**** BEGIN LOGGING AT Fri Feb 16 02:59:58 2007
Feb 16 13:20:27 <VoodooZ> CoreDump|afk: I might be the only one but boot-time is very important as I use my slug as a robot controller which boots often in development (always!)
Feb 16 13:21:07 <VoodooZ> CoreDump|afk: So I would appreciate any optimizations you can come up with or at least alternatives so I can customize my build. 30 seconds is already too long.
Feb 16 16:38:31 <CIA-19> 03bzhou * r5512 10optware/trunk/Makefile: changed UCLIBC_SPECIFIC_PACKAGES to not include libiconv
Feb 16 19:12:58 <hillct> I still don't understand how the ipk packaging step is handled via bitbake. how are the ipkg CONTROL files setup? for example, if I want to use a postinst script, how to I get that added through a bitbake script?
Feb 16 19:32:18 <hillct> oh
Feb 16 19:32:21 <hillct> I think I get it
Feb 16 19:32:31 <hillct> that's sort of cool. sort of scary
Feb 16 19:33:03 <eno> http://www.openembedded.info/wiki/Bitbake_recipes and search for postinst or "inherit update-rc.d"
Feb 16 19:33:15 <eno> exactly, sort of cool, sort of scary
Feb 16 19:45:51 <hillct> oh that's SO much better than what i was doing. i was reading about each class and trying to piece things together
Feb 16 19:46:03 <hillct> thanks
Feb 16 21:17:09 <hillct> the cron package defaults ot placing crontabs into /var/cron/tabs ? Is there any other distro where that's the standard location? What happened to /etc/cron.d ? Isn't that the usual default?
Feb 16 21:17:30 <hillct> oh wait
Feb 16 21:19:20 <hillct> ok. that's the user crontab location
Feb 16 21:19:40 <hillct> there seems to be no /etc/crontab included for that package
Feb 16 21:37:33 <hillct> ok, so it's just not included in the cron package. There should definately be a stub /etc/crontab included in that package. What's the procedure for adding that item since no maintainer is listed in the .bb file ?
Feb 16 21:43:20 <NAiL> hillct: look in the MAINTAINERS file in root if there's anyone relevant to bug
Feb 16 21:46:43 <hillct> yeah. I didn't see a relevent entry there either
Feb 16 21:59:46 <rwhitby> hillct: raise a bug on the OE bug tracker, attach a patch.  if you do that enough, they give you write access.
Feb 16 22:04:59 <hillct> ah
Feb 16 22:05:31 <hillct> will do. jsut as soon as I prove to myself that I know what I'm doing with bitbake recipies
Feb 16 22:06:47 <hillct> how does cross-testing for OE packages work? Is there an existing process or is is mroe ad-hoc?
Feb 16 22:19:33 <marceln> rwhitby: Just changing the page and adding a new summery doesn't work ;-(
Feb 16 22:21:03 <rwhitby> marceln: approved the sites
Feb 16 22:25:17 <CoreDump|home> hi
Feb 16 22:25:58 <CoreDump|home> note-to-future-self: Do not cut power to the slug while altboot is non-functional
Feb 16 22:26:16 <hillct> is this the needed command for cleaning up temp files for a failed build (without deleting the packages data for my new recipie?
Feb 16 22:26:17 <hillct>  bitbake -c clean -b openembedded/packages/foo/foo_x.y.z.bb 
Feb 16 22:26:41 <CoreDump|home> hillct: yes
Feb 16 22:26:46 <hillct> K
Feb 16 22:26:48 <hillct> thanks
Feb 16 22:27:40 <CoreDump|home> np
Feb 16 22:32:57 <hillct> hmm
Feb 16 22:33:13 <hillct> logrotate has a compile-time dependency on popt
Feb 16 22:33:27 <hillct> but the compile phase can't find popt.h
Feb 16 22:33:58 <hillct> shouldn't that just be a matter of defining the include path?
Feb 16 22:34:21 <hillct> is there a defined var for dealing with cross-package include paths?
Feb 16 22:40:49 <CoreDump|home> that would be a question for #oe ;)
Feb 16 22:41:15 <hillct> ah
Feb 16 22:41:25 * hillct reads more of the docs first
Feb 16 22:51:41 <marceln> rwhitby: That doesn't work. It only makes it worse
Feb 16 22:52:22 <rwhitby> marceln: I don't understand the problem then ...
