**** BEGIN LOGGING AT Wed Nov 30 02:59:57 2011 Nov 30 03:25:54 i added a few recipes to task recipe however they're not in image, just in seperate ipks, what did i do wrong? Nov 30 03:29:02 xxiao: add them to IMAGE_INSTALL Nov 30 03:29:18 xxiao: is it using oe-core Nov 30 03:31:57 khem: i'm actually running the legacy oe Nov 30 03:32:19 but strange thing is that it worked for all packages i added in the past, but not today :( Nov 30 03:33:06 xxiao: I do something like IMAGE_INSTALL_append_pn-console-image = " libstdc++" Nov 30 03:33:18 if I want to add libstdc++ to console-image Nov 30 03:33:20 yes it's in IMAGE_INSTALL Nov 30 03:33:48 i created a new task, and a new image, the image has IMAGE_INSTALL = mytask Nov 30 03:34:41 xxiao: I have no way to look what u did Nov 30 03:34:52 maybe i need increae PR number? Nov 30 03:34:58 trying that Nov 30 03:35:07 i added a few packages, and it all worked, until today that is Nov 30 03:35:27 you can just do clean on .bb and .bb Nov 30 03:35:27 need add sysklogd, as syslogd is not a server, i need a syslogd server Nov 30 03:35:38 oh let me try. thanks Nov 30 03:37:45 khem: bump PR made it, i guess you're right i should clean then rebake Nov 30 06:29:41 offtopic: anyone using rpmbuild? Nov 30 06:31:54 Jin|away: you need BB_LOCALCOUNT_OVERRIDE = "" to keep LOCALCOUNT incrementing and to fake values you need to convert sqlite db in cache/bb_pers* Nov 30 06:32:25 morning JaMa|Off Nov 30 06:35:52 moin Nov 30 08:01:13 I m configuring D-Link DWA-125 for ARM.... when I insert a USB dongle.... I m getting .... " " Nov 30 08:01:20 anyone knows how to activate console=ttyS0 during runtime with a need to reboot and have it on the kernel parameter Nov 30 08:01:33 without a need ... Nov 30 08:01:48 but still when I run "iwlist scan" .....I m getting " No scan result" Nov 30 08:02:17 can anyone help why it is not configuring.... Nov 30 08:09:33 good morning Nov 30 08:54:57 mornin Nov 30 08:56:28 moin Nov 30 08:56:38 hi JaMa Nov 30 08:57:19 JaMa|Off: based on my google research you are the one to talk to when it comes to questions regarding AUTOREV, LOCALCOUNT and stuff like that? :) Nov 30 08:57:53 see backlog :) Nov 30 08:58:05 and yes :) Nov 30 08:58:22 indeed let me scroll up :) Nov 30 08:59:25 JaMa|Off: aha.. I thought I just set BB_LOCALCOUNT_OVERRIDE = "" and that then I could do LOCALCOUNT = "xx" in my recipe and it would act as a PRINC, i.e. adding count from the db to my local count Nov 30 08:59:38 so I do have to mess around with the database directly... Nov 30 09:00:39 yes with default BB_LOCALCOUNT_OVERRIDE = "1" you can override LOCALCOUNT from database with value like LOCALCOUNT = "xx" in your recipe Nov 30 09:01:03 but if you want it to do auto increments you have to disable BB_LOCALCOUNT_OVERRIDE Nov 30 09:01:25 but be aware of implications of it for multiple builders not sharing the same persistent database Nov 30 09:02:12 JaMa|Off: well thats exactly my problem - I had to wipe TMPDIR Nov 30 09:02:17 but I need those packages to be upgradable Nov 30 09:02:47 then just copy that database Nov 30 09:03:30 can I enforce LOCALCOUNT but keep the autoincrement? actually that would be a simple solution for builders, in local.conf they could just do LOCALCOUNT_pn = "xx" to sync with each other, and the counts would still get incremented each build, i.e. LOCALCOUNT + database_count or something like that? Nov 30 09:03:44 das bitbake kommt mit Nov 30 09:03:47 oops wrong window the last line Nov 30 09:06:20 no you cannot Nov 30 09:07:01 BB_LOCALCOUNT_OVERRIDE is basicly switch between LOCALCOUNTs from persistent db (where they are auto incremented) and LOCALCOUNT variables where they are overriden with constant value Nov 30 09:07:38 ok, so the "solution" or rather hack would be to either copy the database (if available) or just go into sqlite and set the value in the db? Nov 30 09:08:50 no other ways.. ok Nov 30 09:08:52 thanks Nov 30 09:12:48 yes I do even share db on builder web Nov 30 09:13:16 so if you want to get upgradable path from image downloaded from that builder you can download the db before starting your build Nov 30 09:13:38 so you can start with same LOCALCOUNTS and your feed will be "compatible" then with feeds from where the image was created Nov 30 09:13:51 but that's only "one-time" sync Nov 30 09:14:08 I see, well, too late for downloading build is already done, so I guess I'll fix it with the sqlite command line client Nov 30 09:15:39 be aware that there was also change of format for keys in that db (branch added to key) so if you have very old db (like >1y) then you have to migrate carefully Nov 30 09:17:30 well right now I did a clean build recently, so the fix will be done on a db that is about 1 month old or so Nov 30 09:18:13 Jin^eLD: easiest way to fix it is to update all keys with max(value) + 1 if it's "local" build :) Nov 30 09:18:32 or just some high enough magic number you like Nov 30 09:18:44 what exactly is the benefit of the gitpkgv class? Nov 30 09:19:35 SRCPV is updating PV (build time version), gitpkgv is updating PKGPV (runtime version) Nov 30 09:20:45 and gitpkgv is more stable as it's using real revision count if available Nov 30 09:21:44 I saw that gitpkgv had some pretty high number Nov 30 09:21:45 something like BB_GIT_CLONE_FOR_SRCREV does Nov 30 09:21:50 and I was not sure if it would go back or not Nov 30 09:22:07 it shouldn't unless upstream is rebasing.. Nov 30 09:22:14 so would that be correct? RCREV="${AUTOREV}" PKGV = "1.3.1+gitr${GITPKGV}" PV = "${PKGV}" Nov 30 09:23:05 aem, SRCREV would the first one ;) lost in paste Nov 30 09:23:44 and inherit gitpkgv Nov 30 09:24:06 I read the whole thread from 2009 about the AUTOREV stuff yesterday and it was still not clear to me what to do :)) I thought by now there would be a simple solution Nov 30 09:24:36 what confused me most was the fact that 1.3.1 has no meaning, I'd expect this version string to always have a higher priority than whatever is appendet to it Nov 30 09:25:11 it has 1.3.1+whatever is higher than 1.3.0+whatever Nov 30 09:25:19 unless the later has higher PE Nov 30 09:26:26 I did not touch PE manually and in my case 1.3.1+gitr12+somehash was treated as lower than 1.3.0+gitr76+somehash Nov 30 09:26:55 that's wrong and you should debug it Nov 30 09:27:13 wrong as in opkg problem then? Nov 30 09:27:29 I see Nov 30 09:27:48 if you still have those 2 ipk you should double check the version Nov 30 09:28:11 because I haven't seen this and it should take 1.3.1 as higher for sure.. Nov 30 09:28:20 I don't have the gitr12 one anymore Nov 30 09:28:30 but I could make one of course Nov 30 09:30:49 or you can create dummy .ipk files with such version fields and then try to install them Nov 30 09:31:07 and if it doesn't work those dummy will be usefull for bugreport.. Nov 30 09:31:14 but I belive it will work Nov 30 09:31:47 I'll try Nov 30 10:01:08 Hi all... I want to use autogen.sh from opnenjdk but it failes on the target with Can't locate Autom4te/ChannelDefs.pm in @INC (@INC contains: @autom4te_perllibdir@ - what need I to install? Nov 30 10:23:01 where does pthreads "come from"? I'm trying to find out why pthread_create gives an EPERM on oe-core but not on oe-classic. it doesn't accept my scheduling attributes. Nov 30 10:26:40 I thought from libc, no? Nov 30 10:28:32 Jin^eLD: grepping led me to that thought as well, so I'm trying to find out if there is a difference between my classic and core image in that respect. Nov 30 10:33:10 hmm, update-rc.d does RUNLEVEL=`runlevel | cut -d" " -f2` but there is no "runlevel" command on my system.. where