**** BEGIN LOGGING AT Tue Sep 17 02:59:58 2013 Sep 17 03:00:49 this confirms it... ubuntu is done by a bunch of total imbicles Sep 17 03:04:25 film at 11 Sep 17 03:04:36 was there ever any doubt ? Sep 17 03:04:56 what is it this time? Sep 17 03:15:00 the dinkweedle broke on the dinkwebber Sep 17 03:15:35 what kind of moron disables the middle button by default Sep 17 03:26:00 ubuntu rocks Sep 17 03:26:07 ubuntu is a great OS Sep 17 03:26:07 :p Sep 17 03:30:26 Is there any low level programming information for the beaglbone and the L3G4200D accelerometer/temprature sensor ? It'si2c Sep 17 03:30:32 i2c Sep 17 03:34:14 it just works Sep 17 03:34:21 the bone runs stock Linux Sep 17 03:34:26 nothing special about it Sep 17 03:34:41 took all of 20minutes (compile/install/start) to get it running and visible in Android Sep 17 03:54:27 hey all -- apart from maybe the MAC address, is there an easily accessible GUID on each BBB? Sep 17 04:46:51 hey all -- apart from maybe the MAC address, is there an easily accessible GUID on each BBB? Sep 17 04:48:17 pkh: no certain, but I think there may be a unique ID burned into the system ROM area that's accessible through sysfs. Sep 17 04:49:55 ah, ok. will see if google can come up with something with that as a keyword Sep 17 04:53:21 hmmm, nothing obvious -- guess will use mac for now... Sep 17 05:10:44 OS: Windows 8 Server Datacenter (full installation) (Version 6.2 Build 9200), CPU: 12 x Intel(R) Xeon(R) CPU E5-2630 0 @ 2.30GHz @ 2294 MHz 1536 KB Cache, MEM: 73693 MB, 86% (63296 MB) free, DISKS: total 19.0 Tb, 17.2 Tb free, GFX: NVIDIA GeForce 9800 GTX/9800 GTX+ 512 MB, SCREEN: Generic PnP Monitor, 2560 x 1440 @ 32 bit, AUDIO: Sound Blaster Z, UPTIME: 0 d, 3 h, 47 m Sep 17 05:47:41 hrmm, my kernel module doesn't build with 9 Sep 17 05:47:43 -5 image. Sep 17 05:47:49 "disagrees about version of symbol device_create" Sep 17 05:47:58 at least the headers are the right version this time :D Sep 17 05:58:54 err, never mind. looks like the kernel-dev and kernel-headers packages are garbage again. broken for the third image in a row. hooray. Sep 17 05:59:39 silly kids... packages blah Sep 17 05:59:57 opkg is for nerds, just like manuals. Sep 17 06:00:15 real men don't use distributions, the compile everything from source. Sep 17 06:00:31 and me without my boots Sep 17 06:01:23 it ain't deep enough to need boots Sep 17 06:02:03 ok, jumpin ship to debian. angstrom is too bleeding for me. Sep 17 06:02:30 ewwwwwwwww debian Sep 17 06:02:35 i'll wait for features to stop dropping off and give it another go sometime in the future. Sep 17 06:02:42 you can use OE w/o Angstrom Sep 17 06:05:02 as long as whatever has *any* level of automated testing to see if stuff actually works, I'm for it. Sep 17 06:05:23 this "let the users and devs stumble into the problems" is f'n nuts man. **** BEGIN LOGGING AT Tue Sep 17 06:12:43 2013 Sep 17 07:10:27 a wonderfully JIHADy good morning everyone Sep 17 07:14:34 angstrom too bleeding? no, not really Sep 17 07:14:38 it's quite outdated Sep 17 07:14:58 it's just quite broken, like slackware in 90s Sep 17 07:22:19 AStorm: good morning Sep 17 07:22:25 AStorm: a wonderfully, gurmpy day to you too! Sep 17 07:27:48 LOL Sep 17 07:57:56 hello am enthusiastic with electronic gift beagle bone black with couple of projects can i read various parameters hummidity,temperature of the soil Sep 17 07:58:42 sounds fun Sep 17 07:59:14 sounds dirty fun. Sep 17 07:59:36 yes keesj Sep 17 08:00:15 what tools are you using nodejs android something else? Sep 17 08:00:26 * KotH uses plain old c Sep 17 08:00:41 nothing of that newfangled stuff called node.js ;) Sep 17 08:03:06 but then you're not