**** BEGIN LOGGING AT Thu Sep 03 02:59:57 2009 Sep 03 06:26:39 hello now i am using Angstrom on Beagleboard Sep 03 06:26:48 I want to know some terminal command for mount sd card in Angstrom Sep 03 06:56:52 good morning Sep 03 07:07:04 * * OE Bug 4488 has been RESOLVED (FIXED) by Sep 03 07:07:06 * * libxine-x11 and libxine-fb fail do_configure Sep 03 07:07:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4488 Sep 03 07:30:40 good morning Sep 03 07:40:05 * * OE Bug 3076 has been RESOLVED (FIXED) by Sep 03 07:40:07 * * opie-mediaplayer2 doesn't build on fresh OE/angstrom Sep 03 07:40:09 * * http://bugs.openembedded.net/show_bug.cgi?id=3076 Sep 03 07:52:36 hrw|gone: let us know if you have any comments about Sep 03 07:52:56 !oebug 312 Sep 03 07:52:58 * * Bug 312, Status: NEEDSINFO, Created: 2005-09-08 05:16 Sep 03 07:52:59 * * mickeyl(AT)handhelds.org: xine deserves an update Sep 03 07:53:00 * * http://bugs.openembedded.org/show_bug.cgi?id=312 Sep 03 07:53:20 I think it's finally ready for closure Sep 03 08:06:22 03acid-burn  07org.openembedded.dreambox * re7b275c2e3 10openembedded.git/packages/zope/zope-interface_3.3.0.bb: zope-interface_3.3.0.bb: remove -dbg and -doc from PACKAGES to get rid of OE Warning Sep 03 08:06:22 03acid-burn  07org.openembedded.dreambox * r8c584453a8 10openembedded.git/packages/twisted/twisted_8.2.0.bb: twisted_8.2.0.bb: remove -dbg from PACKAGES to get rid of OE Warning Sep 03 08:06:23 03acid-burn  07org.openembedded.dreambox * r81af9e9f3c 10openembedded.git/conf/machine/dm7025.conf: dm7025.conf: also move gstreamer binaries from /usr/bin/ to jffs2. Sep 03 08:06:24 03acid-burn  07org.openembedded.dreambox * r354828b963 10openembedded.git/packages/enigma2/ (enigma2-plugins.bb enigma2.bb): enigma2.bb,enigma2-plugins.bb: push srcdate, install meta informations inside /usr/share/meta Sep 03 08:24:55 03Koen Kooi  07org.openembedded.dev * r7457196b20 10openembedded.git/recipes/gnome/gnome-keyring_2.26.0.bb: gnome-keyring: fix merge error Sep 03 08:34:31 03Steve Sakoman  07org.openembedded.dev * r6aea479c25 10openembedded.git/recipes/wpa-supplicant/ (5 files in 2 dirs): Sep 03 08:34:31 wpa-supplicant: updated scripts Sep 03 08:34:31 makes gnome networkmanager happy and functional Sep 03 08:34:31 03Steve Sakoman  07org.openembedded.dev * rc420b9d428 10openembedded.git/ (conf/checksums.ini recipes/gnome/libunique_1.1.2.bb): libunique: add 1.1.2 Sep 03 08:34:34 03Steve Sakoman  07org.openembedded.dev * r2882710987 10openembedded.git/ (6 files in 3 dirs): orbit2: restructure recipes to use orbit2.inc, add 2.14.17 Sep 03 08:34:37 03Steve Sakoman  07org.openembedded.dev * rb2d5c4000c 10openembedded.git/ (2 files in 2 dirs): system-tools-backends: add 2.8.1 Sep 03 08:34:40 03Steve Sakoman  07org.openembedded.dev * r5e46d34f5a 10openembedded.git/ (2 files in 2 dirs): inotify-tools: add 3.13 Sep 03 08:34:43 03Steve Sakoman  07org.openembedded.dev * rae751ce3cd 10openembedded.git/ (conf/checksums.ini recipes/libnotify/libnotify_0.4.5.bb): libnotify: add 0.4.5 Sep 03 08:34:46 03Steve Sakoman  07org.openembedded.dev * rcf7f5dab3b 10openembedded.git/ (conf/checksums.ini recipes/libglade/libglade_2.6.4.bb): libglade: add 2.6.4 Sep 03 08:34:51 03Steve Sakoman  07org.openembedded.dev * r94c2e1f513 10openembedded.git/recipes/powervr-drivers/ (libgles-omap3.inc libgles-omap3/cputype libgles-omap3/rc.pvr): libgles-omap3: detect cpu type and don't load modules if no SGX hardware Sep 03 08:34:55 03Steve Sakoman  07org.openembedded.dev * r40db23e33e 10openembedded.git/ (conf/checksums.ini recipes/swfdec/swfdec-gnome_2.26.0.bb): swfdec-gnome: add 2.26.0 Sep 03 08:38:12 I looking in the recipes directory in oe and see that there is a flash-plugin_10.0.22.87.bb. Is that a firefux plugin really that will run on beagle board? Sep 03 09:04:29 good morning Sep 03 09:12:45 florian: good morning Sep 03 09:13:27 good morning Sep 03 09:22:49 hi pb_ Sep 03 09:22:52 hi dth Sep 03 09:26:52 jovox: the adobe-flash one? no, that is for x86 only. Sep 03 09:27:13 you might find that gnash runs on your hardware though. Sep 03 10:10:17 goodmorning Sep 03 10:11:37 is there a way to add a user to a custom image before installing it on the machine? Sep 03 10:13:33 morning Sep 03 10:14:07 Laibsch: no comments to 312 from me Sep 03 10:16:17 IOW, you're not aware of anything left to be done? Sep 03 10:16:48 morning all ! Sep 03 10:17:06 Laibsch: I did not looked at xine for over year if not more Sep 03 10:17:28 Laibsch: I think that better way would be close it as wontfix and start working on newest version instead Sep 03 11:04:32 03Koen Kooi  07org.openembedded.dev * r53f12b0e21 10openembedded.git/ (3 files in 2 dirs): gst-ffmpeg: bump PR on 0.10.7, add 0.10.8 and cleanup EXTRA_OECONF Sep 03 11:04:37 03Koen Kooi  07org.openembedded.dev * rb1cc93c116 10openembedded.git/recipes/xorg-lib/libxext-native_1.0.5.bb: libxext-native: stop providing 'xext' Sep 03 11:58:04 good morning Sep 03 12:00:49 recalcati: hi Sep 03 12:16:25 morning Sep 03 12:44:22 re raster Sep 03 12:45:20 woglinde: bewz! Sep 03 12:48:48 hi raster woglinde Sep 03 12:49:33 yo yo chouimat Sep 03 12:50:49 hmmm need coffee brb Sep 03 13:16:02 hi Sep 03 13:16:34 I am trying to compile kernel according to notes on http://elinux.org/BeagleBoard#Linux_kernel Sep 03 13:17:29 hm where is the oe-specific question? Sep 03 13:17:30 at the stage :make CROSS_COMPILE=arm-none-linux-gnueabi- omap3_beagle_defconfig Sep 03 13:17:39 I get the error : Sep 03 13:17:59 HOSTCC scripts/basic/fixdep Sep 03 13:17:59 gcc: error trying to exec 'cc1': execvp: No such file or directory Sep 03 13:17:59 make[1]: *** [scripts/basic/fixdep] Error 1 Sep 03 13:17:59 make: *** [scripts_basic] Error 2 Sep 03 13:18:17 any resolutions? Sep 03 13:18:26 hm.. looks like Gnash 0.8.3 doesn't build :( Sep 03 13:18:46 kayhan your host gcc is borken install it proberly Sep 03 13:18:54 I am trying to compile on Ubuntu 9.04 Sep 03 13:18:56 kayhan but thats a kernel question not qe Sep 03 13:18:58 ups oe Sep 03 13:19:06 apt-get install gcc Sep 03 13:19:11 