is it supposed to be coming from? Nov 30 10:33:44 ha never mind, found it;) Nov 30 10:40:33 update-rc.d should depend on sysvinit, otherwise one might end up without the "runlevel" utility :P Nov 30 11:29:49 hmm I have a really strange problem Nov 30 11:29:54 http://pastebin.mozilla.org/1389379 Nov 30 11:30:00 it fails to process the license file Nov 30 11:30:08 I am not sure if its a bug in bitbake Nov 30 11:30:11 the COPYING file is there Nov 30 11:30:29 but log still says /dss-oe/angstrom-devel-build-eglibc/work/at91sam9g20ek-angstrom-linux-gnueabi/sysvinit-inittab-2.88dsf-r10/sysvinit-2.88dsf/COPYING could not be copied for some reason. It may not exist. WARN for now. Nov 30 11:31:04 however on another system it just works Nov 30 11:31:08 the "other" is ubuntu Nov 30 11:31:13 on Fedora 14 it doesn't Nov 30 11:33:15 the last operation in the log is cp /dss-oe/angstrom-devel-build-eglibc/work/at91sam9g20ek-angstrom-linux-gnueabi/sysvinit-inittab-2.88dsf-r10/COPYING /dss-oe/angstrom-devel-build-eglibc/work/at91sam9g20ek-angstrom-linux-gnueabi/sysvinit-inittab-2.88dsf-r10/sysvinit-2.88dsf/ Nov 30 11:33:23 that cp works if done on command line Nov 30 11:35:33 hmm, is alsa-state just not brought over to oe-core or are there other mechanisms to handle alsa state files now? Nov 30 11:37:10 * JaMa|Off fsodeviced provides way to handle alsa states Nov 30 11:38:11 Jin^eLD: I've seen similar errors when it's a link to something which doesn't work later or not at all on the build system Nov 30 11:39:16 JaMa|Off: were you able to figure out to what that link could be? Nov 30 11:40:19 JaMa|Off: well, fso is not used here. Nov 30 11:41:03 * stefan_schmidt has a look at some more BSP layers Nov 30 11:49:51 well funny thing is that it seems to be taking the checksum from somewhere else | ERROR: sysvinit-inittab: The new md5 checksum is 94d55d512a9ba36caa9b7df079bae19f Nov 30 11:50:07 but only on my system, if I add that in bbappend as override it would work for me, but on other systems it does not Nov 30 11:50:17 so at least on Fedora something must be going terribly wrong Nov 30 11:53:03 Jin^eLD: I haven't seen it in sysvinit, but I've fixed it in openssl/libxml2/libxslt Nov 30 11:54:00 JaMa|Off: how did you fix it? Nov 30 11:54:16 I looked into my WORKDIR now ant indeed there is a COPYING file there with the wrong checksum Nov 30 11:54:45 see the git log Nov 30 11:54:47 but NOT the one from openembedded-core/meta/recipes-core/sysvinit/sysvinit-inittab Nov 30 11:55:59 I think I found your commit Nov 30 12:06:20 JaMa|Off: thanks, seems you are right, someone is messing aroun with COPYING in the background, I renamed the file ant it worked out Nov 30 14:27:25 do we have a recipe for sudo? Nov 30 14:28:40 hmm, looks like there is Nov 30 14:49:53 hi , is dm355 capable of image processing Nov 30 14:55:17 Guh, can people please use PATCH in their patch emails? Nov 30 14:55:28 [meta-efl ..] etc are not all in my filters :P Nov 30 14:55:42 kergoth: sorry Nov 30 14:56:02 * kergoth adds more common ones to try to catch most of them Nov 30 14:56:11 I'll update my bash aliasses Nov 30 14:56:41 I auto-archive patch mails to a label in gmail, but let non-patch mails hit the inbox, for discussions. so going to review patches is an explicit, separate operation from responding to other mails Nov 30 14:57:04 maybe that's a weird way of doing things, but I do think its useful to be able to identify them :) Nov 30 15:01:00 * kergoth yawns Nov 30 15:41:39 hi ericben Nov 30 15:43:03 JDuke128, you mean like having a DSP for jpeg? Nov 30 15:43:13 like for instance using the TI DSP for jpeg? Nov 30 15:47:44 ok there is definitely something wrong with the LIC_FILES_CHKSUM stuff, I do not know who or what is doing it in the background, but it seems that it does not check