asynchronous-free-cool-in ! Sep 17 08:03:32 *sigh* Sep 17 08:03:39 1) i know how to write async code in c Sep 17 08:03:53 2) node.js is not fully async Sep 17 08:04:14 3) any c code i write is much faster than anything you can write with node.js, even if i dont do async i/o Sep 17 08:04:53 4) why the heck did the "node.js is cancer" blog post get removed, i would like to point people there Sep 17 08:06:54 https://github.com/RobertCNelson/netinstall/commits/master/hwpack/am335x-boneblack.conf I .. don't understand. I just threw netinstall on a microSD card, booted it on my BBB, and ... emmc doesn't show up in the installer, because it's been disabled. Has anyone else used this netinstall tool? Is it supposed to just install over the microSD card used to boot itself? Sep 17 08:07:24 I like c for kernel code but find all that userland stuff interacting in c a major pain Sep 17 08:07:26 KotH: http://pages.citebite.com/b2x0j8q1megb Sep 17 08:07:52 I was expecting to use the microSD to boot the netinstaller, install debian to the eMMC, then reformat the microSD to use as extra storage. Am I totally goofing this up? :P Sep 17 08:07:53 *bookmark* Sep 17 08:07:55 av500: thanks Sep 17 08:08:11 what you need is a call that maps code and kicks off execution in kernel space Sep 17 08:08:19 keesj: yes, writing c can be tedious Sep 17 08:08:32 keesj: for the quick and dirty stuff i use perl Sep 17 08:08:48 bah... quick and dirty is for /bin/sh :P Sep 17 08:08:54 keesj: which is not only faster to write than js but also faster in execution time Sep 17 08:09:47 keesj: there's Lua for that :) Sep 17 08:10:02 ah yes, lua works as well Sep 17 08:10:10 also faster than node.js stuff :) Sep 17 08:10:36 actually, anything is faster to write and execute than node.js, even this abomination called python ;) Sep 17 08:10:59 speed is relative Sep 17 08:10:59 KotH: hey, leave my family out of that! Sep 17 08:11:36 LetoThe2nd: python, not sandworms Sep 17 08:11:46 KotH: "abominations" Sep 17 08:11:49 lua might be the right tool. but it doesn't tell you how to make different components talk to eachother either.(I had to deal with dbus in my previous jobs...) Sep 17 08:12:28 KotH: thank Larry and Sergey Sep 17 08:12:32 anyway. people writing c code are less noisy (they blame themself when it doesn't work) Sep 17 08:13:11 you mean sockets? Sep 17 08:13:13 LetoThe2nd: as a descendant of a bene geserit and the heir of atreides, you have a weird way to refer to your family Sep 17 08:13:55 sockets is only part of the story you also need "naming" and serialisation. I think modern systems like nodejs are just great at that Sep 17 08:13:56 KotH: i'm used to be called "weird" Sep 17 08:14:04 keesj: using dbus for IPC is... i dont have words for that Sep 17 08:14:29 serialization is built in Lua Sep 17 08:14:46 keesj: just to make sure you understand what you are doing: dbus was NEVER intended to be a general IPC system. NEVER EVER. that why it's so horribly slow, and doesnt really work as expected Sep 17 08:14:59 keesj: actually, from it's design it's a poor mans version of the X11 atom system Sep 17 08:15:16 keesj: without the network transparency and the authentication Sep 17 08:15:56 at my previous employer we also combined that with upstart (I am partly to blame and something at night I wake up and scream ) Sep 17 08:16:57 KotH, I'm glad i'm not the only one that thinks using dbus for ipc is a gigantic wtf. Sep 17 08:18:37 c just lacks some better standarisation in that area (communication,threading,gui) . threre are tons's of choices to be made by anyone trying to make something remotly usefull :p. but that is that makes it fun Sep 17 08:18:44 nomel: dbus was