**** BEGIN LOGGING AT Thu Aug 09 02:59:59 2012
Aug 09 06:43:49 * alexhairyman is away: sleepy sleepy time
Aug 09 09:09:53 <CruX|> DEBUG: Executing shell function do_configure
Aug 09 09:09:54 <CruX|> /oe/tmp-eglibc/work/x86_64-nativesdk-oesdk-linux/qemu-nativesdk-0.15.1-r10/qemu-0.15.1/configure: 1123: /tmp/qemu-conf--12036-.exe: Permission denie
Aug 09 09:10:02 <CruX|> where can be problem here ?
Aug 09 10:29:09 <pb_> morning all
Aug 09 10:29:22 <pb_> does anybody happen to have recipes for clutter 1.10?  The ones in oe-core still seem to be 1.8.
Aug 09 10:36:14 <bluelightning> pb_: guacamayo has some I think
Aug 09 10:36:23 <bluelightning> pb_: https://github.com/Guacamayo/meta-guacamayo
Aug 09 10:39:16 <bluelightning> rburton: hi, will we be updating the clutter recipes in OE-Core based on the ones in guacamayo at some point?
Aug 09 10:39:26 <rburton> IT WAS A TRAP
Aug 09 10:39:30 <bluelightning> ;)
Aug 09 10:39:34 <bluelightning> gotcha
Aug 09 10:40:39 <pb_> ah right, thanks
Aug 09 10:40:40 <rburton> yeah, should do
Aug 09 10:42:13 <pb_> ah, and there's even an e{gl|lg} version, awesome
Aug 09 10:42:31 <rburton> hell yeah
Aug 09 10:42:35 <rburton> even works!
Aug 09 10:42:42 <pb_> even better
Aug 09 10:43:02 <rburton> modulo drivers, of course.  the matrix of drivers|platform|level of working is quite scary
Aug 09 10:43:16 <rburton> tf is the guy to talk to if you intend to use clutter+egl+something like beagle
Aug 09 10:43:57 <pb_> I'm actually using clutter+egl+directfb+broadcom
Aug 09 10:44:35 <pb_> (mips not arm though)
Aug 09 10:45:44 <rburton> i'm being warned that the setup in guaca isn't simply upstreamable
Aug 09 10:45:54 <rburton> Tomas Frydrych: because while the loadable backends idea in clutter is great, it does not work if something tries to do a compile-time check for a particular backend
Aug 09 10:45:55 <rburton> Tomas Frydrych: like clutter-gst does
Aug 09 10:45:55 <rburton> Ross Burton: well clutter-gst is wrong
Aug 09 10:45:55 <rburton> Tomas Frydrych: yes
Aug 09 10:45:55 <rburton> Tomas Frydrych: but because of it, we have to have separate x11 and egl packages
Aug 09 10:45:55 <rburton> Tomas Frydrych: which are not parallel installable
Aug 09 10:46:21 <rburton> but that's just a barrier to something being good enough for oe-core, works in guaca fine
Aug 09 10:46:31 <pb_> ah, fair enough.  well, for my purposes, I don't care about clutter-gst or x11 so I think I can ignore that :-)
Aug 09 10:46:46 <rburton> yeah
Aug 09 10:47:46 <pb_> well, might start caring about clutter-gst at some point, but not now.  We don't really have the bandwidth to run video through the texture mapper anyway.
Aug 09 10:49:04 <rburton> clutter-gst can do some interesting magic, but i presume you don't have anything like libva either
Aug 09 10:49:11 <rburton> clutter-gst + vaapi == WIN
Aug 09 10:50:30 <pb_> no, our video decode is all handled by dedicated hardware.  we more or less just put h.264 in at one end, and frames come out at the other end which then get piped into the hardware compositor.
Aug 09 10:51:22 <pb_> unfortunately, the video decoder and the texture mapper don't speak any formats in common, so getting video into a mappable format would require an extra trip through the 2d blitter for every frame in order to do format conversion.
Aug 09 10:52:32 <pb_> we could do that under some circumstances, but we don't have enough bandwidth to sustain two 1080p60 streams through that mechanism reliably.
