**** BEGIN LOGGING AT Mon Sep 16 02:59:59 2013 Sep 16 03:04:42 lol Sep 16 03:08:51 anybody try building a PATA adapter for their BBB? Sep 16 03:21:29 Is there an official repo with up to date dts files for beagleboneblack ? Sep 16 03:27:00 kinda a pain to apply the patches, but http://github.com/beagleboard/kernel is the official repo Sep 16 03:28:09 So, is the "official" image the 06-20-2013 or are the later ones "official" or better? Sep 16 03:28:23 i.e. is there a "stable" branch and a "dev" branch Sep 16 03:32:07 N2TOH PATA<->USB adapter works pretty good :) Sep 16 03:32:38 I know Sep 16 03:36:34 jkridner: thanks Sep 16 03:41:44 ptan: newer images are mostly better, but regressions haven't been 100% avoided. Sep 16 03:54:23 jkridner, but one 1bit problem has been solved...finally :) Sep 16 03:54:49 yeah, koen was mentioning that Sep 16 03:58:11 1bit problem? Sep 16 04:01:42 jkridner, it makes a huge difference for our apps. Sep 16 04:04:44 N2TOH, the external uSD was in 1 bit mode Sep 16 04:07:48 is meta-beagleboard preferred over meta-ti for the BeagleBone Black? Sep 16 04:12:45 what has to be done to patch the uSD interface to support class 10 mode? Sep 16 04:18:05 change 1 line? Sep 16 04:19:40 where does the change need to be made? Sep 16 04:21:29 there is a patch for it Sep 16 04:21:33 and its in the repo Sep 16 04:21:41 (latest patch I saw) Sep 16 04:22:10 ok I should look into that after I'm done with the current task Sep 16 04:24:29 instructions on the groups Sep 16 04:24:55 ka6sox my uSD card is actually ( barely ) faster than the emmc now Sep 16 04:25:37 im getting 17.68M/s reads, emmc gets about 17.55M/s Sep 16 04:26:16 what class of SD card? Sep 16 04:26:22 10 Sep 16 04:27:37 about the fastest thing I have found to date is a high end USB thumb drive I am using with my RPi Sep 16 04:28:15 fastest for writes is external USB HDD I have using a PATA <->USB controller Sep 16 04:28:32 tis the only one ive tried so far, and thats because i boot from it sometimes Sep 16 04:29:08 writes wer aroudn 21.6M/s i think Sep 16 04:29:18 this patriot rage 32Gig stick will do around 40MB/s Sep 16 04:29:27 on the BBB ? Sep 16 04:29:32 no RPi Sep 16 04:30:20 I wanted to see if I could improve the disk IO on the Pi board so I installed a boot loader on the SD card and run the OS from the USB Sep 16 04:30:52 i still netboot, its the most convient for me Sep 16 04:31:32 I can actively backup on the fly while the OS is running. whcih since im experimenting with some thigns it helps out a lot Sep 16 04:32:06 plus, i can cross compile and just drop the executable into location the BBB can get to it Sep 16 04:32:17 ( easily ) Sep 16 04:33:54 sooner or later ill probably get my physical debian machine back up to date, and drop LXDE on it so i can run eclipse on it ( maybe ) Sep 16 05:29:29 what is an inverterd XOR called Sep 16 05:29:50 whhhaat? :) Sep 16 05:30:34 you mean xnor? or something else? Sep 16 05:31:06 i need !(a^b) Sep 16 05:31:13 xnor Sep 16 05:31:24 http://en.wikipedia.org/wiki/XNOR_gate Sep 16 05:32:29 dont' think theres a C operator for it Sep 16 05:33:20 oh so you want to do a bitwise xnor on an int or something? Sep 16 05:33:34 yes Sep 16 05:36:44 if ( !((ButtonState & 2) ^ ( ButtonStateLast & 2 ) )) Sep 16 05:36:52 Im doing button deboucing. Sep 16 05:38:04 well if you are just taking one bit from there i.e. a boolean true/false then xnor is the same op as == Sep 16 05:38:55 no. Sep 16 05:39:01 no quite Sep 16 05:39:18 or maybe it is. Sep 16 05:39:20 ha ha Sep 16 05:39:40 .. :) Sep 16 05:39:48 simple wins Sep 16 05:39:49 :-) Sep 16 05:40:54 and if you do it for the whole word, I guess ~(a^b) would work a whole lot better than with ! Sep 16 05:42:53 just make sure everything is unsigned, I don't even know how ~ and ^ work with signed... Sep 16 