively saying openmoko will never use matchbox-common Nov 15 13:56:21 XorA: that's ok for me. eventually we really have to clean all the x/wm stuff up a bit Nov 15 13:56:40 it's too hard to support multiple window manager / desktop environment configurations in OE right now Nov 15 13:57:03 mickeyl: OE is just too matchbox oriented Nov 15 13:57:12 yeah Nov 15 13:57:16 we need to improve that a bit Nov 15 13:57:27 danger of 90% of devs employed by OH :-D Nov 15 13:57:31 still hoping we can drag some pdaXrom guys in Nov 15 13:57:40 they have support for more than one wm Nov 15 13:57:57 pdaXrom is slowly dieing so hope some come across, hope there isnt too much bad blood Nov 15 13:58:21 yeah Nov 15 13:58:25 mickeyl: using another wm is real easy, just u-a x-window-manager Nov 15 13:58:47 gpe-session-scripts execs x-window-manager, not matchbox-wm Nov 15 13:59:06 xjqian: framebuffer on tosa is better in 2.6.23 indeed Nov 15 13:59:13 yeah, but matchbox-wm should come with its start script then Nov 15 13:59:34 not having this in a seperate package that is shared by other managers Nov 15 13:59:37 mb-wm doesn't need a start script :) Nov 15 13:59:46 oh it does need matchbox-common Nov 15 13:59:50 which contains /usr/bin/matchbox-foo Nov 15 14:00:23 and actually it doesn't really need matchbox-common. it runs fine with it when you suplpy /usr/bin/matchbox-foo Nov 15 14:00:24 gpe-session-scripts, gpe-login, gpe-dm... Nov 15 14:00:32 s/with/without/ Nov 15 14:00:34 why OE is so gpe oriented too... Nov 15 14:00:47 ;d Nov 15 14:00:49 because gpe was the first X based environment Nov 15 14:00:54 don't worry, we can fix that Nov 15 14:01:07 gpe-dm isn't actually gpe specific Nov 15 14:01:16 it has gpe in name Nov 15 14:01:18 koen: right. which makes me think it should be renamed :) Nov 15 14:01:36 I'm amazed o-hand didn't do their usual sed s:gpe:poky:g ; svn import routine yet Nov 15 14:02:23 mickeyl: matchbox-common only has matchbox-session in bindir, and mb-wm doesn't need that Nov 15 14:02:33 wow, running lttng over startup is a bit of an eye-opener Nov 15 14:02:43 koen: without this bindir, the wm doesn't start Nov 15 14:03:03 since x-window-manager links to it Nov 15 14:03:57 ~hail zecke Nov 15 14:03:58 * ibot bows down to zecke and chants, "I'M NOT WORTHY!!" Nov 15 14:04:11 :D Nov 15 14:04:32 * XorA switchs OM to evilWM Nov 15 14:04:59 XorA: Just use ion3 :D Nov 15 14:05:18 stefan_schmidt: thats got a crapper name :-) Nov 15 14:05:28 :) Nov 15 14:05:33 matchbox-common was designed for use with matchbox-wm. If you want to replace mb-common, fine you need to ensure the replacement does everything mb-common does... Nov 15 14:05:37 mickeyl: ah, I see the problem Nov 15 14:06:23 the mb-wm recipe was doing something illegal Nov 15 14:06:32 u-a-ing to another package Nov 15 14:06:56 AFAIK in OE you're only allowed to u-a to yourself Nov 15 14:07:42 Just keep in mind mb has been in OE for a long time (really from before OH was around), its bound to need some cleanup :) Nov 15 14:08:19 koen: matchbox-session? Nov 15 14:08:26 yes Nov 15 14:09:35 RP: pushing the fixes now Nov 15 14:10:27 koen: I have a ton of bugfixes in Poky for nasty cross compiling path issues in Poky building up... Nov 15 14:11:03 koen: Removing the -L and -I options from bitbake.conf CFLAGS and LDFLAGS has let lots of bugs out the woodwork... Nov 15 14:11:11 well, I wanted to get those merged, but hrw said I should wait for the poky autobuilders or something Nov 15 14:11:27 