Aug 09 10:52:49 <rburton> sigh
Aug 09 10:52:59 <pb_> yeah, suck
Aug 09 10:53:00 <pb_> oh well
Aug 09 10:53:07 <rburton> why do they always build hardware with some idiotic design :/
Aug 09 10:53:42 <rburton> so does this hw have hardware planes, or blend regions, or what?
Aug 09 10:57:45 <pb_> yeah, hardware planes
Aug 09 11:00:05 <bluelightning> hmm, some kind of issue with the list server atm
Aug 09 11:00:14 <bluelightning> 5.1.0 - Unknown address error 554-'5.7.1 <openembedded-core@lists.openembedded.org>: Relay access denied'
Aug 09 11:00:42 <pb_> oh, that's a bit unfortunate
Aug 09 11:00:47 <pb_> let me have a look
Aug 09 11:01:26 <bluelightning> thanks
Aug 09 11:03:32 <pb_> hm, I can't see anything obviously wrong from the logs, and that "Relay access denied" message doesn't look immediately like the kind of thing that discovery would be generating.  (The exim terminology is "relay not permitted"; "Relay access denied" sounds like postfix.)
Aug 09 11:03:50 <pb_> Are you sure that error was actually coming from lists.openembedded.org and not some intermediate relay?
Aug 09 11:04:17 <bluelightning> pb_: I'll have a look
Aug 09 11:04:36 <bluelightning> pb_: oops, yes, it was, sorry...
Aug 09 11:04:45 <pb_> righto
Aug 09 11:05:36 <CruX|> could any1 fix this ?
Aug 09 11:05:37 <CruX|> /oe/tmp-eglibc/work/x86_64-nativesdk-oesdk-linux/qemu-nativesdk-0.15.1-r10/qemu-0.15.1/configure: 1123: /tmp/qemu-conf--12036-.exe: Permission denie
Aug 09 11:05:40 <CruX|> thx
Aug 09 12:24:39 <pb_> rburton: the guacamayo clutter.inc seems to do:
Aug 09 12:24:42 <pb_> require ${GUACABASE}/meta/recipes-graphics/clutter/clutter-fpu.inc
Aug 09 12:24:53 <pb_> but I don't appear to have that file
Aug 09 12:25:53 <pb_> oh, wait, I see, it's expecting to find it in oe-core
Aug 09 12:26:05 <pb_> I guess I need to adjust my ${GUACABASE} slightly
Aug 09 12:27:36 <pb_> or maybe I should just copy over the file, that seems like a more wholesome solution.
Aug 09 13:11:44 <JaMa> known issue with oe-core ML?  The error that the other server returned was: 554 554 5.7.1 <openembedded-core@lists.openembedded.org>: Relay access denied (state 13).
Aug 09 13:17:33 <bluelightning> JaMa: oh interesting, so it's not just intel folks then...
Aug 09 13:17:49 <bluelightning> pb_: ^ hmm....
Aug 09 13:38:18 <pb_> hmm indeed
Aug 09 13:38:26 <pb_> JaMa: what server are you talking to exactly?
Aug 09 13:39:18 <pb_> er, oh, that's a bit weird
Aug 09 13:39:32 <pb_> someone seems to have changed the DNS so that lists.linuxtogo.org is now an alias for opal
Aug 09 13:39:47 <pb_> but, apparently, not bothered to set up mail configuration on opal so that it actually does anything useful
Aug 09 13:40:13 <pb_> s/lists.linuxtogo.org/lists.openembedded.org/
Aug 09 13:40:25 <pb_> RP: ping?
Aug 09 13:41:11 <bluelightning> pb_: unfortunately he is on holiday
Aug 09 13:41:17 <pb_> ah, obviously not him then
Aug 09 13:41:30 <pb_> well, maybe not "obviously", but seems unlikely he would do that while on holiday
Aug 09 13:42:03 <bluelightning> right, I would tend to agree
Aug 09 13:42:19 <JaMa> pb_: I don't see much from Delivery notice.. Received: by 10.180.105.130 with SMTP id gm2mr2860106wib.6.1344517818188; after by mx.google.com with ESMTPS id ex20sm2516043wid.7.2012.08.09.06.10.15
Aug 09 13:42:28 <pb_> presumably this isn't something that the TSC had instructed, right?