woglinde : ok Sep 03 13:19:14 apt-get install gcc-4.3 Sep 03 13:19:23 woglinde :ups oe? Sep 03 13:19:52 sorry saw the line rite now Sep 03 13:20:15 I am trying to do this for beagleboard Sep 03 13:20:50 so I have thought that OE people might have walked through this path Sep 03 13:21:35 nope on oe == openembedded we do bitbake linux Sep 03 13:21:53 which builds us a linux-kernel.ipk Sep 03 13:21:58 starting docs for that? Sep 03 13:22:08 www.openembedded.org Sep 03 13:22:14 does it work for Angstrom lin? Sep 03 13:22:38 angstroem is one distro you can build with openembedded Sep 03 13:23:53 will check that ..thanks a lot.. but for now I have to fix this. Sep 03 13:24:48 has anyone sucessfylly built gnash lately? I get a compile error (NetStreamFfmpeg.h:95: error: 'offset_t' does not name a type) Sep 03 13:25:00 jovox hm let me see Sep 03 13:25:42 Reading state information... Done Sep 03 13:25:43 gcc-4.3 is already the newest version. Sep 03 13:26:00 woglinde, I'm building for beagleboard. added it to the x11-image Sep 03 13:26:09 woglinde : I got that msg? Sep 03 13:26:22 seems installed , but may be broken? Sep 03 13:27:03 yes it cannt find the real compiler Sep 03 13:27:09 cc1 is the real one Sep 03 13:27:17 gcc command only a frontend Sep 03 13:28:41 jovox did you an git pull? Sep 03 13:29:03 if I remember correctly koen updated something Sep 03 13:29:20 two days ago since last update Sep 03 13:29:25 I will update Sep 03 13:29:59 kayhan try apt-get install --reinstall gcc-4.3 Sep 03 13:30:16 kayhan what does gcc --version tell? Sep 03 13:31:38 woglinde, did a git pull but still same error Sep 03 13:33:31 just trying Sep 03 13:33:33 my self Sep 03 13:36:06 hi hrw Sep 03 13:36:39 hi rkirti Sep 03 13:36:47 hello woglinde Sep 03 13:36:57 morning everyone Sep 03 13:37:12 hm I could update libdrm Sep 03 13:37:36 woglinde, was that for me? Sep 03 13:37:41 nope Sep 03 13:37:43 only for me Sep 03 13:37:45 oki Sep 03 13:37:47 heh Sep 03 13:37:50 whlie watching the compile process Sep 03 13:38:27 tell me about it.. havn't done much else the last five days :/ Sep 03 13:38:41 emoving old gcc doc directory. Sep 03 13:38:41 Unpacking replacement gcc ... Sep 03 13:38:41 Processing triggers for man-db ... Sep 03 13:38:41 Setting up gcc (4:4.3.3-1ubuntu1) . Sep 03 13:39:30 woglinde : that is after reinstall Sep 03 13:44:52 HOSTCC scripts/basic/fixdep Sep 03 13:44:52 gcc: error trying to exec 'cc1': execvp: No such file or directory Sep 03 13:44:52 make[1]: *** [scripts/basic/fixdep] Error 1 Sep 03 13:44:52 make: *** [scripts_basic] Error 2 Sep 03 13:45:01 and goes on with the error msg Sep 03 13:45:04 kayhan please use pastebin Sep 03 13:45:07 ~pastebin Sep 03 13:45:08 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste Sep 03 13:46:05 kayhan hm try apt-get install --reinstall cpp-4.3 Sep 03 13:46:27 ok.thnx for pastebin btw Sep 03 13:47:07 hi Sep 03 13:47:32 hi Sep 03 13:47:56 has anybody tried running uImage-2.6.29-r44 on omap3evm? Sep 03 13:48:18 hm only few people here has opam3evm Sep 03 13:48:23 most have beagleboard Sep 03 13:48:35 :( Sep 03 13:49:06 woglinde : same error goes on :( Sep 03 13:49:07 I can take one under my roof Sep 03 13:51:11 kayhan ls -la /usr/lib/gcc/i486-linux-gnu/4.3/cc1? Sep 03 13:52:51 http://pastebin.com/m2544829f Sep 03 13:53:36 rwxr-xr-x 1 root root 7393328 2009-03-17 02:25 /usr/lib/gcc/i486-linux-gnu/4.3/cc Sep 03 13:53:49 cc or cc1? Sep 03 13:53:56 cc1 Sep 03 13:53:59 hm Sep 03 13:54:01 strange Sep 03 13:54:23 sorry cannt tell whats broken with your host compiler Sep 03 13:54:32 I'm getting crazy with ads7846 and calibration with pxa270. I'm using qt3.3.4 qith linuxtp. Passing to tslib with qt3 is a good idea? Sep 03 13:54:37 but sure it is broken? Sep 03 13:54:38 hmm Sep 03 13:54:56 qt3 is outdated Sep 03 13:55:05 nad lacks many features qt4 has Sep 03 13:55:38 woglinde : ok i am ready to burn all my time wasted on it.. I would be very happy if you could recommend linux distro and tools to go on with beagleboard Sep 03 13:56:51 kayhan ubuntu works normaly Sep 03 13:56:59 dont know whats wrong Sep 03 13:57:35 woglinde: I'm sorry. I have to move to qt4.5, but not immediately Sep 03 13:57:56 recalcati maybee its a bug in qt3 Sep 03 13:57:58 who knows Sep 03 13:58:02 woglinde: I have to improve touchscreen calibration Sep 03 13:58:17 woglinde, still building?` Sep 03 13:58:22 jovox yes Sep 03 13:58:29 but should be fine i 1 minute Sep 03 13:58:35 NOTE: Running task 2890 of 2901 (ID: 8, /devel/arm/git/openembedded/recipes/gnash/gnash_0.8.5.bb, do_compile) Sep 03 13:58:55 hm note updating cups too Sep 03 14:01:26 when builing gnash? Sep 03 14:01:39 no Sep 03 14:01:58 we have version 1.2 which is old and has security bugs Sep 03 14:04:28 hms the gnash build lasts long Sep 03 14:04:43 yupp Sep 03 14:09:39 hm args gnash cant be build with -j option Sep 03 14:10:21 woglinde, didn't compile for you either? Sep 03 14:10:32 just compiling Sep 03 14:10:40 I looked at the recipe Sep 03 14:10:48 we turn of PARALLEL_MAKE Sep 03 14:10:52 off Sep 03 14:10:58 oki I will Sep 03 14:11:01 hm? Sep 03 14:11:15 cant say if it compiles to the end or not Sep 03 14:11:20 its just compiling Sep 03 14:12:05 I removed the -j option and now running a 'bitbake gnash' again Sep 03 14:12:26 ????? Sep 03 14:12:54 how did you bitbake gnash? Sep 03 14:13:11 just "bitbake gnash" Sep 03 14:13:20 jepp Sep 03 14:13:32 the recipes itself turns of the parallel make Sep 03 14:13:39 ah ok Sep 03 14:13:48 so you dont need to change your local.conf Sep 03 14:13:57 oki I see Sep 03 14:13:59 hm no problems here Sep 03 14:14:02 just builds Sep 03 14:14:09 for beagleboard Sep 03 14:14:12 yes Sep 03 14:14:15 hmm Sep 03 14:14:17 with angstroem Sep 03 14:14:20 Stransge Sep 03 14:14:22 strange Sep 03 14:14:31 try bitbake -c clean gnash Sep 03 14:14:35 bitbake gnash Sep 03 14:14:44 will do Sep 03 14:15:46 this is conna take some time Sep 03 14:15:48 gonna Sep 03 14:16:31 hmm maybe not that bad 1971 of 1982 Sep 03 14:22:27 woglinde, no. It failed again. We must be using diffrent