**** BEGIN LOGGING AT Mon Nov 20 02:59:59 2006 Nov 20 03:30:41 03rwhitby 07org.oe.dev * r0080c426... 10/ (3 files in 3 dirs): slugos-init: Removed ixp4xx_mac module (now compiled into kernel) Nov 20 05:20:07 03pfalcon 07org.oe.dev * rbfbb5ffb... 10/ (1 conf/distro/include/preferred-opie-versions.inc): Nov 20 05:20:07 preferred-opie-versions.inc: Use week assignment for PREFERRED_VERSION_libopie2 Nov 20 05:20:07 * libopie2 is a special package in OPIE. Its interface is mostly stable, Nov 20 05:20:07 and changes are made only to add support for new devices. So, let distro Nov 20 05:20:07 override the version to use. Nov 20 05:20:19 03pfalcon 07org.oe.dev * rfd880ea8... 10/ (6 files in 3 dirs): Nov 20 05:20:19 plinciv 0.16: Recover lost patch and save from extermination. Nov 20 05:20:19 * It may be from year 2003, but still a better start with porting than nothing Nov 20 05:20:19 at all. Nov 20 05:20:19 * And yes, one sweet day we'll have people doing something useful (like Nov 20 05:20:21 porting games), not just dumb kernel hacking ;-). Nov 20 07:24:09 good morning Nov 20 07:56:59 hi Nov 20 08:06:25 BitBake Build Tool Core version 1.6.1, bitbake version 1.6.1 Nov 20 08:06:25 real 6m41.835s Nov 20 08:06:25 user 5m41.480s Nov 20 08:06:25 sys 0m8.990s Nov 20 08:06:34 :) Nov 20 08:06:50 i test the new now :) Nov 20 08:11:07 good morning all Nov 20 08:12:01 good morning koen et al Nov 20 08:12:13 * likewise coffee coffee Nov 20 08:29:33 is the new recipe of xserver-kdrive_git work fine ? Nov 20 08:32:00 morning Nov 20 08:33:17 time to resolve hostap-modules Nov 20 08:36:07 for me this week : xinerama issue & firefox integration Nov 20 08:37:32 i found no info on xinerama & xfbdev , so my boss create a big fb device with the both in the kernel :'] Nov 20 08:37:37 Genesis: firefox 2.0? Nov 20 08:37:43 1.0.7 Nov 20 08:37:51 no time to try 2.0 Nov 20 08:38:15 i should package mozplugger and mplayerplug-in before Nov 20 08:38:52 03hrw 07org.oe.dev * r9bffa04e... 10/ (1 conf/local.conf.sample): local.conf.sample: libsdl-x11 and DISTRO="generic" instead o unmaintained familiar-unstable Nov 20 08:39:33 Genesis: then update it to 1.0.8 atleast Nov 20 08:39:50 ho there is a 1.0.8 stable , i look thx :) Nov 20 08:40:34 not in my tree Nov 20 08:41:41 Genesis: then try to update? Nov 20 08:43:42 i do this ten minutes ago Nov 20 08:43:44 ;) Nov 20 08:44:09 i see a new minimo but no new ff Nov 20 08:44:31 i'd compiled dbud 1.x :) Nov 20 08:44:51 dbus & cairo Nov 20 08:47:49 obergix[work], : is it you that work on robotic ? Nov 20 08:49:55 morning all Nov 20 08:50:30 hi RP Nov 20 08:51:57 hey RP Nov 20 08:54:54 03koen 07org.oe.dev * r697fec81... 10/ (1 packages/enca/enca_1.9.bb): enca: remove -I/usr/include from Makefiles and build native tools without -I Nov 20 09:02:34 I will REMOVE RRECOMMENDS = "hostap-modules" from hostap-utils Nov 20 09:08:14 koen: can you bump angstrom to hostap-utils 0.4.7? Nov 20 09:08:24 hrw|work: I think so Nov 20 09:08:49 PREFERRED_VERSION_hostap-modules ?= "0.4.7" Nov 20 09:09:06 utils Nov 20 09:09:15 angstrom should not use hostap-modules at all Nov 20 09:09:36 angstrom == 2.6 == hostap from kernel Nov 20 09:09:51 ipaq == hh.org == broken hostap in kernel Nov 20 09:10:06 fscking hh.org Nov 20 09:10:46 03hrw 07org.oe.dev * re7c3c58f... 10/ (4 files in 2 dirs): Nov 20 09:10:46 hostap-utils: removed rrecommending hostap-modules Nov 20 09:10:46 - most of our target devices run kernels newer then 2.6.14 so they have Nov 20 09:10:46 hostap modules from kernel as kernel-module-hostap* Nov 20 09:10:46 - hostap modules and utils are now both added into dependency of task-base Nov 20 09:10:46 for machines with pcmcia or pci feature Nov 20 09:13:24 03koen 07org.oe.dev * r7f11ce9c... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: prefer hostap-utils 0.4.7 Nov 20 09:17:45 03rpurdie 07org.oe.dev * r56b8494c... 10/ (6 files in 2 dirs): classes: Clean up anonymous function whitespace Nov 20 09:29:09 so when will OE build for PS3? Nov 20 09:31:14 mithro: when someonme adds a apprpriate machine.conf and site file :) Nov 20 09:32:17 koen: site? its powerpc and we have this arch Nov 20 09:32:34 even better Nov 20 09:34:03 http://www.linuxdevices.com/news/NS6057216598.html - someone? Nov 20 09:34:16 it's not IBM Cell ? Nov 20 09:34:58 Genesis: fedoracore ppc works on it Nov 20 09:35:14 ho oki Nov 20 09:35:20 "Presentations that are commercial advertisements or sales pitches are not appropriate for this conference." Nov 20 09:35:26 bye-bye mvista ;) Nov 20 09:38:41 someone remember when Easter is in 2007? Nov 20 09:39:20 8.04.2007 Nov 20 09:39:22 8+9 april Nov 20 09:39:28 20070408 I meant Nov 20 09:39:46 * koen hugs www.mozilla.org/projects/calendar/caldata/Dutch_Holidays_2006-2011.ics Nov 20 09:40:49 thx for link Nov 20 09:42:19 ics is cool stuff :) Nov 20 10:25:36 wtf does librcf_0.4.bb set IPKG_ARCHS? Nov 20 10:25:59 heh Nov 20 10:42:33 03rpurdie 07org.oe.dev * rd7191428... 10/ (1 packages/librcf/librcf_0.4.bb): librcf: Remove insane IPKG_ARCHS from the .bb file. This is a setting for conf/, not a .bb file Nov 20 10:42:38 03rpurdie 07org.oe.dev * reb2dfd86... 10/ (65 files in 7 dirs): Convert IPKG_ARCHS -> PACKAGE_ARCHS, IPKG_EXTRA_ARCHS -> PACKAGE_EXTRA_ARCHS in preparation for deb handling Nov 20 10:44:05 now we need to update docs Nov 20 10:44:29 RP: please send mail to announce/devel about reb2dfd86 change Nov 20 10:44:29 and hug sed Nov 20 10:46:15 hrw|work: ok Nov 20 10:47:52 ipkg should be able to handle .debs as well, right? Nov 20 10:48:43 koen: So I'm told... Nov 20 10:49:05 So in theory we could set up a feed with debs and ipkg could use that Nov 20 10:49:25 and the apt fanboys can install apt+dpkg and work with the same feeds Nov 20 10:49:33 koen: that remind me other thing - do we plan to ship dpkg/apt for angstrom? Nov 20 10:49:43 koen: Maybe... :} Nov 20 10:50:13 koen: dpkg will complain about archs - isnt it? Nov 20 10:50:46 dpkg can be tricked into not-complaining Nov 20 10:51:25 * RP had to hack dpkg-deb to get rid of warnings about extra fields in control files :-/ Nov 20 10:51:52 hrw|work: support apt/dpkg in angstrom is 'planned' Nov 20 10:52:06 koen: ok Nov 20 10:54:31 good morning Nov 20 10:54:53 Good morning Nov 20 11:06:00 morning all Nov 20 11:17:44 hi Liam Nov 20 11:20:56 hey Richard Nov 20 11:25:04 * gerwinin is looking for a manual on how to install grub on a cf Nov 20 11:33:25 gerwinin: cf as /dev/hdX? Nov 20 11:34:03 hrw in my target system it is /dev/hda but on my host system it is /dev/sdb1 Nov 20 11:36:15 gerwinin: grub-install /dev/sdb does not work? Nov 20 11:58:11 http://people.freedesktop.org/~krh/akamaru.git/ Nov 20 11:58:17 that would rock on a touchscreen Nov 20 12:24:19 RP: any plans for s/IPKG_INSTALL/PACKAGE_INSTALL/g ? Nov 20 12:24:36 koen: That would be next on the list Nov 20 12:24:40 03koen 07org.oe.documentation * r4caf2890... 