designed to used as a notification system of hardware events into x11 apps (mostly destkop environments). and that's IMHO the only thing you should use it, if at all Sep 17 08:19:21 nomel: beside, the ICCCM (the document that standardizes the high level part of X11 communication proto) say "Thou shalt not have any IPC system beside me" Sep 17 08:19:37 seems like that should be moved into the kernel, no? Sep 17 08:20:22 I looks at a set of alternatives (zeromq, binders, .....) Sep 17 08:20:23 definitly! Sep 17 08:20:29 i looked Sep 17 08:20:44 I do like the binders approach Sep 17 08:20:54 keesj: zeromq is quite nice imho. unles you want to do point-to-point communication only. then it's a bit weird to set up :) Sep 17 08:20:54 binders full of busses? Sep 17 08:21:06 binders full of chocolate! Sep 17 08:22:08 no binder is quite lean , no buses (only p2p) the rest is built on top of that and it offers nice features and guaranties Sep 17 08:22:32 Hi! It looks like some touchscreen / adc bugs have been introduced through the latest 3.8 kernel patches that make the LCD touchscreens unusable. I see that Zubair Lutfullah has submitted the patches. Can you please tell me if I should contact him directly or what is the procedure in such case? Sep 17 08:22:49 fix it, send patch Sep 17 08:23:23 (but not usable outside of a single host) Sep 17 08:23:59 I am afraid that I don't have the experience to fix it Sep 17 08:24:09 but you can bisect it right? Sep 17 08:24:16 there's directory called Documentation which contains documentation and there's some file called 'submitting patches' Sep 17 08:26:24 basically you don't write directly to author, but to the corresponding mailing list Sep 17 08:26:48 each part has such mailing list/maintainer, so check MAINTAINERS file Sep 17 08:26:51 so kasii_ happy with your code? Sep 17 08:26:59 a ok, so i have to find the list and write there about the problems i discovered? Sep 17 08:27:04 or scripts/get_maintainer.pl path/to/the/borken/source.c Sep 17 08:27:25 doesn't looks like systemd is smart enough to handle restarting a service with sucessfully started dependencies Sep 17 08:27:54 Anguel: basically you want to find a preferred way of submitting bugs, each project has it's own rules Sep 17 08:28:16 Anguel: so read something about it Sep 17 08:30:08 ynezz: I am not 100% sure where the bug comes from, just see that in Angstrom 2013.06.20 everything works ok, and in later images it does not. I also see that some patches have been released in the meantime. I will have to look where to post that feedback. Sep 17 08:31:13 but Zubair is our gsoc student and should be here right? Sep 17 08:36:08 this is probably their mailing list? https://lkml.org/lkml/2013/7/17/450 Sep 17 08:37:14 Anguel: lkml and linux-arm-kernel Sep 17 08:37:32 Anguel: mail should go on both, with linux-arm-kernel being the more important Sep 17 08:37:52 Anguel: and CC the maintainer of the code Sep 17 08:38:05 Anguel: for more info see the patch submission guidelines in Documentation/ Sep 17 08:38:35 KotH: thank you, can I submit bug reports in the same way? Sep 17 08:39:56 Anguel: sure Sep 17 08:40:05 ok, i will have a look Sep 17 08:40:09 thanks Sep 17 08:41:51 I doubt, that someone would care about Angstrom kernel there Sep 17 08:41:56 :) Sep 17 08:44:30 nobody sane is going to waste his time with some custom kernel, so do not expect much help Sep 17 08:45:10 basically you want to compile your own kernel from vanilla sources Sep 17 08:46:02 once you have done that, it's easy to bisect the right broken commit which introduced this issue for you Sep 17 08:46:18 google for 'git bisect kernel' Sep 17 08:47:41 I understand, that it might