sources Sep 03 14:22:55 armv7a-angstrom-linux-gnueabi/gnash-0.8.3-r5 right? Sep 03 14:24:24 ??? Sep 03 14:24:32 0.8.5 is latest Sep 03 14:24:37 are you using stable? Sep 03 14:24:39 or dev Sep 03 14:29:22 stable Sep 03 14:29:55 okay Sep 03 14:30:18 then request for stable that the newe version should be default Sep 03 14:31:03 I gotta do my git homework I guess :) Sep 03 14:32:00 I'm not very familiar with git an only followed the http://elinux.org/BeagleBoardAndOpenEmbeddedGit guide so far. But I will spend some time on figuring it all out Sep 03 14:34:03 http://wiki.openembedded.net/index.php/GitPhraseBook Sep 03 14:34:14 http://wiki.openembedded.net/index.php/Stable Sep 03 14:34:38 Thanks! Sep 03 14:36:00 which branch is the latest? HEAD? Sep 03 14:36:21 dev Sep 03 14:36:33 each branch has its own head Sep 03 14:36:35 oki Sep 03 14:37:59 can't see any branch named dev. Is it org.openembedded.dev or origin/org.openembedded.dev? Sep 03 14:40:43 yes Sep 03 14:40:47 origin/org.openembedded.dev Sep 03 14:40:52 didnt you read the links? Sep 03 14:41:59 only the second. I will go through the first one as well Sep 03 14:47:54 ~seen khasim Sep 03 14:47:58 i haven't seen 'khasim', rkirti Sep 03 15:00:04 bitbake is not included in the origin/org.openembedded.dev branch? Sep 03 15:00:29 so, I figured out how to build konqueror-embedded. Is that useful info for anyone here? :) Sep 03 15:00:52 Jotheberlock bitbake konqueror-embedded? Sep 03 15:01:05 woglinde: won't work out of the box ;) or didn't for me anyway Sep 03 15:01:21 whats the problem provide patches Sep 03 15:01:27 (among other things it seems to configure itself with a malfunctioning libtool) Sep 03 15:01:55 I don't know bitbake well enough to know the 'right' way to fix it. I ended up doing things like copying the main cross-libtool's libtool in after the configure stage ran Sep 03 15:02:18 I was hoping the Opie maintainer-guy would be in-channel but I don't see him; I know he's going to work on it at some point and figured I could be of some help Sep 03 15:03:36 so its only a libtool problem Sep 03 15:03:41 that should be easy to fix Sep 03 15:05:08 woglinde, how come bitbake is part of the stable/2009 but not in dev? Sep 03 15:05:22 bitbake is its own tool Sep 03 15:05:31 which formerly was hosted at berlios Sep 03 15:06:07 Ok, it makes sense. So it was simple moved out? Sep 03 15:06:08 Hi all Sep 03 15:08:26 jovox: because stable/2009 is working with this version of bitbake and I (and other maintainers of stable branch) do not plan to support any newer version Sep 03 15:08:53 ah right Sep 03 15:08:55 thanks hrw Sep 03 15:09:26 jovox: you can of course use any newer version and it will probably work. but no warranty at all that it will Sep 03 15:10:36 I'm discovering OpenEmbedded, and I would like to ask just one little question : I was using buildroot, and I would like to test Openembedded. My first tests with it (reading the Getting_started webpage) give me a big system with X11, gtk, udev, etc. I'm using a mini2440, with a angstrom-2008.1 Distro base. Could anyone give me some steps about making a very small OS, like a basic configuration of buildroot would give me ? Sep 03 15:10:37 which is fine for me at the moment since I'm just doing some research Sep 03 15:11:32 damunix: there are many images available in OE metadata - check recipes/images/ directory and choose one of them Sep 03 15:11:47 micro-image or micro-base-image are the smallest one iirc Sep 03 15:12:19 when angstrom-beagleboard-demo is very big beast Sep 03 15:14:06 hrw: so to generate a basic image, I just would have to use 'bitbake micro-image' ? Sep 03 15:14:48 yes Sep 03 15:15:16 damunix: you can also do 'bitbake micro-image console-image minimal-image' and choose one of them for your next play Sep 03 15:16:50 hrw : can you explain me a little bit more ? it would give me 3 rootfs images ? Sep 03 15:17:57 yes Sep 03 15:18:34 damunix: OE builds software into packages which can be then used to create rootfs images (in many formats) Sep 03 15:19:29 damunix: so if you want image which has busybox+udev/mdev+sysvinit you takes micro(-base)-image. if you want x11 then you take x11-image (which is X11 server + matchbox environment + some GPE stuff to configure system) Sep 03 15:20:15 and if you want something to show complete desktop with some apps you can take angstrom-beagleboard-demo which results in >200MB image with x11+e17+gimp+mplayer+abiword+many other applications Sep 03 15:20:26 hrw: ok thanks for this informations. A thing that is strange for me : what is DISTRO for ? Because you say that (if I understand) that image/ folder is location of packages, but what would Angstrom for example will do ? Sep 03 15:20:33 I have to do build with buildroot one day Sep 03 15:20:56 woglinde: also, a directory config problem (it expects there to be a ../kdelibs/admin), and I had to grab a font it expected but opie didn't install in order to get it to start Sep 03 15:21:22 Jotheberlock that could be fixed with RDEPENDS line Sep 03 15:21:25 damunix: DISTRO value selects distribution (as OE is buildsystem capable of building many distributions for devices) which config selects toolchain and applications versions, variants etc Sep 03 15:21:38 hrw: no what I'm looking for is generate a really little image, just with qt-embedded package Sep 03 15:21:48 damunix: distro selects glibc/eglibc/uclibc, gcc3.4/4.1/4.2/4.3/4.4 etc Sep 03 15:22:02 damunix: which device you are targetting? Sep 03 15:22:05 ok Sep 03 15:22:16 mini2440 (s3c2440 device) Sep 03 15:22:20 ah right Sep 03 15:22:25 wvga screen? Sep 03 15:23:26 damunix: make a copy of recipes/images/console-image.bb and add qt-embedded packages names to IMAGE_INSTALL variable - then build this image and test on device Sep 03 15:30:35 http://elinux.org/BeagleBoardAndOpenEmbeddedGit Sep 03 15:30:43 started to apply these Sep 03 15:30:51 hope this one works Sep 03 16:00:35 03Michael 'Mickey' Lauer  07org.openembedded.dev * r7b21597db9 10openembedded.git/ (2 files in 2 dirs): libpersistence: new