koen: Feel free to merge, last night's autobuild worked Nov 15 14:11:29 Eh Nov 15 14:11:33 NOT ION3 Nov 15 14:11:41 hi all. is this the place for a newbie ? Nov 15 14:11:57 Quick, run! Nov 15 14:12:25 justri: Depends what kind of newbie ;-) Nov 15 14:13:11 rp: i'm trying to compile lmsensors-apps, but it fails (no provider for libsysfs). ist this a known problem? Nov 15 14:13:44 RP, probably what I'm trying to report. Nov 15 14:14:39 !oebug 3304 Nov 15 14:14:41 * * Bug 3304, Status: NEW, Created: 2007-11-15 05:46 Nov 15 14:14:41 * * B_Linuz(AT)yahoo.com: Bitbake complains about missing deps due to changes to the repo. Nov 15 14:14:42 * * http://bugs.openembedded.org/show_bug.cgi?id=3304 Nov 15 14:14:50 Probably not the best bug report ever. Nov 15 14:15:48 RP: should I just merge the staging layout change, or send a mail to oe-devel first? Nov 15 14:16:18 koen: We need to prepare things first i.e. merge all the fixes from poky Nov 15 14:17:03 can I somehow add a compile flag to a package using autotools ? I need to manually add -I{STAGING_INCDIR} Nov 15 14:17:04 koen: OE will need a good sedding to remove references of things like ${STAGING_DIR}/${HOST_SYS} Nov 15 14:17:48 justri: edit recipe and change libsysfs to sysfsutils Nov 15 14:17:54 B_Lizzard: It really needs some patches to fix things ;-) Nov 15 14:18:01 justri: or wait 10 minutes and pull/update Nov 15 14:18:04 yeah Nov 15 14:18:08 hey all, is there something more up to date than http://www.openembedded.org/wiki/PackagedStaging? Nov 15 14:18:27 gilligan_: extend CFLAGS ? Nov 15 14:18:37 gilligan_: EXTRA_CFLAGS += Nov 15 14:18:57 hrw: thanks a lot! Nov 15 14:19:16 03koen 07org.oe.dev * re89c8857... 10/ (1 packages/matchbox-wm/matchbox-wm_1.2.bb): mb-wm: u-a to itself, not another package Nov 15 14:19:20 03koen 07org.oe.dev * rb21b578d... 10/ (1 packages/matchbox-common/matchbox-common_0.9.1.bb): mb-common: use u-a to make the link to x-wm, do not depend on mb-wm to do it Nov 15 14:20:19 hrw: ah good.. will try that Nov 15 14:20:54 RP: I can fix these, though it might take time :) Nov 15 14:21:27 03xora 07org.oe.dev * r901d93e1... 10/ (3 files in 3 dirs): Nov 15 14:21:27 openmoko-dialer2_svn.bb, sane-srcrevs.inc : 3306 really is the last Nov 15 14:21:27 working version of dialer for a while. Nov 15 14:22:50 XorA: do we need to catch up in anything w/ the mb-wm and mb-common ow? Nov 15 14:22:53 s/ow/now/ Nov 15 14:23:35 hrw: seems like that is not passed through Nov 15 14:24:23 mickeyl: you should probably u-a openmoko-session to x-wm if you want to stay bug compatible :) Nov 15 14:26:48 hmm Nov 15 14:28:00 We really shouldn't link session scripts to x-wm, though Nov 15 14:28:06 yes Nov 15 14:28:09 that's what i was thinking Nov 15 14:30:09 koen@lieve:/OE/org.openembedded.dev/packages$ grep '${STAGING_DIR}/${HOST_SYS}' . -rn | wc -l Nov 15 14:30:09 108 Nov 15 14:30:23 koen@lieve:/OE/org.openembedded.dev/packages$ grep '${STAGING_DIR}/${HOST_SYS}' . -rn | grep qt |wc -l Nov 15 14:30:23 34 Nov 15 14:30:56 mickeyl: I think that should be us fine Nov 15 14:32:04 RP: s:${STAGING_DIR}/${HOST_SYS}:${STAGING_NATIVE_DIR}:g ? Nov 15 14:33:08 florian: did Jose Manrique Lopez de la Fluente ever contribute to GPE? Nov 15 14:35:31 fix-configure.patch no longer applies to evas_cvs Nov 15 14:35:48 and is probably not needed Nov 15 14:35:55 zecke: I'm not sure if he produced any code, but he was present from time to time and we discussed a lot of stuff. Why do you ask? Nov 15 14:36:07 as