**** BEGIN LOGGING AT Tue Jun 21 02:59:56 2011 Jun 21 07:38:53 good morning Jun 21 07:53:43 Hi, I have obtained bitbake from git in order to use the layers feature but i'm now seeing a weird error parsing one of the .bb files, i have seen a few other pastbins of this from google but can't find a solution. any ideas? http://pastebin.com/u3e8Tfs7 Jun 21 07:59:03 mangomake_: I've got a fix for this; but I would suggest using the 1.12.0 tag until it's fixed Jun 21 07:59:53 bluelightning: k, i was trying git becasue i could get the bblayers to work with 1.12.0 - trying to recall specifc problem (was last firdya) Jun 21 08:00:02 s / could / coud not Jun 21 08:00:42 It was not expanding ${THISDIR} Jun 21 08:01:49 mangomake_: hmm Jun 21 08:14:00 mangomake_: well, you could apply my fix then: http://pastebin.com/BGZcCrf8 Jun 21 08:44:25 Hi, should oe-core have recipes for squashfs and cramfs? These image types are included in image_types.bbclass, but recipes to provide squashfs-tools-native and cramfs-native seem to be missing. Jun 21 09:18:20 melo hosed again, groan Jun 21 10:41:54 hmmm does the oe website work fine? Jun 21 10:42:17 I cannot reach it Jun 21 10:42:37 no, melo is broken Jun 21 10:43:23 ok Jun 21 10:46:11 can someone update the topic to include a notice to that effect? Jun 21 10:53:51 I've not enough details to update the topic Jun 21 10:54:15 people who have the details should do it Jun 21 10:55:05 just "melo broken, website and git server down" would probably be enough Jun 21 10:55:31 I don't imagine any more details are necessary Jun 21 10:55:38 so, please go ahead and set that in the topic Jun 21 10:55:46 ah I cannot set the topic Jun 21 10:55:53 it say I'm not a channel operator Jun 21 11:15:30 I find amazing that even though OE gained momentum and some reputation we still suffer form outages... Jun 21 11:15:44 yeah, it is a bit sad Jun 21 11:15:55 I'm not quite sure what's going wrong with melo, this has happened a few times in the last week or two Jun 21 11:16:31 the firens @ T* or Inte* could ship some obsolete rig, isn't? Jun 21 11:16:36 it would be nice if some of the larger corporate users of OE could contribute some resources, but this doesn't seem to be happening Jun 21 11:16:47 heh Jun 21 11:17:11 yeah, you'd hope that TI, Intel and Mentor could somehow scrape together one spare computer between the three of them Jun 21 11:17:17 but, I guess this is the global recession, times are hard Jun 21 11:17:47 for me, I'm used to pay for my hobbies, np Jun 21 11:17:58 but is insane situation :p Jun 21 11:20:07 according to yahoo finance, the three of them have a combined market cap of about $250 billion dollars. Jun 21 11:20:48 but, well. Jun 21 11:21:10 wouldn't want them to have to sell kergoth or anything Jun 21 11:23:11 :) Jun 21 11:24:30 I know you guys are just kidding around, but do you know whether anyone tried asking TI, Intel or Mentor whether they'd consider donating a server to the OE cause? Jun 21 11:25:28 pb_: we'll have to ask for alms the e.v members Jun 21 11:25:45 I don't know, actually. I do know that all three of those companies have seats on the TSC, though, so you'd hope they would be aware of the situation. Jun 21 11:25:56 thhp: a discussion was started long ago, we have a sponsor page iirc Jun 21 11:26:16 But, as to whether anybody has actually made a direct request, I'm not sure. Jun 21 11:26:41 indeed, so straight, I