05:42:56 when in doubt, check the generated code Sep 16 05:46:05 when in doubt, run around in circles and scream. Sep 16 05:49:22 panto: I'll do that if I ever find myself with signed ints and bitwise ops in the same room.. Sep 16 05:49:28 hasn't happened in a while Sep 16 05:51:47 Vaizki: thanks Sep 16 05:51:49 that worked! Sep 16 05:52:03 problem is i actually started with a & Sep 16 05:52:09 thining that is what i wanted Sep 16 05:52:19 but it was'nt Sep 16 06:54:04 /quit Sep 16 07:06:30 http://hothardware.com/News/Wrap-That-Rascal-With-A-USB-Condom/ Sep 16 07:06:36 ooookayyyy :D Sep 16 07:07:27 yeah so the 2.0 version is a cable with a DPDT switch to disconnect the DATA lines, problem solved! Sep 16 07:07:55 I hope you patented that before spilling it here Sep 16 07:08:48 patents are useless, unless you have unlimited funds to pay the trial lawyers Sep 16 08:09:51 hello folks. Sep 16 08:15:49 hi Sep 16 08:16:37 how to set time in beagle bone black angstrom distribution ? Sep 16 08:17:03 askos: http://derekmolloy.ie/automatically-setting-the-beaglebone-black-time-using-ntp/ Sep 16 08:18:12 and how to log in the root directory in prompt mode? Sep 16 08:18:43 ? Sep 16 08:19:34 mean that when my angstrom is loaded. i want to go to prompt mode black screen. how to do that? Sep 16 08:20:11 open a serial console or ssh conenction? Sep 16 08:20:21 yes Sep 16 08:20:36 well so do that. Sep 16 08:21:11 in terminal mode it is not accepting the opkj command Sep 16 08:21:28 i have no idea what you mean. Sep 16 08:21:30 .o0(#linuxbeginners) Sep 16 08:21:39 a wonderfull JIHAD everyone! Sep 16 08:21:47 either you are on a console, or you are not. Sep 16 08:22:02 and "not accepting" is not a valid error message. Sep 16 08:22:43 KotH: my fremens will be coming soon ta venue near you sir. Sep 16 08:22:56 i am on simple desktop of angstrom distribution on bb black. and i want to enter in console mode where i can install the sofwares Sep 16 08:23:21 *sigh* Sep 16 08:23:23 LetoThe2nd: cool! i always wanted to meet them face to face Sep 16 08:24:08 askos: waht are you talking about? do you have a monitor+keyboard conencted to your bbb and are using that? or are you sitting in front of your laptop, with the bbb connected over usb? or ethernet? or... Sep 16 08:24:56 KotH: bring chocolate, then :) Sep 16 08:25:00 i have connected my bb black with monitor and keyboard and mouse and facing the beaglebone window Sep 16 08:26:09 strg-alt-f1 then probably. like on any other sane distribution. Sep 16 08:26:51 what is strg Sep 16 08:26:57 ctrl, sorry. Sep 16 08:27:29 or just open a terminal app, afaik its a pretty standard gnome on there. Sep 16 08:28:31 ok good. is there is important to update the image of anstrom window or not.? Sep 16 08:29:20 bad idea. rather reflash it. Sep 16 08:30:54 ok thanks . i am totally MS window user and newbie to linux system. so thats why installing software in linux is headache for me Sep 16 08:31:14 there's lots of documentation out there. go and read it. Sep 16 08:31:28 askos: and: better practise on your desktop first. Sep 16 08:31:56 ok thank you for your kind help! Sep 16 08:46:19 LetoThe2nd: you should know be by now: i always have chocolate with me :) Sep 16 08:47:48 it helps that the command isn't 'opkj', but 'opkg' Sep 16 08:48:22 koen: why are you still here?!? Sep 16 08:56:11 LetoThe2nd: why shouldnt he be here? Sep 16 08:56:47 because... well, BECAUSE! Sep 16 09:00:54 https://fbcdn-sphotos-d-a.akamaihd.net/hphotos-ak-prn2/1269806_10151872306967661_2084341184_o.jpg Sep 16 09:01:09 theres a chocolate fish for anyone who can see what is wrong with that packet Sep 16 09:01:44 the Microchip TCP/IP stack is annoyingly incomplete Sep 16 09:02:12 KotH: are you aware of chocolate fish culture? Sep 16 09:02:21 the total length looks evil ;) Sep 16 