recipe; gobject serialization library Sep 03 16:00:44 03Michael 'Mickey' Lauer  07org.openembedded.dev * r2438d3f2f7 10openembedded.git/recipes/freesmartphone/mickeydbus_git.bb: mickeydbus: add more RDEPENDES Sep 03 16:01:04 mickeyl: can you take mickeydbus out of task-tools-cli? Sep 03 16:01:22 mickeyl, did you add -pprint to the deps? Sep 03 16:01:33 Tartarus: yep Sep 03 16:01:48 k Sep 03 16:02:09 hrw: mickeydbus is only in task-cli-tools-python Sep 03 16:02:28 do you want it out of the recipe? Sep 03 16:02:36 yes Sep 03 16:03:25 hmm, you could've done that on your own Sep 03 16:03:30 committing... Sep 03 16:03:50 task-tools-cli is your toy so... Sep 03 16:04:18 oh well, don't be shy :) Sep 03 16:04:31 anyways, i'm on it Sep 03 16:06:50 does PRERERRED_VERSION_foo in local.conf override the DEFAULT_PREFERENCE in foo.bb? Sep 03 16:07:22 yes Sep 03 16:07:36 default_preference is only a default :-} Sep 03 16:07:57 hi mickeyl Sep 03 16:08:14 pb__: yeah, I must be doing something wrong then ... Sep 03 16:09:10 cbrake: it'd be worth checking with bitbake -e to make sure nothing is clobbering your PREFERRED_VERSION. some of the conf/include files are a bit gung-ho about setting what they think are appropriate preferred ersions. Sep 03 16:09:33 PREFERRED_VERSION_virtual/kernel="git" Sep 03 16:09:43 hm, that doesn't sound right Sep 03 16:09:56 the preferred version needs to actually match what the recipe would declare as ${PV} Sep 03 16:10:08 PREFERRED_PROVIDER_virtual/kernel="linux-omap3" Sep 03 16:10:17 I doubt that really is just "git" in this case, or, if it is, the recipe is probably a bit broken. Sep 03 16:10:18 ahh, ok Sep 03 16:10:53 heya pb__ Sep 03 16:11:23 mickeyl: I just noticed that your oe.dev boycott seems to be over :-) Sep 03 16:12:10 pb__: yes, unfortunately i had to stop that due to a higher goal, namely helping the SHR guys to bring the branches fso/milestone5.5 and shr/import nearer to mainline. Sep 03 16:12:34 at the end of the day bringing the project forwards won over my pride Sep 03 16:13:04 mickeyl: hi - that's good news! Sep 03 16:13:05 mickeyl: ah, I see. well, I am glad that you are back, no matter what the reasons :-} Sep 03 16:13:17 :) Sep 03 16:13:52 but we shouldn't let that undermine our resolve to appoint a tsc, I think this is still an important task. Sep 03 16:14:41 if necessary I will lock the doors at oedem and nobody will be allowed to leave until this has been accomplished. Sep 03 16:14:42 i fully agree. this is high on my wish list Sep 03 16:14:48 good strategy Sep 03 16:14:52 i fully support that .) Sep 03 16:15:28 03Koen Kooi  07stable/2009 * rf7e2a23839 10openembedded.git/recipes/xscreensaver/ (xscreensaver.inc xscreensaver_5.07.bb): Sep 03 16:15:28 xscreensaver: fix packaging and QA Sep 03 16:15:28 Signed-off-by: Koen Kooi Sep 03 16:15:28 Acked-by: Philip Balister Sep 03 16:15:28 Acked-by: Khem Raj Sep 03 16:16:23 I suggest tsc to be discussed on first day Sep 03 16:16:38 as on second we (woglinde and me) have to leave earlier Sep 03 16:17:22 anyway - do we have some wiki page for agenda? Sep 03 16:18:16 hrw: I try what you tell me, but it is compiling the Xorg package Sep 03 16:18:30 which isn't necessary Sep 03 16:18:42 damunix: xserver-xorg or libx11? Sep 03 16:18:57 xorg-lib, xorg-proto, etc Sep 03 16:19:20 damunix: some components builds also some x11 support which will not land in non-x11 images Sep 03 16:19:44 http://wiki.openembedded.net/index.php/Oedem/2009 - who will start? Sep 03 16:21:01 hrw: now it's gtk :), so there will not be in my rootfs ? Sep 03 16:21:54 * mckoan is worried about swine flu during OEDEM2009 Sep 03 16:22:00 pb__: PREFERRED_VERSION_virtual/kernel = "2.6.31-rc8+r43+git3777b1e9ffb6e0901be93cbe201ae8e96003179b" failed but PREFERRED_VERSION_linux-omap3 = "2.6.31-rc8+r43+git3777b1e9ffb6e0901be93cbe201ae8e96003179b" worked. Does that sound right? Sep 03 16:22:09 mckoan why? Sep 03 16:22:14 cbrake: yes Sep 03 16:22:25 cbrake: virtual/kernel is virtual... Sep 03 16:22:37 hrw: ok, that makes sense Sep 03 16:22:47 woglinde: airport, aircraft, train, people Sep 03 16:22:50 virtual only takes package Sep 03 16:22:59 mckoan swine flu isnt that bad Sep 03 16:23:08 only some people suffers from it Sep 03 16:23:21 mckoan: swine flu is often nicer then nomral flu Sep 03 16:23:29 you shouldnt git crazy by the media hype Sep 03 16:23:31 get Sep 03 16:24:50 your optimism is comforting Sep 03 16:25:13 * mckoan is booking a flight Sep 03 16:25:29 mckoan: when you land/leave? Sep 03 16:26:55 hrw: I don't understand why Ryanair advertise 0 euro and at last are 80 Sep 03 16:27:13 mckoan: 80? lot... Sep 03 16:27:40 mckoan: Milano->Stansted or Bergamo->Stansted? Sep 03 16:28:08 ok, Milan (Bergamo) :D Sep 03 16:28:36 hrw: Milano (Bergamo) BGY - London Standsted, arrival: 6nov depart: 8nov Sep 03 16:29:03 79.95€ indeed Sep 03 16:29:21 mckoan: departing at 18:45? Sep 03 16:29:34 mckoan: from stansted I mean Sep 03 16:29:41 yes, the last flight Sep 03 16:29:55 good, so you will train travel with woglinde and me Sep 03 16:30:05 fine Sep 03 16:30:16 wee have 18:30 or something Sep 03 16:30:32 when you arrive at london? Sep 03 16:30:38 17:15 Sep 03 16:30:46 also 6 nov like you Sep 03 16:31:11 me at 8:00 time to sighteeing Sep 03 16:31:44 I lack morning time for it Sep 03 16:32:11 I kove UK, the more I stay the better I feel Sep 03 16:32:24 s/kove/love Sep 03 16:32:43 I was few times in UK but always too short Sep 03 16:32:50 and was too busy each time Sep 03 16:33:04 last time I did not even left hotel for two days Sep 03 16:33:19 cbrake: yes, that sounds right. you can't set versions on virtuals, only on concrete packages. Sep 03 16:33:46 hrw do we want book youthostel? Sep 03 16:33:46 cbrake: conceptually, it doesn't make much sense to set a version on a virtual if you don't know what package it will resolve to. Sep 03 16:33:59 http://www.yha.org.uk/find-accommodation/east-of-england/hostels/Cambridge/index.aspx Sep 03 16:35:21 args hms credit card again Sep 03 16:36:29 2 Bed(s) in shared single sex rooms B&B - Adults £15.95 Sep 03 16:36:30 * mckoan confirmed flight booking Sep 03 16:36:42 would be the cheapest Sep 03 16:37:24 hrw: now it is compiling alsa package .... Sep 03 16:37:24 hm Sep 03 16:37:27 1 Bunk Room 2 beds B&B £38.95 Sep 03 16:37:53 1 Bunk Room 4 beds B&B £69.95 Sep 03 16:37:59 if some more want to join Sep 03 16:38:48 bb in few Sep 03 16:39:37 hm dinner now Sep 03 16:39:38 till later Sep 03 16:40:29 I'll take care about where to stay later, anybody could suggest any hotel in ML ? Sep 03 16:41:34 further detail about the event would help the choice Sep 03 16:44:31 have a nice rest of the day Sep 03 16:44:52 03Michael 'Mickey' Lauer  07org.openembedded.dev * rbc6ce2f8b8 10openembedded.git/recipes/tasks/ (task-cli-tools-python.bb task-cli-tools.bb): task-cli-tools[-python]: factor python-based cli tools out of common recipe to improve build time for those who don't need/want any of the python-based ones. Sep 03 17:16:41 have a nice rest of day Sep 03 17:17:55 what's cli-tools, the standard versions of the busybox stuff, or more special purpose than that? Sep 03 17:26:00 doesn't look like the standard versions of busybox stuff, seems to be just a grab-bag of miscellaneous utils. Sep 03 17:26:22 screen, nano, tcpdump, htop, socat, ... Sep 03 17:26:30 I don't even know what the last two of those are. Sep 03 17:39:38 pb___: hi Sep 03 17:41:17 pb___: should I go ahead with ABI_LAYOUT change ? Sep 03 18:16:29 http://wiki.openembedded.net/index.php/Gcc_issues_in_Intrepid_and_later Sep 03 18:16:35 I am trying to apply these Sep 03 18:16:44 but I cant find local.conf to change Sep 03 18:16:59 could someone please tell me its location? Sep 03 18:17:48 huh local.conf is what you define Sep 03 18:17:55 uh, you create it when you set up your build directory. if you don't know where it is, nobody else can tell you Sep 03 18:18:01 unless somebody has their crystal ball handy Sep 03 18:20:01 * rkirti observes most newbies have local.conf at the same place where local.conf.sample is -> ${OEDIR}/openembedded Sep 03 18:20:09 ugh Sep 03 18:20:20 most newbies should RTFM Sep 03 18:20:36 the getting started guide on the wiki tells them what to do Sep 03 18:21:28 most people should be careful about F words. Sep 03 18:21:46 kayhan: it means read the *fine* manual Sep 03 18:21:51 oh sure Sep 03 18:23:09 i was careful. I considered the most appropriate response, and chose one harsh enough, yet succinct and not angry or aggressive in any way, to ensure that it might actually be absorbed Sep 03 18:23:14 there are also some other variants like read the *fantastic* manual but i will cut it here , I am trying to do something. Be a person please Sep 03 18:23:36 how many times should we tell you that *YOU* created the local.conf when you set up your build? Sep 03 18:23:40 we aren't psychic Sep 03 18:23:53 if you're using something other than stock OE, you're going to have to tell us that Sep 03 18:24:17 generally you create a 'conf' directory in a build area, put a new local.conf in conf/, and build Sep 03 18:24:33 GettingStarted tells you this, as I just said 4 minutes ago Sep 03 18:24:59 * khem thinks thicker skin is needed to work in opensource :) Sep 03 18:27:57 * kergoth mulls over possible apis for some sort of pure python bbclass Sep 03 18:27:59 kergoth: I saw a infinite hang in bitbake yesterday. When I have something in ASSUME_PROVIDED = " blah" and then do bitbake blah Sep 03 18:28:11 impressive Sep 03 18:28:34 its not called for but hang is not desired Sep 03 18:28:39 hehe Sep 03 18:28:44 i think that would be a good candidate for a bug in the bts, so it's not forgotten Sep 03 18:29:02 * kergoth won't have time to work on it today Sep 03 18:30:54 * kergoth wonders where in the build process he should inject emission of a PN<->PROVIDES mapping Sep 03 18:32:25 kergoth: right after parse I would say Sep 03 18:32:58 meh, thats what i was thinking, but i hate using anonymous python for anything but metadata changes. just feels messy. we need a better hook for global operations Sep 03 18:34:00 or you can add a new pass something like gendep Sep 03 18:35:12 for config based stuff, you can use the ConfigParsed event with an event handler, that's great for some things Sep 03 18:35:21 but recipe stuff... no clean method Sep 03 18:48:04 * khem s bicycle got stolen last night :( Sep 03 18:48:34 me is sad but has to attend this 4 hour long meeting where customer will grill its a QBR Sep 03 18:56:11 doh :( Sep 03 19:56:23 03Koen Kooi  07org.openembedded.dev * r19e6e8e30a 10openembedded.git/recipes/linux/ (3 files in 3 dirs): linux-omap: move ppoll patch to files/ and exchange it with the queued version Sep 03 19:56:25 03Koen Kooi  07org.openembedded.dev * r4f7a0cb818 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.org:openembedded into org.openembedded.dev Sep 03 19:56:25 03Koen Kooi  07org.openembedded.dev * r5366c51056 10openembedded.git/recipes/xorg-lib/ (5 files in 2 dirs): pixman git: add some more NEON and fastpath patches Sep 03 19:58:14 re Sep 03 19:58:42 khem what it pity Sep 03 20:00:23 meh Sep 03 20:00:43 we really need a *clean*, *explicit* mechanism for accessing info about other recipes from a recipe, rather than having to emit tons of crap into tmpdir Sep 03 20:01:29 though, actually, i really only want classes to be able to do that... so.. time to think about what a pure python bbclass could do.. Sep 03 20:01:30 hmm Sep 03 20:02:03 things like 'which PV of yzx will be'? Sep 03 20:02:18 re hrw Sep 03 20:02:27 yeah, or the stuff in pkgdata, or what recipes PROVIDES what things, or.. Sep 03 20:02:37 hrw some mind about the youth hostel? Sep 03 20:04:06 woglinde: as long it will not be too far from oedem and safe enough to leave bag with clothes I am fine Sep 03 20:04:56 i got shlibs based missing dep checking to work, but it has to read in all of pkgdata/* (not the runtime bits, though), and had to emit/read extra info into the persist_data database Sep 03 20:05:07 hrw we could take the 2 bed rooms Sep 03 20:05:20 hm I will have my bag with me anyway Sep 03 