the source file that is inside the package, but that it checks some COPYING that is linked to the system; so when building on a different system the build fails Nov 30 15:48:12 GNUtoo , i want to edit pixels of video Nov 30 15:48:23 is it possible on dm355 ? Nov 30 15:48:35 i will add effects on video at runtime Nov 30 15:48:50 ah ok, no idea Nov 30 15:49:09 first identify the software that could do that Nov 30 15:49:10 i need fastest way to edit pixels Nov 30 15:49:31 concurrent Nov 30 15:49:40 DSP can handle that or i need OpenCL really? Nov 30 15:49:57 and i donno any embedded device having OpenCL capable now. Nov 30 15:50:53 JDuke128: Its probably best if you had a DSP Nov 30 15:51:08 DM368 ? Nov 30 15:51:16 i ve leopardboard it has DM368 inside Nov 30 15:51:34 JDuke128: Nope. That doesn't have a DSP either. Maybe OMAP3 like Beagleboard Nov 30 15:51:56 i ve OMAP3 too Nov 30 15:52:04 OMAP3 can handle ? Nov 30 15:52:09 Yes. Nov 30 15:52:29 hmm , whats the best embedded device that has huge DSP and low cost ? Nov 30 15:52:45 panda board? Nov 30 15:52:46 OMAP3 is big for me i dont need OpenGL ES Nov 30 15:53:08 i just need concurrent pixel calc. Nov 30 15:53:17 and edit 2d pixels Nov 30 15:54:17 Beagleboard is probably the cheapest TI board. Nov 30 15:54:54 If you algorithm can use Neon then you don't need DSP. Nov 30 15:55:07 Its all a tradeoff Nov 30 15:55:34 neon is slow? for adding 1080p video effects Nov 30 15:55:57 1080p60? Nov 30 15:56:04 1080p30 Nov 30 15:56:18 Might work but the ARM will spend a lot of cycles on it Nov 30 15:56:21 for example at runtime , video effects i will add Nov 30 15:58:03 like , adding rain effect on 1080p30 at realtime Nov 30 15:58:35 That's tough. Nov 30 15:58:56 If the rain effect is static, then maybe you could blend it with the video Nov 30 15:59:12 which device can do it ? dm368 ? Nov 30 15:59:31 or OMAP4 ? Nov 30 15:59:49 OMAP4 should be able to do much better than DM368 Nov 30 15:59:59 Dual core A9 vs. Arm9 Nov 30 16:01:04 hmm Nov 30 16:01:22 DM368 can add rain effect at runtime ? Nov 30 16:01:45 realtime Nov 30 16:01:45 How many MHz does the Algorithm take? Nov 30 16:02:01 You need to profile it first. Nov 30 16:02:10 static image i will add Nov 30 16:02:14 or blend Nov 30 16:03:07 hmm, I have a library that links vs libxml2, libxml2 is in DEPENDS of the packge, still the library did not get installed on the target, why could this be? Nov 30 16:04:06 JDuke128: DM368 does support blending in hw. Nov 30 16:04:08 the control file says Depends: libxml2 (>= 2.7.8) Nov 30 16:04:11 should be fine.. Nov 30 16:07:07 it seems its updated to a newer version, then more packages get updated, then during that libxml2 becomes orphaned and is autoremoved again Nov 30 16:27:01 SidH_ , thanks alot and what about editing pixels ? can we do manually edit pixels ? Nov 30 16:53:26 Hi there, the meta/recipes-graphics/xorg-lib/libx11-trim_1.4.4.bb:SRC_URI[md5sum] needs to be updated, right? Nov 30 16:56:18 never mind, it has been noticed already (just saw the corresponding patch on the oe-core ml) Nov 30 17:45:54 bluelightning, did I fix the wiki for you? Nov 30 17:56:22 ka6sox: sorry, but it's still prompting me for a captcha... Nov 30 17:57:15 bbl Nov 30 17:57:48 I guess it doesn't like him...I need to read the logs and see whats up :( Nov 30 17:59:06 hi zecke Nov 30 17:59:20 ho Nov 30 17:59:50 zecke, there are a number of things that were on the server with your uid on them...do you need them? Nov 30 18:16:16 florian, ping? Nov 30 18:17:01 ka6sox: pong Nov 30 18:22:54 khem, ping? Nov 30 18:28:53 ka6sox: is that with a public folder? or on a build machine? Nov 30 18:29:40 I think this was related to the build stats. Nov 30 18:30:01 on the front facing server. Nov 30 18:32:52 sgw: there? Nov 30 18:34:42 otavio: physically