09:02:28 nope, i'm not aware of the chcolate fish culture Sep 16 09:03:09 lol. it is evil Sep 16 09:03:10 :-0) Sep 16 09:03:17 i had not picked up that teh devil was in my code Sep 16 09:03:32 we have in NZ, chocolate fish Sep 16 09:03:41 chocolate in teh shape of fish Sep 16 09:04:06 in the NZ IT world, when oen of your team members found your bug Sep 16 09:04:09 mrpacket_: shouldnt the dst mac addr be ff's when sending multicast? Sep 16 09:04:11 you had to pay them a chocolate fish Sep 16 09:04:20 KotH: you get a box of chocolate fish Sep 16 09:04:28 almost! Sep 16 09:04:44 hmm? Sep 16 09:04:51 you picked up the problem Sep 16 09:04:57 just did'nt quite get it wrong Sep 16 09:04:59 wright Sep 16 09:05:16 well.. i've never seen multicast :) Sep 16 09:07:17 what would it take to play audio out of gpio pins? Could it be done without writing a kernel module? Sep 16 09:10:19 mrpacket_: what else is wrong? Sep 16 09:11:24 isaacbw: no. Sep 16 09:11:51 isaacbw: gpio pins are just not able to output analog signals. Sep 16 09:11:54 LetoThe2nd: there is a possibility withouth writing a kernel module: the PRU Sep 16 09:12:27 KotH: if it can access soma at tleast 22kHz DAC, then yes. Sep 16 09:12:34 plain digital gpio - no. Sep 16 09:12:39 LetoThe2nd: think pwm :) Sep 16 09:12:56 LetoThe2nd: if you toggle the pins with 200MHz, any audio device will act as low pass filter Sep 16 09:13:24 sure, go dykstra. you will not be able to output different aplitudes, though. Sep 16 09:13:40 you will Sep 16 09:13:46 (which is essential for anything close to human audible audio.) Sep 16 09:14:11 1bit DAC work this way Sep 16 09:14:20 so do class d amplifiers Sep 16 09:14:49 hm, Sep 16 09:16:27 LetoThe2nd: how would I go about playing audio from a (white) beaglebone? Sep 16 09:16:56 would I need dedicated audio hardware which I send audio data via gpio? Sep 16 09:17:12 ic2, whatever Sep 16 09:17:22 isaacbw: thats the usual way. Sep 16 09:17:35 isaacbw: and if you use an usb soundcard, probably no driver is needed. Sep 16 09:17:56 a driver is always needed Sep 16 09:18:15 but luckily, for most devicees there are already drivers around Sep 16 09:19:20 KotH: sorry for the long post Sep 16 09:19:21 "Both Ethernet and FDDI frames have a 48 bit destination address field. In order to avoid a kind of multicast ARP to map multicast IP addresses to ethernet/FDDI ones, the IANA reserved a range of addresses for multicast: every ethernet/FDDI frame with its destination in the range 01-00-5e-00-00-00 to 01-00-5e-ff-ff-ff (hex) contains data for a multicast group. The prefix 01-00-5e identifies the frame as multicast Sep 16 09:19:22 , the next bit is always 0 and so only 23 bits are left to the multicast address. As IP multicast groups are 28 bits long, the mapping can not be one-to-one. Only the 23 least significant bits of the IP multicast group are placed in the frame. The remaining 5 high-order bits are ignored, resulting in 32 different multicast groups being mapped to the same ethernet/FDDI address. This means that the ethernet layer a Sep 16 09:19:22 cts as an imperfect filter, and the IP layer will have to decide whether to accept the datagrams the data-link layer passed to it. The IP layer acts as a definitive perfect filter." Sep 16 09:21:18 so ka Sep 16 09:21:27 domo Sep 16 09:21:33 so, as it sits Sep 16 09:21:39 the microchip TCP/IP stack is broken. Sep 16 09:21:47 its is free Sep 16 09:21:53 fix it :) Sep 16 09:21:54 and the uP is $3.50 Sep 16 09:22:32 if you're ever thinking of putting a uC like PIC on the wide internet you're insane Sep 16 09:22:41 theres a distint lack of 'fixes' on the interweb Sep 16 09:22:43 private networks ok Sep 16 09:22:53 put it out there and you're toast Sep 16 09:22:54 panto: whys that? Sep 16 09:23:21 one could argue your more secure than many