Aug 09 13:42:40 <bluelightning> pb_: no, I know nothing about that...
Aug 09 13:42:45 <pb_> bizarre
Aug 09 13:42:49 <bluelightning> very
Aug 09 13:42:55 <pb_> I don't even really know who has the keys to the oe.org dns.  let me investigate.
Aug 09 13:43:00 <bluelightning> florian: any ideas?
Aug 09 13:43:10 <pb_> Registrant Name:Philip Balister
Aug 09 13:43:18 <pb_> so, I guess Crofton is the man who would know
Aug 09 13:43:24 <bluelightning> ah ok
Aug 09 13:43:34 <pb_> in the meantime, I suspect that sending mail to openembedded-core@lists.linuxtogo.org will probably work
Aug 09 13:44:19 <bluelightning> ok, will try resending one of my earlier messages using that address
Aug 09 13:48:02 <pb_> right, there you go
Aug 09 13:48:39 <JaMa> yes that worked, thanks
Aug 09 13:48:45 <bluelightning> pb_: thanks
Aug 09 13:55:17 <pb_> well, I guess the moral of that story is that you shouldn't let RP go on holiday, it seems that every time he does something bad happens.
Aug 09 13:55:44 <bluelightning> it's not the first time that sentiment has been expressed... ;)
Aug 09 14:05:32 <rburton> haha
Aug 09 14:13:10 <khem> pb_: ask Crofton|work 
Aug 09 14:13:22 <Crofton|work> rofl
Aug 09 14:13:32 <Crofton|work> what is the issue?
Aug 09 14:13:49 <khem> Crofton|work: some DNS thing. I think 
Aug 09 14:14:01 <khem> ltg ml is resolving to opal
Aug 09 14:14:42 <Crofton|work> nothin gin oe.org has changed
Aug 09 14:14:45 <Crofton|work> hmmm
Aug 09 14:14:59 <Crofton|work> I did get some email from dotster
Aug 09 14:15:07 <Crofton|work> I wonder if they did something dumb
Aug 09 14:15:45 <khem> rschus: around? I got some openjdk question for you
Aug 09 14:16:19 <pb_> yes, exactly.  if you do "host lists.openembedded.org" you currently get 140.211.169.152, which is wrong
Aug 09 14:16:30 <pb_> and, consequently, none of the oe lists work
Aug 09 14:17:01 <Crofton|work> lloks likeonly a ui change
Aug 09 14:17:06 <bluelightning> I guess at least it's still pointing at an OE server, just the wrong one...
Aug 09 14:17:14 <pb_> it should be 188.40.83.200
Aug 09 14:17:24 <pb_> bluelightning: heh, true
Aug 09 14:18:06 <Crofton|work> this new interface is awful
Aug 09 14:18:09 <Crofton|work> wtf
Aug 09 14:28:46 <Crofton|work> ka6sox ping
Aug 09 14:29:17 <ka6sox> morning
Aug 09 14:29:29 <Crofton|work> something is up with dns
Aug 09 14:29:43 <ka6sox> so I see
Aug 09 14:29:45 <Crofton|work> I think dotster has gone mad
Aug 09 14:29:50 <ka6sox> agreed
Aug 09 14:30:02 <Crofton|work> does this make any send for mail.oe.org  	
Aug 09 14:30:02 <Crofton|work> 66.96.161.157
Aug 09 14:30:02 <ka6sox> let me poke a minute
Aug 09 14:30:28 <ka6sox> is that ltg?
Aug 09 14:30:34 <ka6sox> its not us
Aug 09 14:30:35 <pb_> no
Aug 09 14:30:37 <Crofton|work> I do not think so
Aug 09 14:30:51 <ka6sox> we have 140.211's
Aug 09 14:30:55 <pb_> that seems to be some random host in the USA, as far as I can tell
Aug 09 14:30:59 <Crofton|work> they added things like imap, smtp, ftp, pp, email etc
Aug 09 14:31:01 <pb_> ltg is 188.40.83.200
Aug 09 14:31:15 <Crofton|work> ok so mail.openmebedded.org shuold point to ltg?