don't think so Jun 21 11:28:01 * ant_work is bombed daily by that 'clouds-computing' offers Jun 21 11:28:15 It might be worth a go -- clearly all three companies see OE as being a part of their future(s), and they're all dedicating developer time Jun 21 11:28:35 A server is small change by comparison, I would have thought Jun 21 11:29:51 Yeah, you'd think. I guess it would be a struggle to spend much more than $10k on the sort of server we'd need. Jun 21 11:31:30 and the would have sysadm for free ;) Jun 21 11:31:32 * Jay7 may manage some servers for $500/month :) Jun 21 11:31:40 :p Jun 21 11:31:43 ;) Jun 21 11:32:04 s/free/peanuts/ Jun 21 11:32:25 (do not scare them) Jun 21 11:32:27 +- 5 servers are not critical now :) Jun 21 11:35:29 hm, lamawithonel_ should really get a grip on his client Jun 21 11:36:53 do Mentor/Intel or TI using some CDN? Jun 21 11:37:07 thhp: about your issue with uclibc, this sounds strange Jun 21 11:37:24 once for I'm pretty sure I've read successful reports Jun 21 11:37:27 it would be easy to place oe site to CDN then Jun 21 11:37:33 then because the overrides look sane Jun 21 11:37:37 and it will be always online Jun 21 11:37:58 git is other thing, we may just place some mirrors around Jun 21 11:38:01 joho Jun 21 11:38:08 git server still down? Jun 21 11:38:17 melo is down afaik Jun 21 11:38:52 uh Jun 21 11:38:59 so worst case Jun 21 11:39:34 dunno about that, I can think of worse cases Jun 21 11:39:48 tsunami? Jun 21 11:40:01 asteroid? Jun 21 11:40:03 right, or it could have been consumed by a plague of locusts Jun 21 11:40:33 or accidentally converted to antimatter by some kind of rogue physics experiment in the vicinity Jun 21 11:40:50 the current situation sounds fairly tame by comparison Jun 21 11:41:22 64 bytes from melo.openembedded.org (140.211.169.165): icmp_req=1 ttl=45 time=218 ms Jun 21 11:41:28 hehe Jun 21 11:41:57 ant_work: well, yeah, but last time I looked the load average was 830 Jun 21 11:42:08 830? Jun 21 11:42:10 muahah Jun 21 11:42:11 did you dare to login? Jun 21 11:42:18 patchwork again? Jun 21 11:42:29 pb_: now your office may be infested by bugs... Jun 21 11:42:39 ah drat, didn't think of that Jun 21 11:44:19 my cousins says (tm) that just typing the IP is enough to be doomed Jun 21 11:44:37 ah, gutted. curtains for me then. Jun 21 11:45:25 did you even type the URL? gosh Jun 21 11:45:35 lol Jun 21 11:45:49 * pb_ implodes Jun 21 11:47:37 * woglinde fetches the mop Jun 21 11:49:26 * ant_work sits now unwillingly in front of a Win2000 PC thinking about how to secure it's access to local sql server (how to protect the server :p) Jun 21 11:49:58 ant_work: install xenserver and place all windows machines inside Jun 21 11:50:06 this will protect you :) Jun 21 11:50:34 ant win2000? Jun 21 11:50:40 this PC runs some strange business data-keeping rotten app Jun 21 11:50:50 yeah virtualize it Jun 21 11:50:57 from sql server side.. may be just revoke access rights from everybody except ms sql server user Jun 21 11:50:58 makes easy backups Jun 21 11:51:28 I'm evaluating open-source alternatives... Jun 21 11:51:36 btw Jun 21 11:51:45 no chance against management Jun 21 11:51:55 we have translator MS SQL <-> PostgreSQL Jun 21 11:52:17 translator? Jun 21 11:52:22 or convertaor? Jun 21 11:52:30 translator Jun 21 11:52:34 works like proxy Jun 21 11:53:11 seems there is no english version of site.. Jun 21 11:53:56 