20:05:28 it was pretty sweet, i made a hello world recipe, linked the binary against ncurses, and watched it error out at packaging time saying i need to add a dep on ncurses Sep 03 20:05:34 hrw unfornatly there is the creditcard problem again :( Sep 03 20:05:41 otherwise I would rent it Sep 03 20:05:48 03Koen Kooi  07org.openembedded.dev * rb9feafad3e 10openembedded.git/recipes/linux/ (5 files in 4 dirs): linux-omap git: update to rc8, start adding patches from the overo tree Sep 03 20:05:49 woglinde: gash... Sep 03 20:06:07 woglinde: your bank do not provide you with visa or mastercard? Sep 03 20:06:23 hrw I dont want one Sep 03 20:17:17 jo floroan Sep 03 20:57:25 bye Sep 03 20:57:41 bye hrw Sep 03 22:03:15 kergoth: Have you considered using the persist db stuff in bitbake? Sep 03 22:03:34 kergoth: API can be expanded as needed ;-) Sep 03 22:03:50 i used it for what i just did, was thinking about moving some of the pkgdata stuff to it Sep 03 22:03:54 :) Sep 03 22:05:09 hm springpython? Sep 03 22:05:16 used it in a couple other features i worked on over the past couple weeks, too, RP Sep 03 22:05:32 thinking about moving my abiversions stuff to it Sep 03 22:06:16 kergoth: cool. Did you need to extend the api? Sep 03 22:06:49 perhaps, may need more columns in the tables. will see, i'll just inherit & extend PersistData if i need to, can always push things back in Sep 03 22:06:58 * RP tried to extend it recently and just hit those wierd python errors Sep 03 22:07:09 might be useful to make it easier to store/retrieve lists to/from it also at some point Sep 03 22:07:17 right, agreed Sep 03 22:07:44 hm Sep 03 22:07:56 maybee a look springpython is worth for such a task Sep 03 22:07:57 s/return/yield/ in getValue, let the value passed to set be a non-string iterable, whatever.. needs more thought Sep 03 22:08:22 is springpython an ORM thing? Sep 03 22:08:29 * kergoth really needs to learn more database stuff Sep 03 22:08:52 woglinde: is that part of standard python? Sep 03 22:09:01 rp no Sep 03 22:09:12 such as java spring is not part of java Sep 03 22:09:21 but I made a workin debian package already Sep 03 22:09:23 woglinde: there is the problem then :/ Sep 03 22:09:39 * RP doesn't fancy making bitbake depend on java Sep 03 22:09:48 lol Sep 03 22:09:57 we can always pull in copies of extra deps into our tree as fallback in general, not sure about that package in particular Sep 03 22:09:57 http://springpython.webfactional.com/1.0.0/reference/html/index.html Sep 03 22:10:09 heh Sep 03 22:10:41 rp is sqlite part of normal python? Sep 03 22:10:51 woglinde: With python 2.5, yes Sep 03 22:10:56 pfff Sep 03 22:11:11 woglinde: hence why we used it Sep 03 22:11:32 awesome, my first (mostly) full build from MVL6 with shlibs dep checking enabled shows 8 failures Sep 03 22:11:53 NOTE: Task failed: tcptraceroute-1.5beta6-r0 has no dependency upon libpcap Sep 03 22:11:54 NOTE: package tcptraceroute-1.5beta6-r0: task do_package: failed Sep 03 22:12:01 kergoth: is the shlibs code anywhere the rest of us can play? Sep 03 22:12:23 kergoth: and did you ever get to share the staging package fetcher code? Sep 03 22:12:31 the first incarnation is on my pastey, but right now it requires a re-run of all the setscenes of the deps, probably have to move that to parse time Sep 03 22:12:34 * RP never did get around to writing it Sep 03 22:12:38 http://kergoth.pastey.net/124606 Sep 03 22:12:55 RP: i think they'll probably let me push it now, its in my queue Sep 03 22:13:06 * kergoth has a large backlog of stuff to push still Sep 03 22:13:23 unfornatly remote object with springpython dont support pyro namserver yet so I need to use pyro directly Sep 03 22:13:32 also, that first incarnation reads in all of pkgdata/* for each do_package, which is ugly, probably going to use persist_data for that as well Sep 03 22:13:38 but you can see its pretty straightforward Sep 03 22:14:02 had to add an extra check to make sure dep_pkg isn't in this recipe's PACKAGES, for cross-package deps within a single recipe Sep 03 22:16:54 kergoth: Althernatively, you just add a reverse mapping to pkgdata Sep 03 22:17:04 but I really hate pkgdata, it sucks Sep 03 22:17:07 yeah, that would work too. either way. Sep 03 22:17:19 agreed, thats why i stopped before doing that and thought about persist_data instead :) Sep 03 22:17:31 cool :) Sep 03 22:17:53 It is relatively simple which is nice Sep 03 22:18:54 hmm, should probably only enable this checking when using --as-needed, otherwise it'll want you to add direct deps which aren't necessary, the indirect deps are fine if it doesn't -really- use that lib Sep 03 22:19:52 * kergoth got MV to assign him 5 of the tasks on his mvl6/OE todo list for this sprint :) Sep 03 22:20:54 :) Sep 03 22:21:25 kergoth: Did you ever look at pkghistory.bbclass ? Sep 03 22:21:42 nope, haven't seen that one Sep 03 22:21:50 * RP wonders if that should have used persistdata Sep 03 22:22:04 I kind of like the games with symlinks Sep 03 22:22:09 we really need to teach bitdoc to extract file comment headers from classes Sep 03 22:22:16 so we know what we _can_ pull into INHERIT Sep 03 22:23:38 kergoth: http://git.pokylinux.org/cgit.cgi/poky/tree/meta/classes/packagehistory.bbclass Sep 03 22:24:22 cute, i like it Sep 03 22:25:48 RP is it possible to add the size Sep 03 22:25:50 i was thinking, as a part of some auditing of classes/, it may be a good idea to separate classes which recipes inherit directly vs those intended to be pulled in via INHERIT, just to make it slightly less confusing Sep 03 22:26:04 woglinde: yes, I wanted to add things like that Sep 03 22:26:16 and log when the dependencies change Sep 03 22:26:29 kergoth: Yes, that would be good Sep 03 22:27:06 * RP should throw together a "Poky updates for OE.dev" branch Sep 03 22:27:11 also, we have some .inc's intended to be pulled in via an include from the distro, but there's fuzziness here, since the INHERIT classes also go into the global metadata. should look that over.. some maybe should be switched Sep 03 22:27:54 At