10/ (1 usermanual/usermanual.xml): usermanual/usermanual.xml: adjust ipkg archs to latest metadata Nov 20 12:25:00 gerwinin: is that OE server 1U or 2U? Nov 20 12:25:09 koen: The idea is to try and sync up those changes from poky so package_deb.bbclass will work Nov 20 12:25:35 koen: Once that works, we can look at the deb images... Nov 20 12:25:36 IPKG_INSTALL is for rootfs_deb.bbclas :) Nov 20 12:26:00 koen: Syncing that is an aim though :) Nov 20 12:26:20 koen: Also things like IMG_PKGTYPE Nov 20 12:26:27 Koen 1u Nov 20 12:26:47 gerwinin: ok, I'll start lobbying for some rackspace :) Nov 20 12:27:12 Koen: I want to make a seperate branch for epia based boards Nov 20 12:27:20 why a branch? Nov 20 12:27:28 branching is nonsense for most stuff Nov 20 12:27:52 Koen: so that it can be developed without interfering in the arm and fic stuff Nov 20 12:28:07 that's the whole point of OE Nov 20 12:28:13 gerwinin: think twice before creating branch Nov 20 12:28:27 hrw/koen: it was more a proposal Nov 20 12:28:40 thinking OE is arm only sounds like emte style FUD Nov 20 12:29:16 hrw/koen: okay nevermind as long as we get some good mirrors than all is okay with me Nov 20 12:29:26 gerwinin: after few months with .oz354x branch I think about building OZ 3.5.5 from .dev Nov 20 12:29:47 hrw/koen: is one server enough ? Nov 20 12:30:07 gerwinin: for source mirror? Nov 20 12:30:12 gerwinin: The key is more a good distro setup and tags Nov 20 12:30:16 more is better, but I think I can only host one here Nov 20 12:31:00 RP: I am busy with that now , got most things running for most epia based boards Nov 20 12:32:35 RP: but I am bit stuck now where they grep the /dev/proc and load the xserver from that Nov 20 12:33:12 gerwinin: Can you detect an epia system based on cpuinfo before that? Nov 20 12:33:30 RP: that is what I am looking at now Nov 20 12:33:49 or make Xfbdev the default rather than Xipaq or whatever it is atm Nov 20 12:34:26 Yeah the thing is that in most cases they don't define a default so I am busy doing that Nov 20 12:36:15 XSERVER=Xipaq Nov 20 12:36:15 if [ -f /usr/bin/Xfbdev ]; then Nov 20 12:36:15 XSERVER=Xfbdev Nov 20 12:36:16 fi Nov 20 12:36:16 if [ -f /usr/bin/Xepson ]; then Nov 20 12:36:16 XSERVER=Xepson Nov 20 12:36:17 fi Nov 20 12:36:19 if [ -f /usr/bin/Xorg ]; then Nov 20 12:36:21 XSERVER=Xorg Nov 20 12:36:23 fi Nov 20 12:36:46 not very pretty, but it should work Nov 20 12:37:37 from http://projects.linuxtogo.org/plugins/scmsvn/viewcvs.php/trunk/base/xserver-common/X11/Xserver?rev=8834&root=gpe&view=auto Nov 20 12:38:13 RP: syncing xserver-kdrive-common in poky with gpe's xserver-common would be usefull as well Nov 20 12:41:32 koen: I guess one of the gpe guys could need to do that... Nov 20 12:41:40 I'd take patches to poky's Nov 20 12:43:10 ah, cool Nov 20 12:43:13 wasn't sure of that Nov 20 12:43:28 * koen adds to TODO Nov 20 12:44:01 koen: As long as it doesn't change the behaviour of poky - mallum will watch those changes ;-) Nov 20 12:44:11 speaking of OE being ARM only .... Nov 20 12:44:24 Has anyone looked into adding PS3 support to OE? Nov 20 12:44:27 RP: if it changes behaviour it's only due to bugfixes Nov 20 12:44:43 koen: bugfixes should be ok Nov 20 12:44:45 RP: e.g. using a correct DPI Nov 20 12:45:25 RP: if a certain freelancer working on a phone doesn't beat me to it Nov 20 12:45:26 koen: hmm :-/. You might have to send the patch to mallum ;-) Nov 20 12:53:12 03rpurdie 07org.oe.dev * r01b05c1b... 10/ (1 classes/base.bbclass): base.bbclass: whitespace cleanup, remove unneeded stage addtask call Nov 20 12:54:14 RP: does do_stage(){ foo } still work in recipes? Nov 20 12:54:23 koen: yes Nov 20 12:55:12 gerwinin: I sent a request for rackspace to the uni Nov 20 12:55:21 gerwinin: lets see if they grant it Nov 20 12:55:31 koen: It just removes that task from being visible to bitbake = parsing speedup (populate_staging is the task visible to bitbake) Nov 20 12:56:05 Crofton: ps3 is ppc machine so it only need machine config and kernel Nov 20 12:56:29 RP: aren't you affraid the coffee manufacturers of the world will come after you? Nov 20 12:56:42 parse time is too short to make coffee nowadays Nov 20 12:56:47 koen: :) Nov 20 12:57:16 koen: 18 seconds for poky is way too short for that :) Nov 20 12:58:25 hey zecke Nov 20 12:58:29 RP: hey, use python2.5 Nov 20 12:58:32 hi zecke Nov 20 12:58:36 RP: use cProfile for profiling :) Nov 20 12:58:50 zecke: What's the different? Just speed? Nov 20 12:58:54 RP: tell me how I broke poky :) Nov 20 12:59:01 RP: cProfile is the difference Nov 20 12:59:12 zecke: You didn't, I screwed up ;-) Nov 20 12:59:19 RP: you suck :) Nov 20 12:59:29 zecke: There was a "return" in an anonymous method ;-) Nov 20 12:59:30 Shoragan: I'm in favor of having a BitBake DEB Nov 20 12:59:45 hehe, so I messed up as well :) Nov 20 12:59:56 Shoragan: with 1.8 we will require a module to be compiled Nov 20 13:00:02 kergoth: 2006-11-20 14:25:06 1Gm7KP-00082b-21 kergoth.com [207.210.106.158] No route to host Nov 20 13:00:10 zecke: No, it was only in poky ;-) Nov 20 13:00:12 kergoth: 2006-11-20 14:25:06 1Gm7KP-00082b-21 == clarson@kergoth.com R=dnslookup T=remote_smtp defer (113): No route to host Nov 20 13:00:14 zecke: we have bitbake deb Nov 20 13:00:39 zecke: I'd be interested in your thoughts on the genereateDot function... Nov 20 13:00:39 hrw|work: sure, but the new one will be for different archs Nov 20 13:00:40 zecke: I sent updates to it for Shoragan yesterday Nov 20 13:01:06 zecke: then Shoragan will change one entry in control file.. Nov 20 13:01:08 hrw|work, my wife will kill me if I purchase a PS3 for "testing" :) Nov 20 13:01:17 Crofton: how much is ps3 btw? Nov 20 13:01:35 not sure, rouughly 700usd Nov 20 13:01:36 $800 a few hours ago Nov 20 13:02:04 it should make an interesting platform for software radio Nov 20 13:02:16 zecke: With the getVarFlag optimisation I made, feeder and handle are at the top of the profile graph again :-/ Nov 20 13:02:19 ~change 800 usd to eur Nov 20 13:02:28 800.00 United States Dollar (USD) makes 623.265 Euro (EUR) (from http://www.xe.com/) Nov 20 13:02:32 RP: finally! Nov 20 13:02:36 too much for me Nov 20 13:03:24 RP: it is always funny how easy things are :) Nov 20 13:04:11 zecke: :) Nov 20 13:04:27 zecke: Did you see my mails to oe-devel? Nov 20 13:04:38 RP: yes, didn't read it though Nov 20 13:04:43 RP: now browsing commit logs Nov 20 13:05:11 zecke: There's a profile in there and I'd appreciate your comments on the generateDotGraph function Nov 20 13:05:29 (which isn't committed yet until its beend discussed) Nov 20 13:05:58 RP: where is generateDotGraph? I didn'ts ee that in the mail Nov 20 13:07:24 zecke: http://lists.linuxtogo.org/pipermail/openembedded-devel/2006-November/000811.html Nov 20 13:07:58 zecke: i.e. http://www.rpsys.net/openzaurus/temp/bitbake_depgraph.patch Nov 20 13:10:08 * RP -> food Nov 20 13:12:53 RP: tough stuff, specially the DEPENDS stuff :) Nov 20 13:13:00 I'm unsure about the consequences Nov 20 13:13:18 RP: we should search for join/split and for loops in bitbake Nov 20 13:25:11 ~change 398 usd to eur Nov 20 13:25:16 398.00 United States Dollar (USD) makes 310.089 Euro (EUR) (from http://www.xe.com/) Nov 20 13:32:35 hi Nov 20 13:32:46 hey CoreDump|home Nov 20 13:34:10 hrw|work: ping Nov 20 13:37:12 zecke: taskdata and runqueue are bad offenders for loops :-/ Nov 20 13:37:27 zecke: no real joins or splits though Nov 20 13:37:41 RP: sure, but more like does it search for stuff it seldomly finds Nov 20 13:38:06 zecke: not really Nov 20 13:38:21 zecke: Its also reasonably fast, world builds excepted Nov 20 13:38:50 RP: I'm sadly out of time until friday Nov 20 13:38:53 (it only operates on the data it actually uses so in most cases can drop things) Nov 20 13:39:18 zecke: np, I know the feeling... Nov 20 13:39:26 03koen 07org.oe.dev * r6b50eeff... 10/ (1 classes/mozilla.bbclass): mozilla bbclass: inherit pkgconfig Nov 20 13:39:45 zecke: Any thoughts on the DotGraph - we have one person requesting we retain the DEPENDS code... Nov 20 13:39:48 RP: my OE contribution this week is installing edgy, and asking sysadmin to open anoncvs, svn, mtn port Nov 20 13:40:08 RP: likewise summed up my thoughts as well Nov 20 13:40:18 RP: I'm not too deep in DEPENEDS/RDEPENDS but I like it/am used to it Nov 20 13:40:25 CoreDump|home: pöng Nov 20 13:40:40 RP: another 28 seconds faster Nov 20 13:40:40 CoreDump|home: only spitz has own config Nov 20 13:40:54 As time goes on, I suspect there will be dependencies that don't show up in DEPENDS/RDEPENDS Nov 20 13:40:59 real 1m16.556s Nov 20 13:40:59 user 1m8.323s Nov 20 13:40:59 sys 0m5.579s Nov 20 13:41:10 zecke: cool :) Nov 20 13:41:22 20 only, but still :) Nov 20 13:41:43 hrw|work: actually c7x0 and akita have their own udev as well Nov 20 13:41:45 RP: hehe, I always think I'm cutting down the numbers and then you come :) Nov 20 13:42:15 zecke: We've both made significant contributions ;-) Nov 20 13:42:33 thanks :) I will put my name first (stupid ego) Nov 20 13:42:56 from thursday to monday we have almost lost a minute in parsing time Nov 20 13:43:35 zecke: I also just dropped an extra addtask from base.bbclass which might give us another speedup in OE (it was already missing in poky) Nov 20 13:43:44 