yes, mentally not so sure right now! Nov 30 18:34:51 sgw: ok Nov 30 18:34:59 sgw: ping me when you can talk, heh Nov 30 18:35:21 otavio: what's up, I will do my best, just a bit sick today. Nov 30 18:35:24 sgw: i am doing the initramfs system and want to talk to you about it Nov 30 18:36:23 sgw: first, the base. https://github.com/OSSystems/oe-core/commit/7d24fbb2ca6f8c3e9ae81ec0cd62dcba7fe7c9c0 Nov 30 18:36:40 sgw: I am working on modules now but base seems flexible enough for me, for now Nov 30 18:38:24 so this is for initramfs and sysvinit, not systemd or am I confusing 2 orthognal ideas? Nov 30 18:38:52 sgw: only initramfs Nov 30 18:40:31 sgw: the hook system allow for fancy features Nov 30 18:40:52 sgw: i am doing a module that, if you give: shell=after:ext3 Nov 30 18:41:05 it will open a shell just after finishing ext3 module ;-) Nov 30 18:45:15 hey. are there any resources there regarding psplash? Nov 30 18:45:35 i need to add a different psplash backend Nov 30 18:47:00 the project URL is now giving a 404, so... where are the docs now? Nov 30 18:48:05 CMoH-notebook: meta-angstrom has psplash recipes Nov 30 18:48:12 CMoH-notebook: you might look at them Nov 30 18:48:32 CMoH-notebook: oe-core is the base for that one also, otavio does angstrom bbappend? Nov 30 18:49:10 CMoH-notebook: are you trying to generate a new image.h file? Nov 30 18:49:24 Or change the something else? Nov 30 18:49:48 no; i want to use the splash infrastructure that exists there, but to have my own psplash/psplash-write programs that use a totally custom LCD Nov 30 18:50:19 sgw: not sure Nov 30 18:50:22 i have the code to write stuff on the LCD, but it's not via a kernel framebuffer device Nov 30 18:50:27 CMoH-notebook: you mean the init? Nov 30 18:50:35 CMoH-notebook: oe-core doesn't has it yet Nov 30 18:51:09 CMoH-notebook: not on initrd Nov 30 18:51:17 so i figured if i could take a look at the psplash API, i could write a my-psplash alternative that draws what needs to be drawn on the LCD during initscripts Nov 30 18:51:21 CMoH-notebook: you can do it in regular system thought Nov 30 18:51:50 CMoH-notebook: not sure psplash is the way to go but ... Nov 30 18:52:21 what would you suggest? Nov 30 18:55:16 CMoH-notebook: really not sure ... I've found some other boot splash alternative a while a go. Fedora uses a X11 based but this also works without X AFAIK Nov 30 18:55:48 CMoH-notebook: but if possible, try to keep it simple as it stay a small amount of time on screen Nov 30 18:56:39 CMoH-notebook: I didn't start to play with splash in oe-core since the products that need it still runs on classic oe, but will do for sure Nov 30 18:56:40 thanks otavio; i've managed to find the sources using the meta-angstrom layer Nov 30 18:56:46 CMoH-notebook: good Nov 30 18:57:07 i'll take a look at them and see if they fit with my endeavor Nov 30 18:57:17 sgw: looked at the code? Nov 30 18:57:20 I was just going to point you to http://git.yoctoproject.org/cgit/cgit.cgi/psplash/ Nov 30 18:59:46 well, for now it seems the idea of using the whole of psplash, except the actual backend (the psplash program) fits with what i want Nov 30 19:00:03 i.e. reading commands from the pipe and reacting by drawing on the LCD Nov 30 19:01:07 this is great! Nov 30 19:01:11 thanks otavio, sgw Nov 30 19:01:59 otavio: yes, other than the get_bootparam always returning 0, and the "reverse sensing" of enabled in the shell (returning 0), seems OK to me, would you propose replacing the existing init-* with modules and using this infrastructure? Nov 30 19:02:46 sgw: the get_bootparam was mangled on debug here; this was just to show to you for initial discussion Nov 30 19:02:55 sgw: later yes Nov 30 19:03:17 sgw: I intend to talk to their authors and ask for them to rewrite to use it Nov 30 