evas seems to be building fine Nov 15 14:37:34 florian: I wonder why he thinks he deserves special treatment by mailing me privately :) Nov 15 14:37:41 hm..doesn't matter what I do.. I don't get the additional include statement to be put through to the autotools build Nov 15 14:38:03 zecke: uh? Nov 15 14:39:30 evas_cvs builds fine without the patch Nov 15 14:39:51 So please remove that from evas.inc Nov 15 14:41:17 B_Lizzard: no Nov 15 14:41:23 Ok :) Nov 15 14:41:26 B_Lizzard: submit a patch to add maxdate to patch Nov 15 14:41:28 B_Lizzard: it needs maxdate added rather Nov 15 14:41:36 Why? Nov 15 14:41:43 Since it builds fine without the patch Nov 15 14:41:44 :/ Nov 15 14:41:49 because I want to build it for 20061213 date Nov 15 14:41:58 Then do that :) Nov 15 14:42:00 B_Lizzard: and it will not build without it Nov 15 14:42:02 Not everyone wants that Nov 15 14:42:15 Or just comment it out, whatever. Nov 15 14:42:24 koen: No, HOST_SYS is STAGING_DIR_HOST Nov 15 14:42:53 I just thought that telling you here is faster than clogging up the bugtracker Nov 15 14:43:04 B_Lizzard: proper way is adding maxdate to patch Nov 15 14:43:09 koen: but its not as simple as that, each use case needs to be checked to see if it still means that or if it means ${STAGING_DIR_HOST}${layout_prefix} or something else entirely Nov 15 14:43:34 Ok... Nov 15 14:44:10 I'll search how to do that and give you a patch or whatever Nov 15 14:44:12 :) Nov 15 14:44:47 please file a bugreport instead of trying to push patches on IRC Nov 15 14:44:51 B_Lizzard: s/patch=1/patch=1;maxdate=YYYYMMDD/ where YYYYMMDD is date when it was usable last time Nov 15 14:45:01 Oh Nov 15 14:45:02 OK Nov 15 14:45:05 Thanks a lot :) Nov 15 14:45:19 time to RFC 'only patches that went into bugzilla may be applied' Nov 15 14:46:03 Yeah, but isn't that a bit bureaucratic? Nov 15 14:46:08 mickeyl: what do you think about adding PROVIDES="python-pygtk2" into python-pygtk? Nov 15 14:46:17 B_Lizzard: it is Nov 15 14:46:22 :/ Nov 15 14:46:23 B_Lizzard: but we work that way Nov 15 14:46:30 I guess I'll have to live with it. Nov 15 14:46:31 it gives people the chance to review patches Nov 15 14:46:43 so we can stop stuff like DEPENDS = "libiconv" getting in Nov 15 14:46:58 shouldn't $STAGING_INCDIR be exported in a devshell session ? Nov 15 14:46:58 Hum, what was the last date it worked? Nov 15 14:47:04 hehe Nov 15 14:47:48 hrw: ok w/ me Nov 15 14:47:51 Cause I also need to fix all those damn dangling dependancy problems Nov 15 14:48:10 And post a proper patch Nov 15 14:49:04 RP: http://www.openembedded.org/~koen/OE/sed-my-staging.diff.txt Nov 15 14:49:20 B_Lizzard: few deps errors are already fixed ;) Nov 15 14:49:28 B_Lizzard: pull/update and restart Nov 15 14:49:33 OK Nov 15 14:49:40 and file one bug per package Nov 15 14:49:45 we hate mega-bugs Nov 15 14:50:05 * * OE Bug 3299 has been RESOLVED (FIXED) by stefan(AT)datenfreihafen.org Nov 15 14:50:07 * *  cvs recipe for navit Nov 15 14:50:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3299 Nov 15 14:50:17 * * OE Bug 3304 has been created by B_Linuz(AT)yahoo.com Nov 15 14:50:19 * * Bitbake complains about missing deps due to changes to the repo. Nov 15 14:50:21 hehe Nov 15 14:50:21 * * http://bugs.openembedded.org/show_bug.cgi?id=3304 Nov 15 14:50:24 hrw: any probs with ettercap-ng? Nov 15 14:50:26 Took it a bit Nov 15 14:50:43 mickeyl: http://bugs.openembedded.org/show_bug.cgi?id=3303 - can you? Nov 15 14:50:51 dcordes: I do not run it