other O/S's Sep 16 09:23:23 :-0 Sep 16 09:23:33 as you might have 1% of the code to go wrong! Sep 16 09:23:35 because the amount of memory and processing required for a resilient tcp/ip implementation is not in the PIC range Sep 16 09:23:42 it's not about security Sep 16 09:23:51 yes, you can overun them Sep 16 09:23:52 :-) Sep 16 09:23:53 it's about buffering requirements Sep 16 09:24:11 are you saying 128k is'nt enough Sep 16 09:24:17 :-) Sep 16 09:24:19 mrpacket_, that's DOA in the commercial space Sep 16 09:24:33 cause who knows what goes on the network you're putting it Sep 16 09:24:40 it's not the 1980's anymore Sep 16 09:24:42 panto: do you remember the smalles webserver ever? a PIC16C84 serving data over a SLIP interface :) Sep 16 09:24:45 this particualr thing is'nt goign ont eh internet Sep 16 09:24:53 panto: it was on the intarwebz! Sep 16 09:25:02 but they do go on large corpoprate networks Sep 16 09:25:05 mrpacket_, that's what I am talking about 'private network' Sep 16 09:25:08 mrpacket_, bzzzt Sep 16 09:25:12 and *touch* wood Sep 16 09:25:17 we've not had too many problems Sep 16 09:25:22 other than when peopel do dumb things Sep 16 09:25:26 you assume that the security in the corporate network is perfect Sep 16 09:25:39 i assuem nothing about corporate networks ecuriety Sep 16 09:25:39 :-) Sep 16 09:25:40 you only need one infected pc on it, and you're toast Sep 16 09:25:44 * KotH would trust a corporate network less than he trusts the internet Sep 16 09:25:59 Having spent the last 10 years of my IT life as a security / network architect Sep 16 09:26:01 i can tell you Sep 16 09:26:07 some scary scary storys Sep 16 09:26:10 most corporate firewalls keep the worm infected machines _in_ not out Sep 16 09:26:13 I believe you Sep 16 09:26:32 i can tell you also, that some corporates go the other direction Sep 16 09:26:34 that's why you'd better set aside another ethernet segment just for those devices Sep 16 09:26:44 they lock their network down so much, the users can't do any work Sep 16 09:27:02 and have a really, really good router with firewalls and QoS Sep 16 09:27:04 panto: normally thats waht we'll do. Sep 16 09:27:15 we'll run seperate Vlans Sep 16 09:27:19 mrpacket_, true, but that doesn't apply if the box is compromized Sep 16 09:27:27 *compromised Sep 16 09:27:31 mrpacket_, not enough Sep 16 09:27:41 a coupel of environmetns, we actually have implemented seperate switching Sep 16 09:27:49 if the device is not VLAN aware on the h/w level you'll get killed Sep 16 09:28:13 jsut because it was *easier* than tippping a broken lcient network Sep 16 09:28:17 most of the times it's better to eat the cost and put something with a real tcp/ip stack in there Sep 16 09:28:25 panto: disagree Sep 16 09:28:28 when you have to put in 1000's Sep 16 09:28:29 * panto shrugs Sep 16 09:28:36 your funeral, not mine Sep 16 09:28:43 panto: my bank balance, not yours Sep 16 09:28:49 true again Sep 16 09:28:55 we agree to disagree Sep 16 09:29:07 got to run for a bit Sep 16 09:29:07 its why i have 4 super yachts, 6 jets and 14 palaces Sep 16 09:29:22 one of those statemetns was a lie Sep 16 09:29:25 can you tell which one. Sep 16 09:29:35 the bank balance one? Sep 16 09:29:39 lol. Sep 16 09:29:47 that was a game for KotH Sep 16 09:30:00 l8r Sep 16 09:30:10 other thing. I dont' do misison critical Sep 16 09:30:14 or health Sep 16 09:30:27 or stuff neeing saftey Sep 16 09:30:33 aka, pryo! Sep 16 09:30:55 you want to see the performance that goes into doing pryo on an IP network Sep 16 09:30:56 :-) Sep 16 09:41:45 panto: I wish it was possible to bet online on such projects :) One could get rich quickly... Sep 16 09:42:11 Would the BBB run a N64 Emulator? Sep 16 10:02:06 ynezz: you want "betting-cape" Sep 16 10:20:26 can anyone help me with some question about