**** BEGIN LOGGING AT Sun Mar 11 02:59:58 2012 Mar 11 03:18:02 etheretic, you can contact me ... I have several propasal for you Mar 11 03:19:17 whoa Mar 11 03:19:33 i can run ARM code on x86? Mar 11 03:21:12 i did an enormous doubletake Mar 11 03:23:46 panicking_mike: i'm not a programmer. used arm-based acorn comps from 92 to 2002, loved riscos. see apotential in the bb as the engine in an eat-your-heart-out-garmin contender, playing movies and musicalong with providing opencpn functionality (primarily). Mar 11 03:25:19 I'm working on Android ICS on my reference design Mar 11 03:28:19 I have seen opencpn Mar 11 03:28:26 I think that is possible to move on Android Mar 11 03:28:31 I don';t know if it running Mar 11 03:28:41 I have done some work in the past with naivionics Mar 11 03:28:45 and geonav Mar 11 03:34:44 android = google. Google =sp00ks. they're even bragging about it along with fac3b00k. erk. Mar 11 03:35:41 stay assembler. it's verifiable. Mar 11 03:43:18 etheretic: assembler is only verifiable if you have a known trusted machine Mar 11 03:43:34 and if you have a known trusted machine, why not compile on it Mar 11 03:45:05 arm code has been expended since arm7/stongarm - but it's still risc. fast. Mar 11 03:45:21 expanded, even. Mar 11 03:46:13 huh? Mar 11 03:46:25 if someone tampers with your assembler, you Mar 11 03:46:30 're just as screwed Mar 11 03:47:30 etheretic, can you explain better? what do you mean? Mar 11 03:57:30 panicking_mike: interpretable languages are way more hackable than down-on-the-silicone langs - or am i mistaken? Mar 11 04:03:01 hint @ https://www.youtube.com/watch?v=tiRUOzJ-FrE Mar 11 04:16:24 case settled? 8) Mar 11 04:20:11 http://toughlet.com/ is onto it, but it lacks IP67ness, GPS and an NMEA port. Mar 11 04:22:20 vast market. Mar 11 05:29:55 woohoo Mar 11 05:30:41 guess who's got a working "OS" (okay, raw unprotected memory image that writes an 'h' to the serial port because I can't handle the FIFO yet to write the 'ello world') Mar 11 05:31:21 sounds basic. Mar 11 05:31:30 you don't say? Mar 11 05:31:51 did. Mar 11 05:32:55 it took the entire day, because I had to wade through 4000 pages of TRM to figure out how the UARTs worked Mar 11 05:34:53 traumatic realistic manual? Mar 11 05:35:18 http://www.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=spruh73c&fileType=pdf this. Mar 11 05:36:13 it's something about an async r/t chip thing... Mar 11 05:38:39 XMPPwocky: did you ever come across the amulet async arm proc? Mar 11 05:41:04 Ubuntu question - does xserver-xorg-video-omap3 not work when VRFB rotation is enabled? My LCD is "tall screen" and I want widescreen, but this driver loads without VRFB and does not load with it Mar 11 05:48:07 etheretic: can't say I have Mar 11 05:50:03 XMPPwocky: it was something furber messed around with. there was silicone, afaik. Mar 11 07:53:02 <_av500_> CalcProgrammer2: vrfb is evil Mar 11 07:53:55 is there a better way to rotate the screen? Mar 11 07:54:33 <_av500_> use the sgx like android :) Mar 11 07:55:39 is this even possible through X11? if so I'm interested to see how it works, I almost have SGX working (modules loaded, just need to figure out how to keep X from undoing my fbset changes and stick at 16 bit) Mar 11 08:00:19 CalcProgrammer2: talk to stskeeps, e.g. on #mer Mar 11 08:38:41 Ok, so I should be close to making SGX work but still giving me trouble, I have the three modules loaded and I installed the gfx_rel_es5.x libraries on the