**** BEGIN LOGGING AT Mon Jun 06 02:59:57 2011 Jun 06 08:26:41 morning all Jun 06 08:59:20 morning all Jun 06 09:00:39 hi RP__ Jun 06 14:36:34 gm all Jun 06 14:38:15 hi zenlinux Jun 06 14:39:52 hi zenlinux Jun 06 14:39:54 good discussion about the PRIORITY metadata tag. I agree it should be a distro thing Jun 06 14:40:02 morning RP__ and bluelightning Jun 06 14:40:02 * bluelightning concurs Jun 06 14:55:31 Morning All Jun 06 14:55:41 hi sgw Jun 06 14:56:39 RP__: Afternoon to you Jun 06 14:58:12 sgw: Just looking at the pull request Jun 06 15:00:37 RP__: great, just pushed a quick update based on Phil's email. Jun 06 15:31:11 RP__: I was looking at the live image as an IMAGE_FSTYPE over the weekend and there are 2 classes related to live, bootimg, which generates live and boot-directdisk which generates the live directdisk, both are currently scripted from bbclass and inherited by the various image recipes. Therefore, what ever I do, I think needs to address both of these. They also both use syslinux. Jun 06 15:35:34 RP__: I think part of my question, is being able to optionally inherit a class (ie bootimg.bbclass) vs pulling all of that code into image_types.bbclass, maybe I am missing a subtle bitbake mechanism or ...? Jun 06 15:47:25 morning Jun 06 15:52:24 sgw: All good questions Jun 06 15:52:46 sgw: Could we turn the operation of those classes into a script we could run? Jun 06 15:52:50 hi fray Jun 06 15:53:06 sgw: i.e. make this more of a command like "make an ext2 image" is Jun 06 15:55:37 RP__: yes, that's what one thing I was looking at, part of that was some of the VARIABLES and defaults are different between directdisk and live, and could I contain that in such a way as to not pollute the other, possibly a python() setting the defaults and then calling the embedded shell script. Jun 06 15:56:14 RP__ I'll hopefully get something sent out today ont he directory ownership.. assuming I don't melt.. (supposed to up to 35C/95F today (and tomorrow) Jun 06 15:56:38 RP__: can we leave the scripts in the .bbclass and inherit so as to not make image_type a very large bbclass with shell scripts? Jun 06 15:57:04 sgw: I was wondering about a script in scripts/ Jun 06 15:57:15 sgw: Taking appropriate parameters as needed Jun 06 15:57:24 fray: sometimes (not always) coastal fog is a good thing ;-) Jun 06 15:57:30 sgw: but I haven't looked at exactly what is getting passed around Jun 06 15:57:37 heh Jun 06 15:57:38 RP__: Ok, I will look at that. Jun 06 15:57:47 26C on Friday was bad enough here Jun 06 15:59:02 it's already 26C here.. (if not warmer) Jun 06 15:59:19 and did I mention the 80% relative humidity?! Jun 06 16:01:18 fray: ick. ick ick ick. Jun 06 16:01:34 ya.. this is not normal for June... Jun 06 16:01:40 (this is August weather) Jun 06 16:02:04 now last week we had a couple of days where the highs were 10C/50F Jun 06 16:07:09 is there a separate IRC channel for bitbake? #bitbake seems empty Jun 06 16:21:54 RP__ in a recipe if I do: FILES_${PN} = "${localstatedir}/foo" where "foo" is a directory, what happens? Jun 06 16:22:05 does it walk foo? or does it just find and include it? Jun 06 16:22:10 (as a directory) Jun 06 16:23:09 Hmm.. it looks like it walks it.. Jun 06 16:30:44 zenlinux: no, there's not... I've looked for one too. I think the best place is either #poky or Jun 06 16:30:47 #oe Jun 06 16:31:16 argh @ enter key difference