the I2C capabilities of beaglebone black? :) Sep 16 10:20:35 ask Sep 16 10:20:38 dont ask to ask Sep 16 10:21:25 hi i want to connect 18bpp TFT lcd to Begalebone black Sep 16 10:21:47 then do it Sep 16 10:22:03 which mode i have to use and how can i set timing parameters like HSYNC and VSYNC Sep 16 10:22:15 I'm back Sep 16 10:22:53 I want to control 24 servos via I2C, Can I figure I2C to send a message to all slaves at a time? Sep 16 10:27:22 just to verify, the lipo charger circuit on the bbb is *only* for single cell operation? Sep 16 10:27:36 panto: your the on duty troll tongith Sep 16 10:27:45 we had a vote while you were gone. Sep 16 10:28:11 I demand to see the vote tallies Sep 16 10:28:20 panto: 19 Sep 16 10:28:21 that was it Sep 16 10:28:44 geepers how can we answer the if they dissapear Sep 16 10:30:11 this si interesting Sep 16 10:30:21 theres another way of opening a udp socket Sep 16 10:34:33 C question Sep 16 10:35:14 the learning curve this week has been steep Sep 16 10:35:15 :-) Sep 16 10:41:14 mrpacket_: i can tell you, doing mission critical or medical stuff with ip is.. PAINFULL Sep 16 10:41:42 oh i knwo Sep 16 10:42:54 Hello, can I integrate the bonescript functions in a c progamm? Sep 16 10:43:09 I searched it by google without good results. Sep 16 10:43:48 I want to give parameters to the bonescript functions and get the return values. Sep 16 10:44:55 Are there c or c++ functions like System or asm for javascript / bonescript? Sep 16 10:45:18 And if I'm in the wrong forum please tell me :) Sep 16 11:18:22 mrpacket_: actually anything mission critical that involves software is painfull ^^' Sep 16 11:35:39 koen: I tried tweaking the touchscreen in the kernel but without success. Now I found some recent driver patches regarding the ADC that seem to have been applied: https://lkml.org/lkml/2013/7/20/104 I wonder if it could be possible that some of them have introduced the touchscreen bug or something? Sep 16 11:50:00 Anguel: has it ever worked? Sep 16 11:54:19 SpeedEvil: I don't know. IMHO a touchscreen should not be a big problem for a properly implemented ADC. Other people maybe ignore the problem. I don't know. Noone replied to my postings in the beagleboard group :( Sep 16 11:54:49 hi friends.. i bought a new beaglebone black.. when i connect it for the first time to my windows 7 pc, my usb disk security deleted auturun.inf and start.htm files.. how can i recover it ? please help me... Sep 16 11:56:59 Madhuborn2win: Which anti-virus are you using? Turn it off, go to beagleboard.org and follow the instruction to reflash the image Sep 16 11:58:16 SpeedEvil_: Those guys are happily selling their LCDs although they seem to know that the touchscreen simply does not work. Sep 16 11:59:10 Do they actually say the touchscreen works? Sep 16 12:00:09 thank u so much Anguel :) :)... Sep 16 12:00:59 Madhuborn2win: what anti-virus by the way? Sep 16 12:01:29 Speedevil: "Follow the instructions given by the Auto Calibration utility to calibrate the LCD4 touchscreen." http://beagleboardtoys.info/index.php?title=BeagleBone_LCD4 Sep 16 12:02:03 Investigate what version of various stuff was current at that time. Install that version. See if it works. Sep 16 12:02:17 If that works - then get bisectual on its ass. Sep 16 12:04:36 Speedevil: I think I already tested that but will now retest :) Sep 16 12:06:02 Also - have you contacted them abou tthis? Sep 16 12:06:23 Speedevil: It probably depends on how they understand "works". If my cursor sometimes jumps 100 px away if I touch a point on the screen I say it does NOT work Sep 16 12:06:33 ah Sep 16 12:06:41 Yes, 4D Systems told me they know about the problem Sep 16 12:06:44 Considered a hardware fail? Sep 16 12:07:25 and it was certainly some bug in the kernel. then I contacted them again and