Nov 15 14:51:07 OK, I'll post a bug and a fix for each package I do. Nov 15 14:52:11 Are missing packages to be considered removed? Nov 15 14:52:22 hrw: please wait. jeremy is revamping this atm in a broader scope Nov 15 14:52:28 opie-usermanager no longer exists, for instance. Nov 15 14:52:31 hrw: he also just applied for commit access Nov 15 14:52:35 so I'd like him to do that Nov 15 14:52:50 koen: No, needs more attention than that Nov 15 14:53:19 RP: I know, but anything that now has '${STAGING_DIR_HOST} ' can get committed, right? Nov 15 14:53:20 mickeyl: I do not touch python in OE - I point you to bugs rather Nov 15 14:53:26 hrw: sure. Nov 15 14:53:35 koen: --with-aspell-prefix=${STAGING_DIR_HOST} should be ${STAGING_DIR_HOST}/${layout_prefix} Nov 15 14:53:42 B_Lizzard: opie-usermanager was removed Nov 15 14:53:58 iirc Nov 15 14:53:59 ${STAGING_DIR}/${HOST_SYS}/share -> ${STAGING_DATADIR} Nov 15 14:54:04 Uh huh Nov 15 14:54:20 libtheora also depends upon itself, btw Nov 15 14:54:21 :) Nov 15 14:54:24 koen: At the moment we have something easy to grep for and I'd prefer to leave it like that and fix it properly Nov 15 14:54:59 And fails to build Nov 15 14:55:04 B_Lizzard: and has crappy recipe Nov 15 14:55:13 Yeah, but builds anyways. :) Nov 15 14:55:55 Whoa Nov 15 14:56:00 It really *is* crappy. Nov 15 14:56:29 B_Lizzard: will you provide nicer one? Nov 15 14:56:44 I think http://www.infradead.org/~dwmw2/email.html should be taught in school Nov 15 14:56:53 Well, it probably needs updating and some DESCRIPTION and stuff Nov 15 14:57:05 I'll look into newer versions, see how those build Nov 15 14:57:23 Does anything actually *depend* upon libtheora? Nov 15 14:58:18 It's not in the DEPENDS of any package Nov 15 14:59:14 Actually, only linphone and libxine *might* be using it. Nov 15 14:59:36 Is the online repository browser down or just being awfully slow? Nov 15 15:00:54 I'm having a touchscreen problem on a non-OE distribution and would like to look for possible solutions there... Nov 15 15:01:53 DataBeaver: and using OE is not the answer you want to here? Nov 15 15:02:05 DataBeaver: use the tslib test utilities Nov 15 15:02:07 morning Nov 15 15:03:31 zecke: The problem is in the interaction between tslib and the Xorg tslib driver. Specifically, due to the variance module holding up samples, I'm not getting ButtonRelease events until the next tap on the screen. Nov 15 15:04:04 DataBeaver: but that is Xorg question then :) Nov 15 15:04:39 DataBeaver: and if you have analysed that far you probably can think of two solutions already? Nov 15 15:06:08 So far I've thought of (1) asking ts_read for more than one sample at a time or (2) implementing a select-read-loop in the X driver. However, it turns out that ts_read can't handle partial reads very well and the X server already does selecting on the fd. Nov 15 15:06:33 I could of course patch tslib so that it does support partial reads, after which option (1) would work. Nov 15 15:10:54 The Kdrive tslib driver seems to just call ts_read as long as it gets events, which would suggest that either it needs a patched tslib or there's still some details I'm not aware of... Nov 15 15:16:17 Ah, there we go. The ts_open function takes a nonblock parameter, and if I enable that, I can use while(ts_read(...)