between UK & US keyboard layouts :/ Jun 06 16:31:48 ok, just making sure Jun 06 16:42:40 Just an FYI -- I sent a quick RFC on the directory ownership stuff to the oe-core list Jun 06 16:55:53 Hi there Jun 06 16:56:28 RP__: I notice you didn't take the Qt translation splitting patch ... did you find any issue with it? Jun 06 16:57:36 otavio: you might notice that it's in my CONSOLIDATED pull request from last evening. Jun 06 16:58:41 sgw: I thought this wasn't merged from it ... from what I see on the tree most of it is already in ... am I wrong? Jun 06 16:59:59 otavio: you are correct, I have not checked the tree this morning, was going by email. Jun 06 17:00:24 otavio: or lack there of, I am sure RP will comment on it shortly. Jun 06 17:01:41 sgw: good; no problem. I was curious to see if it has any issue so I can start working on it and get it merged asap Jun 06 17:02:10 sgw: specially for those class changes is better to get them merged so we can test and get feedback faster Jun 06 17:03:16 * otavio is finishing a from scratch build Jun 06 17:03:31 otavio: I'll reply to the email when I get a moment Jun 06 17:03:42 RP__: no problem Jun 06 17:11:53 Whats the magic to get this thing to built pseudo_git instead of pseudo_? Jun 06 17:15:02 fray: PREFERRED_VERSION? Jun 06 17:15:15 but whats the format? Jun 06 17:15:21 to specify 'git'? Jun 06 17:15:28 fray: what does PV say in the recipe? Jun 06 17:15:44 s/SRCPV/%/ Jun 06 17:15:48 PV = "1.0+git${SRCPV}" Jun 06 17:15:55 1.0+git% ? Jun 06 17:16:00 correct Jun 06 17:16:11 ok, thats the piece I was missing thanks Jun 06 17:17:21 PREFERRED_VERSION_pseudo = "1.0+git%" Jun 06 17:17:33 thats what I have.. still not working (did I spell it wrong?) Jun 06 17:18:06 are you looking at pseudo or pseudo-native? Jun 06 17:18:11 crap.. thats it Jun 06 17:18:34 IMHO there REALLY needs to be an easier way to specify "build this version" (on the command line) Jun 06 17:18:56 now it's working.. thanks Jun 06 17:19:03 -b means build this file... Jun 06 17:19:17 although it does also short circuit all the dependencies Jun 06 17:19:19 ahh Jun 06 17:19:29 that would have been fine in this case Jun 06 17:19:46 hmmm...getting a ghostscript md5 checksum error when building core-image-lsb Jun 06 17:19:58 known issue? Jun 06 17:20:26 zenlinux: I saw that on the autobuilder, but only external, I have not dug it out yet. Jun 06 17:20:43 sgw, ok Jun 06 17:26:39 sgw, a cleanall and rebuild of ghostscript worked this time. :/ Jun 06 17:28:05 Well that's strange, wonder if it grabbed it from the Autobuilder? Jun 06 17:28:23 RP__: I got your reply but didn't get what you meant for some of your comments Jun 06 17:28:33 RP__: please take a look when you can Jun 06 17:31:11 zenlinux: looks like ghostscript.com is giving bad downloads, and its possible the autobuilder has a bad one at this point. Jun 06 17:31:24 ah, that's not nice Jun 06 17:31:39 no, not at all, I need to fix that Jun 06 17:31:49 pidge, ping Jun 06 17:47:42 zeddii: ping Jun 06 17:48:28 'sup ? Jun 06 17:49:43 zeddii: just to confirm, I can grab the latest SRCREV without taking the Friday set correct? Jun 06 17:50:28 sgw: pong Jun 06 17:50:31 correct. you'd only loose any bug tracking info or history, but your tree will be correct. Jun 06 17:51:32 Bug tracking is important, I guess I will have verify the commit messages Jun 06 17:51:59 pidge: I did a little