Feb 16 22:54:32 <marceln> You checked  "http://www.nslu2-linux.org/wiki/Site/AllRecentChanges" as well
Feb 16 22:55:29 <marceln> This is the second time this week that i had problems with spam on a website.
Feb 16 22:59:27 <CoreDump|home> rwhitby: you're in phase0 as well =) I got a good chuckle out of the OM wiki where you rub that fact into the face of other folks hehe
Feb 16 23:34:13 <CoreDump|home> *sigh*
Feb 16 23:34:26 <CoreDump|home> I _really_ need to get the parts for the serial cable
Feb 16 23:34:56 <CoreDump|home> it's kinda a pain to debug altboot w/o
Feb 16 23:36:46 * CoreDump|home says so and heads off to reflash his slug...
Feb 16 23:42:53 <CoreDump|home> muwahaha, first official altboot boot of a slug
Feb 16 23:46:01 <rwhitby> cool
Feb 16 23:46:08 <hillct> what's altboot?
Feb 16 23:46:31 <CoreDump|home> a flexible boot-manager
Feb 16 23:46:42 <CoreDump|home> rwhitby: which mtdblock was / again?
Feb 16 23:46:51 <hillct> replacement for RedBoot or on top of RedBoot?
Feb 16 23:47:14 <CoreDump|home> hillct: neither, it's a userspace helper replacing /linuxrc
Feb 16 23:47:15 <rwhitby> CoreDump|home: mtdblock 4 on a slug I think
Feb 16 23:47:34 <hillct> oh, cool
Feb 16 23:47:47 <hillct> so it lets you tweak the init process?
Feb 16 23:48:17 <hillct> more easily
Feb 16 23:48:35 <CoreDump|home> It lets you choose which partition / device to boot from
Feb 16 23:48:40 <rwhitby> replaces turnup
Feb 17 00:12:12 <CIA-19> 03bzhou * r5513 10optware/trunk/make/erl-yaws.mk: erlang yaws: 1.66 -> 1.68
Feb 17 00:13:04 <mwester> CoreDump|home: mtd block 2 on the dsmg600 and nas100 systems, if that makes any difference for what you're doing...
Feb 17 00:15:27 <VoodooZ> is it still going to be possible to use the old method or none at all though?
Feb 17 00:17:48 <CoreDump|home> VoodooZ: You are free to completely ignore it. It won't affect your boot-time in that case at all
Feb 17 00:17:54 <CoreDump|home> usage is optional
Feb 17 00:18:27 <CoreDump|home> mwester: thanks, but I found a better way to get to the partition (ie: I stole it from turnup ;) )
Feb 17 00:18:54 <VoodooZ> cool. Just curious. I'll definitely try it first though. I might love it. Keep up the excellent work.
Feb 17 00:19:40 <VoodooZ> rwhitby: I was thinking.. if I was to install angstrom (LE), do I have to change the FIS for proper endianess?
Feb 17 00:19:43 <CoreDump|home> =) it will be some time before it's finished
Feb 17 00:23:31 <CoreDump|home> n8
Feb 17 00:47:42 <hillct> well, aparently the ansewer to my 'can't find popt.h' problem is:
Feb 17 00:47:44 <hillct> CFLAGS_prepend = "-I${STAGING_INCDIR} "
Feb 17 00:48:06 <hillct> yet it doesn't seem to pass the added CFLAGS to the compile process
Feb 17 01:16:07 <hillct> is there a way to force .bb files to be re-parsed when a build is done? I don't see any pattern to when my recipie file is re-parsed (regardless of file modifications)
Feb 17 01:35:04 <rwhitby> VoodooZ: our kernel will read FIS in either endianness
Feb 17 02:00:20 <hillct> rwhitby: any thoughts on why CFLAGS_prepend = "-I${STAGING_INCDIR} " might not work in a .bb file? I'm trying to get my package to pickup headers from the dependency package popt ?
Feb 17 02:02:25 <hillct> it just isn't being picked up
Feb 17 02:02:48 <rwhitby> no idea
Feb 17 02:02:54 <CIA-19> 03osas * r5514 10optware/trunk/make/asterisk14.mk: asterisk14: r53528 -> r55097 upstream fix for gsm codec
Feb 17 02:16:23 <hillct> I've been trying all afternoon to try and get this build and it just won't work
Feb 17 02:32:20 <VoodooZ> rwhitby: very nice. I'll try that soon...
**** ENDING LOGGING AT Sat Feb 17 02:59:57 2007