==1). Nov 15 15:17:30 Thanks for tolerating me with this, I won't bother you more for now. Nov 15 15:17:45 hehe you are welcome... ;) Nov 15 15:34:04 * * OE Bug 3293 has been RESOLVED (INVALID) by Nov 15 15:34:06 * *  opie-taskbar-images fails to build Nov 15 15:34:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3293 Nov 15 15:35:19 koen: does maemomapper supports non-gps usage? Nov 15 15:36:34 hrw: define non-gps usage :) Nov 15 15:37:01 you can input address as well, and scroll around the maps if that's what you mean Nov 15 15:37:10 yes Nov 15 15:37:48 if 64M card will be enough Nov 15 15:38:09 heh.. I forgot that maemo == sharprom Nov 15 15:38:19 dependency hell Nov 15 15:40:01 well, it's tough. if you really want something coherent and exciting, you have a hard time making it flexible on multiple systems as well Nov 15 15:40:05 i start to understand all these things Nov 15 15:40:53 there's a difference between 'hard time' and 'choosing to ignore the issue' Nov 15 15:41:27 nokia upgrades 2 libs on the system (no API or ABI changes) and requires a reflash for it Nov 15 15:42:01 so people with a 2 weeks old system can't install apps, because they lack the 2 libraries Nov 15 15:42:19 and nokia refuses to put those libs updates in the online repo as well Nov 15 15:42:21 not to mention not-existing locale support Nov 15 15:48:45 and maemo.org being connected to world via 33600 modem Nov 15 15:51:45 the .install file is a neat idea, though Nov 15 15:51:53 (if it actually worked) Nov 15 15:51:57 (which it doesn't) Nov 15 15:52:49 they got few nice ideas Nov 15 15:52:57 but only got.. Nov 15 15:59:31 good that atleast dropbear server can be installed Nov 15 15:59:51 Failed to fetch file:/OE/build/tmp-new/angstrom/deploy/glibc/deb/all/./openmoko-common2_0.1.0+svnr2679-r0_all.deb Hash Sum mismatch Nov 15 16:00:02 but the sha256 sum matches .... Nov 15 16:01:03 is sfdisk the best to partition a disk from a shell script? Nov 15 16:01:22 what's wrong with regular fdisk? Nov 15 16:02:05 koen: last time I tried it I had difficult automating it (no user interaction) Nov 15 16:02:17 * cbrake looks again ... Nov 15 16:06:46 looks like sfdisk is still the only one that can be used noninteractively Nov 15 16:08:26 hmm, parted is another option Nov 15 16:08:44 cbrake: http://svn.o-hand.com/view/poky/trunk/scripts/poky-addptable2image?rev=1734&view=markup Nov 15 16:08:54 cbrake: ugly but it does work... Nov 15 16:10:18 RP: interesting, so you can pipe commands into fdisk -- thanks Nov 15 16:11:18 cbrake, that frightens me :) Nov 15 16:12:14 Crofton: that is the problem with Linux -- you can do some truely scary things :-) Nov 15 16:13:40 heh, this works: echo ",,83" | sfdisk /dev/mmcblk0 Nov 15 16:13:49 we must be discussing feeding fdisk a bunch of numbers and entries, and then hoping like heck that it didn't destroy some other random device... Nov 15 16:14:33 mwester: nod, I think I'll stick with my sfdisk -- seems pretty clean Nov 15 16:15:11 the above command creates one partition using the entire disk Nov 15 16:15:56 cbrake: I had the contraint of making it match a real zaurus when I wrote that fdisk script... Nov 15 16:16:11 Yep, sfdisk is way better IMO. (I implemented an auto-fdisk script for Unslung on the NSLU2 that piped text into fdisk - it worked (usually) but I was too worried about it to actually add it to the distro). Nov 15 16:16:16 mwester: Nobody would be mad enough to run that script as root, right? :) Nov 15 16:18:15 RP: nod, well at least fdisk gives you the option when you need it Nov 15 16:24:04 * * OE Bug 3304 has been RESOLVED (INVALID) by koen(AT)openembedded.org Nov 15 16:24:06 * *  Bitbake