CoreDump|home: feel free to make them machine related Nov 20 13:44:00 CoreDump|home: btw, I'm thinking about using .dev for 3.5.5 Nov 20 13:44:14 zecke: Most of the reaming getVarFlag calls are for "content" Nov 20 13:44:38 RP: who is paying us for that shit anyway? Nov 20 13:44:42 hrw|work: ohhh that would be a lot of trouble =) But we'd find out which changes didn't make it from .oz to .dev =) Nov 20 13:44:52 zecke: nobody :-( Nov 20 13:44:57 I'm not even getting postcards Nov 20 13:45:24 zecke: Not even one? Nov 20 13:45:27 zecke: hehe Nov 20 13:45:28 RP: one ;) Nov 20 13:45:36 CoreDump|home: and we base on old soft Nov 20 13:45:36 zecke: ;-) Nov 20 13:45:47 RP: but is like sending them yourself Nov 20 13:46:08 RP: Do you know "Club Mate" Nov 20 13:46:24 zecke: I just didn't want to be accused of breaking the licence ;-) Nov 20 13:46:26 hrw|work: old but well-working soft =) The user doesn't notice an older system base. He only looks out for applications Nov 20 13:46:30 zecke: no... Nov 20 13:46:38 RP: it is a beverage Nov 20 13:46:39 hrw|work: anyway, its fine with me Nov 20 13:47:05 RP: http://wiki.whatthehack.org/index.php/Club-Mate Nov 20 13:47:29 RP: for the dot code, I really won't look into the patch before friday. Commit if you like, we will work it out Nov 20 13:48:16 zecke: We can probably pull the DEPENDS/RDEPENDS out of taskdata and have three typres of output meaning we keep what we have but also gain the full task graph Nov 20 13:48:25 zecke: I'll look at it some more Nov 20 13:49:19 RP: Ah and I hate you. This BBHandler and BBTasks optimisation was in the new parser already :) Nov 20 13:49:32 RP: you took one of the things that make it different :) Nov 20 13:49:36 zecke: ;-) Nov 20 13:50:00 zecke: I hadn't looked the difference but it makes sense... Nov 20 13:51:26 BTW: It would be nice if one could fix parse errors :) Nov 20 13:51:42 real 0m40.183s Nov 20 13:51:42 user 0m34.353s Nov 20 13:51:43 sys 0m3.410s Nov 20 13:51:47 still faster :) Nov 20 13:52:26 zecke: Getting closer in speed though :} Nov 20 13:52:32 RP: We should create another OE repository with 10.000 dummy bb files Nov 20 13:52:46 RP: and I lack the DEPENDS calculation and have parse errors Nov 20 13:53:42 my bb is 6min40 Nov 20 13:53:46 We'll get the parse errors fixed sooner or later. Merging them into poky is a nightmare though Nov 20 13:54:25 Genesis: send me a postcard, try the new parser Nov 20 13:54:41 s/a/your/ :( Nov 20 13:54:54 hopes no one believes I really hate RP Nov 20 13:55:03 * zecke hugs RPs and leaves torwards the work Nov 20 13:55:25 zecke, : i use svn syncing this morning Nov 20 13:56:36 where is this new parser ^^ Nov 20 13:56:46 zecke: give me command used to get parse errors and I will fix them Nov 20 13:57:40 cd bitbake-parser/standalone Nov 20 13:57:41 make Nov 20 13:57:54 ./bitbake-helper /your/bb/files/packages/*/ Nov 20 13:57:56 How to raise my bloodpressure: "if os.access(os.path.join(os.path.join(tmpdir, "stamps"),"do_packages"),os.R_OK):" in package)ipk.bbclass Nov 20 13:57:57 ./bitbake-helper /your/bb/files/packages/*/*.bb Nov 20 13:58:09 RP: hehe Nov 20 13:59:05 zecke: and where in bitbake svn it is? Nov 20 13:59:12 trunk Nov 20 13:59:20 argos-wlan:~/Projects/OpenEmbedded/parser/trunk/bitbake-parser/standalone koen$ ./bitbake-helper ~/Projects/OpenEmbedded/org.openembedded.dev/packages/*/*.bb Nov 20 13:59:20 -bash: ./bitbake-helper: Argument list too long Nov 20 13:59:21 trunk/bitbake-parser Nov 20 13:59:29 koen: "quotest" Nov 20 13:59:42 zecke: fetched Nov 20 13:59:43 koen: it ain't DOS, nor Linux/GNU/HURD Nov 20 14:01:57 In fact that whole section of code in package_ipk looks like its totally useless. We write out a stamp, then delete it Nov 20 14:09:05 Genesis: did you see the new release of gnash? Nov 20 14:10:33 Genesis: use time, use profiling, send us the output Nov 20 14:10:49 03koen 07org.oe.dev * rea1b8730... 10/ (1 packages/tinymail/tinymail_svn.bb): tinymail: fix parse errors Nov 20 14:10:55 03koen 07org.oe.dev * raf4fef7c... 10/ (1 packages/angstrom/task-angstrom-x11.bb): task-angstrom-x11: fix parse error Nov 20 14:11:50 03rpurdie 07org.oe.dev * rc647d363... 10/ (1 classes/package_ipk.bbclass): package_ipk.bbclass: Simplify os.path.join calls (from poky) Nov 20 14:11:57 03rpurdie 07org.oe.dev * rd6303c5e... 10/ (1 classes/kernel.bbclass): kernel.bbclass: unset MACHINE otherwise things break (from poky) Nov 20 14:12:39 RP: we should unset it in sanity? Nov 20 14:13:01 koen, : i wait for it and they do it Nov 20 14:13:20 koen, ; i complained because there was no stable release ;) Nov 20 14:13:24 zecke: Or maybe in oerunmake or something. I just know it breaks kernel builds badly Nov 20 14:13:35 RP: and binutils Nov 20 14:13:42 but pb with this last release is no one care about stabilise the stable release Nov 20 14:14:26 zecke: btw, I want to only call sanity once per bitbake run, just after conf file parsing. Can we do it in a backwards comptible way if we add a new task? Nov 20 14:14:37 zecke, : where is this trunk , Nov 20 14:14:43 not a new task. A new event Nov 20 14:14:59 RP: the issue is os.environ['MACHINE'] Nov 20 14:15:24 RP: a) remove export flag b) remove it from environ Nov 20 14:15:45 zecke: Its not exported. The problem comes when you do MACHINE=x bitbake Nov 20 14:16:07 RP: then it should be exported Nov 20 14:16:13 it comes from inheritFromOS Nov 20 14:16:26 and is in our bitbake.conf data/dict Nov 20 14:16:27 zecke: Ah, I didn't know we did that internally... Nov 20 14:16:44 We should nuke its export flag then Nov 20 14:16:55 RP: and remove it from os.environ Nov 20 14:17:02 zecke: yes Nov 20 14:17:38 only matchesenv is set, so killing it from os.environ is enough Nov 20 14:22:12 zecke: 1st part pushed Nov 20 14:22:38 03hrw 07org.oe.dev * rcf74582c... 10/ (25 files in 13 dirs): many packages: fix parse errors Nov 20 14:28:53 182 parse errors left Nov 20 14:36:34 less Nov 20 14:36:45 I'm pushing another set Nov 20 14:36:53 "Only 182 ones left in rcf74582c2f1f66b88ee6661c6cdf7de75b0df961" Nov 20 14:37:04 ah Nov 20 14:37:16 160 left Nov 20 14:37:22 03hrw 07org.oe.dev * rf063ce04... 10/ (25 files in 15 dirs): many packages: fix parse errors #2 Nov 20 14:38:46 * mrz80 is away: Metaphysically I'm still here; physically, I'm elsewhere Nov 20 14:38:48 161 actually Nov 20 14:39:23 nope, spurious awk bug Nov 20 14:42:06 http://lists.linuxtogo.org/pipermail/openembedded-devel/2006-November/000821.html Nov 20 14:42:26 03hrw 07org.oe.dev * r01d82c54... 10/ (16 files in 2 dirs): kernels: fix parse errors Nov 20 14:42:35 awesome :) Nov 20 14:43:10 from 459 to ~140 Nov 20 14:44:16 Can we nuke the libxt depends from pango? The patches to autoconf should remove that problem? Nov 20 14:44:36 what's libxt? Nov 20 14:47:52 コエン Nov 20 14:50:31 'X toolkit intrinsics' Nov 20 14:50:33 hmmm Nov 20 14:50:47 koen: Its something we don't need, its only there due to broken standard autoconf Nov 20 14:51:16 Genesis: so why is your bitbake so slow? Nov 20 14:51:45 i donno Nov 20 14:52:16 i'm in 1.6.3 Nov 20 14:52:30 i've no ram in this macine Nov 20 14:52:50 $ free Nov 20 14:52:50 total used free shared buffers cached Nov 20 14:52:50 Mem: 515940 447176 68764 0 34708 236416 Nov 20 14:53:00 perharps that. Nov 20 14:53:20 Genesis: see above, do a profiled run Nov 20 14:59:50 RP: if that's the case, remove it Nov 20 15:01:36 107 left Nov 20 15:02:28 hrw|work: how fast is lexing these files? Nov 20 15:02:30 03hrw 07org.oe.dev * rfa629947... 