**** BEGIN LOGGING AT Sun Oct 23 02:59:57 2011
Oct 23 15:29:46 <asslowashell> does anyone have a virtual machine image or liveCD or something with an already set up open embedded environment for gumstix?  
Oct 23 15:31:46 <asslowashell> the amount of time i have already burned trying to get it installed on my ubuntu its beyond my downttime threshold, I want to code not spend days configuring and installing 
Oct 23 15:31:52 <Crofton|work> not that I know off
Oct 23 15:31:59 <asslowashell> I see, thats unfortunate
Oct 23 15:33:26 <asslowashell> well, I'll just hang around.  maybe somebody knows one
Oct 23 15:35:09 <Crofton|work> make one :)
Oct 23 15:39:12 <asslowashell> I want one so that I don't have to waste more time trying to get it installed, making one myself defeats that purpose
Oct 23 15:39:53 <asslowashell> I had time in high school to blow weeks on compiling a stage 1 of gentoo
Oct 23 15:40:41 <asslowashell> not these days
Oct 23 15:41:33 <asslowashell> looking into using crosstool-ng instead, if its quick and easy to get running
Oct 23 15:45:59 <asslowashell> as a sidenote- i got to roughly task 350 out ~1200 with a compile error for dbus, after already having to manually fetch zlib and alter some md5s due to mismatches for some unexplained reason
Oct 23 17:21:55 <CMoH> hey. i'm having a strange problem with bitbake-1.12.0: I have an image with IMAGE_INSTALL="A"; A has RDEPENDS=B, which has RDEPENDS=C. However, I can see A and B in the final image, but not C
Oct 23 17:24:01 <CMoH> however, making A RDEPENDS on B and C pulls C in the image
Oct 23 17:24:08 <CMoH> how can this happen?
Oct 23 17:25:10 <bluelightning> CMoH: it sounds like a bug if that's really what's happening
Oct 23 17:26:09 <CMoH> i can only think of a bug, or that the recipe for B is somehow "broken"
Oct 23 17:27:01 <bluelightning> CMoH: you don't have a require / inherit in recipe B that's setting RDEPENDS to a different value?
Oct 23 17:27:18 <CMoH> only update-rc.d
Oct 23 17:27:25 <bluelightning> hmm
Oct 23 17:27:59 <bluelightning> I'm guessing you didn't add the RDEPENDS some time after it was built either?
Oct 23 17:28:37 <CMoH> nope
Oct 23 17:28:48 <CMoH> i cleaned all: image, A B and C
Oct 23 17:28:53 <bluelightning> are you able to pastebin recipe B
Oct 23 17:28:55 <bluelightning> ?
Oct 23 17:30:34 <CMoH> http://pastebin.com/5dytKJHA
Oct 23 17:31:05 <CMoH> however, if i run bitbake -g i can see A depending on B and B depending on C in pn-depends.dot, as well as in the other dependencies
Oct 23 17:33:26 <bluelightning> CMoH: what about the pkgdata for recipe B, does that show the dependency?
Oct 23 17:33:52 <CMoH> pkgdata?
Oct 23 17:34:20 <CMoH> ah, found it
Oct 23 17:36:09 <CMoH> or did not; the only pkgdata i found only lists packages in recipes (tmp/pkgdata/...)
Oct 23 17:36:15 <CMoH> no dependencies for any of the recipes
Oct 23 17:36:47 <CMoH> ok, now it looks better
Oct 23 17:36:50 <CMoH> no dependency 
Oct 23 17:37:06 <CMoH> just for update-rc.d
Oct 23 17:37:38 <kergoth> bitbake -e B | grep \^RDEPENDS
Oct 23 17:38:52 <bluelightning> CMoH: ^ what he said
Oct 23 17:39:05 <CMoH> RDEPENDS="media-ctl"
Oct 23 17:39:14 <CMoH> RDEPENDS_karbon-cam-setup="update-rc.d"
Oct 23 17:39:28 <CMoH> and there's another for the -dev package
Oct 23 17:39:40 <bluelightning> ah, right... so the second is blowing away the first
Oct 23 17:39:44 <bluelightning> unfortunate :/
Oct 23 17:39:51 <bluelightning> ah, wait a sec
Oct 23 17:40:08 <CMoH> hmm; weird
Oct 23 17:40:11 <bluelightning> CMoH: you need to set RDEPENDS_${PN} not just RDEPEDNS in your recipe
Oct 23 17:40:27 <bluelightning> that ought to fix it
Oct 23 17:40:54 <bluelightning> s/not just/instead of/
Oct 23 17:42:05 <CMoH> the weirder thing is recipe A also uses the same structure (only RDEPENDS, no RDEPENDS_${PN}), also with update-rc.d for another initscript
Oct 23 17:42:23 <CMoH> although it is more complex, more packages etc
Oct 23 17:42:36 <kergoth> wonder if we shouldn't inject a couple of the recipe sanity checks from recipe_sanity into something which is always run, rather than being opt-in
Oct 23 17:43:05 <kergoth> you should never set RDEPENDS for more than one reason. as youv'e already discovered, classes will assume the package specific, but beyond that, RDEPENDS affects all packages you emit, including -dev, -doc, -dbg, which is unlikely to be what you want :)
Oct 23 17:43:18 <CMoH> kergoth, i'd love to hear more on the sanity checks opt-in :D:D
Oct 23 17:43:54 <bluelightning> kergoth: yes, and as shown here it interacts badly with _append
Oct 23 17:43:54 <kergoth> INHERIT += "recipe_sanity" in local.conf, then bitbake -c recipe_sanity foo -- I'm not sure if it's in oe-core/meta-oe though. just checks for some things like the rdepends/rdepends_${PN} thing
Oct 23 17:44:19 <kergoth> bluelightning: heh, the behavior of append/prepend relative to override variables is confusing all around :\
Oct 23 17:44:26 * kergoth gets food
Oct 23 17:45:12 <CMoH> thanks to both
Oct 23 17:48:10 <CMoH> is it hard to migrate from 2011.03-maintenance to oe-core and meta-oe?
Oct 23 17:48:49 <CMoH> by the way - are there any resources that describe what the difference is between these two?
Oct 23 17:57:03 <bluelightning> CMoH: http://www.openembedded.org/wiki/OpenEmbedded-Core
Oct 23 17:58:55 <CMoH> thanks again
Oct 23 18:01:58 <CMoH> it certainly looks cleaner that the branch we're using now
Oct 23 18:05:32 <CMoH> however, the getting started page talks mostly of bitbake-1.8.x and what seems to me to be the old openembedded,  not the oe-core 
Oct 23 18:05:40 <CMoH> why is that?
Oct 23 18:07:27 <CMoH> i mean, is the wiki behind? or is oe-core not recommended for stable releases?
Oct 23 21:27:36 <Jin^eLD> anyone tried to build mdadm for powerpc recently?
Oct 23 21:32:22 <woglinde> good nite
**** ENDING LOGGING AT Mon Oct 24 02:59:56 2011