Aug 09 14:31:23 <pb_> lists.openembedded.org actually, but yes
Aug 09 14:31:26 <ka6sox> somebody forgot to pay the domain?
Aug 09 14:31:30 <pb_> I don't think mail.openembedded.org is used for anything
Aug 09 14:31:31 <Crofton|work> no
Aug 09 14:31:35 <ka6sox> j/k
Aug 09 14:31:38 <Crofton|work> dotster changed some software arund
Aug 09 14:31:47 <ka6sox> this looks like a hijack from dotster
Aug 09 14:31:59 <Crofton|work> yeah
Aug 09 14:32:02 <Crofton|work> I am annoyed
Aug 09 14:32:09 <pb_> afaict, they have added a wildcard record so <anything>.openembedded.org resolves to opal
Aug 09 14:32:19 <Crofton|work> yeah, lists went away 
Aug 09 14:32:22 <pb_> and, presumably, deleted the previously-existing record for lists
Aug 09 14:32:33 <Crofton|work> ok, what should lists be? a, cname, mx?
Aug 09 14:32:38 <ka6sox> is now a time to move the ML?
Aug 09 14:32:47 <pb_> doesn't matter, A is fine
Aug 09 14:32:50 <ka6sox> A
Aug 09 14:33:05 <Crofton|work> A to ltg ip
Aug 09 14:33:09 <ka6sox> right
Aug 09 14:33:10 <pb_> yes please
Aug 09 14:35:04 <Crofton|work> mirrors is pointing at some de server
Aug 09 14:35:15 <Crofton|work> 82.197.159.157
Aug 09 14:35:39 <ka6sox> amethyst
Aug 09 14:35:43 <Crofton|work> ok
Aug 09 14:35:55 <Crofton|work> everythign else is pointed at a 140 machine
Aug 09 14:36:01 <ka6sox> opal
Aug 09 14:36:35 <ka6sox> garnet and agate too
Aug 09 14:36:40 <ka6sox> oh and emerald
Aug 09 14:36:50 <ka6sox> (from my list)
Aug 09 14:39:36 <Crofton|work> ok, lists is na A for lrtg now
Aug 09 14:39:42 <Crofton|work> will that fix things?
Aug 09 14:39:47 <ka6sox> eventually
Aug 09 14:40:07 <ka6sox> dns will propigate and this will be sorted in a couple of hours.
Aug 09 14:40:30 <Crofton|work> keep me posted
Aug 09 14:40:35 <Crofton|work> I am really pissed off
Aug 09 14:42:40 <ka6sox> Crofton|work, what are the TTL values?
Aug 09 14:43:16 <Crofton|work> doesn't say, don't think they are setable from the user interface
Aug 09 14:43:38 * Crofton|work notes that if anyone wants to talk to me about where we do dns, I am listening
Aug 09 14:43:58 <Crofton|work> dotster is going to lose a bunch of my business when I have time to deal with it
Aug 09 14:46:04 <ka6sox> for oe.org we can handle it.
Aug 09 14:46:07 <ka6sox> easily
Aug 09 14:50:30 <arimeh> Hi, how can I get to the cross compiler location in OE?
Aug 09 14:50:48 <arimeh> I am using arm-oe-linux-gnueabi version 4.4.2
Aug 09 14:51:00 <arimeh> I want to upgrade the version to a newer one
Aug 09 14:51:16 <arimeh> Which recipe should I look for?
Aug 09 14:51:48 <bluelightning> arimeh: the recipe is gcc-cross
Aug 09 14:52:04 <bluelightning> arimeh: it sounds like you are using quite an old version of OE though
Aug 09 14:52:32 <arimeh> Yes, I am using a specific git branch
Aug 09 14:52:52 <arimeh> But perhaps I can just adopt the cross-compiler, I don't want to change other stuff at this point
Aug 09 14:53:04 <bluelightning> should be possible
Aug 09 14:53:11 <arimeh> ok thanks, let me take a look
Aug 09 14:56:59 <ka6sox> pb_, if this doesn't clear in a few hours ping me?