10/ (43 files in 17 dirs): another set of parse errors fixing Nov 20 15:02:34 03rpurdie 07org.oe.dev * r8f5e674e... 10/ (1 classes/package_ipk.bbclass): package_ipk.bbclass: Fix invalid syntax added earlier Nov 20 15:02:45 03rpurdie 07org.oe.dev * r8c0da7de... 10/ (8 files in 2 dirs): pango: Remove DEPENDS on (lib)xt. OE's autoconf has a patch which should fix this problem and avoid this DEPEND (from poky) Nov 20 15:02:56 time ~/devel/bitbake/bitbake-trunk/bitbake-parser/standalone/bitbake-helper */*.bb 2>&1|grep Error|wc 107 322 5285 Nov 20 15:02:58 real 0m0.974s Nov 20 15:03:00 user 0m0.382s Nov 20 15:03:03 sys 0m0.147s Nov 20 15:03:18 that is good right? Nov 20 15:03:23 very Nov 20 15:03:36 sadly we spend most of the time in python Nov 20 15:04:38 psokolovsky_: we are finally at a point where parsing actually shows up in a log on initial parsing Nov 20 15:07:22 <_slacker_> hi all, exists any repositorie with ARM pre-compiled packages?? Nov 20 15:07:22 _slacker_: http://www.angstrom-distribution.org/repo/ Nov 20 15:07:57 real 0m0.566s Nov 20 15:07:58 user 0m0.244s Nov 20 15:07:58 sys 0m0.293s Nov 20 15:08:06 ~7k packages per second Nov 20 15:11:50 85 files left Nov 20 15:11:52 03hrw 07org.oe.dev * rb0bcbd55... 10/ (23 files in 17 dirs): another set of parse errors fixing Nov 20 15:12:30 koen: bad news, I'm unlikely to get down this number Nov 20 15:12:34 flex --fast might be faster Nov 20 15:12:45 perl recipes are nightmare Nov 20 15:12:48 require ${@bb.data.getVar('P', d , 1).replace('-native-', '_')}.bb Nov 20 15:12:55 but we spend 30 seconds in python afterwards Nov 20 15:12:59 how to fix it for parser? Nov 20 15:13:07 require "${@bb.data.getVar('P', d , 1).replace('-native-', '_')}.bb" is correct? Nov 20 15:14:12 hrw|work: the lexer script might need fixing Nov 20 15:14:37 hrw|work: it is ugly, but it looks correct Nov 20 15:15:11 FILENAME ([a-zA-Z_./]|{REF})(([-+a-zA-Z0-9_./]*)|{REF})* Nov 20 15:15:14 needs some extras Nov 20 15:15:35 start by adding an '@' Nov 20 15:15:46 bitbakescanner.l Nov 20 15:19:23 03hrw 07org.oe.dev * r7fcf5ee3... 10/ (22 files in 17 dirs): another set of parse errors fixing Nov 20 15:19:27 when I get a backtrace from a python app running on my target machine, the file paths always include the reference to ${B} from the host Nov 20 15:19:46 why does that happen? Nov 20 15:20:57 tkp: not enough sed magic in OE Nov 20 15:20:57 zecke: 65 errors left Nov 20 15:21:10 koen: nice, I hope I will land the parser this weekend Nov 20 15:24:55 koen: you meant 38 - right? Nov 20 15:25:18 export BUILD_CFLAGS += "-I${S}/include -DHAVE_CONFIG_H" Nov 20 15:25:25 it is not a bug for me Nov 20 15:25:44 export EXTRA_CFLAGS += " -DLINUX_2_6=1" Nov 20 15:25:47 same Nov 20 15:26:00 03hrw 07org.oe.dev * r218b565d... 10/ (30 files in 10 dirs): another set of parse errors fixing Nov 20 15:26:52 34 left Nov 20 15:26:57 perl and += Nov 20 15:27:20 hrw|work: write me a mail Nov 20 15:27:21 03hrw 07org.oe.dev * rec50d77c... 10/ (3 files in 3 dirs): another set of parse errors fixing Nov 20 15:28:16 RP: one could consider evaluating only RDEPENDS,DEPENDS as a speedup Nov 20 15:30:56 huhu Nov 20 15:31:00 zecke: sent to OEML Nov 20 15:38:10 zecke: Considering them where? Nov 20 15:38:26 RP: As operation on the Syntax Tree Nov 20 15:38:48 RP: e.g. define that anonfuncs can not change (R)DEPENDS/ Nov 20 15:39:03 RP: we could partially evaluate the tree to get information :) Nov 20 15:39:17 tomas without and 'h'... hmm Nov 20 15:39:22 hi tf :) Nov 20 15:39:26 hi zecke Nov 20 15:40:08 zecke: stop scaring tf away! Nov 20 15:40:11 zecke: Its not just DEPENDS/RDEPENDS we'd need to consider. Who's to say which variables are involved overall? :-/ Nov 20 15:40:43 RP: AST-> evaluate DEPENDS,RDEPENDS Nov 20 15:40:58 RP: 1st) find leaves assigning getting RDEPENDS Nov 20 15:41:06 but it gets tricks with inlined python code again :) Nov 20 15:41:28 but basicly you will try to evaluate the tree in a greedy way Nov 20 15:41:43 koen: I was about to propose the great Free Software Berlin hike tour Nov 20 15:44:38 Hi! Nov 20 15:44:40 zecke: Good news! Nov 20 15:44:56 yes? Nov 20 15:45:11 reagrding parsing ;-) Nov 20 15:45:13 was afk Nov 20 15:45:58 psokolovsky: did it get faster for you? Nov 20 15:46:14 didn't try yet, will do Nov 20 15:50:25 http://wiki.laptop.org/images/e/ee/AP1_12.jpg - looks nice Nov 20 15:51:05 * mrz80 is back (gone 01:12:20) Nov 20 15:54:04 yeap , happy children :) Nov 20 15:54:18 who said lucky :] Nov 20 16:14:13 umm now what was the name of the merge program w/ GUI for mt merge? Nov 20 16:15:47 which one? Nov 20 16:16:00 CoreDump|home: meld, kdiff3, vim, emacs Nov 20 16:16:02 vimdiff, xxdiff, kdiff3, emacs, meld, FileMerge.app? Nov 20 16:16:14 I prefer FileMerge.app Nov 20 16:16:25 zecke: indeed Nov 20 16:16:33 koen, thanks for being an ass: "Nov 20 04:52:13 koen thinking OE is arm only sounds like emte style FUD" i am glad that you remember that i used bitbake+oe to build an x86 distro no wonder OE is degrading Nov 20 16:16:35 =) Thank, kdiff it has been Nov 20 16:16:51 koen: sometime the arrows confuse me Nov 20 16:20:14 I should have syntax fixes for three more bb files, once I get to the machine to do the commit Nov 20 16:21:36 hrw|work: looks like the perl bb files are a plot to write python code that looks like perl Nov 20 16:21:54 zecke: ;D Nov 20 16:23:34 cu Nov 20 16:31:57 03rpurdie 07org.oe.dev * r6219faf6... 10/ (4 files in 2 dirs): classes/package_*: Add IMAGE_PKGTYPE (from poky) Nov 20 16:32:45 a+ Nov 20 16:38:03 later Nov 20 16:43:04 hmh. I want to add a package to my gpe-image, so I added 'BOOTSTRAP_EXTRA_RDEPENDS += "lrzsz"' to my local.conf - however, even though the lrzsz package is built (an .ipk is generated), it does not show up in tmp/rootfs/usr/bin, nor in the final jffs2 image ? Nov 20 16:43:20 I did do 'bitbake -c clean gpe-image', but that didn't help Nov 20 16:44:17 BOOTSTRAP_* is deprecated Nov 20 16:49:35 koen: so what is the recommended way I should modify my local.conf file to include an extra package ? GPE_EXTRA_RDEPENDS? Nov 20 16:51:23 koen, can you have a look at patch at http://bugs.openembedded.org/show_bug.cgi?id=1586 ? Nov 20 16:57:13 koen: nevermind, apparantly I had to do "bitbake -c clean task-bootstrap" as well. Nov 20 16:58:29 hmm. if bootstrap is deprecated, how come the latest version of familiar-unstable.conf still contains 'PREFERRED_PROVIDER_task-bootsrap = "task-bootstrap"' ? Nov 20 17:01:07 IvoC: because familiar is deprecated as well Nov 20 17:02:06 psokolovsky: that looks about right Nov 20 17:02:25 ah. so what is the current non-deprecated distro ? Just opie ? Nov 20 17:02:32 opie isn't a distro Nov 20 17:02:41 try 'generic' Nov 20 17:03:03 ok I will Nov 20 17:03:30 koen, thanks, will commit sson, or I'm afraid going to have lots of conflicts with these parsing cleanups ;-) Nov 20 17:12:04 morning Nov 20 17:12:09 hey ggilbert_ Nov 20 17:12:35 * CoreDump|home sync his tree Nov 20 17:12:57 Anyone object to setting the minimum version requirement of .dev to 1.6.2? Nov 20 17:13:09 03coredump 07org.oe.oz354x * re1faccd9... 10/ (3 files in 3 dirs): zaurusd: Fix Poodles asound.state Nov 20 17:13:11 (for bitbake) Nov 20 17:13:13 03coredump 07org.oe.oz354x * r86093070... 10/ (3 files in 3 dirs): base-files: Change fstab for Poodle Nov 20 17:13:18 03coredump 07org.oe.oz354x * r91d43666... 10/ (1 packages/udev/udev_084.bb): udev: Bump PR Nov 20 17:13:20 03coredump 07org.oe.oz354x * r25e98f77... 10/ (1 packages/udev/udev_084.bb): udev: Set PACKAGE_ARCH = "${MACHINE} to make udevs behaviour identical on all devices Nov 20 17:13:25 03coredump 07org.oe.oz354x * r7d3ec3a4... 10/ (1 packages/altboot/altboot_1.1.0-wip.bb): altboot: Do not show SD-Boot option on Poodle Nov 20 17:13:29 03coredump 07org.oe.oz354x * r023d06e1... 10/ (1 conf/distro/openzaurus-3.5.4.2.conf): openzaurus-3.5.4.2.conf: Set PREFERRED_VERSION_altboot = "1.1.0" Nov 20 17:13:33 03coredump 07org.oe.oz354x * rb6ecf7c8... 10/ (1 conf/machine/poodle.conf): poodle.conf: Add altboot to BOOTSTRAP_EXTRA_DEPENDS & BOOTSTRAP_EXTRA_RDEPENDS Nov 20 17:13:37 03coredump 07org.oe.oz354x * r8a2741bf... 10/ (1 packages/linux/linux-openzaurus.inc): linux-openzaurus.inc: Add CMDLINE_ROOT_poodle for native root-on-SD Nov 20 17:13:44 03coredump 07org.oe.oz354x * r9eb98b7e... 10/ (3 files in 3 dirs): zaurus-updater-reloaded: Apply a change to the machine setup as suggested by hrw Nov 20 17:13:50 03coredump 07org.oe.dev * r87fdd2d3... 10/ (9 files in 6 dirs): altboot: Do not show SD-Boot option on Poodle Nov 20 17:15:23 RP: no, please go ahead Nov 20 17:15:33 RP: 1.6.x is marked 'stable' :) Nov 20 17:18:06 03rpurdie 07org.oe.dev * r8fc2e240... 