device, /etc/init.d/rc.pvr start worked but still getting errors Mar 11 08:39:21 if I run eglinfo, I get Mar 11 08:39:51 Couldn't load library libGL.so: libGL.so: cannot open shared object file: No such file or directoryUnable to initialise egl Mar 11 08:40:54 <_av500_> CalcProgrammer2: using sgx to rotate x11 means to have the final x11 buffer rendered through sgx Mar 11 08:41:04 <_av500_> not sure the usual xx1+3d covers that Mar 11 08:41:31 still trying to get sgx up at all, I have the screen in normal portrait orientation and rotation off Mar 11 08:42:46 if I run services_test, I get a bunch of OK's then it fails at PVRSRVEnumerateDeviceClass() returns 0 display devices() FAIL - PVRSRV_ERROR_NO_DC_DEVICES_FOUND Mar 11 08:47:48 I purchased 2 DVI-D capes for 2 beaglebones, which the capes will not show up on the new monitor equipped with DVI port and the beaglebone do in fact work. Does anyone know of known issues with the capes? I also used radio shack small mounts on the beagle bones too. Please advise, the new guy. Mar 11 09:49:59 hi there, I've some problem here with running dsp apps on BBxm , I get such error while loading cmemk module: FATAL: Error inserting cmemk (/lib/modules/2.6.32/kernel/drivers/dsp/cmemk.ko): Bad address Mar 11 09:50:35 I googled a lot but didn't figure out what exactly to do ! Mar 11 09:51:08 I'v set bootargs like this: setenv bootargs console=${console} vram=${vram} mapfb.mode=dvi:${dvimode} omapfb.debug=y omapdss.def_disp=${defaultdisplay} root=${mmcroot} rootfstype=${mmcrootfstype} mem=80M@0x80000000 128M@0x88000000 Mar 11 09:51:56 and cmemk options when loading is this: phys_start=0x83700000 phys_end=0x85900000 allowOverlap=1 Mar 11 09:53:04 I do not really understand what is the problem with! I'll happy if anyone give me a help on this, or a good how to which could solve this Mar 11 09:54:15 ( dsplink and lpm_omap3530 modules are already loaded) Mar 11 09:58:55 any idea? Mar 11 10:13:24 * zhn ping! :D Mar 11 10:21:31 <_av500_> zhn: its sunday Mar 11 10:21:58 <_av500_> people dont spend all their time on irc waiting for you Mar 11 10:22:55 _av500_: I didn't except people waiting for me dude, I just was wondered how silent the channel is :D ;) Mar 11 10:24:44 <_av500_> zhn: but do some math Mar 11 10:26:04 <_av500_> 80M < 0x3700000 Mar 11 10:26:13 <_av500_> err, > Mar 11 11:16:00 etheretic, Are you there? Mar 11 12:05:30 panicking_mike: aye. Mar 11 12:06:48 Hi, i'm interesting of the usage of beagleboard. Because I can provide same technologies and help Mar 11 12:06:59 on new design. Right now I'm working on Android Mar 11 12:52:58 the navbox i'm thinking of would run navigatrix. Mar 11 17:25:09 gah i hate daylights savings Mar 11 17:25:53 question regarding the usb ether gadget.. does it show up as a ethx in linux on the gadget ? Mar 11 18:01:53 SGXInit: Mismatch in client-side and KM driver build options; extra options present in the client-side driver: (0x1). Please check sgx_options.h Mar 11 18:02:30 I take it this means I need to build SGX modules with DEBUG=1? Mar 11 18:44:57 Is this any way to receive analog inputs in Beaglebone under ubuntu Mar 11 18:46:26 ? Mar 11 18:50:26 <_av500_> usb Mar 11 18:50:35 <_av500_> oops Mar 11 18:50:47 <_av500_> thought you meant analog video Mar 11 18:51:21 you'd have to convert to digital first wouldn't you? Mar 11 18:52:35 I want to read analog inputs AIN0-AIN6 from header in c program or some script Mar 11 18:54:04 googling "beaglebone ain0" produces