http://etersoft.ru/content/view/154/208/ Jun 21 11:54:10 in russian, use favourite translator :) Jun 21 11:54:33 but seems they are supporting only some subset of apps right now Jun 21 11:54:51 etersoft is company which providing services to smooth transition from windows to linux Jun 21 11:58:37 I wrote 10yrs ago the queries and th ephp code for migrating msaccess -> mysql Jun 21 11:58:49 imagine there are still employees preferring the old app Jun 21 11:59:00 (most of them tbh) Jun 21 11:59:35 the situation is even worse for the accounting apps Jun 21 12:03:27 that translator was born to resolve problem with accounting apps :) Jun 21 12:03:40 and wine@etersoft as well Jun 21 12:03:58 it was specially patched wine version able to run that apps Jun 21 12:04:37 now almost mainlined Jun 21 12:05:08 well.. I should go now Jun 21 12:06:11 ? Jun 21 12:31:22 Anyone have recommendations for having oe target images include the rootfs changes made by any _postinst rules of the installed packages? As I am targetting squashfs, this can't be done at runtime on first boot. Jun 21 12:32:57 I don't think there's any single silver bullet for that. You probably have to consider each postinst individually. Jun 21 12:34:31 In theory you could probably do something with qemu to get all the postinsts run under emulation but I'm not sure how practical that would really be. Jun 21 12:34:43 Might be an interesting project to try, anyway. Jun 21 12:36:39 I was thinking the same. Maybe capture changed files into a tarball which could be wrapped into the image by a recipe included only for certain image types. Jun 21 12:40:33 Yeah, or just pack the image into an .ext3, boot it under qemu-system, and then transform it into your desired output format. Jun 21 12:41:11 I guess something along those lines could probably be made to work. There is a fair amount of qemu scripty stuff in oe-core already for the unit tests and you might perhaps be able to borrow some of that. Jun 21 12:41:42 ok, thanks for the pointers. I'll have a look. Jun 21 13:46:13 jo mickeyl Jun 21 14:11:20 hi kergoth Jun 21 14:11:26 hi mickey|office Jun 21 14:11:42 hey pb_ Jun 21 15:06:45 anyone know where to dig up really old versions of sysvinit? Jun 21 15:06:48 ka6sox-away: ka6sox-work around ? Jun 21 15:07:05 hi khem and kergoth Jun 21 15:07:15 hey yo Jun 21 15:08:19 ka6sox-away: melo is really mellow Jun 21 15:08:45 kergoth: tsx-11, maybe? Jun 21 15:08:47 kergoth: debian would be the place I would say Jun 21 15:08:59 Hi pb_ Jun 21 15:09:07 how old do you want? Jun 21 15:09:20 hi khem Jun 21 15:10:10 I'm experimenting with building an oe-like system entirely out of local git repositories, no fetching, so am attempting to dig up / create useful git repositories for all the core apps in the system Jun 21 15:10:20 turns out the savannah svn repo for sysvinit only goes back to 2.88 Jun 21 15:10:27 so i figured i'd see waht i could dig up, for historic value Jun 21 15:11:14 can anyone point me to how either ubuntu or debian create their kernel-headers packages? Jun 21 15:11:18 http://snapshot.debian.org/package/sysvinit/ goes about as far back as i've been able to find so far Jun 21 15:11:46 mm, 2.57 is pretty old Jun 21 15:12:12 indeed Jun 21 15:12:35 there's not much real tangible value from this, I just think it's interesting to have history available for software where possible :) Jun 21 15:12:38 I think back