10/ (1 conf/bitbake.conf): bitbake.conf: Add MACHINE into the path for the CACHE (when set). Allow machines/distro to override IMAGE_ROOTFS_SIZE_ext2. Add DEPLOY_DIR_DEB. (from poky) Nov 20 17:18:11 03rpurdie 07org.oe.dev * r7e6ad84a... 10/ (1 conf/sanity.conf): sanity.conf: Set minimum bitbake version to 1.6.2 Nov 20 17:19:10 parsing is indeed a lot faster Nov 20 17:19:28 with bitbake trunk Nov 20 17:20:18 is it normal to get a bunch of alignment traps when starting a GTK application? Nov 20 17:20:20 Should we have a 1.8 series based on the current 1.7 I wonder?... Nov 20 17:20:37 cbrake: No Nov 20 17:20:41 RP: I was wondering about that as well Nov 20 17:21:00 RP: zecke made some statements about a roadmap yesterday iirc Nov 20 17:21:32 koen: he was asking about where the C parser fitted in... Nov 20 17:22:37 aaaah, right Nov 20 17:23:51 is there any description on how to finetune the scripts/settings in /etc/ of an image that will be created? (I created a distro and a rootfs for my target platform but need to tweak some settings, else it is not booting/operating correctly) Nov 20 17:24:16 hrw said something about "finetune confs" as I understood it's supposed to be some file but I could not find it Nov 20 17:24:35 koen: What needs doing to bitbake trunk to make it suitable for common use? the dotgraph is one. Anything else? Nov 20 17:25:08 Maybe having a threading algorithm which would support rm_work better? Nov 20 17:26:00 -os.sleep(5) Nov 20 17:26:17 RP: that would be great. rm_work is great for quick image builds in RAM if it deletes leftovers fast enough. Nov 20 17:26:51 pH5: I've been hoping someone would write the threading algorithm ;-) Nov 20 17:26:53 rm_work has lost a lot of its usefullness due to threading Nov 20 17:27:11 since during a glibc compile you will fill up your disk anyway Nov 20 17:29:33 ~curse collies SD controller Nov 20 17:29:40 May you be reincarnated as a Windows XP administrator, collies SD controller ! Nov 20 17:35:19 03rpurdie 07org.oe.dev * r630c8f4d... 10/ (7 files in 2 dirs): apt 0.4.46 -> 0.4.46 (sync with changes in poky) Nov 20 17:36:13 0.4.46 -> 0.4.46 Nov 20 17:36:16 * koen wonders Nov 20 17:38:37 Hello Nov 20 17:44:20 koen: 0.6.46 -> 0.6.42.2 :-/ Nov 20 17:45:24 03koen 07org.oe.dev * r3d939eee... 10/ (1 packages/esound/esound-gpe_20020817.bb): esound-gpe: move to the new gpe location, with a slightly confusing date thanks to the wonders of svn Nov 20 17:45:25 mtn: renaming packages/apt/apt_0.6.46.bb to packages/apt/apt_0.6.46.2.bb Nov 20 17:45:29 now I see :) Nov 20 17:46:02 could someone donate me 4 weeks of spare time so I can get .deb feeds + ipkg working? Nov 20 17:47:43 :> Nov 20 17:47:57 koen: Hi. fontconfig seems to be where the unaligned traps are taking place. Nov 20 17:48:07 aaah Nov 20 17:48:16 Program received signal SIGBUS, Bus error. Nov 20 17:48:16 0x40066584 in FcValueCanonicalize () from /usr/lib/libfontconfig.so.1 Nov 20 17:48:24 ~lart fontconfig Nov 20 17:48:25 * ibot eats fontconfig and falls over dead Nov 20 17:48:30 :) Nov 20 17:48:58 koen: I don't know if all the problems are into fontconfig, because after the first SIGBUS, well Nov 20 17:49:39 But, as most of programs generating alignment traps where gui programs, it's probably that most problems are into fontconfig Nov 20 18:21:55 hi all Nov 20 18:22:37 hi Nov 20 18:23:09 can someone help me out with a little problem I have with devshell & configure.ac? Nov 20 18:23:48 I know nothing about devshell, but I did write a couple of configure.ac scripts Nov 20 18:24:03 ... as my attempts to use uinput for keyboard & ts events failed (mostly) I decided to give opie a try ... Nov 20 18:24:26 so I added the following in my configure.ac: AC_CHECK_HEADER([opie2/odevice.h],[AC_DEFINE(ODEVICE, 1, [opie device information class available])],[]) Nov 20 18:24:49 ... but ./configure never finds the file in the devshell environment Nov 20 18:25:39 the file is staged in ./staging/arm-linux/include/opie2 Nov 20 18:26:14 any idea what I do wrong? Nov 20 18:27:08 I am not so experienced in OE, I do not know for sure what variables it is setting/replacing to make autoconf find the stuff where it is supposed to be; when creating my script I added options to search for headers/libs in specific paths Nov 20 18:27:43 so in my bb file I just passed the parameter pointing to the directory where it should look for my stuff; in your case.. you could try setting CFLAGS accordingly before the header check Nov 20 18:27:44 Jin^eLD: how can I do this - is there a macro, or an env var? Nov 20 18:27:58 do you want a "clean" solution or just a hack to test? Nov 20 18:28:27 Jin^eLD: for a moment a hack is ok - once I see this approach works I'll have to clean it up of course Nov 20 18:29:18 well, simply try adding staging/arm-linux/include to your CFLAGS Nov 20 18:29:38 then the header check should pass Nov 20 18:29:51 it is also a good idea to look at config.log Nov 20 18:30:02 because sometimes some header will include some other crap that is defined elsewhere and so on Nov 20 18:30:46 CFLAGS is already set accordingly: CFLAGS=-I/stuff/tmp/org.openembedded.oz354x-collie/staging/arm-linux/include -fexpensive-optimizations -fomit-frame-pointer -frename-registers -O2 Nov 20 18:32:54 moment, phone Nov 20 18:35:43 3hi andersee Nov 20 18:39:02 re Nov 20 18:39:07 pgfeller: what does config.log say? Nov 20 18:39:44 Jin^eLD: you were right ... it can not find some other .h files: opie2/odevicebutton.h:23:21: qpixmap.h: No such file or directory Nov 20 18:40:08 pgfeller: there you go... Nov 20 18:40:52 Jin^eLD: yep ... will add those path(s) to CFLAGS for the moment - thanks for the help. Nov 20 18:41:05 no problem Nov 20 18:41:06 :) Nov 20 19:02:29 morning Nov 20 19:02:35 hrw: Morning Nov 20 19:03:12 hi hrw Nov 20 19:03:30 hi all Nov 20 19:03:32 hrw: you said I had to edit "confs" to finetune the resulting /etc/ in my rootfs... I can't find it Nov 20 19:03:47 ~lart linux for not allowing to set ondemand as main cpufreq governor Nov 20 19:03:48 * ibot executes killall -TERM linux for not allowing to set ondemand as main cpufreq governor Nov 20 19:04:35 Jin^eLD: configs of applications Nov 20 19:04:58 oh, that's what you meant :) I thought there was a global etc finetune configuration Nov 20 19:08:26 woglinde: howdy Nov 20 19:08:39 * andersee is a bit slow today Nov 20 19:09:45 andersee no prob Nov 20 19:09:48 me is sleepy too Nov 20 19:09:56 only 3 hours sleep Nov 20 19:10:41 * andersee had a luxurious 4 hours Nov 20 19:10:53 uh why only 4? Nov 20 19:10:53 is there something I need to know if I want to check for opie2/odevice.h with configure in oe's devshell; I always have errors in config.log (like qt2/include/qglobal.h:648: error: parse error before "c" ...) Nov 20 19:11:02 I need a nap Nov 20 19:11:10 woglinde: up late working Nov 20 19:11:11 pgfeller her is not #opie Nov 20 19:11:25 andersee I had to to some assigements for today Nov 20 19:12:25 woglinde: yep - I know (sorry); I'll try again in #opie - but it is very silent there .... Nov 20 19:12:50 andersee: have a good nap. we will keep it quiet here. Nov 20 19:12:58 hehehe Nov 20 19:13:25 pgfeller if it silenet theire than it will be silient about this topic here too Nov 20 19:13:42 ~hail 50M/s from source mirror Nov 20 19:13:44 * ibot bows down to 50M/s from source mirror and chants, "I'M NOT WORTHY!!" Nov 20 19:13:56 hrw: which one? Nov 20 19:14:01 likewise: localhost Nov 20 19:14:13 hrw: haha, ok Nov 20 19:14:54 likewise: if everything will go properly it will be available in ~month Nov 20 19:15:56 with some luck we might have another one next month as well Nov 20 19:16:55 I should bribe the snt to get gbit enabled :) Nov 20 19:20:40 D'oh Nov 20 19:20:54 ipkg list_installed | grep dbg | xargs ipkg remove Nov 20 19:21:05 * koen forgot the '| grep dbg' Nov 20 19:21:11 ;DDDD Nov 20 19:21:13 tne *one* time ipkg is fast Nov 20 19:21:36 it has 'if remove all then rm -rf /'? Nov 20 19:21:47 ~summon zecke Nov 20 19:21:49 apt takes out 20 clean, identical-looking phones, some extra hands, and pretends to be a telemarketer for a large corporation, so he gets delivered a phonelist containing zecke's coordinates Nov 20 19:22:08 hrw: probably "don't parse lists" Nov 20 19:22:19 koen: remove does not parse Nov 20 19:24:38 psokolovsky: ping Nov 20 19:35:50 koen: I'm interested in providing a mirror as well, but I would like to know what traffic to expect. Nov 20 19:36:39 likewise: with the current arrangement the primary source mirror gets hammered, the rest will virtually have no traffic Nov 20 19:36:48 unless you instruct OE otherwise Nov 20 19:37:08 03lrg 07org.oe.dev * r1f713602... 