Circuitco for more info who actually tests the drivers. Sep 16 12:07:30 then i got no reply Sep 16 12:08:39 i have the circuitco lcd4 and the 4d systems lcd, both have the same issue. it could be my BBB in theory but as i said, they confirmed the error Sep 16 12:09:11 I also saw a youtube video about lcd7 and the same issue can be seen there Sep 16 12:11:01 afaik circuitco just uses the lcd with touchscreen from newhaven which are pretty good. it must be some issue on their boards or actually a software problem or the adc is not capable to go an accurate measurement because of some bouncing issue Sep 16 12:24:39 * georgem logs a strange phenomenon at Monday 07:24am. Happy, relaxed to be at work. Sep 16 12:26:29 15:26 at work: happy to go home soon Sep 16 12:35:57 <_vicash_> Hi. I updated a debug version of u-boot.img with the -DDEBUG option turned on in the u-boot code v2013.04 with all the patches needed as per the Angstrom build. However it doesn't boot anymore and gives this screen error. I am trying to debug using UART but nothing happens. Please help. http://pastebin.com/8kxTPEJN Sep 16 12:36:46 <_vicash_> I had installed this on the eMMC part of the BBB so it seems like the system is now just not trying to boot from any other source. Sep 16 12:40:35 <_vicash_> anyone ? Sep 16 12:42:02 this is not hotline, if you look at topic it says "be patient" Sep 16 12:42:10 english patient Sep 16 12:42:57 <_vicash_> ynezz: i asked the question on Friday as well. Sep 16 12:43:11 <_vicash_> ynezz: i have been patient for 2 days waiting for someone to respond Sep 16 12:43:53 then maybe nobody knows the answer? Sep 16 12:43:53 well, some questions might never be answered... *jedihandmove* Sep 16 12:44:17 _vicash_: which uboot is that? Sep 16 12:44:20 mainline? Sep 16 12:44:57 <_vicash_> av500: u-boot v2013.04 the default Angstrom build u-boot Sep 16 12:45:10 <_vicash_> for the beaglebone Sep 16 12:45:18 from where? Sep 16 12:45:33 <_vicash_> this is the version that gets built when you build the Angstrom from scratch using setup-scripts Sep 16 12:45:49 ah Sep 16 12:45:58 and if you build from scratch, does this uboot work? Sep 16 12:46:02 before you made changes Sep 16 12:46:44 <_vicash_> Well I had placed the non-debug version on the SD card and booted many times from it for a month. But I cannot say if the u-boot used was the default one from the eMMC or the one from the SD card Sep 16 12:46:54 <_vicash_> the kernel and OS were definitely from the SD card Sep 16 12:52:58 <_vicash_> av500: is there a way to force a boot without the eMMC ? Sep 16 12:55:21 sure Sep 16 12:55:24 hold the boot button Sep 16 12:55:28 when powering up Sep 16 12:55:33 or kill MLO on the emmc Sep 16 12:56:28 <_vicash_> av500: i do hold the boot button for about 5 minutes but nothing happens. it had worked a month ago hwen i had updated it. on the UART screen I just see the character 'C' being printed every second or so.. Sep 16 12:56:50 <_vicash_> av500: how do i access the eMMC when the system doesn;t boot itself. the debug u-boot.img is on the eMMC !! Sep 16 12:59:02 ? Sep 16 12:59:08 C means no MLO on sd card Sep 16 12:59:37 or card not good Sep 16 12:59:53 <_vicash_> oh cool.. let me retry with an SD card with a default image then :) Sep 16 13:00:27 <_vicash_> av500: thanks for the tip regarding 'C' . that doesnt seem to be documented anywhere Sep 16 13:01:17 the TRM might mention it Sep 16 13:03:53 <_vicash_> you mean SRM ? Sep 16 13:04:18 TRM Sep 16 13:04:25 the cpu manual Sep 16 13:04:37 the SRM is a bit to thin for that Sep 16 13:04:56 <_vicash_> ok cool. i will download the CPU manual Sep 16 13:23:36 Hey folks! I'm trying to write down a report describing some of the problems i found throught my GSOC with BeagleBoard (my project is Beagle-ROS). One of the things that gave quite