http://www.gigamegablog.com/2012/01/22/beaglebone-coding-101-using-the-serial-and-analog-pins/ Mar 11 18:54:31 <_av500_> W1N9Zr0: googling is unfair cheating Mar 11 18:55:16 It doesn't work in Ubuntu. folder tsc is empty in Ubuntu. I googled for 2 hours before asking Mar 11 18:59:36 well that page says why it doesn't work on ubuntu, looks like you have to track down the kernel modification that makes it work in angstrom Mar 11 18:59:41 that or use an i2c adc Mar 11 19:07:37 Can I take angstrom kernel as is and put it to ubuntu? I mean uImage Mar 11 19:16:59 not as far as i know Mar 11 19:17:01 Vic999: you can, but some things will work while others won't Mar 11 19:21:41 thanks. I will try to compile it from git Mar 11 19:39:43 anyone used usb gadget ethernet Mar 11 19:39:45 ? Mar 11 19:42:39 I've used g_ether on the beagleboard... but it's been a while Mar 11 19:43:01 thurbad: does it show up as a separate ethernet device in linux ? Mar 11 19:43:10 (on the actual board) Mar 11 19:43:25 or is it simply a bridge Mar 11 19:43:32 guess it can be both Mar 11 19:45:00 on the actual board I think it appeared as usb0 or eth0 f that's what you're asking Mar 11 19:45:15 thurbad: yea thats what i was wondering Mar 11 19:45:28 is it a piece of sw in addition to the kernel ? Mar 11 19:45:36 been about a year and a half since I've actually used it Mar 11 19:45:47 thurbad: not important. Mar 11 19:45:49 ;) Mar 11 19:45:51 it's requires loading a kernel module Mar 11 19:45:59 and thats it ? Mar 11 19:46:01 or two Mar 11 19:46:47 well, you also have to configure it based on how you connected it Mar 11 19:47:01 hmm Mar 11 19:47:06 what board did you do it on? Mar 11 19:47:11 beagle-xm ? Mar 11 19:47:41 classic beagle c4 Mar 11 19:47:46 k Mar 11 19:47:55 should be the same tho Mar 11 19:48:04 the xm has a dedicated ethernet, so it's not needed as much Mar 11 19:48:41 thurbad: if your "pc" or "tablet" lacks ethernet you get a usb->ethernet converter Mar 11 19:48:51 provided you bridge it Mar 11 19:49:28 yes you can use usb-ethernet, but that doesn't use the g_ether connection Mar 11 19:49:31 also the performance may be better than the one on ehci Mar 11 19:49:34 the g_ether is for otg Mar 11 19:49:57 hmm Mar 11 19:50:12 so if you want to be a usb-ethernet client how would you do that. Mar 11 19:50:21 usb-ethernet requires a driver for the usb device Mar 11 19:50:30 hmm Mar 11 19:50:36 depends on the device you use Mar 11 19:51:30 so gadget != client device Mar 11 19:51:46 or gadget is gadget & client Mar 11 19:51:47 I've used a couple from trendnet successfully Mar 11 19:52:16 I /think/ gadget refers to a mode on the otg usb port Mar 11 19:52:28 not 100% on that Mar 11 19:52:33 thurbad: i thought gadget = client mode Mar 11 19:52:54 at least thats what i am reading of this http://www.linux-usb.org/gadget/ Mar 11 19:53:22 The API makes it easy for peripherals and other devices embedding GNU/Linux system software to act in the USB "device" (slave) role. Mar 11 19:53:36 the fact is that otg is a controller capable of operating in a slave mode. Mar 11 19:53:45 but also as host Mar 11 19:54:27 possible, I've never used it on anything but the beagle's otg port Mar 11 19:54:34 ;) Mar 11 19:54:57 thurbad: thanks anyways now i think i can "try" it ;) Mar 11 19:55:15 just had to check if i needed any special user space stuff to get it going Mar 11 20:06:11 * mranostay waves Mar 11 20:12:10 hmm i wonder if i can loop the ethernet gadget to the ehci on the same board Mar 11 20:18:13 Question: Do U guys own a beaglebone cape dvi-d and if so, are you have any issues getting it to work? Sorry for my intrusion into your discussion. Mar 11 20:22:10 no discussion to intrude ;) Mar 11 20:22:19 sorry i own a bone but it has no cape Mar 11 20:23:18 No problem. Thank U. Mar 11 20:25:31 hmm not to bad 3.6MB/s transfering from my laptop to my beagleboard over usb with g_ether Mar 11 20:25:34 me likes Mar 11 20:25:44 and damn easy to use too Mar 11 20:28:52 even get same performance as the on board xm ethernet.. Mar 11 20:29:04 <-- likes this Mar 11 21:00:20 so what is the general procedure to get opkg update working when all the fetching in opkg update fails Mar 11 21:02:02 Depends why it fails Mar 11 21:02:42 damn things missing default gateway.. Mar 11 21:02:46 <-- blames himself Mar 11 21:14:50 hmm Mar 11 21:14:53 interesting Mar 11 21:15:07 why does usb outperform the ether gadget Mar 11 21:15:17 maybe my cable is shit Mar 11 21:15:26 or my pc device driver is shit Mar 11 21:15:49 Hello Mar 11 21:15:53 i see no reason why the omap ehci should outperform the otg in g_ether mode Mar 11 21:16:09 other than crappy linux drivers or something like that Mar 11 21:16:20 Anyone have working wince7 nap? Mar 11 21:16:24 Bsp Mar 11 21:17:34 The demo image supplyed with the board is completely useless. Mar 11 21:18:12 id say the same for wince7 Mar 11 21:18:24 bbl Mar 11 21:18:26 wife time Mar 11 21:19:17 I have not used Linux b4, to start now would set me back for years :( Mar 11 21:20:12 wince for arm not so free :/ Mar 11 21:20:43 For some reason I got myself mislead about the demo image vs demo bsp Mar 11 21:22:33 If I know that bsp cost extra $500 I would not buy it Mar 11 21:58:27 damm so much ppl Mar 11 21:59:39 we try Mar 11 22:00:50 can anyone help me? i want to know if the communication between the bone and the XM is easy Mar 11 22:01:14 how flexible it is Mar 11 22:01:44 what do you want to communicate? Mar 11 22:02:36 it is a called a network :) Mar 11 22:02:56 process video on the bone but its comming from the XM Mar 11 22:03:17 i dont know anything about this right now, is this possible? Mar 11 22:04:32 a lot is possible, but not nessecarily practical Mar 11 22:04:56 you could use ethernet, but why would you do that when the XM is the higher end board Mar 11 22:05:30 because im trying to share the load between boards Mar 11 22:05:53 heh that doesn't work well with 10/100 lines :) Mar 11 22:06:16 * mranostay vaguely remembers OpenMosix Mar 11 22:07:40 the MX has not enough power to do everything so i want to use the bone to help the MX out with some extra tasks like video processing Mar 11 22:08:05 you should probably keep high bandwidth tasks on the xm Mar 11 22:08:37 move the lower/slower stuff to the bone, or look into a higher end cpu all together, like the dual core pandaboard Mar 11 22:09:11 say that i need to process 4 streams with all main tasks the MX can do 2 and the bone would do the other 2 Mar 11 22:09:42 the pandaboard can be expanded too? Mar 11 22:16:24 shot in the dark here, but would anyone know how to set up GPIO interrupts? Mar 11 22:17:41 http://elinux.org/EBC_Exercise_03_gpio_Polling_and_Interrupts Mar 11 22:20:36 thanks, reading Mar 11 22:28:29 in the link you provided, the sample program that demonstrates gpio interrupts uses "POLLPRI" event. Mar 11 22:29:13 can you elaborate on how this works? Is this form of polling considered an interrupt? Mar 11 22:31:53 or are there any prebuilt images that have gpio