10/ (4 files in 3 dirs): mx31ads: Fixed kernel oops and upgraded to 2.6.19rc6 Nov 20 19:37:40 likewise: so you could point your local.conf to the local mirror to get fast downloads :) Nov 20 19:38:05 koen: what options are there to mirror OE. Is rsync in place? Nov 20 19:38:21 wget -r :) Nov 20 19:39:21 I plan to introduce own-mirror.bbclass which will use SOURCE_MIRROR_URL variable Nov 20 19:39:41 btw - progear can get bios flashed under linux ;D Nov 20 19:39:54 progear? Nov 20 19:40:21 likewise: 5 years old webpad which I play with recently Nov 20 19:40:32 likewise: Do you mean to generate the mirror or use it in OE? Nov 20 19:40:33 likewise: transmeta tm3200 (x86) based Nov 20 19:40:56 * hrw want bitbake_qa/bittest.. Nov 20 19:41:07 RP: both Nov 20 19:41:17 * ggilbert_ perks his ears up Nov 20 19:41:20 likewise: to use is simple Nov 20 19:41:35 ggilbert_: bitbake trunk had -c fetchall support Nov 20 19:41:42 oooooooh? Nov 20 19:41:58 RP: I have to think how to get it ignore compatible_machine Nov 20 19:42:29 RP: did fetchall fetch really everything from the network, i.e. so that subsequent builds did not need a network in place? Nov 20 19:42:32 ggilbert_: It also touches the .md5 files if a download would be attempted so it also solves the problem of when to get rid of archives Nov 20 19:42:39 RP: Nifty. Nov 20 19:42:52 I'll give it a try tonight Nov 20 19:42:59 ggilbert_: and it works nice - I grabbed world during last week Nov 20 19:43:01 ggilbert_: You need bitbake trunk Nov 20 19:43:03 since that basically solves my issues Nov 20 19:43:26 RP: does it also fetch all patches etc? Nov 20 19:43:27 fetch all, then delete everything older than 2 months or something Nov 20 19:43:38 ggilbert_: yes :) Nov 20 19:44:07 hrw: For compatible machines, add a variable which disables the skip package in base.bbclass Nov 20 19:44:21 nice trick Nov 20 19:44:59 Who wrote the code for it? Nov 20 19:45:25 ggilbert_: I ripped the fetcher code up a bit ;-) Nov 20 19:46:48 Well, if I'm ever in Leeds ( I think ti was Leeds), I owe you your drink of choice :) Nov 20 19:47:18 ggilbert_: we will have to not get him overdrinked Nov 20 19:47:24 ggilbert_: Newcastle but in the right country :) Nov 20 19:47:37 hehe Nov 20 19:48:10 likewise: It fetches everything, just like a normal bitbake run would Nov 20 19:48:42 RP: By everything, what do you mean? Nov 20 19:49:09 All machines? All packages? Nov 20 19:49:31 err I guess I should say all overrides Nov 20 19:49:40 since it isn't necessarily a machine Nov 20 19:49:48 ggilbert_: It takes a list of packages so bitbake ope-image -c fetchall would everything for that opie image as per the current local.conf Nov 20 19:50:10 ah cool. So I still need to account for overrides on my own Nov 20 19:50:26 Or iterate over machines/distros Nov 20 19:50:43 RP: this make very long list of them Nov 20 19:51:02 very few packages have machine specific downloads Nov 20 19:52:40 There is probably a case for writing something to simply iterate over every .bb file. Since the fetcher code is ioslated, this should be trivial to do though and any fetcher can benefit from what I've done... Nov 20 19:53:35 I can see someone writing bitfetch that does that... Nov 20 19:54:28 03lrg 07org.oe.dev * ra9cc7fa2... 10/ (5 files in 4 dirs): mx21ads: Initial support for the Freescale IMX21ADS Nov 20 19:55:09 thats worth looking for me Nov 20 19:55:30 but now I try my SOURCE_MIRROR = "1" change Nov 20 19:56:22 koen, apparently efika's are shipping Nov 20 19:56:43 NOTE: package linux-openzaurus-2.6.18-r5: task do_fetch: started Nov 20 19:56:51 for MACHINE="omap5912osk" Nov 20 19:57:06 ? Nov 20 19:57:21 Crofton: ignore this Nov 20 19:57:43 hrw: The question is does it build any other kernel providers though? :-/ Nov 20 19:58:06 RP: I did 'bitbake -cfetchall linux-openzaurus' Nov 20 19:58:16 hrw: ah, right :) Nov 20 19:58:34 hrw: I guess world will do that too Nov 20 19:58:38 but I will look into bitfetch Nov 20 19:58:51 RP: yeah the overrides isn't a big deal really. It'd be nice, but I can easilly sort that Nov 20 19:58:54 RP: sure, but I have ~80 fetch fails sodont want to wait Nov 20 19:59:15 hrw: np, I understand :) Nov 20 20:00:43 03rpurdie 07org.oe.dev * r1a8b8b46... 10/ (1 conf/machine/mx21ads.conf): mx21ads.conf: Fix IPKG_EXTRA_ARHCS reference Nov 20 20:00:43 ggilbert_: I guess you have code for that already anyway? Nov 20 20:01:07 Probably. Nov 20 20:01:12 I haven't looked at it for a while Nov 20 20:01:20 but it's not terribly difficult Nov 20 20:01:38 Kernels are the only place I can think of where the patches are common Nov 20 20:01:50 ggilbert_: Anything simply doing a bitbake world -c fetchall would be a big improvement over what we have atm... Nov 20 20:01:58 pretty much Nov 20 20:02:11 You'd better add a -k into that :) Nov 20 20:02:18 What does -k do? Nov 20 20:02:21 (continue upon failure) Nov 20 20:04:50 RP: thanks for that, looks like i had copied from mx31 before the change was made. cheers :) Nov 20 20:05:10 lrg: ;-) Nov 20 20:09:30 Crofton(_): that's good news Nov 20 20:11:31 our projects are still approved Nov 20 20:12:52 * koen foresees a sudden burst in powerpc support in OE Nov 20 20:12:58 heh Nov 20 20:13:16 good preparation for when PS3 supply stabilizes Nov 20 20:13:20 :) Nov 20 20:13:59 http://www.powerdeveloper.org/playstation.php Nov 20 20:14:25 crofton *g* Nov 20 20:16:14 hey gerwinin Nov 20 20:16:25 hey Koen Nov 20 20:16:52 Koen: my mini-itx distri is ready now Nov 20 20:17:03 gerwinin: cool Nov 20 20:17:12 gerwinin: you know where the bugtracker is :) Nov 20 20:17:20 * koen will try to merge some patches tomorrow Nov 20 20:17:49 Koen: yes , i am busy now with creating the website and stuff Nov 20 20:17:53 ~lart opencvs Nov 20 20:17:53 * ibot DoSes opencvs Nov 20 20:18:46 Koen: the only thing I would like to do is to create a programm to write the bootloader to the cf because that stays tricky Nov 20 20:22:12 gerwinin: mini-itx with a via/x86 processor? Nov 20 20:23:45 likewise: yes Nov 20 20:24:52 gerwinin: which board? Nov 20 20:25:33 gerwinin: I flashed linuxbios in a M10000 board. Boots the linux kernel in only a few secs after powerup. Nov 20 20:26:06 likewise : all via cn700 based boards, all jetway boards, all gigabyte boards Nov 20 20:26:25 mreimer, pong, if still online ;-) Nov 20 20:26:33 likewise: that is a cle266 board Nov 20 20:26:53 likewise: how did you do that ? Nov 20 20:27:01 mreimer, thanks for h2200 update, sorry that I didn't send email that I did trivial updates... Nov 20 20:27:42 likewise: they sent me today a dom I still need to start playing arround with that Nov 20 20:28:50 gerwinin: dom? Nov 20 20:30:55 likewise: Disk on module Nov 20 20:31:09 koen: Disk on module Nov 20 20:32:09 gerwinin: cf with bigger plug? Nov 20 20:33:20 Koen: yes :) Nov 20 20:36:17 <[g2]> anybod run monotone on a SMP system ? Nov 20 20:36:26 <[g2]> s/anybod/anybody/ Nov 20 20:36:32 yes (3x) Nov 20 20:36:41 opteron, athlonmp and c2d Nov 20 20:36:50 <[g2]> koen does it only use one processor ? Nov 20 20:36:52 it's singlethreaded, so you won't notice Nov 20 20:37:09 <[g2]> that's pretty silly no ? Nov 20 20:37:17 as njs said "it only use one cpu so you can use the other to watch a movie while it's pulling" Nov 20 20:37:22 * [g2] see 50% loading on the dual core Nov 20 20:38:05 <[g2]> koen it'd be way faster if one processor did I/O and the other the computations Nov 20 20:38:31 <[g2]> basically you could probably remove a lot of the I/O waiting Nov 20 20:38:41 <[g2]> (by read-ahead) Nov 20 20:39:08 