in those days it wasn't so common to archive old versions, the sites generally only kept the "latest" online Jun 21 15:12:53 that is, there's value in the pure scm based build, just not in the resurrection of historic things Jun 21 15:12:56 * kergoth nods Jun 21 15:13:47 hmm, i should check old sources to other old distros Jun 21 15:13:48 e.g. slackware Jun 21 15:13:50 yeah, sunsite (which was the primary site for sysvinit originally) only has 2.83 Jun 21 15:14:03 tsx-11 seems offline, but the mirrors have 2.64 Jun 21 15:14:19 wow, sunsite brings back memories Jun 21 15:14:24 from when i first started getting into linux Jun 21 15:14:46 yeah, I was slightly surprised to find that sunsite.unc.edu:/pub/linux is still online Jun 21 15:15:24 hah. ftp://ftp.slackware-brasil.com.br/historic/slackware-2.0.1/slacksrc/a/sysvinit/SysVinit-2.4.tar.gz, ftp://ftp.slackware-brasil.com.br/historic/slackware-2.0.1/slacksrc/a/sysvinit/SysVinit-2.50.tar.gz Jun 21 15:15:25 nice. Jun 21 15:15:38 heh, the 2.64 readme says "You should use at least this version if you upgrade to the Linux 2.0 kernel" Jun 21 15:15:48 oh, very good Jun 21 15:16:12 I've been having networking issues today.. I can't pull from git.openembedded.org -- is anyone else having this problem? Jun 21 15:16:20 (I can ping it) Jun 21 15:16:31 it's down Jun 21 15:16:54 thanx Jun 21 15:17:02 I'll stop trying to diagnose it here then.. :) Jun 21 15:17:49 not entirely sure what's wrong with it, needs someone with more privs than me at the hosting centre to sort it out. Jun 21 15:18:29 earlier today melo was accepting logins but had a load average of 830. now it doesn't seem to be allowing login at all. Jun 21 15:18:43 ouch Jun 21 15:18:54 sounds like what happens on my servers when spamassassin goes insane.. Jun 21 15:22:29 ah ha, okay, slackware 1.1.2 had sysvinit 2.4, which is in the 2.0.1 source archives. i really doubt i'm going to manage to find anything older than that :) Jun 21 15:22:33 2.4 was 14-05-1993 Jun 21 15:22:35 nice Jun 21 15:22:48 cool Jun 21 15:22:57 nearly 20 years of history, that's not too bad Jun 21 15:23:23 course there's missing intermediate versions, but oh well, its something anyway. can always graft other bits in if i find them later Jun 21 15:23:34 oh my god Jun 21 15:40:36 * pb_ stabs the git fetcher Jun 21 15:40:43 kergoth: what are you going to use it in I wonder Jun 21 15:41:01 kergoth: some atari Jun 21 15:41:06 hehe Jun 21 15:41:28 or just collecting for your museum Jun 21 15:41:40 * pb_ hands kergoth a PDP-11 Jun 21 15:42:26 * khem saw family of PDPs at computer history museum few weeks back Jun 21 15:49:24 khem the one in the mountain? Jun 21 15:56:53 d'oh.. accidently sent a private response to the oe-core mailing list.. whoops.. Jun 21 15:56:58 stupid reply-to.. ;) Jun 21 15:59:47 woglinde: after reading http://gcc.gnu.org/onlinedocs/gcc/Preprocessor-Options.html#index-isystem-930 I'm even more confused Jun 21 16:01:24 hah Jun 21 16:01:28 http://www.oldlinux.org/Linux.old/bin-src/sysvinit.tar.Z Jun 21 16:01:28 - sysvinit 2.0, it seems Jun 21 16:01:35 rock Jun 21 16:01:45 yggdrasil is on that site Jun 21 16:01:50 * kergoth gets all nostalgic Jun 21 16:01:55 sls? Jun 21 16:01:56 heh Jun 21 16:02:24 prerelease version of caldera 1.0, hah Jun 21 16:05:16 so I have a fix for the failure of bitbake master with oe-dev... it's just an avoidance of calling explode_deps on None