least the distro .inc files are clearly in one place (conf/distro/include) Sep 03 22:28:52 yeah, that's true Sep 03 22:30:01 kergoth: You don't use floating SRCREV much do you? Sep 03 22:30:12 not much, no Sep 03 22:30:20 * RP is seeing cache corruption issues with it Sep 03 22:30:34 The magic "SRCREVINACTION" sneaks into PV in the cache files Sep 03 22:30:56 That should be internal to bitbake only and I can't see how its escaping Sep 03 22:31:21 also, its only the automated builder that sees this in general Sep 03 22:31:26 so I can't reproduce Sep 03 22:31:37 installed bitbake and required environment in ubuntu, did apt-get to gcc 4.1 and symlinked , tried to bitbake console image : Sep 03 22:31:49 got this error : Sep 03 22:32:22 gcc: error trying to exec 'cc1': execvp: No such file or directory Sep 03 22:32:40 huh, i'll have to play with it. i never really looked at the autorev/srcrev stuff at all, it's one of those areas that got introduced after I left that I haven't yet taken the time to get comfortable with Sep 03 22:33:13 kergoth: DISTRO="moblin-bleeding" in poky is a lovely toy for it Sep 03 22:33:35 kergoth: builds all the git repo tips for a set list of packages Sep 03 22:33:57 ah, cute Sep 03 22:34:13 has real world use too :) Sep 03 22:36:16 anyone? Sep 03 22:36:23 NOTE: Task failed: python-2.6.1-ml5 has no dependency upon ncurses ... huh, guess it must have built a curses module or something Sep 03 22:36:38 kayhan: what recipe? it sounds like a problem with your build machine, not OpenEmbedded Sep 03 22:37:22 beagleboard Sep 03 22:37:28 *sigh* Sep 03 22:37:50 kayhan I already told you that your ubuntu installation is some how broken Sep 03 22:37:52 beagleboard is not a recipe Sep 03 22:38:02 no need to ask the same questions again Sep 03 22:38:14 03Khem Raj  07org.openembedded.dev * r8b452a5ae1 10openembedded.git/recipes/libidl/ (libidl-native_0.8.13.bb libidl_0.8.13.bb): Sep 03 22:38:14 libidl_0.8.13.bb: Add new recipe Sep 03 22:38:14 * Add native and target recipes for libidl 0.8.13 Sep 03 22:38:14 Signed-off-by: Khem Raj Sep 03 22:38:16 03Khem Raj  07org.openembedded.dev * r558f6d4436 10openembedded.git/ (51 files in 24 dirs): (log message trimmed) Sep 03 22:38:19 OpenEmbedded: Switch to using linux-uclibceabi and linux-gnu for TARGET_OS Sep 03 22:38:21 * Bump the ABI_LAYOUT. Build from scratch will be needed. Sep 03 22:38:23 * Replace using TARGET_OS from linux->linux-gnu. Sep 03 22:38:25 * Replace using linux-uclibcgnueabi->linux-uclibceabi. Sep 03 22:38:27 * Add 'eabi' to DISTRO_FEATURES (only for minimal and micro). Sep 03 22:38:29 * Use eabi and BASE_PACKAGE_ARCH to compute real TARGET_OS. Sep 03 22:38:42 so following http://wiki.openembedded.net/index.php/Gcc_issues_in_Intrepid_and_later page is useless Sep 03 22:39:13 doing apt-get to gcc 4.1 might fix, I had tought Sep 03 22:39:53 kayhan: Can you compile things outside OE? Sep 03 22:40:09 will try with a hello_world.c Sep 03 22:40:14 kayhan: It does look like your compiler is somehow badly broken Sep 03 22:40:51 things were easier in slack days : Sep 03 22:40:54 :) Sep 03 22:44:31 rp I already told him to do apt-get install --reinstall gcc-4.3 and cpp-4.3 Sep 03 22:44:36 so there is more broken Sep 03 22:45:53 did so Sep 03 22:46:18 Time for sleep for me :) Sep 03 22:46:21 'night all Sep 03 22:46:23 night Sep 03 22:46:27 good nite Sep 03 22:46:32 thanks Sep 03 22:47:18 nite rp Sep 03 22:47:38 kayhan is reinstalling your ubuntu no option? Sep 03 22:48:22 this will be 3rd buddy Sep 03 22:48:43 first had 64 bit, had to remove it because python-sycho or something like that Sep 03 22:48:54 didnt support 64 bit which is required for OE Sep 03 22:49:06 so installed 32 bit Sep 03 22:49:17 over and over Sep 03 22:49:52 python-psycho is only for speed up Sep 03 22:49:57 64 bit works Sep 03 22:50:05 I have it running at work Sep 03 22:50:41 over and over Sep 03 22:50:54 I will install it again I guess Sep 03 22:51:00 sorry for the first line Sep 03 22:51:10 typo Sep 03 22:51:52 I am just looking for a proper setup to do things on beagleboard Sep 03 22:54:42 *sigh* Sep 03 22:55:43 install the packages from koen from angstroem distribtuion and be happy Sep 03 22:55:52 you never have to hassle with bitbake Sep 03 22:56:09 need to compile kernel Sep 03 22:56:49 or use whatever linux distribution you want, if it mets the openembedded requirements Sep 03 22:58:48 hm why the manual lacks this part Sep 03 22:59:48 it decodes into two ways Sep 03 23:00:05 OE way or Code Sorcery way Sep 03 23:00:39 use the code sourcery toolchain Sep 03 23:00:46 less pain for you Sep 03 23:00:47 did try Sep 03 23:00:54 it was also problematic Sep 03 23:01:10 did install code sourcery Sep 03 23:01:23 and this morning I was fed up with it Sep 03 23:01:39 went to OE page Sep 03 23:02:30 the distribution have to be python at least 2.5 and a host compiler Sep 03 23:04:27 will re-install ubuntu :( Sep 03 23:04:33 will go over it again Sep 03 23:04:42 thats all for tonite Sep 03 23:05:23 got a job to go today 4 hrs left for sleeping Sep 03 23:05:28 thanks all Sep 03 23:05:33 thanks woglinde Sep 03 23:05:39 nite Sep 03 23:06:48 woglinde: can you start a build from scratch for some distro you care if you have cycles Sep 03 23:06:59 with the latest patches I bumped in Sep 03 23:09:52 khem ???? Sep 03 23:10:57 woglinde: I committed a rathar painful change which requires a complete rebuild Sep 03 23:11:14 I have tested quite a lot but then it was only me Sep 03 23:11:59 ah Sep 03 23:12:03 this discussion Sep 03 23:12:05 yes Sep 03 23:12:09 abi change Sep 03 23:12:22 no prob Sep 03 23:12:35 thank you Sep 03 23:12:43 hm for me Sep 03 23:12:52 dont know what the other says Sep 03 23:13:16 I believe its a good change that will eventually solve many problems users have Sep 03 23:18:45 sure Sep 03 23:19:01 let us see if koen shows up Sep 03 23:19:10 but I think he is at sleep now Sep 03 23:19:11 hehe Sep 03 23:19:29 khem: what was the original need for the change, out of curiosity? Sep 03 23:19:43 kergoth hm was discussed at