koen: Who looks after gstreamer in .dev? Nov 20 20:39:38 RP: tmbinc and me most of the time Nov 20 20:39:53 koen: Have you looked at what's in poky? Nov 20 20:40:25 RP: "makes things faster" "ooops, makes it crash" "makes things faster" "ooops, another crash" Nov 20 20:40:58 the commit messages made me stay away from it Nov 20 20:41:00 koen: really? :-/ Nov 20 20:41:28 I got the impression it was stable now... Nov 20 20:41:30 we had a commit message about "making devil noises" Nov 20 20:42:09 RP: apart from liboil gst is working pretty neat on my devices Nov 20 20:42:35 koen: Do we need all these versions? Nov 20 20:42:44 RP: no we don't Nov 20 20:43:03 RP: .dreambox should have the best gst integration Nov 20 20:43:11 Hum, I think I've found the exact place where the missaligning is happening in fontconfig, but I'm starting to suffer a headache with this code Nov 20 20:43:42 <[g2]> koen you play with streaming much ? Nov 20 20:43:52 [g2]: nope Nov 20 20:44:08 [g2]: my efika would be part of the streaming experiments Nov 20 20:44:33 <[g2]> koen do you have efika installed ? Nov 20 20:44:38 hi all ! Nov 20 20:44:39 still waiting.... Nov 20 20:44:42 hey cyrilRomain Nov 20 20:44:55 <[g2]> koen how about skpye ? Nov 20 20:45:02 <[g2]> or skype Nov 20 20:45:05 <[g2]> :) Nov 20 20:45:07 [g2]: eeeeeeeevil Nov 20 20:45:30 <[g2]> lol.... why so eeeeeevil ? Nov 20 20:45:31 most people I know abroad do IRC or jabber :) Nov 20 20:48:07 koen: I got phoneme-advancd to compile against OZ 3.5.4.2 :) Nov 20 20:49:36 koen: As I understand it, I can merge gst-fluendo*, gst-ffmpeg* and gst-meta* Nov 20 20:49:52 koen: poky's gst-plugins also look a bit more advances that .devs Nov 20 20:50:56 hvontres|poodle: in OE? Nov 20 20:51:03 Ekiga is not my favourite Nov 20 20:51:19 I prefer my own client or kphone Nov 20 20:51:35 RP: I don't really trust cwiis' patches to gst Nov 20 20:51:36 gerwinin: I posted all work on the linuxbios mailing list. Nov 20 20:51:49 gerwinin: I am searching my old harddisk for the source build tree... Nov 20 20:52:19 gerwinin: ekiga is crap Nov 20 20:52:35 gerwinin: I want to use telepathy(-sip) Nov 20 20:52:40 koen: yeah I agree Nov 20 20:52:59 is D-P the hotkey for 'load service menu' on 3200? Nov 20 20:53:18 Koen: isn't ekiga from the guy that takes his girlfriend to all the shows ? Nov 20 20:53:27 gerwinin: yes Nov 20 20:53:46 koen: Ok, I just won't merge Nov 20 20:53:59 Koen: she is ugly as well :| Nov 20 20:54:05 lol Nov 20 20:55:04 koen:not yet. I did use the cross tools from my OE build. Nov 20 20:55:19 koen: I started a while ago on a sip-server in c but is more complicated as I thought Nov 20 20:55:42 psokolovsky: your commits were perfect. the commit messages were good because they alerted me to things that needed attention. Nov 20 20:55:46 hvontres|poodle: the build system looked to much like it did ~2 years ago Nov 20 20:55:53 so far it passes the basic tests but the qt awt libs dont work yet Nov 20 20:56:08 hvontres|poodle: and I didn't want to relive that horror Nov 20 20:56:27 koen: hehe Nov 20 20:57:06 the worst part: I'm not even allowed to distribute the patches that remove all gcc 2.95.3-isms.... Nov 20 20:57:18 mreimer, Ok, nice to hear! ;-) Nov 20 20:57:47 psokolovsky: it might be awhile before I can test h2200 changes, because my wife is using my h2200 now Nov 20 20:57:55 hvontres|poodle: if you can make a .bb for it, I can see if some patches are still needed Nov 20 20:58:51 mreimer, ok, np, maybe some other h2200 folks will do that. (koen: hint ;-) ) Nov 20 21:00:46 maybe Nov 20 21:00:53 but now it's time to watch NCIS Nov 20 21:00:55 later all Nov 20 21:05:51 psokolovsky: does hx4700 audio work for you on 2.6.18? Nov 20 21:07:07 mreimer, didn't try yet, sorry. h4000 (uda1380) works, though - with familiar. somehow, EABI kernel with angstrom had issues on h4000 too ;-( Nov 20 21:07:47 psokolovsky: it's not working for me on non-EABI familiar Nov 20 21:08:50 mreimer, so, something wrong did happen in .18, as there were report for .17-hh2 working... Nov 20 21:08:50 I didn't compile any alsa utils so far, but cat /dev/urandom > /dev/dsp did make a sound on my hx4700 (angstrom eabi 2.6.18-hh0) Nov 20 21:12:13 pH5, alsu utils went x.13 in OE.dev, so I suspect them, too Nov 20 21:12:32 hellp Nov 20 21:12:46 i install pdaxrom r121 with uboot and want to uninstall Nov 20 21:12:50 but nothing works Nov 20 21:12:51 psokolovsky: I'm still using alsa 1.0.11 Nov 20 21:13:01 when i plug in ac adaptater Nov 20 21:13:13 camousse: this is not a distro support channel Nov 20 21:13:28 pdaxrom boot, and any combinaison of keys doesn't work Nov 20 21:13:31 hum Nov 20 21:13:33 sorry Nov 20 21:13:39 but i'm pretty desperate Nov 20 21:14:12 sorry Nov 20 21:14:44 camousse: we know that, but zauruses are less then 20% of OE devices Nov 20 21:15:30 camousse: join #openzaurus Nov 20 21:15:55 okay thanks Nov 20 21:16:49 hrw: managed to login into my newly created distro for the first time.. finally :> Nov 20 21:16:50 thanks again for helpig out Nov 20 21:17:01 s/helpig/helping/ Nov 20 21:17:59 Jin^eLD: great Nov 20 21:18:20 dropbear does not seem to get going though Nov 20 21:18:48 Jin^eLD: you have /dev/urandom and /dev/random? Nov 20 21:19:00 it generated a key upon first startup I think Nov 20 21:19:18 urandom I guess Nov 20 21:19:33 at least nothing comes out when catting /dev/random Nov 20 21:21:02 hi hrw Nov 20 21:21:14 hi woglinde Nov 20 21:21:28 hrw: why did you ask? do I have to set sometihng? Nov 20 21:23:57 pH5, psokolovsky: do your hx4700s running 2.6.18 resume properly? I just got an all-white screen and no serial prompt back Nov 20 21:24:18 Jin^eLD: there was someone with nonworking dropbear - lack of /dev/random /dev/urandom Nov 20 21:24:44 dropbear: ../sysdeps/unix/sysv/linux/check_pf.c:68: make_request: Assertion `sizeof (req) - ((size_t) &((struct req *)0)->pad) == 3' failed. Nov 20 21:24:44 Aborted Nov 20 21:24:47 that's what I get Nov 20 21:25:13 hmm, wait, maybe I screwed up.. I think I forgot to add my kernel modules Nov 20 21:25:17 glibc 2.5? Nov 20 21:25:44 I fixed that by going to eabi, but I thought someone added a patch to glibc for the old abi Nov 20 21:26:12 I think 2.5, yes Nov 20 21:26:13 PREFERRED_VERSION_glibc ?= "2.5" Nov 20 21:26:22 arm? Nov 20 21:26:26 yes, arm9 Nov 20 21:26:43 abi or eabi? Nov 20 21:26:53 uhm :) Nov 20 21:27:06 I am not so good with hardware, so give me a hint :> Nov 20 21:27:06 what is DISTRO_TYPE Nov 20 21:27:19 release Nov 20 21:27:39 I meant TARGET_OS? Nov 20 21:27:48 :) Nov 20 21:27:49 :) Nov 20 21:27:51 linux Nov 20 21:28:04 mreimer, so, far, I haven't seen issues w/resume on h4000/hx4700. but I did too little testing to say it doesn't have Nov 20 21:28:21 the TARGET_OS var? Nov 20 21:28:30 psokolovsky: ok. I'll try to debug it when I have more time Nov 20 21:28:38 ok Nov 20 21:28:39 Crofton: I was creating a new distro from scratch for some custom hardware, koen and hrw helped me to setup the distro file Nov 20 21:28:39 TARGET_OS = "linux" Nov 20 21:28:51 ok Nov 20 21:29:01 so what is this abi/eabi thing? Nov 20 21:29:16 there is a problem with the old abi in glibc 2.5 Nov 20 21:29:28 I thought someone had added a patch to glibc, but maybe not Nov 20 21:29:48 what would you suggest, how should I proceed? Nov 20 21:29:51 I have one (from the libx-ports list), but I haven't been able to test it Nov 20 21:30:13 change TARGET_OS t olinux-gnueabi :) Nov 20 21:30:23 er linux-gnueabi Nov 20 21:30:42 ok Nov 20 21:30:53 what does it effectively do? Nov 20 21:31:09 it uses the newer eabi tool chain Nov 20 21:32:12 so I will have to recompile the whole toolchain and everything? Nov 20 21:32:19 yes :( Nov 20 21:32:28 what was the SMP setting again? :) Nov 20 21:33:15 hopefully I have some time to try and test the glibc patch over the next few weeks Nov 20 21:34:34 ...no honestly, I heard SMP support is in, right? Nov 20 21:34:34 :) Nov 20 21:34:50 I don't know that answer Nov 20 21:35:33 03crofton 07org.oe.dev * r72a42924... 10/ (3 files in 2 dirs): omniorb, omniorb-native : Fix bb file syntax and style. Nov 20 21:35:39 03crofton 07org.oe.dev * r71deae69... 