within native.bbclass and nativesdk.bbclass Jun 21 16:05:30 do I need to send a patch for review to the ml or should I just commit the fix? Jun 21 16:05:49 oe-core does something similar already Jun 21 16:07:37 heh, cool, H.J Lu's bootroot images are there…ahh, the good ol' days Jun 21 16:47:58 GNUtoo: ping Jun 21 16:48:16 pong Jun 21 16:48:49 So I exchange a few email with Steve regarding the libertas_tf and firmware. Jun 21 16:48:58 ok Jun 21 16:49:03 and? Jun 21 16:49:09 So the driver is GNU GPL 2+ Jun 21 16:49:27 While you will be never be able to get the source code of the firmware from Marwell. Jun 21 16:49:36 You can only get the binary version. Jun 21 16:49:40 thats normal Jun 21 16:50:05 yes usually it's the case Jun 21 16:50:08 So, OLPC does have the right to use it and it publishes it publicly. (the old version) Jun 21 16:50:17 what about the new version Jun 21 16:50:18 ? Jun 21 16:50:42 But he doesn't know if you are allowed to use it in other consumer products. Jun 21 16:50:48 There will be no new versions. Jun 21 16:50:53 ok Jun 21 16:50:57 and what about the driver Jun 21 16:51:03 there are no signed-off-by Jun 21 16:51:15 It looks like cozybit didn't even wanted to create it, but that was still done as OLPC was huge deal for them. Jun 21 16:51:16 that can be a problem if someone try to mainline that code Jun 21 16:51:46 He is going to that, but not soon. He is going to do that till the end of this year. Jun 21 16:52:09 ok nice Jun 21 16:52:35 So he send me OLPC contact details regarding the usage of the firmware. Jun 21 16:52:44 So more news later. Jun 21 16:52:52 ok thanks a lot Jun 21 16:53:23 I don't remember exactly, but I think OLPC doesn't user libertas_tf anymore? Jun 21 16:53:46 did they ever use the tf version? Jun 21 16:53:56 I think they used the hard mac version Jun 21 16:54:12 I don't know. Jun 21 16:54:18 (libertas_usb and libertas_sdio) Jun 21 16:54:25 sdio would be for newer models Jun 21 16:54:42 I've no idea if they are out or not Jun 21 16:54:56 like the one based on marvell soc Jun 21 16:55:02 So libertas_tf looks very much dead, just let's hope the latest version get upstream. Jun 21 16:55:21 yes Jun 21 16:58:56 re crofton Jun 21 17:00:40 hi Jun 21 17:01:19 how was the flight? Jun 21 18:02:39 ka6sox-farfarawa: around ? Jun 21 18:11:41 sorta... Jun 21 18:12:20 khem sorta Jun 21 18:12:35 ka6sox-farfarawa: melo is really down Jun 21 18:12:52 really really? :D Jun 21 18:12:58 okay let me get on the console. Jun 21 18:13:05 no ssh nothing Jun 21 18:13:21 let me get on the console and see whats up before I reboot. Jun 21 18:13:37 jas...let me get out of this multiplexer and into that. Jun 21 18:13:43 k Jun 21 18:14:17 I wonder if there is some crazy patch that killed it Jun 21 18:15:40 whats happening is that both apache and the github hook are basically spawning processes... Jun 21 18:15:52 and OOMing the machine Jun 21 18:16:30 now, there may be an underlying FS issue...so I'm going to fsck -f before I restart it. Jun 21 18:19:06 hmmm ok Jun 21 18:19:14 this time its not pw Jun 21 18:31:04 ka6sox-farfarawa: limit apache forks number Jun 21 18:32:18 not sure how to deal with hooks.. Jun 21 18:32:36 iirc, we have disabled all access to git except ssh Jun 21 18:32:53 and hooks are starting on push Jun 21 18:33:05 but I see no large pushes last time Jun 21 18:40:06 Jay7, nginx FTW Jun 21 18:41:53 limit connections there then :) Jun 21 18:42:15 Jay7, I varnish. Jun 21 18:42:56 hehe :) Jun 21 18:43:14 I've said about CDN today Jun 21 18:43:18 okay now that the multiplexer I was working on is fixed...