interrupts configured already? Mar 11 22:43:15 from the linux man: POLLPRI: "High-priority data may be read without blocking." Mar 11 22:44:14 In the example, I don't see how interrupts are being implemented. The example still uses the poll method on the gpio every "POLL_TIMEOUT" seconds Mar 11 22:44:35 it's not a poll like a spin loop wait poll Mar 11 22:44:47 the os suspends the thread untill there's an interrupt Mar 11 22:44:55 then returns from poll Mar 11 22:45:33 ah ic. is there any kernel configuration required to use this method of interrupts? Mar 11 22:45:46 i don't know, but i'm about to try it on my ubuntu image Mar 11 22:48:29 hmm, I need a pana with 4G of ram Mar 11 22:49:29 So if we need near real-time notification without blocking, would this method still work without taking all the CPU time when setting the POLL_TIMEOUT to a very small time? Mar 11 22:52:25 mranostay: openmosix is painful Mar 11 22:53:13 Crofton: a panda not a pana i guess Mar 11 22:53:33 btw why do you need 4 gigs of ram Mar 11 22:53:35 you put the poll loop in a separate thread and set a long timeout Mar 11 22:53:37 unsolo_: it is dead too :) Mar 11 22:53:55 mranostay: i buildt some openmosix clusters back in the day Mar 11 22:53:58 your main thread will be block free, and your poll thread will only wake up when there's an interrupt Mar 11 22:54:10 but distcc and povray where a lot better at distributing the load than mosix Mar 11 22:54:31 Ah that makes sense. Any luck testing on your image? Mar 11 22:55:13 trying to find which gpio i should test with Mar 11 22:55:30 i was going to try it on the user button, but that's owned by some other driver Mar 11 22:55:34 * unsolo_ wonders why someone is polling interrupts off gpio instead of enabling gpio irq Mar 11 22:55:42 unsolo_, yes Mar 11 22:55:47 thanks for the asm help last night Mar 11 22:55:52 mru got it Mar 11 22:55:58 hehe Mar 11 22:56:07 <-- sucs at asm Mar 11 22:56:14 yeah Mar 11 22:56:24 the trick is the clobbers or whatever Mar 11 22:56:29 why do you need 4 gigs of ram ? Mar 11 22:56:38 tell gcc you write the value it uses in the section you want to time Mar 11 22:56:51 unsolo, if you have a reference for gpio irq implementation, can you link it? we have been looking for documentation but unable to find anything Mar 11 22:57:00 because some guy on the panda list is wondering why they have such little ram Mar 11 22:57:07 Crofton: if i got mru right one should maybe stick asm outside of c Mar 11 22:57:13 that also does the trick Mar 11 22:57:25 yes, but I am trying to time C :) Mar 11 22:57:29 xD Mar 11 22:57:35 so that defeats the purpose Mar 11 22:57:54 adsfd: is this the bone or the beagle _ Mar 11 22:57:56 haha Mar 11 22:58:01 basically, inline gets tricky getting the constraints right Mar 11 22:58:06 the board or the bone is more correct./ Mar 11 22:58:07 beagle - xm Mar 11 22:58:23 adsfd: using gpios in linux on xm is simple Mar 11 22:58:25 intrinsics will give you issues with register scheduling Mar 11 22:58:29 regular asm lets you control all details Mar 11 22:58:35 gpio request gpio set direction Mar 11 22:58:45 and then use the gpio to irq translate Mar 11 22:58:47 request ireq Mar 11 22:58:51 -e Mar 11 22:58:56 and voila Mar 11 22:58:59 can you elaborate on the irq translate? Mar 11 22:59:01 even works for the gpio on the tps Mar 11 22:59:27 adsfd: its a function that translates the gpio pin number to a linux irq number so you can request it Mar 11 23:00:02 what is "watching" for the irq request? Mar 