be quite hard with BBB (you would need to patch your kernel anyway, otherwise it won't boot) Sep 17 08:48:26 but you don't want to spam lkml with 'it was working with Angstrom.2013.x' but it's broken 'in 2013.9' Sep 17 08:48:58 you need to isolate the problem to bare minimum, possibly with just vanilla kernel tree Sep 17 08:49:37 then your chances for some meaningful answer and probably help/fix are high :) Sep 17 08:50:26 but as keesj already suggested, that guy might be idling here so it could be a nice shortcut to slap him with some trout Sep 17 08:50:39 :) Sep 17 08:50:52 ynezz: I will try to post zubair and if he wants he can look, i mean those guys are working with the code, I don't have experience and resources to start from scratch Sep 17 08:52:48 anyone using a wifi adapeter with their beaglebone black Sep 17 08:53:33 Anguel: well, it migh also be the situation, that he has fixed the ADC and the broken is the touchscreen driver now? Sep 17 08:54:47 Anguel: as it was based on previously broken code from the ADC (or maybe the touchscreen calibration)? So now the ADC is fixed but touchscreen broken... Sep 17 08:55:34 it's possible, will look what code they have changed Sep 17 08:55:43 anyone using a wifi usb dongle Sep 17 08:56:08 trivia: insmod exits before the module has loadede Sep 17 08:56:43 rmedia_: there are lots of people. check the beagleboard forum. Sep 17 08:57:06 do you need the power supply for it to work because mine isnt detected Sep 17 08:58:20 how do you know it's not being detected? Sep 17 08:58:25 does dmesg show anything? Sep 17 08:58:35 when you plug it in? Sep 17 08:58:36 ifconfig doesnt show wlan0 Sep 17 08:58:57 ok, so it's not being used as a network device Sep 17 08:59:04 but does the os know the device exists? Sep 17 08:59:13 how can i check? Sep 17 08:59:25 did you install drivers? Sep 17 08:59:37 according to the info online, it should work with my model out of the box Sep 17 09:00:04 check dmesg after you plug it in. Sep 17 09:00:34 and, this info is for your os and your beaglebone? Sep 17 09:00:50 err...your distribution Sep 17 09:01:00 that im not sure Sep 17 09:01:09 i just plugged it in and ssh into it Sep 17 09:01:11 lemme make sure Sep 17 09:02:22 ok its supposed to work on angstrom Sep 17 09:02:32 what distro does bbb come shipped with? Sep 17 09:02:41 angstrom Sep 17 09:03:07 i just found this... Sep 17 09:03:09 WIFI Adapters NOTE: It is highly recommend that you use DC power when running these dongles due to the current requirements of the dongles. Sep 17 09:03:29 http://circuitco.com/support/index.php?title=WiFi#WiFi_With_rtl8192cu_.28Edimax.29 Sep 17 09:03:49 did you "opkg install wireless-tools" Sep 17 09:04:18 and install the firmware? Sep 17 09:04:25 no im oging to do that now Sep 17 09:08:38 Thanks nomel i hope this works Sep 17 09:09:00 i also need to make a backup of the os after i upgrade and install the drivers Sep 17 09:09:23 do you have a link for making an image of my existing BBB? Sep 17 09:29:47 sorry, my internet provider cut the connection :( Sep 17 09:50:58 Hi all. So, this image wouldn't work on the white? - http://www.armhf.com/index.php/boards/beaglebone-black/ Sep 17 09:51:05 What Ubuntu image can I use for the white? Sep 17 09:51:29 narcos: bbw and bbb are different hardware Sep 17 09:51:46 narcos: using the image of one on the other might let the magic smoke out Sep 17 10:00:37 still looking if anyone knows -- does each beaglebone black have a GUID? barring any other option, going to use MAC Sep 17 10:01:34 afaik it doesnt Sep 17 10:01:41 CPU might have an ID Sep 17 10:01:44 have a loot at the TRM Sep 17 10:01:56 OMAPs have, dunno about AM335x Sep 17 