complains about missing deps due to changes to the repo. Nov 15 16:24:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3304 Nov 15 16:25:04 * * OE Bug 3279 has been RESOLVED (FIXED) by Nov 15 16:25:06 * *  gpsdrive SRC_URI changed Nov 15 16:25:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3279 Nov 15 16:29:20 koen: got it Nov 15 16:29:47 koen: maemo mapper fetched 29MB of maps (~5000 parts) for needed region Nov 15 16:30:22 you should be able to select which zoom level you want in the download dialog Nov 15 16:30:30 Anyone, if I wanted to play around with the current package up staging stuff, where should I look? Nov 15 16:30:33 which is hidden somewhere Nov 15 16:31:06 not hidden even Nov 15 16:31:26 atleast not in mm 1.4.7 Nov 15 16:31:29 iirc it's under 'manage maps' Nov 15 16:31:59 2.0.x freaked me out at first Nov 15 16:32:18 it defaults to idiot mode where 'up' is your driving direction instead of north Nov 15 16:34:22 its 'car mode' Nov 15 16:35:08 bbl Nov 15 16:38:52 hi florian Nov 15 16:45:29 * cbrake finishes his first C# program -- seems to work pretty well Nov 15 16:59:59 Sending and flashing rootfs image (128256 kB)... Nov 15 17:00:00 100% (128256 of 128256 kB, avg. 5953 kB/s) Nov 15 17:00:13 I also love the 6MB/s flashing of the n800 Nov 15 17:02:59 koen nice Nov 15 17:18:26 bye Nov 15 17:18:35 ~ping crofton Nov 15 17:18:36 pong crofton Nov 15 17:19:21 stefan_schmidt, pong Nov 15 17:19:50 er Nov 15 17:19:56 steliosk, pong Nov 15 17:21:28 Crofton|home : Send me the invoices you have from the OEDEM Nov 15 17:21:45 working on it right now Nov 15 17:34:25 IT2008 is pretty neat Nov 15 17:34:39 it's almost as responsive as openmoko :) Nov 15 17:35:34 Which is presumably good :) Nov 15 17:37:34 actually, no Nov 15 17:38:15 but the OMG SHINY TRANSLUCENT PANEL makes up for it Nov 15 17:40:06 koen is a sucker for cool design features Nov 15 17:40:15 yeah Nov 15 17:40:58 Ok, things depend on ettercap, ettercap doesn't exist, but ettercap-ng does, do I do "PROVIDES = ettercap" or do I change the DEPENDS of all those things to ettercap-ng? Nov 15 17:41:01 Crofton|home: http://linuxdevices.com/news/NS7500261933.html Nov 15 17:41:17 I mean do PROVIDES = ettercap on ettercap Nov 15 17:41:19 -ng Nov 15 17:41:22 :/ Nov 15 17:42:47 ? Nov 15 17:43:21 changing the depends would probably be better Nov 15 17:43:27 hmm, the shape shifting reptile Nov 15 17:43:38 but the smart thing would be drop that into the lap of the person that deleted ettercap :) Nov 15 17:43:46 :D Nov 15 17:44:34 heh, I hadn't seen the reptile crack until after I posted the above Nov 15 17:46:47 Is ${XSERVER} valid anymore? Nov 15 17:47:11 only in package-space Nov 15 17:47:19 Should I set that to virtual/xserver Nov 15 17:47:21 ? Nov 15 17:47:22 in recipe space you'd want virtual/xserver Nov 15 17:47:22 ok Nov 15 17:47:30 ok Nov 15 17:47:40 thanks :) Nov 15 17:47:57 fixing packages, one by one :) Nov 15 18:04:23 osso-gnome-vfs2 was moved to obsolete, but packages "osso-thumbnail", "hildon-home" and "task-maemo" still depend on it. Nov 15 18:04:31 Course of action? Nov 15 18:04:58 I'm not sure these packages *really* require it, though Nov 15 18:08:15 Dammit a lot of things have changed since I last built world Nov 15 18:08:25 Back then it actually *worked* :D Nov 15 18:11:19 osso-gnome-vfs2 was moved to obsolete, but packages "osso-thumbnail", "hildon-home" and "task-maemo" still depend on it. Course of action? :) Nov 15 18:12:21 A whole load of things fail