11 23:00:07 read the arch/arm/mach-omap2/bard-omap3beagle.c on how to get the irq Mar 11 23:00:15 We have only worked with interrupts on atmega/similar chips Mar 11 23:00:20 and aren't familiar with linux interrupts Mar 11 23:00:23 Crofton: you could write then whole thing in asm Mar 11 23:00:29 reading now* Mar 11 23:00:32 what you're timing is a function call Mar 11 23:00:43 and a trivial one at that Mar 11 23:00:45 question: I should set up VMSA before I deal with the interrupt vectors, yes? Mar 11 23:00:58 *after Mar 11 23:03:19 since I can at least have some basic debugging printouts Mar 11 23:03:50 adsfd: call gpio_to_irq(pin) Mar 11 23:04:00 to get the returned irq number of that pin Mar 11 23:04:49 adsfd: http://osdir.com/ml/beagleboard/2011-05/msg00729.html Mar 11 23:05:22 adsfd: that should contain everything you need Mar 11 23:05:37 and interrupt handling happens in kernel space Mar 11 23:06:38 so from our userspace program, is there a function used to jump to when an interrupt occurs? Mar 11 23:07:10 adsfd: fastest thing is to either do a blockign ioctl untill it happens or wait using epoll/poll on a char device. Mar 11 23:07:43 unsure if you can map gpio to userspace directly then poll on that Mar 11 23:08:17 but in either case you need a "kernel" module that registers for that interrupt Mar 11 23:08:41 which should also handle it unless its something silly that doesnt require any form of realtime Mar 11 23:08:51 but this is not omap specific Mar 11 23:08:53 its linux Mar 11 23:10:01 there is a way to export gpio directly to userspace and you can in a thread or process scan that but it will be slow Mar 11 23:10:11 you can also do that with memory space Mar 11 23:10:17 but you cannot export irq's Mar 11 23:10:32 so why is it that you said to use irq instead of the non blocking poll Mar 11 23:10:43 adsfd: its faster and more proper Mar 11 23:11:12 We need it in our userspace programs Mar 11 23:11:15 polling a pin every n msec generates a certain load compared to never Mar 11 23:11:34 never meaning hw irq occurs otherwise nothing happens Mar 11 23:12:06 the dm3730 has 6 gpio banks mapping 6x32 irq's into 6 hw irq pins Mar 11 23:12:47 those 6 hw irq pins map into the dm3730's total of 96 hardware interrupts Mar 11 23:13:10 this is where the linux irq handler is called Mar 11 23:13:13 in kernel space Mar 11 23:13:53 it's normal to even use kernel space irq to detect headset insertion. Mar 11 23:14:16 so what is it you are doing and its maybe somewhat easier to tell you what you should do Mar 11 23:17:13 adsfd: if you have to have everything in userspace (which makes me think this is some proprietary stuff) http://groups.google.com/group/beagleboard/browse_thread/thread/27995ebeb765695d Mar 11 23:17:26 we have 3 pir sensors and 1 accelerometer that will be connected to 4 gpio pins. We need to activate different cameras based on what gpio pin is high as fast as possible Mar 11 23:17:28 but you can allways add a small opensource driver in kernelspace Mar 11 23:17:43 adsfd: move to kernelspace.. Mar 11 23:18:05 and have the camera code do an epoll on which camera to run Mar 11 23:18:19 or hack linux camera code and have it swap it in kernel space Mar 11 23:18:27 depends on the speed you desire Mar 11 23:19:39 we are using the uvc driver for our usb web cameras. We are running gstreamer for recording Mar 11 23:19:42 btw wouldnt swapping cameras every ms make you go crazy ? Mar 11 23:20:12 we aren't- we are recording as soon as the