Aug 09 14:58:58 <pb_> ok, cool, will do
Aug 09 14:59:14 <arimeh> bl, like you said, my branch is quite old it seems. it does not have 4.6.0 gcc. the latest one is 4.5.0 it seems. 
Aug 09 14:59:27 <pb_> I'm getting the right answer now so I think it probably will start working soon enough
Aug 09 14:59:38 <arimeh> I guess I need to siwtch to something newer, hopefully things should not break
Aug 09 14:59:49 <pb_> ttl 1 hour by the looks of it
Aug 09 15:32:05 <Crofton|work> is this better lists.openembedded.org.	3600	IN	A	188.40.83.200
Aug 09 15:35:46 <bluelightning> messages seem to be getting through so I think it's already propagating
Aug 09 15:51:43 <bluelightning> Crofton|work: users in #
Aug 09 15:51:50 <bluelightning> er
Aug 09 15:51:55 <bluelightning> UK keyboard fail
Aug 09 15:52:34 <bluelightning> Crofton|work: users in #angstrom are reporting that angstrom-distribution.org is failing now, could that be related?
Aug 09 16:58:46 <Crofton|work> yeah
Aug 09 17:06:42 <Crofton|work> https://twitter.com/#!/search/dotster
Aug 09 17:11:14 <bluelightning> Crofton|work: oh, so they mangled lots of people's configs it seems
Aug 09 17:11:17 <bluelightning> good times...
Aug 09 17:11:29 <pb_> bizarre
Aug 09 17:11:37 <Crofton|work> yeah
Aug 09 17:12:21 <Crofton|work> they have fucked up royally
Aug 09 17:12:27 <Crofton|work> oe website is down
Aug 09 17:13:01 <pb_> oh, that's odd, it was working earlier today
Aug 09 17:13:08 <pb_> maybe that was before they broke it
Aug 09 17:13:27 <Crofton|work> yeah, I feel liek they must have screwed somehtin gup trying to sort out the dns mess
Aug 09 17:13:53 <bluelightning> often with such problems the scramble to fix can just make the issue worse :(
Aug 09 17:14:30 <pb_> yeah
Aug 09 17:14:32 * pb_ go home now
Aug 09 17:14:33 <pb_> later all
Aug 09 17:14:58 <Crofton|work> this is going to motivate me to send out the voting proposal for spi affiliation
Aug 09 17:15:16 <Crofton|work> so I can talk about seeing if they can hold domains and provide dns
Aug 09 17:15:25 <Crofton|work> becuase all the .com places suck
Aug 09 17:15:33 * Crofton|work is hopping mad at this point
Aug 09 17:15:52 <gmc> pff, /me has spend the day trying to make angstrom not use systemd and other bloat like that.. time to build my own distro i guess
Aug 09 17:16:27 <gmc> btw, oe site is not down, you just have to remove the www. from url's it redirects to :)
Aug 09 17:26:09 <bluelightning> gmc: creating a new distro does seem to me to be an easier approach; angstrom is systemd-focused these days
Aug 09 17:28:01 <gmc> yes, systemd pulls in a large amount of bloat
Aug 09 17:34:08 <bluelightning> gmc: well one person's bloat is another person's feature if you actually want systemd, but it definitely pulls in a lot I agree
Aug 09 17:34:36 <gmc> bluelightning: of course, but it's thoroughly woven into the distro, so if you don't want it, it's not a suitable distro :)
Aug 09 17:35:07 <gmc> i've been looking at some other distro's too, specifically micro.. but that seems to be a bit out of date with current oe-core
Aug 09 17:37:59 <bluelightning> gmc: yes I noticed that last time I looked at it, you might want to poke pb_ about it tomorrow if you're keen to use it
Aug 09 17:39:06 <Crofton|work> gmc, do you need binary feeds?