(multiple corefiles clogging it up) Jun 21 18:43:24 ya, CDN good. Jun 21 18:43:25 I'm sure Intel should use some CDN Jun 21 18:43:31 yup Jun 21 18:43:36 we may ask to place OE site into CDN Jun 21 18:43:40 I can work on melo now Jun 21 18:43:52 at least our site will be always accessible Jun 21 18:45:40 ya, right now I will put it back online..but I will have to schedule outage to convert from apache to nginx/varnish Jun 21 18:46:25 first...fsck -f / Jun 21 18:47:23 you already have outage Jun 21 18:47:28 use it :) Jun 21 18:47:36 reuse even :) Jun 21 18:48:12 "network/server outage and how to reuse it" Jun 21 18:48:20 for dummies :) Jun 21 18:48:30 good idea to write own book :) Jun 21 18:50:46 can someone send an email to the ML...I'm kinda busy now. Jun 21 18:55:21 ka6sox-farfarawa: email about services state? Jun 21 18:55:33 Jay7, FS not happy Jun 21 18:55:38 Jay7, yes Jun 21 18:55:52 f[*s]ck :) Jun 21 18:56:10 I'll send to oe-devel Jun 21 18:56:21 and yocto too if you would... Jun 21 18:56:28 what services affected? www, git, what else? Jun 21 18:56:37 www, git, patches Jun 21 18:56:37 yes, yocto is ok Jun 21 18:56:41 ok Jun 21 18:56:43 ka6sox-farfarawa: that means disk is broken? Jun 21 18:56:50 FS is not happy. Jun 21 18:56:56 ka6sox-farfarawa: or just the filesystem is broken? Jun 21 18:56:59 FS Jun 21 18:57:03 ok Jun 21 18:57:12 its old and crufty...needs to be started over. Jun 21 18:57:18 and just transfer data. Jun 21 18:58:18 xm create -c melo Jun 21 18:58:24 er...not gonna work. Jun 21 18:59:13 khem please look @ parsemail and see if its choking? Jun 21 19:01:54 yey! repos online again heh Jun 21 19:05:13 ka6sox-farfarawa: can you say time expected when melo will be up again? Jun 21 19:06:58 Does Machine configuration override Distro configuration in OE? Or viceversa? Jun 21 19:07:20 iirc, yes Jun 21 19:07:44 but it is easy to check :) Jun 21 19:08:22 it does. machine should always have final say over distro. machine generally defines capabilities and hardware requirements which aren't negotiable, whereas the distro is more general / policy Jun 21 19:08:51 ka6sox-farfarawa: mail sent Jun 21 19:09:42 kergoth, ok, thanks Jun 21 19:10:10 Jay7, its sorta up now... Jun 21 19:10:26 ka6sox-farfarawa: reply to it then :) Jun 21 19:10:27 but I will need to schedule a time (maybe tonight) when we convert it. Jun 21 19:10:44 * joelagnel is having a hard time figuring out which kernel version is chosen for the angstrom-2008.1 distro and a beagleboard Jun 21 19:11:06 It seems that builds chose 2.6.32, but I can't see this in the /conf for either the distro or the machine Jun 21 19:11:14 ka6sox-farfarawa: can't you copy VM, setup it properly and then replace old? :) Jun 21 19:11:42 well.. not just replace, but merge :) Jun 21 19:12:49 Any ideas? :) Jun 21 19:13:26 joelagnel: check linux/* recipes for default preference Jun 21 19:15:55 Jay7, just found: Jun 21 19:15:55 linux/linux-omap_2.6.29.bb:DEFAULT_PREFERENCE_beagleboard = "1" Jun 21 19:16:58 joelagnel: run bitbake virtual/kernel and check which version will be built :) Jun 21 19:17:19 Jay7, sure I was about to do that, seeing is believing ;) Jun 21 19:17:28 Just trying to understand the mechanics of how it works, a bit