gpio goes high Mar 11 23:20:20 i see Mar 11 23:20:39 and all the camera streams are running at the same time ? Mar 11 23:20:49 over the usb i mean Mar 11 23:20:53 up to 3. we are able to do this up to ~5fps Mar 11 23:21:04 (usb bandwidth issue) Mar 11 23:21:15 adsfd: use the otg Mar 11 23:21:20 and get one extra Mar 11 23:21:23 is that on another bus? Mar 11 23:21:30 yup Mar 11 23:21:32 awesome Mar 11 23:21:55 you have been most useful person for our project lol. more than our advisor. thanks Mar 11 23:21:57 you can even attach an extra hub to it provided the hub gets power Mar 11 23:22:10 you need to configure the otg into host mode Mar 11 23:22:16 and make yourselves a custom cable Mar 11 23:22:36 btw Mar 11 23:22:49 if you start the stream only when needed Mar 11 23:22:58 you can get more ..fps Mar 11 23:23:09 but imho there is no realtime requirement here Mar 11 23:23:30 because gstreamer setup and uvc initiation is waay slower than polling irq pins every 1ms Mar 11 23:23:35 i see. Mar 11 23:24:00 so the nonblocking polling method would suffice since the bottleneck is at our other components Mar 11 23:24:12 so you could get from 1ms and down to maybe 0.1 ms or less using a kernel driver but that wouldnt help the 500ms it takes to do the gstreamer + uvc Mar 11 23:24:27 one way to get another camera onto the beagle is to actually use the camera port on it Mar 11 23:24:39 leopard imaging has a few to sel Mar 11 23:24:40 sell. Mar 11 23:25:42 so imho in 640x480@30fps one can easely have 3 running cameras on the beagle. and it would severely help you if those usb cameras did on chip encoding Mar 11 23:26:13 actually with on chip encoding h264 or similar id bet maybe 8 or 9 cameras could be possible Mar 11 23:26:48 while polling, we would get segmentation faults at 1ms Mar 11 23:27:00 4 cams on otg through hub, 4 cams on EHCI through onboard hub.(remeber ethernet also uses that hub) Mar 11 23:27:07 1 camera on the camera interface Mar 11 23:27:26 adsfd: that would be your code beeing crap Mar 11 23:29:24 Taking a look at the code, must be the fopen in our while loop. yep Mar 11 23:33:48 jay6981: http://christwire.org/2012/03/lillyhammer-tv-show-reveals-the-corrupt-alcoholic-culture-of-socialist-norway/ Mar 11 23:34:00 <-- is having a blast reading it Mar 11 23:36:45 the sad thing is that people actualyl believe that shit Mar 11 23:36:54 hehe Mar 11 23:37:04 wow, that guy's not beating around the bush Mar 11 23:38:26 notice that the X mafia boss is mentioned as an American Businessman Mar 11 23:38:36 isn't norway one of the richest countries in the world (per capita)? Mar 11 23:38:58 mru: last time i checked my account id didn't feel that way Mar 11 23:39:15 there is a huge difference between money and value for money Mar 11 23:39:20 yeah, tell me about it Mar 11 23:39:21 a lot of oil subsidies iirc Mar 11 23:39:24 * mru lived in oslo for a while Mar 11 23:39:55 <-- has lived in oslo for 26 years of his 32 Mar 11 23:39:59 christwire? please tell me that is a joke Mar 11 23:40:18 the joke is you can't tell Mar 11 23:40:21 still, I was not poor by any definition Mar 11 23:40:32 mru: indeed Mar 11 23:40:42 jay6981: i was raised baptist so i know people believe this shit :) Mar 11 23:40:53 there are people with beliefs so ridiculous it is impossible to tell them from satire Mar 11 23:41:19 yeah, when you hit INT_MAX it wraps around Mar 11 23:41:36 jay6981: its good that they write satire for us non belivers Mar 11 23:42:08 mru: whenever something