10:03:06 ...Device Part Number (Unique JTAG ID)... Sep 17 10:03:51 cheers, will google around with that Sep 17 10:16:36 pkh: read the TRM Sep 17 10:16:44 pkh: 9.3.15 device_id Register (offset = 600h) [reset = 0x] Sep 17 10:17:07 was hacking my way though it -- thanks for the pointer Sep 17 10:17:30 hmm Sep 17 10:17:30 not sure Sep 17 10:17:37 says part number Sep 17 10:17:42 and 16bit only Sep 17 10:17:54 I guess its to identify the part Sep 17 10:21:24 i've been chewing on the concept and am leaning toward MAC -- that way if I end up migrating to another platform I can guarantee it'll have to have network, therefore MAC, any hardware-based GUID is likely to be BBB-secific Sep 17 10:22:23 well, its all HW specific Sep 17 10:22:30 nothing garantees a MAC Sep 17 10:22:34 BB XM has no MAC Sep 17 10:22:48 its generated from CPU ID Sep 17 10:22:50 or from /dev/random Sep 17 10:23:19 how random is /dev/random on the xm? Sep 17 10:23:37 about this random |<----------------------->| Sep 17 10:27:07 I have now sent an e-mail to Zubair regarding the touchscreen bug, let's see if he can give some hints where the problem comes from. Sep 17 10:55:20 Hi Sep 17 10:58:17 Is it possible to connect two external hdd to the beaglebone black without an usb hub? Sep 17 11:01:36 mcl024: if you have to ask, then "no" :) Sep 17 11:03:48 Ok!? With an external powered usb hub? Sep 17 11:04:37 KotH: i got my multicast to work! Sep 17 11:04:45 the "powered" or not entirely depends on the harddisks. Sep 17 11:11:05 mrpackethead: \o/ Sep 17 11:12:34 Anguel: carefull with the mac: on embedded systems the mac address is often generated at boot up from some random source Sep 17 11:12:49 Anguel: it's neither guaranteed to be stable nor to be unique Sep 17 11:13:04 KotH: Why me? :) Sep 17 11:13:10 er.. Sep 17 11:13:12 Anguel: sorry Sep 17 11:13:21 KotH: no problem Sep 17 11:13:32 post-lunch-brain-not-working-syndrome Sep 17 11:13:43 KotH: was just a matter of reading more doco Sep 17 11:13:52 and trying something Sep 17 11:13:52 pkh: carefull with the mac: on embedded systems the mac address is often generated at boot up from some random source Sep 17 11:13:58 pkh: it's neither guaranteed to be stable nor to be unique Sep 17 11:14:04 bugger Sep 17 11:14:10 mrpackethead: it's always about reading more docu :) Sep 17 11:14:43 now i need to uderstand UUID's and how to create them Sep 17 11:14:47 pkh: if you want a stable and unique mac, you have to check the hardware whether it provides one and if it doesn, use one of those serial eeprom devices that come with a mac Sep 17 11:14:57 was quite proud of myself Sep 17 11:14:58 pkh: maxim sell's them for cheap Sep 17 11:15:18 i went from not knowing how to program sockets Sep 17 11:15:24 to doing something useful in almost no time. Sep 17 11:15:31 mrpackethead: i hope you have a copy of stevens? Sep 17 11:16:28 being the CTO is not conducive to also programming Sep 17 11:16:32 interesting, might sidestep the sisue completel -- nice Sep 17 11:16:38 i need to be isoalted from teh world to program Sep 17 11:16:44 the concentration required is pretty had out Sep 17 11:17:31 thx Sep 17 11:18:01 mrpackethead: being a CTO means usually that you should not touch and device at all. much less work on one Sep 17 11:19:26 pkh: http://www.maximintegrated.com/datasheet/index.mvp/id/3748 Sep 17 11:19:40 pkh: maxim has more of that kind, even complete authentification solutions Sep 17 11:19:55 thats the one I was reading through. Sep 17 11:20:09 pkh: look at the one wire solutions too Sep 17 11:20:13 mrpackethead: it's the same here Sep 17 11:20:15 does look ideal, and 1-wire so simple to plumb in Sep 17 11:20:30 mrpackethead: i usually