Jun 21 19:17:33 was it 2.6.32? Jun 21 19:17:46 Jay7, will check and let you know, thanks Jun 21 19:17:53 there is linux-headers also Jun 21 19:17:59 and headers may be .32 Jun 21 19:18:13 final kernel version is .32 AFAIK, but don't know Jun 21 19:18:46 angstrom 2008.1 doesn't specify a preference, neither does the beagleboard machine, so there's a default getting into virtual/kernel somewhere Jun 21 19:21:10 ooh gotcha Jun 21 19:21:38 conf/beagleboard.conf prefers linux-omap-psp as provider for kernel Jun 21 19:21:50 and linux-omap-psp has a highest version of only .32 Jun 21 19:22:05 Could that be it? Jun 21 19:22:17 look like Jun 21 19:22:28 Jay7, cool :) Jun 21 19:25:31 pb_, you were right. Jun 21 19:25:41 but I'm still going to convert it. Jun 21 19:35:09 ka6sox-farfarawa: okay cool Jun 21 19:36:09 FS needs to be replaced and brought up to the current standard. Jun 21 19:37:54 is openembbeded.org being ddosed? Jun 21 19:38:06 yes Jun 21 19:38:09 webpage is quite unusable Jun 21 19:38:55 is there any form of webarchive or alternative mirror? Jun 21 19:39:36 dockside_, I'm trying to isolate the sources. Jun 21 19:39:36 dockside_: google cache? Jun 21 19:39:42 googlecache is good. Jun 21 19:40:11 didn't think of that Jun 21 19:40:13 thanks Jun 21 19:40:17 site should be mirrored or go into CDN.. Jun 21 19:40:42 what manner of hosting is there at present? Jun 21 19:40:49 pretty good... Jun 21 19:41:17 dockside_, apache sucks at the loads we are seeing...tonight I'll varnish. Jun 21 19:41:27 ah, Jun 21 19:41:29 yeah Jun 21 19:41:45 apaches connection scaling is not acceptable Jun 21 19:41:49 nope Jun 21 19:41:57 probably go to nginx too. Jun 21 19:42:08 since that takes like 10% of the RAM. Jun 21 19:42:17 :) Jun 21 19:42:48 that combination has survived 5 /.'ings Jun 21 19:42:57 :D Jun 21 19:42:59 on another project on the same server. Jun 21 19:43:05 (well same dom0 anyways) Jun 21 19:43:29 well, if it surivived /. that well Jun 21 19:43:38 it takes quite a ddos to compare Jun 21 19:43:52 dockside_, we have Gb access in the location we are. Jun 21 19:44:07 this is a local server performance issue. Jun 21 19:44:14 it *will* be rectified. Jun 21 19:44:27 nice :) Jun 21 19:44:44 don't let me disturb then and gl Jun 21 19:45:13 will wait till .eu is asleep and .us is gone home. Jun 21 19:45:18 i would love to help but i feel that is out of my domain Jun 21 19:45:18 khem, ping Jun 21 19:46:36 ya, I need more help too...maybe I recruit some folks I know who have the experience :D Jun 21 19:50:24 * Jay7 should upgrade his experience with nginx too Jun 21 19:58:28 only 40Mb now... Jun 21 20:21:15 ka6sox-farfarawa, Jay7 and others: thanks for your dedication on resolving the server issue! Jun 21 20:21:54 I've created a custom toolchain/sdk. In my task-MY_TOOLCHAIN-toolchain-target.bb file I have listed directfb in RDEPENDS_${PN}, but the header files are still not present in the SDK Jun 21 20:22:28 I have followed the guide here: http://docs.openembedded.org/usermanual/html/commonuse_qte_sdk.html Jun 21 20:23:06 Hmm... Am I supposed to list directfb-dev? Jun 21 20:25:05 I'll try that Jun 21 20:28:43 denix, glad we can at least come to a conclusion...now to implement. Jun 21 20:53:20 denix: I'm just "walking around", all real work is done by ka6sox-farfarawa :) Jun 21 21:13:42 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rdf3e3de4e4 