a lot of trouble was the I2C bus i was wondering if some of you experienced something similar Sep 16 13:24:38 vmayoral: what trouble? Sep 16 13:24:45 while working with I2C2 (P9_19, P9_20) and https://www.sparkfun.com/products/11084 i had a lot of trouble detecting the sensor Sep 16 13:24:59 vmayoral: why did you have trouble? Sep 16 13:25:24 good night zz_ka6sox :) Sep 16 13:25:43 the thing is that the sensor was detected just in some cases (and just if i used really short cables) Sep 16 13:25:57 it was really really sensitive to that Sep 16 13:26:07 i gave it a try with other sensors as well Sep 16 13:26:25 https://www.sparkfun.com/products/11282 and https://www.sparkfun.com/products/11295 Sep 16 13:26:28 same behaviour Sep 16 13:26:38 in both the BB and the BBB Sep 16 13:27:24 i also tried it with I2C1 (P9_17, P9_18), same again Sep 16 13:27:32 i read as much as i could about it Sep 16 13:27:50 tried it with external pull up resistors also Sep 16 13:28:11 well, depends on the frequency, the termination, and the cable length of course. Sep 16 13:28:49 if you limit it to 100kHz or such, it should be relatively stabel :) Sep 16 13:29:27 and pullups are a must anyways. 10k or such. Sep 16 13:31:06 i was using 10k pullups Sep 16 13:31:16 and what frquency? Sep 16 13:31:24 and the I2C freq was 100 KHz Sep 16 13:31:45 i thought it could be the sparkfun breakout modules though Sep 16 13:31:59 i have an evalulation board from invensense Sep 16 13:32:03 have you checked the init sequence with a scope/logic analyzer then? Sep 16 13:32:03 and it works fine with that Sep 16 13:33:01 no, i did not do that Sep 16 13:33:39 would be the next step then. if a) the bone does not send the correct commands b) the sparkfun thing garbles it or c) the sensor just does not answer. Sep 16 13:34:02 sounds good, i'll try that Sep 16 13:34:41 as i was saying i was a bit troubled because the Invensense eval module that i ordered worked fine. I also ordered a MPU9150 breakout board from Sparkfun (same chip) Sep 16 13:34:46 and i'll give a try as well Sep 16 13:35:16 if the same sensor doesn't work then Sparkfun might have something to do with it Sep 16 13:35:33 <_vicash_> av500: Thank you for your help. That worked. I used the default image and booted from teh SD card and it booted. Need to figure out how to fix the u-boot.img debug mode now. Sep 16 13:37:06 LetoThe2nd: thanks for your suggestion, i'll give it a try Sep 16 13:37:26 vmayoral: good luck Sep 16 14:53:20 HI all , I want to buy beagle device. But can any one tell me does it has browser which supports javascript and html5 and runs video without delay Sep 16 14:53:38 i mean smothly Sep 16 14:54:05 <_vicash_> Has anyone had success setting up bluetooth on BBB with another device such as a smart phone ? Sep 16 15:18:02 hi everyone Sep 16 15:18:23 i have problem with beagleboard c4 Sep 16 15:18:59 problem in booting process for matlab functionality Sep 16 15:20:50 every time process come to 7 of 9 Sep 16 15:47:52 turns out that the touchscreen works perfectly in image 2013.06.20, so there must have been some bug introduced in the later updates :( Sep 16 15:57:48 Hello. Has anyone had any experience with TheDeck distro on their BBB? Sep 16 15:58:38 I've purchased the micro-SD preloaded with The Deck, but I'm uncertain how to actually install it Sep 16 15:59:53 as in the distro is already installed? Sep 16 16:00:34 to boot from the uSD card you need to push the button near the uSD socket while applying power Sep 16 16:00:34 I believe so. I purchased it fromhere:https://specialcomp.com/beagleboard/thedeck.htm Sep 16 16:01:16 The very first one, item #42125 Sep 16 16:03:31 I'll give that a shot @N2TOH Sep 16 16:04:52 * N2TOH is still trying to figure out how to get the BBB to boot from the internal flash after pushing said