because of these three :( Nov 15 18:17:13 Oh damn. Nov 15 18:19:20 Someone? Nov 15 18:19:32 I'm trying to fix bugs here. Nov 15 18:19:34 :/ Nov 15 18:20:02 I'm around but I have no idea, but I Nov 15 18:20:10 'll try to offer some moral support :) Nov 15 18:23:59 :D:D Nov 15 18:24:05 ;) Nov 15 18:24:17 Hello sir Nov 15 18:24:55 May'st thou provide me with guidelines so that I may fix some bugs pertaining to the OE recipe repository and the items held within? Nov 15 18:25:48 Oh what the fuck Nov 15 18:25:59 I'll bring the package back in, see if it builds. Nov 15 18:26:08 :) Nov 15 18:27:43 Aha Nov 15 18:28:20 Suck me sideways! Nov 15 18:28:36 This gets deeper yet, Barry! Nov 15 18:28:46 :) Nov 15 18:28:50 :) Nov 15 18:34:43 A whole load of things are broken... Nov 15 18:35:05 Whenever it downloads something from svn, it keeps thinking I want revision 1 Nov 15 18:35:06 :/ Nov 15 18:35:30 And adds a nice -r 1 to the svn commandline :/ Nov 15 18:40:19 B_Lizzard: there is a patch for bitbake to fix that I think ... Nov 15 18:40:28 Hmmm Nov 15 18:40:50 I now know that angstrom imports some good known dates and revisions... Nov 15 18:40:56 I'll try updating bb Nov 15 18:41:31 B_Lizzard: http://pastebin.ca/775755 Nov 15 18:42:35 Thanks a lot, cbrake! :) Nov 15 18:42:40 B_Lizzard: sure Nov 15 18:46:43 koen's busy, seems like. Nov 15 18:52:05 * * OE Bug 3305 has been created by hans.ullrich(AT)loop.de Nov 15 18:52:07 * * white keys do not work correctly (akita) Nov 15 18:52:09 * * http://bugs.openembedded.org/show_bug.cgi?id=3305 Nov 15 18:58:17 re Nov 15 19:18:02 damnation Nov 15 19:18:18 this new fangled router only seems to support portforwarding from the outside Nov 15 19:31:49 re koen Nov 15 20:07:35 he crofton Nov 15 20:07:43 re Nov 15 20:07:51 snowing today Nov 15 20:07:59 hehe Nov 15 20:08:05 put the ski on Nov 15 20:08:05 flurries really, maybe global warming is over .... Nov 15 20:19:09 lustig Nov 15 20:19:15 iptv ueber multicast Nov 15 20:19:17 http://www.ard-digital.de/index.php?id=14029&languageid=1 Nov 15 20:19:19 ups Nov 15 20:19:21 sorry Nov 15 20:25:40 03florian 07org.oe.dev * r479d1369... 10/ (1 packages/gpephone/quickdial_svn.bb): quickdial: Add svn recipe Nov 15 20:25:44 03hrw 07org.oe.dev * r353324dd... 10/ (1 packages/lm_sensors/lmsensors-apps_2.10.1.bb): lmsensors-apps: fix DEPENDS Nov 15 20:25:50 03hrw 07org.oe.dev * r876cf349... 10/ (1 packages/tasks/task-gpe-sectest.bb): task-gpe-sectest: dropped as it is now in task-gpe recipe Nov 15 20:25:56 03hrw 07org.oe.dev * r29a3e4fb... 10/ (1 packages/librcf/librcf_0.4.bb): librcf: drop bogus RDEPENDS Nov 15 20:26:01 03hrw 07org.oe.dev * r6bd60192... 10/ (1 packages/python/python-pygtk_2.10.4.bb): python-pygtk: PROVIDES python-pygtk2 for compability Nov 15 20:26:07 03hrw 07org.oe.dev * re0561985... 10/ (1 packages/libogg/libtheora_0.9+1.0alpha7.bb): libtheora: do not depend on itself Nov 15 20:26:12 03hrw 07org.oe.dev * r0366fffb... 10/ (1 packages/gpsdrive/gpsdrive_2.10pre2.bb): gpsdrive: fix SRC_URI - close #3279 Nov 15 21:07:17 steliosk: I'm getting there; now I've got a build failure on popt Nov 15 21:18:10 Crofton: 70 degrees here today, global warming isn't over yet Nov 15 21:18:28 heh Nov 15 21:18:50 36 F here Nov 15 21:18:54 We could use some rain, though . . . Nov 15 21:19:18 we are having issues with global drying also Nov 15 21:19:27 ~lart trolltech Nov 15 21:19:28 * ibot takes a large goose feather pillow and swings it wildly in trolltech's direction, hitting