nip of the ghostscript tarball which was a bad download. Jun 06 17:52:25 I see. Anyone know what the issue with gs's download was? Jun 06 21:22:16 pidge: ping Jun 06 21:25:19 sgw: pong Jun 06 21:25:42 pidge: can we fire up another mut build, or are you working the AB? Jun 06 21:26:11 nope, we can fire up a mut. I have some AB stuff to do, but it can be later Jun 06 21:27:23 hi, i made a patch to build n450 arch (in meta-intel) Jun 06 21:28:09 build was broken Jun 06 21:29:11 but I use angstrom setup script to build my images, so i don't know where to report this patch Jun 06 21:29:29 captainigloo: was it the same as this: Jun 06 21:29:38 http://pastebin.com/n4KTshGy Jun 06 21:30:31 tomz: no Jun 06 21:31:17 captainigloo: hmm, can you compare your environments with angstrom setup vs oe setup? That might give us a hint Jun 06 21:31:44 here the patch : http://paste.ubuntu.com/620289/ Jun 06 21:32:49 atom-pc.conf was not found, and i need to add imagetype to be bzImage, otherwise make zImage in error Jun 06 21:33:03 tomz: why is initramfs-live-install failing, have you looked into it? Jun 06 21:33:49 captainigloo: is it because you don't have meta-yocto in your bblayers.conf file maybe? Jun 06 21:33:56 sgw: oebb.sh get meta-intel from git://git.yoctoproject.org/meta-intel,master,HEAD Jun 06 21:34:23 right, atom-pc.conf is in meta-yocto (from poky) Jun 06 21:34:23 sgw: i don't have meta-yocto at all Jun 06 21:34:39 :) Jun 06 21:34:45 sgw: a little bit: Jun 06 21:35:05 so all the live image builds are failing with the above "Nothing RPROVIDES 'initramfs-live-install'" error Jun 06 21:35:05 (11:38:34 PM) tomz: bisected to commit 52295fa3deef3b0374b99829626d524cefae6001 - Improve handling of 'all' architecture recipes and their interaction with sstate Jun 06 21:35:36 sgw: i made a diff with crownbay which i build succesfull and it usesconf/machine/include/tune-atom.inc Jun 06 21:36:11 sgw: so far, the line Jun 06 21:36:37 sgw: +TARGET_ARCH = "all" Jun 06 21:37:08 sgw: actually, later chagned to "allarch" is what I've narrowed it down to Jun 06 21:38:03 sgw: but the value doesn't matter, just setting TARGET_ARCH there at all seems to cause the problem Jun 06 21:39:07 captainigloo: there's a MAINTAINERS file in meta-intel that should tell you where to send meta-intel patches Jun 06 21:39:54 tomz: if i understand correctly it builds fine in yocto env, it's only angstrom related ... Jun 06 21:40:09 captainigloo: seems crownbay defines things that are in tune-atom.inc also, tomz, I guess meta-intel needs to be self contained, but that means code-duplication between meta-intel and meta-yocto Jun 06 21:40:27 captainigloo: oh, ok, then no matter Jun 06 21:40:50 but meta-intel is just the BSP isn't it ? Jun 06 21:41:06 so it should only give rule to build kernel and xorg drivers Jun 06 21:41:30 but i'm new to OE so i should missed something Jun 06 21:42:16 tomz: I wonder if that is then causing the COMPATIBLE_HOST check to fail Jun 06 21:43:11 sgw: dunno, will have to get back to it later - gotta run now to a softball game Jun 06 22:01:50 is Beth on IRC? Jun 06 22:23:32 fray: pidge Jun 06 22:24:04 * Jay7 is connecting people ;) Jun 06 22:24:17 hello Jun 06 22:26:05 fray: pidge is Beth :) Jun 06 22:26:08 fray: pong Jun 06 22:26:21 ahh I see.. sorry, I already found her elsewhere.. ;) Jun 06 22:26:26 :) **** ENDING LOGGING AT Tue Jun 07 02:59:57 2011