button Sep 16 16:32:46 <_vicash_> Hi. What is the recommended way of building the u-boot.img file in Debug mode with -DDEBUG turned on ? Should I just patch the Angstrom build scripts in setup-scripts/sources/meta-beagleboard or is there another way ? Sep 16 16:34:29 N2T0H you still here? Sep 16 16:34:56 yeah Sep 16 16:35:24 im not sure where that button you referred to is Sep 16 16:39:11 wait I think I found it. Do I just hold the button down as I plug in power? Sep 16 16:39:52 yeah that's all it takes Sep 16 16:40:05 trying it now Sep 16 16:40:43 intern_ why is the headless version of thedeck so much more then the other version? Sep 16 16:41:19 what do you mean? so much more what? Sep 16 16:42:10 [42125] is $32 Sep 16 16:42:26 [42124] only $16 Sep 16 16:43:15 I'm not sure honestly. One is for the BBB the other is for the BBxM, dont know what the difference is Sep 16 16:46:40 btw booting from the SD, if it did, I dont think worked Sep 16 16:47:19 whenever I open up the START html file, I get a "problem loading page" Sep 16 16:47:52 it takes a few minutes the first time Sep 16 16:48:13 are you running the BBB with a monitor, or via the USB interface? Sep 16 16:48:31 via USB Sep 16 16:48:46 oh, I haven't tried that yet Sep 16 16:50:15 http://youtu.be/98kbOKuInv4 Sep 16 16:51:18 Ive watched that video a few times lol, whenever he searches for the tar.gz file for TheDeck it comes up with an unknown directory for me Sep 16 16:52:40 I'm downloading it now, I might take a look at it later, after the BBB finishes installing ubuntu desktop on the other uSD card Sep 16 17:01:53 this is tricky because I'm not sure if I've booted to the SD card or not. They both seem to have the same files Sep 16 17:18:34 N2TOH: Okay so I've booted from the preloaded SD, and I see a disk image file on it that isnt on the BBB's on-board memory. I'm assuming thats what I need to boot up TheDeck Sep 16 17:22:20 intern_, um from what I understand if you push had hold for about 3 seconds, the BBB should be able to boot and run the OS from the uSD card Sep 16 17:24:03 Yeah, it only boots up like normal, just boots the information thats on the SD card. Assuming we are talking about the same SD button on the BBB Sep 16 17:24:28 The BB Getting Started folder pops up like normal, just with a few more files and folders Sep 16 17:24:51 yeah the single button near the sd card slot, the reset and other button are on the other side on the board Sep 16 17:25:14 Gotcha. Yeah, nothing special happens when I boot from that button Sep 16 17:25:40 does it boot the OS from the uSD card? Sep 16 17:26:17 I don't think so. The Windows auto-play pops up,and it opens the BB Getting Started folder, from which I would normally open the START html file Sep 16 17:26:59 hmm, I will have to try the USB cable after I am done installing stuff on the BBB Sep 16 17:27:15 it takes about an hour to install the GUI Sep 16 17:27:40 yeah, the place where i work ordered the parts last week, and we werent sent the micro-HDMI like we ordered Sep 16 17:28:09 first start up was this morning, tethered via the USB. I'm starting to think the HDMI is the way to go Sep 16 17:28:46 if nothing else it can rule out one option Sep 16 17:28:59 how long are you holding the button down? Sep 16 17:29:26 true. Now you use the HDMI correct? Once you boot up the BB, does the on-board linux distro start automatically? like a normal computer would? Sep 16 17:30:53 yes I use the uHDMI, it took a few tries to get it to boot from the card. but now it always seems to default to the uSD card. Sep 16 17:31:22 now all I have to do it supply power and it boots automatically Sep 16 17:33:20 Seems a lot easier than trying to boot via USB. I need to get my hands on that HDMI cable Sep 16 17:40:50 LetoThe2nd: i did tried whwat you suggested. Sep 16 17:41:04