19:03:27 sgw: but i will send few modules too Nov 30 19:03:37 sgw: first is a debug one Nov 30 19:05:41 sgw: in shell, 0 is success and != 0 is error Nov 30 19:05:54 sgw: so this is ugly but don't blame me Nov 30 19:10:08 re Nov 30 19:11:21 otavio: yes I know about ugly shell 0 issue! Nov 30 19:11:59 sgw: but seems flexible enough? Nov 30 19:12:50 otavio: I think you are on a good track and it seem flexible yes, I need to review it with install/live situations, that would really be a good Proof of Concept if you can convert and run the install and live with them as modules. Nov 30 19:13:33 sgw: I don't intend to port those as I don't use them; I will wrote the new ones I need and post together Nov 30 19:13:47 sgw: I will add handle to mdev and udev onto base too Nov 30 19:16:11 sgw: but mdev and udev will be optional; if they're not found it will keep running so if you want, then you can not use them Nov 30 20:20:27 Hi, I have a problem with some uclibc functions. I get "undefined reference to `pipe2'" if I try to compile qt4-embedded. Also udev 171 doesn't compile. There I get undefined reference to "accpet4()", "timerfd_create()", ..., . My uclibc version is the git version of oe-core backported to classic oe and I build for avr32. Why I get these errors? Nov 30 20:22:44 because avr32 on uclibc dont hast accept4 and pip2 implemented Nov 30 20:27:32 but they are defined in the headers. any chance to get them working? Nov 30 20:29:06 definied in the header doesnt mean implemented Nov 30 20:29:13 dont know Nov 30 20:29:24 I didnt saw this much patches vor avr32 Nov 30 20:29:47 maybee you can try something similiar to arm Nov 30 20:30:55 hmm avr Nov 30 20:31:33 there are some backports Nov 30 20:31:43 I have not yet proposed to oe-core Nov 30 20:31:57 since there is another bug in uclibc that is at large atm Nov 30 20:32:24 khem uhm which one? Nov 30 20:32:24 thats the git version Nov 30 20:33:02 woglinde: http://lists.uclibc.org/pipermail/uclibc/2011-November/046028.html Nov 30 20:33:21 woglinde: it comes with a merge of uclibc prelink branch Nov 30 20:34:06 ah prelink Nov 30 20:37:46 yes, avr... seems to me that there aren't many people using it. khem, so is it possible to get the accept4(), pipe2(), ..., functions? Nov 30 20:38:14 Splash: sure. check if kernel implements those syscalls Nov 30 20:42:17 if this is the correct file, then it doesn't seem to be implemented. http://git.kernel.org/?p=linux/kernel/git/stable/linux-stable.git;a=blob;f=arch/avr32/kernel/syscall_table.S;h=6eba53530d1c552ea9813a16419a679c27075415;hb=HEAD Nov 30 22:19:40 khem, pingu? Nov 30 22:54:00 kergoth: did you once post your mirror setup for OE builds? Nov 30 23:53:25 gm all, long time no see Nov 30 23:53:54 likewise: hey Nov 30 23:53:59 hope all is well Nov 30 23:54:00 hey likewise Nov 30 23:54:03 long time indeed! Nov 30 23:54:07 you kind of disappeared Nov 30 23:58:09 hiya likewise Dec 01 00:00:01 yes, non-Linux stuff... happens to me once in a while :) Dec 01 00:00:22 lwIP (light weight TCP/IP) related work was harassing me Dec 01 00:00:45 next project, OE again, phew :) Dec 01 00:01:23 so thanks guys for keeping it going Dec 01 00:01:47 how's life on the channels? Seem quite busy on the ML/IRC. Dec 01 00:01:55 its pretty scary when I know more about OE than the Devs working for me:P Dec 01 00:17:35 cbrake: https://bitbucket.org/kergoth/dotfiles/src/1e0e3df34f6f/oe/classes/mirrors.bbclass is what I use, personally. Might add a few extra, but that covers the bases in my experience Dec 01 00:18:13 cbrake: drops the crapload of location based mirrors, as there's little point to checking them all, they're either down or they're up and in sync most of the time. better to leverage archival mirrors Dec 01 00:25:00 nite **** ENDING LOGGING AT Thu Dec 01 02:59:58 2011