Aug 09 17:39:27 <gmc> oh no, actually i think i'd be better of not using angstrom but building my own, as i want a very minimal image anyway.. system doesn't even have a screen or anything
Aug 09 17:40:15 <gmc> it's just that this project was using angstrom when i took over, and i'm now refactoring to use oe-core as opposed to classic oe
Aug 09 17:42:43 <gmc> Crofton|work: i don't think so, i don't even know what binary feeds are :)
Aug 09 17:43:09 <gmc> i guess that means binary packages? it's going to be a static rootfs image
Aug 09 17:43:17 <Crofton|work> well, I'd start with oe-cor and add layers until you can make what you want
Aug 09 17:43:42 <gmc> well toolchain and kernel are building fine now (3.4 kernel)
Aug 09 17:47:03 * mr_science enforces more test discipline
Aug 09 18:24:07 <gmc> for another project i'd be happy to have systemd or whatever automagic stuff handle hotplug events and all that
Aug 09 18:25:25 <JaMa> khem: I spoke too soon, bash still suffers Illegal instrictions when gold is used on armv4t..
Aug 09 18:26:13 <gmc> what is SHR's focus btw, distro-wise?
Aug 09 18:26:30 <JaMa> gmc: smartphones
Aug 09 18:26:53 <gmc> ah, interesting
Aug 09 19:44:41 <gmc> it'd be nice if there was a way to generate a graph of all dependencies between tasks and config files btw.. 
Aug 09 19:45:01 <kergoth> that doesn't make sense
Aug 09 19:45:13 <kergoth> the files loaded and the tasks generated are two independent sets of information
Aug 09 19:46:53 <gmc> yes and no :)
Aug 09 19:47:13 <gmc> i think.. it's all still abit overwhelming for me, the way distro's are defined and built
Aug 09 19:47:38 <gmc> and then, how it is determined what to build / install for an image
Aug 09 19:47:56 <gmc> having a helicopter overview of the files involved would help me a lot i believe
Aug 09 19:47:58 <kergoth> for the latter, use bitbake -g
Aug 09 19:48:10 <kergoth> that gives you the dependency / runtime dependency information
Aug 09 19:48:17 <gmc> yep, i know of that one, and used that a lot in figuring out why stuff is included in images for eg
Aug 09 19:48:25 <kergoth> for files loaded, you probably want seebs's bitbake variable / file tracking patches
Aug 09 19:48:26 <gmc> nice to import in gephi as well
Aug 09 19:49:14 <kergoth> http://patches.openembedded.org/project/bitbake/list/
Aug 09 19:49:26 <kergoth> http://patches.openembedded.org/patch/34093/ being the 0/2 part of the series
Aug 09 19:49:52 <kergoth> lets you see what config files got loaded in what order from where, and also exactly how the variables got set to what they were set to
Aug 09 19:50:04 <kergoth> so if some variable is set to foo, you can see exactly what file set it that way
Aug 09 19:51:30 <gmc> ah that's interesting
Aug 09 19:52:03 <kergoth> it's most helpful in figuring out why something seemingly inexplicable is happening, but the file loading bits might be helpful
Aug 09 19:55:09 <gmc> i will give that a try! thanks!
Aug 09 19:55:42 <kergoth> np
Aug 09 19:56:04 <gmc> i really like the new oe-core btw, got a bit tired of waiting for bitbake to parse recipes each time in classic :)
Aug 09 19:56:29 * kergoth nods
Aug 09 22:02:24 <carldani> Dumb question: I read a few OpenEmbedded docs, but now I'm quite confused. Apparently recipes for stuff outside OE-core are maintained somewhere... but no central repository exists.
Aug 09 22:02:59 <kergoth> http://www.openembedded.org/wiki/LayerIndex
Aug 09 22:03:52 <carldani> I'm the upstream maintainer of the flashrom utility (flashrom.org), and there was a flashrom OE recipe in the past. Where would a flashrom recipe fit best in this Layer ecosystem?
Aug 09 22:03:56 <carldani> kergoth: thanks!
Aug 09 22:04:25 <kergoth> meta-oe tends to be the place where recipes go if no other layer is more appropriate
Aug 09 22:46:18 <SorenHolm> Regarding compile-time dependencies. I've seen random recipes fail do_install. I suspect it might be because 'install' is not availavle yet. Which default dependencies are declared if any - coreutils-native should be a compiletime dependency always right?
**** ENDING LOGGING AT Fri Aug 10 02:59:58 2012