happens to my car. i go to the dentist or look into buying a house i feel poor Mar 11 23:42:45 and ofc when i was a student Mar 11 23:43:04 biggest problem is that the state has more than 50% of the workforce Mar 11 23:43:22 is it that bad? Mar 11 23:43:29 mru: its horrible Mar 11 23:43:45 I mean is the figure really that high? Mar 11 23:43:46 past 10 years things have only gotten worse on that end Mar 11 23:43:49 so it IS a socialist nightmare Mar 11 23:43:51 of course that figure is terrible if true Mar 11 23:43:53 mru: i think its higher Mar 11 23:44:16 jay6981: well, there are reasons I didn't stick around Mar 11 23:46:01 if you include hospitals etc i think the number is above 1.5 million Mar 11 23:46:14 and only about 800 000 in private sector Mar 11 23:46:28 which means those 800 000 are turning the wheel around Mar 11 23:47:10 and the rest? unemployed or retired? Mar 11 23:47:14 The US is doing its damnedest to go the same direction. Mar 11 23:47:18 wat Mar 11 23:47:22 to young to old to lazt or some other shit Mar 11 23:47:36 * XMPPwocky rings the libertarian circlejerk alarm Mar 11 23:47:41 ka6sox: yeah but don't have 50 million working for the feds yet :) Mar 11 23:48:05 scandinavia is a bit too socialist for my liking Mar 11 23:48:07 this channel is getting too real can we go back to r-pi bashing? :) Mar 11 23:48:19 mranostay: sure Mar 11 23:48:23 rpi has to be a communist plot! Mar 11 23:48:34 i talked to a guy earlier today who was thinking of ordering a r-pi Mar 11 23:48:59 unsolo_: some of my friends are Mar 11 23:49:02 but the r-pi is lacking cool io so hes probably ordering up a bone.. Mar 11 23:49:05 i'm ashamed of them :) Mar 11 23:49:39 friends don't let friends order rpi Mar 11 23:49:51 here are a few other options Mar 11 23:50:41 http://www.origenboard.org/ http://www.nvidia.com/object/carma-devkit.html http://www.fxitech.com/products/ Mar 11 23:50:46 oh dear Mar 11 23:50:58 origen is a disaster Mar 11 23:51:05 and anything nvidia spells trouble Mar 11 23:51:23 but honestly if you want to hack around with some 3v3 io go buy a beaglebone Mar 11 23:51:40 if you want a nice little fanless pc buy a beagleboard Mar 11 23:51:56 i guess the thing i think suck is the lack of 1080p on the beagle.. Mar 11 23:52:01 panda then Mar 11 23:52:08 mru: A9 sux Mar 11 23:52:11 lies Mar 11 23:52:14 a9 is great Mar 11 23:52:20 a15 is even better of course Mar 11 23:52:31 mru: id rather get a A15 or A7 Mar 11 23:52:39 a7 is a joke Mar 11 23:52:40 or both together Mar 11 23:52:46 well, as a solo core it's a joke Mar 11 23:52:55 well i was thinking 8 cores Mar 11 23:53:01 can't have that Mar 11 23:53:02 whats wrong with A9? Mar 11 23:53:08 mru: you can Mar 11 23:53:13 unsolo_: not easily Mar 11 23:53:15 but not in a single core Mar 11 23:53:22 but on a single die Mar 11 23:53:30 (if someone ever does it) Mar 11 23:53:31 a15 only goes to 4x Mar 11 23:53:45 amba allows for 8 way for both of them Mar 11 23:53:46 you can build a cluster of 4 quads Mar 11 23:54:01 oO Mar 11 23:54:13 i thought the number of ways was tied to the number of cores Mar 11 23:54:27 anyhow A9 sux Mar 11 23:54:30 compared Mar 11 23:54:32 it does not Mar 11 23:54:40 a9 is a fine core Mar 11 23:54:44 sure, it's not perfect Mar 11 23:54:53 it is a fine core and its very not TI available Mar 11 23:54:53 but it does not suck by any stretch of imagination Mar 11 23:55:00 in small numbers Mar 11 23:55:05 different matter Mar 11 23:55:12 omap4460 is a fine chip Mar 11 23:55:19