10openembedded.git/recipes/linux/ (2 files in 2 dirs): Jun 21 21:13:42 Add new linux-kexecboot_git-nokia900-meego recipe Jun 21 21:13:42 This is a linux-kexecboot for the nokia900, Jun 21 21:13:42 A classic kexecboot recipe is not used because Jun 21 21:13:42 this one has patches(inside the git tree of the kernel) Jun 21 21:13:43 for proper shut down,and a lot of other machine specific patches. Jun 21 21:13:43 Signed-off-by: Denis 'GNUtoo' Carikli Jun 21 21:33:03 GNUtoo: hi Jun 21 21:33:16 is kexecboot working ok on n900? :) Jun 21 21:33:31 not ok Jun 21 21:33:37 still the 1min delay Jun 21 21:33:42 ah Jun 21 21:33:55 03Tom Rini  07master * rb6838a476a 10openembedded.git/recipes/udev/ (udev.inc udev/cache udev_124.bb): Jun 21 21:33:55 udev: Fix cache problem Jun 21 21:33:55 When /dev/shm is re-mounted we need to re-create the files that end up Jun 21 21:33:55 in /etc/udev/saved.*. In addition, udev 124 wasn't making an initial Jun 21 21:33:56 saved.uname file. Jun 21 21:33:56 Signed-off-by: Tom Rini Jun 21 21:33:58 that's bad.. Jun 21 21:34:31 yes Jun 21 21:34:57 can I push that (along with an update to the inc file of kexecboot) http://patchwork.openembedded.org/patch/5077/ Jun 21 21:35:29 not sure about '-e' removal Jun 21 21:35:46 can't remember why we leave it as is and override in machine recipes Jun 21 21:35:57 the -e is an error Jun 21 21:36:05 it doesn't work for lzma Jun 21 21:36:13 ah...and lz fails for me Jun 21 21:36:22 *xz Jun 21 21:36:28 better to ask Andrea about this Jun 21 21:36:33 so I guess I will want to push my version of my patch Jun 21 21:36:54 Jun 17 17:36:16 we have to change linux-kexecboot.inc to reflect lzma->xz dependency Jun 21 21:37:21 I have nothing against that patch Jun 21 21:37:22 when talking about: Jun 21 21:37:23 Jun 17 17:32:53 GNUtoo: http://patchwork.openembedded.org/patch/5077/ Jun 21 21:37:37 mine is at: Jun 21 21:37:57 http://patches.openembedded.org/patch/6073/ Jun 21 21:38:21 #5077 fails for me Jun 21 21:38:38 but #6073 works fine Jun 21 22:45:18 ka6sox-work: tell me Jun 21 22:45:22 I am here now Jun 21 23:00:34 03Matthew McClintock  07master * r9e3be665b6 10openembedded.git/recipes/valgrind/valgrind_3.6.1.bb: Jun 21 23:00:34 valgrind: Fix shebang issues when compiling valgrind on old bash Jun 21 23:00:34 valgrind is doing some odd substitutions on the shebang for some Jun 21 23:00:34 scripts. This overrides the perl location and uses one that Jun 21 23:00:34 should work. This fixes building valgrind on centos 5.6 Jun 21 23:00:35 Signed-off-by: Matthew McClintock Jun 21 23:00:36 Signed-off-by: Tom Rini Jun 21 23:01:12 well that was quick Jun 21 23:01:46 Tartarus: this is my irc handle by the way Jun 21 23:01:58 msm: Hey there :) Jun 21 23:17:30 Is there anything I can do about "No GNU_HASH section" QA errors on binaries provided by the vendor, other than overriding/disabling those QA stages? Jun 21 23:22:45 housel: If you can't fix them, you need to skip them then Jun 21 23:23:10 INSANE_SKIP = True in the recipe, iirc Jun 21 23:23:35 or better INSANE_SKIP_${PN}-subpackage-that-contains-the-bad-bits = True Jun 21 23:24:13 ok, thanks Jun 21 23:52:35 khem, ping? Jun 22 01:49:52 housel: you should fix the QA error Jun 22 01:49:57 which package is that ? Jun 22 01:59:27 It's my own private/local package, based on binary GPU drivers from Freescale Jun 22 01:59:32 So I can't fix it **** ENDING LOGGING AT Wed Jun 22 02:59:57 2011