10/ (1 packages/tasks/task-ossie.bb): task-ossie : Fixx bb file syntax and style. Nov 20 21:35:44 thanks for the glibc hint Nov 20 21:36:03 I went throught the same problem a few weeks ago Nov 20 21:36:46 PARALLEL_MAKE = "-j 4" Nov 20 21:36:50 for local.conf Nov 20 21:36:54 let's see if it works Nov 20 21:57:49 l8r, everyone Nov 20 21:58:01 RP: bitbake parsing flies Nov 20 21:58:08 or fly Nov 20 21:58:40 hrw: :) Nov 20 21:59:01 I hate skippackage Nov 20 22:00:11 03pfalcon 07org.oe.dev * ra245c698... 10/ (17 files in 9 dirs): (log message trimmed) Nov 20 22:00:11 h5xxx: Rename machine to h5000. Nov 20 22:00:11 * Use naming scheme consistent with other iPaq devices, where "series" name Nov 20 22:00:11 is used. Nov 20 22:00:11 * h5000 is series consisting of h5100, h5400, and h5500 models, each having Nov 20 22:00:12 common architecture, but slightly different builtin peripherals. Each model Nov 20 22:00:13 may also have few marketing nicknames, differing in last digit (sometimes 2 Nov 20 22:00:39 Who did rev #10000? I tell ya, who did 10k? ;-) Nov 20 22:00:43 psokolovsky: ipaq xxx0 = retail, xxx5 = oem iirc Nov 20 22:01:08 hrw, who cares for marketing tricks ;-) Nov 20 22:01:25 hrw, better, count authoritatively who did 10k ;-) Nov 20 22:04:33 monotone: warning: restriction excludes directory 'packages/linux/linux-jlime-jornada7xx-2.6.19rc6' Nov 20 22:04:33 monotone: misuse: invalid restriction excludes required directories Nov 20 22:04:35 means? Nov 20 22:05:47 Kristoffer, command? Nov 20 22:06:16 hm. Nov 20 22:06:36 RP: need to find out how to override skipping without change to OE/base.bbclass Nov 20 22:07:08 mt commit packages/linux/linux-jlime-jornada7xx-2.6.19rc6/jornada7xx.patch Nov 20 22:07:08 monotone: warning: restriction excludes directory 'packages/linux/linux-jlime-jornada7xx-2.6.19rc6' Nov 20 22:07:08 monotone: misuse: invalid restriction excludes required directories Nov 20 22:07:21 Kristoffer: mtn add dir? Nov 20 22:07:40 shouldnt dir be added if I successfully added a patch that exists within dir? Nov 20 22:07:51 hrw: As an exception is raised, I'm not sure you can... Nov 20 22:08:00 ~kill SkipPackage Nov 20 22:08:07 * ibot shoots a super-inverse graviton gun at SkipPackage Nov 20 22:08:25 monotone: skipping packages/linux/linux-jlime-jornada7xx-2.6.19rc6, already accounted for in workspace Nov 20 22:08:25 monotone: adding packages/linux/linux-jlime-jornada7xx-2.6.19rc6/defconf_jlime to workspace manifest Nov 20 22:08:25 monotone: skipping packages/linux/linux-jlime-jornada7xx-2.6.19rc6/jornada7xx.patch, already accounted for in workspace Nov 20 22:08:34 RP: I can make easy fetching by adding one variable and 3 lines to base.bbclass, but thats not quite good Nov 20 22:08:44 Kristoffer: which monotone you are using? Nov 20 22:09:00 mt commit packages/linux/linux-jlime-jornada7xx-2.6.19rc6/jornada7xx.patch Nov 20 22:09:00 monotone: warning: restriction excludes directory 'packages/linux/linux-jlime-jornada7xx-2.6.19rc6' Nov 20 22:09:00 monotone: misuse: invalid restriction excludes required directories Nov 20 22:09:00 let me check Nov 20 22:09:20 hrw: Why not? Nov 20 22:09:20 mt == <0.26 Nov 20 22:09:20 monotone 0.26 (base revision: 4342565107f26ceda955b66c66b5b7ec152f314e) Nov 20 22:09:30 Kristoffer: s/mt/mtn then? Nov 20 22:09:43 anyway I do not have idea Kristoffer Nov 20 22:09:54 hrw, :) will google for it Nov 20 22:10:31 RP: SOURCE_MIRROR_URL for own-mirrors.bbclass already created (will add it into base soon) Nov 20 22:10:42 RP: SOURCE_MIRROR_FETCH = "1" for fetching? Nov 20 22:11:16 hrw: yes Nov 20 22:12:42 Kristoffer: 'mtn commit packages/linux/linux-jlime-jornada7xx-2.6.19rc6' and upgrade to mtn 0.31 Nov 20 22:15:32 koen, it worked if I commited all files at same time. Nov 20 22:16:15 RP: http://pastebin.ca/253598 Nov 20 22:16:27 RP: what do you think? Nov 20 22:17:36 hrw: looks good to me Nov 20 22:18:33 RP: I will have to think about how to add PRE_MIRRORS in base.bbclass if SOURCE_MIRROR_URL is set Nov 20 22:18:49 03kristoffer 07org.oe.dev * r5fa17126... 10/ (1 packages/linux/linux-jlime-jornada7xx_2.6.19rc6.bb): Nov 20 22:18:49 linux/linux-jlime-jornada7xx_2.6.19rc6.bb: Addition of new kernel ver. Nov 20 22:18:49 * Adding new kernel version which greatly improves jornada7xx Nov 20 22:18:49 support. Nov 20 22:18:54 03kristoffer 07org.oe.dev * r6f53cb92... 10/ (1 conf/distro/jlime-donkey.conf): Nov 20 22:18:54 conf/distro/jlime-donkey.conf: Minor changes to distro file Nov 20 22:18:54 * Updated glibc 2.4 -> 2.5 to make it more arm friendly Nov 20 22:19:00 03kristoffer 07org.oe.dev * r57c7ff1b... 10/ (4 files in 3 dirs): Nov 20 22:19:00 linux/linux-jlime-jornada7xx-2.6.19rc6* : Addition of new kernel package Nov 20 22:19:00 * Addition of new kernel packages with greatly increased Nov 20 22:19:02 jornada7xx support. Nov 20 22:19:03 * Addition of jornada7xx.patch Nov 20 22:19:05 * Addition of new defconf Nov 20 22:20:00 hmm Nov 20 22:20:10 ah Nov 20 22:24:34 cu Nov 20 22:24:39 hrw: I'd be tempted to just override base.bbclass Nov 20 22:25:07 'night hrw Nov 20 22:25:08 will look into this tomorrow Nov 20 22:39:36 hrw, really gone? Nov 20 22:43:12 03lenehan 07org.oe.dev * rf8321aa1... 10/ (32 files in 2 dirs): Nov 20 22:43:12 perl: Replace the require in the native modules with a simpler version that Nov 20 22:43:12 names the actual module instead of using python. Should help with the Nov 20 22:43:12 current parse errors that is apparently occurring with new versions of Nov 20 22:43:12 bitbake. The python version was helpful when messing with lots of these Nov 20 22:43:13 things but serves no purpose now. Nov 20 22:43:17 03lenehan 07org.oe.dev * r9ae94be0... 10/ (1 packages/gmp/gmp_4.2.1.bb): Nov 20 22:43:19 gmp 4.2.1: Disable for sh by setting DEFAULT_PREFERENCE since the resulting Nov 20 22:43:21 library cannot be used. See bug #1605. Nov 20 23:27:05 I just cant figure out why bitbake keeps pulling glibc 2.3.x instead of glibc 2.5 that is defined in distro file Nov 20 23:28:50 mickeyl: to whom do we send a key for monotone write access for the openprotium guys? Nov 20 23:30:03 rwhitby: cc' koen and me, please Nov 20 23:30:24 so koen@linuxtogo.org + mickeyl@linuxtogo.org Nov 20 23:31:02 mickeyl: HopsNBarley is the requestor. They've built some new NAS firmware based on SlugOS for a commercial NAS solution. Nov 20 23:31:22 (and they want to contribute the changes back into OE, and build from OE directly in the future( Nov 20 23:31:23 rwhitby: pretty cool. HopsNBarley has been hanging around here for long. The name rings a bell. Nov 20 23:32:38 HopsNBarley: it'd be good to work together on making SlugOS use more of the task-based work that Angstrom has trailblazed. Nov 20 23:32:57 ah, there you are. Nov 20 23:33:24 yeah, i like what i see in the tasks based work. i have yet to migrate to that. Nov 20 23:33:56 mickeyl, name probably rings a bell cause i've bugged you and koen so much. thanks for all the help! Nov 20 23:34:10 HopsNBarley: np :) Nov 20 23:34:28 HopsNBarley: you can send your monotone key directly - no need to go through me. I'll send you your nslu2-linux ssl cert. Nov 20 23:34:32 we are not always in our best shape helping, but we try hard. really Nov 20 23:35:03 HopsNBarley, mickeyl: maybe OpenProtium can do a demo at FOSDEM 07 ? Nov 20 23:35:09 i think you guys were awesome. i try to give back and answer newb questions here too. Nov 20 23:35:24 cool :) Nov 20 23:35:47 rwhitby: that'd be nice. if they have already something demoable, we'd really like to see those guys @ our stand Nov 20 23:37:11 mickeyl: HopsNBarley could take along an NSLU2 as well ;-) Nov 20 23:37:20 rwhitby: yeah! Nov 20 23:37:40 we cut our teeth on the nslu2, first with unslung, and pretty quickly openslug after that. Nov 21 02:25:17 ...if monotone is already using a full cpu, then splitting out io onto another cpu is not going to make any difference, because obviously io is not where it is spending time :-) Nov 21 02:51:20 Hey, hows it going? **** ENDING LOGGING AT Tue Nov 21 03:00:02 2006