trolltech and sending trolltech flying into the closet Nov 15 21:20:46 * mwester thinks Crofton must be referring to the newly-appearing Georgia Desert Nov 15 21:26:56 hmm Nov 15 21:28:24 lorn in the toolchain for neo device arent the qtopia header and libs Nov 15 21:28:35 so crosscompilig for qt is pain in the ass Nov 15 21:29:31 no, its a cross toolchain Nov 15 21:30:14 yeah but only deliver the gcc didnt make so much sense Nov 15 21:30:48 yes it does Nov 15 21:31:11 now I am stucked and can not test the qt4 backend of midpath in qemu-neo Nov 15 21:31:12 we havent released an sdk for it yet Nov 15 21:32:47 *sigh* Nov 15 21:33:40 qtopia hcanges too much to add it to the toolchain Nov 15 21:34:10 okay Nov 15 21:34:24 sorry Nov 15 21:34:27 I understand now Nov 15 21:34:29 ljp: the platform should be stable ;) Nov 15 21:34:37 I was only a little frustraded Nov 15 21:34:45 ljp: we talk about header files + lib Nov 15 21:35:04 qtopia headers and lib do not belong in the toolchain Nov 15 21:35:31 the libs are in rootfs Nov 15 21:35:37 but not the headers Nov 15 21:35:40 that killsme Nov 15 21:35:50 ljp: I see you think toolchain == compiler, SDK = gui + toolchain + qtopia libs... Nov 15 21:36:10 zecke: correct Nov 15 21:36:45 ljp: I don't know if I agree. will think about it later :) Nov 15 21:37:13 * kergoth agrees with ljp on that terminology Nov 15 21:37:16 zecke: sounds about right I think toolchain is bare minimum needed to create a C application Nov 15 21:37:18 toolchain hardly ever needs updating between qtopia versions Nov 15 21:37:33 zecke hehe no sdk is a vmware image so you are stucked on vmware even if you have native linux Nov 15 21:37:42 wow..! kergoth agrees with me! Nov 15 21:37:56 kergoth? Nov 15 21:37:58 i know, i was scared too Nov 15 21:38:10 oh Nov 15 21:38:19 woglinde_: as og 4.3.0 the sdk iso contains a qtopia.iso which contains toolchain+ sdk so you can install to host Nov 15 21:38:19 another sign of the apocalypse Nov 15 21:38:20 hehe Nov 15 21:38:31 I like kergoth too much to steal his mojo Nov 15 21:38:36 :P Nov 15 21:39:14 okay so I will wait Nov 15 21:39:32 and have a good reason to play my games Nov 15 21:39:49 instead of doing freesoftware thinkies *g* Nov 15 21:40:09 for now, you just have to compile qtopia Nov 15 21:40:39 *g* hm I only had to copy the headers of qtopia Nov 15 21:40:43 but I am to lazy Nov 15 22:07:46 koen|awa2: hmm. how can i add keys nowadays. i can't seem to log in to ewi Nov 15 22:16:14 hi Nov 15 22:16:34 I did a fresh monotone pull after removing /tmp Nov 15 22:16:49 I only built console-image Nov 15 22:17:00 now I tried straight gpsdrive Nov 15 22:17:12 it fails Nov 15 22:17:14 No package 'gtk+-2.0' found Nov 15 22:17:42 shouldn't the dependencies come out automatically Nov 15 22:17:44 ? Nov 15 22:19:05 checking for arm-angstrom-linux-gnueabi-pkg-config... no Nov 15 22:19:19 checking for pkg-config... /home/oe/build/tmp/staging/i686-linux/bin/pkg-config Nov 15 22:19:36 should I file a bug or it is too obvious? Nov 15 23:21:41 Is there a way to get bitbake -e to evaluate for a specific package? Nov 15 23:47:58 Tartarus: I used to use the showdata in bitbake -i (It seems to be broken for me though) peek works though Nov 15 23:58:21 thanks Nov 16 00:05:04 * * OE Bug 3306 has been created by autobuild(AT)openembedded.org Nov 16 00:05:06 * * mpeg2dec-0.4.0b-r4-do_compile Nov 16 00:05:08 * * http://bugs.openembedded.org/show_bug.cgi?id=3306 **** ENDING LOGGING AT Fri Nov 16 02:59:58 2007