the ml Sep 03 23:19:51 ah, k, i'll read the thread Sep 03 23:20:05 * kergoth 's not doing very well at keeping his inbox empty :) Sep 03 23:20:38 so good nite Sep 03 23:21:13 we really need to break up some of our classes and the functions/tasks in them. what a mess :) Sep 03 23:25:00 kergoth: there was disjoint between machine and abi and distros Sep 03 23:25:42 it was spread around Sep 03 23:26:02 and it was quite confusing at times which machines choses which abi etc. Sep 03 23:26:29 kergoth: toolchains configure differently depending upon what triplet you use in subtle ways Sep 03 23:26:40 and its important to get the triplet right Sep 03 23:27:01 so I tried to bridge the gap with the patch Sep 03 23:27:41 * khem drives home Sep 03 23:37:32 how can I alter package name for a library when debian names are enabled? Sep 03 23:39:41 ah, PKG_${PN} Sep 03 23:40:20 you can also set DEBIAN_NOAUTONAME_, from a quick glance at debian.bbclass, it seems Sep 03 23:41:56 kergoth: thanks, I'll try it Sep 03 23:43:01 denix: it appears you can set DEBIANNAME_ to set a custom package name, as well Sep 03 23:43:03 heh Sep 03 23:44:44 hmm, I wonder why nobody uses it, seems like people prefer python function with PKG_ override if DEBIAN_NAMES is defined... Sep 03 23:44:56 weird Sep 03 23:45:02 people like doing things the hard way? Sep 03 23:45:08 too lazy to read the bbclass? Sep 03 23:45:10 both? Sep 03 23:45:12 :) Sep 03 23:45:58 03Christopher Friedt  07org.openembedded.dev * r7676ddac22 10openembedded.git/recipes/cups/cups.inc: Sep 03 23:45:58 cups.inc: export DSOFLAGS Sep 03 23:45:58 * This fixes PR 5304 Sep 03 23:45:58 * Makes dash happy. Sep 03 23:45:58 Signed-off-by: Khem Raj Sep 03 23:46:05 :) Sep 03 23:47:04 * * OE Bug 5304 has been RESOLVED (FIXED) by raj.khem(AT)gmail.com Sep 03 23:47:06 * * cups-1.2.12 fails, No GNU_HASH in the elf binary, libcups.so.2 libcupsimage.so.2 Sep 03 23:47:08 * * http://bugs.openembedded.net/show_bug.cgi?id=5304 Sep 03 23:53:34 has anyone considered running an OE distro on an xbox? Sep 03 23:54:04 i saw some place selling xboxes for $58 AUD last night... Sep 03 23:59:05 grg: that's almost stock x86 - shouldn't be hard Sep 03 23:59:23 yeah Sep 03 23:59:24 meh, i need to get better at databases, figuring out how to do the tables and all Sep 03 23:59:29 anyone know any good resources on the subject? Sep 04 00:00:05 avoid anything with the workd UML in it Sep 04 00:00:34 heh, i want to learn about just plain and simple databases, not object oriented design, or ORMs (yet) Sep 04 00:08:58 sql? Sep 04 00:10:28 most oo databases are really just hierarchical databases in disguise Sep 04 00:10:53 SQL is a little strange, but it's not too hard once you wrap your head around it Sep 04 00:11:06 I should say "relational databases" Sep 04 00:11:10 heh, I remember the times when RDBMS' were not SQL... Sep 04 00:11:15 heh Sep 04 00:11:22 That wasn't too long ago :) Sep 04 00:12:32 yeah, only about 20 years ago... :) Sep 04 00:12:47 kergoth, I learned using a book called "Linux Database", which has almost nothing to do with Linux :) Sep 04 00:13:06 I can lend it to you next time I'm in Tempe, if you are not in a hurry. Sep 04 00:13:25 denix, back in the days when big iron ruled the database world. Sep 04 00:13:46 :) Sep 04 00:56:30 cminyard: okay, that'd be cool Sep 04 02:09:49 lots of applications have toolbars with no text Sep 04 02:10:05 they have tooltip information upon mouseover Sep 04 02:10:19 but what is one to do if they only have a touchscreen? Sep 04 02:11:05 Is there a good policy for telling X that the mouse is at a location, without activating a tap? Sep 04 02:12:07 I can drag the stylus from nowhere, onto an icon, then release. That does show the tooltip, but its not really intuitive for users. Sep 04 02:15:57 * m4t wants to tackle his udev woes tonight Sep 04 02:31:59 okay, 'Starting udev/etc/rcS.d/S03udev: line 50: can't create /tmp/uname: Read-only file system' Sep 04 02:32:12 that is without 'root=/dev/hde2 rw' Sep 04 02:32:14 ie. ro Sep 04 02:33:35 Starting udev Sep 04 02:33:35 m4t: /etc/default/udev Sep 04 02:33:44 denix alright Sep 04 02:33:52 denix , udev hangs with root rw Sep 04 02:34:21 until the timeout is reached, then console becomes unresposnive, probably because the necessary devices doesnt exist Sep 04 02:34:43 the above message says your rootfs is ro Sep 04 02:34:52 yea i changed it Sep 04 02:34:53 Starting udevroot: mount: mounting /dev/hde2 on / failed: No such file or directory Sep 04 02:34:54 to rw Sep 04 02:35:03 udev hangs if its rw, udev errors out if its ro Sep 04 02:35:23 not sure why udev hangs Sep 04 02:35:25 after udev is done 'hanging' it can't mount the root fs. and console dies Sep 04 02:35:28 yea Sep 04 02:35:42 maybe i should try mdev? really i'd like udev Sep 04 02:36:08 mounting blah on / failed - I also see it, but it's harmless Sep 04 02:36:19 are you sure it's udev? Sep 04 02:36:32 do you have getty on the right serial port? Sep 04 02:36:34 i think so. i dont know why it waits the full 30s for devices to settle Sep 04 02:36:39 denix yes, ttyS0 Sep 04 02:37:17 udev may take a long time to start the first time, that's why there is cache Sep 04 02:37:29 i.e. devcache Sep 04 02:38:05 denix, hrm Sep 04 02:38:10 i dont think it ever properly starts Sep 04 02:38:14 i wonder if its logging anywhere Sep 04 02:38:34 * m4t change udev_log="debug" Sep 04 02:39:01 Do you see the message "populating devcache"? Sep 04 02:39:31 that should come after udev started and initialized all devices, unless you disable devcache Sep 04 02:40:34 denix i've not seen that message Sep 04 02:41:06 looks like udev is started before syslog Sep 04 02:41:30 hmm, then it sounds like it can't settle some devices for you... Sep 04 02:44:54 tmpfs is mounted no probs Sep 04 02:47:29 sysfs.sh runs fine Sep 04 02:54:38 heh okay i pretty much typed out the init script by hand Sep 04 02:54:42 udev is 'settling' Sep 04 02:59:21 heh so /dev is blank Sep 04 02:59:36 even after it settles **** ENDING LOGGING AT Fri Sep 04 02:59:57 2009