work with my headphone on, so i dont hear people talking around me Sep 17 11:20:58 KotH: I need to know enough about lots of things so i can make good judgement calls Sep 17 11:21:23 Problem is i have a team of very clever but not highly expereicend engineers Sep 17 11:21:36 keeping a lid on it, is hard sometimes Sep 17 11:21:41 mrpackethead: yeah.. the view from 10'000m ;) Sep 17 11:21:49 mrpackethead: ugh... Sep 17 11:21:51 lol Sep 17 11:22:09 mrpackethead: same here.. too many people of my age, who have not seen enough of the world out there :-( Sep 17 11:22:49 mrpackethead: actually, the worst part of it is, that there are only 2 people here i can actually learn from :-(( Sep 17 11:23:11 i hink you can learn from everyone Sep 17 11:23:16 if you try hard enough Sep 17 11:23:42 if you have the time for that Sep 17 11:23:59 i have hardly the time to learn the stuff i need for the work i'm doing Sep 17 11:24:19 why do you think i'm going to lectures again? Sep 17 11:37:37 mrpackethead: do you know why a hatched ground plane results in less deformation of a PCB than a closed ground plane? Sep 17 11:39:19 hm, wild guess I'd say is the spacing in the copper relieves stress Sep 17 11:39:53 the coper planes are usualy in a middle layer, there is nowhere for the copper to go to relieve its stress Sep 17 11:40:00 if you have 30% copper in one area, and 100% in another, you have different thicknesses Sep 17 11:40:09 hatching reduces that Sep 17 11:40:22 ahh.. Sep 17 11:40:26 that might be it Sep 17 11:40:41 but why dont these french morons write that in their document? Sep 17 11:40:42 if you manage your copper fill, it isn't really a problem Sep 17 11:41:00 I haven't seen hatched copper in a long time except on flex Sep 17 11:41:12 same here Sep 17 11:41:29 wait... not true. OGD1 has a hatched plane Sep 17 11:41:42 pkh: DS2401 Sep 17 11:41:47 probably a cargo cult one Sep 17 11:42:15 +1 Sep 17 11:42:20 it looks 1337 Sep 17 11:42:50 Russ: space seems to be full of paper producing industries that have been worshiping aliens for generations Sep 17 11:43:31 i even have different minimu distances for via on the top/bottom vs mid layers Sep 17 11:43:37 how that should ever work out, i dont know Sep 17 11:43:55 its very common to have tighter tolerances on inner layers Sep 17 11:44:03 oh.. and dont bring blind vias up, they are not allowed Sep 17 11:44:04 usually because they have thinner copper Sep 17 11:44:38 KotH: I guess they are afraid of them blowing up Sep 17 11:44:38 yes, but i cannot make the spacing on a mid layer smaller than on the top and bottom layer for a via. they are multilayer by definition Sep 17 11:44:42 goign all jihad on you Sep 17 11:44:54 * KotH will soon go jihad on the french Sep 17 11:45:14 you wouldn't want to pay for blind/buried vias anyway Sep 17 11:45:24 we do that often Sep 17 11:45:42 av500: looks like it would solve the issue we've been discussing, but the other (DS2502) seems to also allow storage of a few bytes of infromation as well... might be ueful (for the extra 80c.) Sep 17 11:45:52 ah, every time I considered it, my thought was "screw that" Sep 17 11:46:01 pkh: sure Sep 17 11:46:03 but as i said, on this space project we are not allowed to use them Sep 17 11:46:19 Russ: if you can get around it, it's better to not use them. but sometime you cannot Sep 17 11:46:40 Russ: and as we have specialized on small and space restricted designs, we quite often cannot get around them Sep 17 11:47:05 well yes, certain components force your hand too Sep 17 11:47:37 luckily I've never needed them Sep 17 11:48:14 Russ: like the last miniature project we've done: msp430 including power supply from 24v, rs485 driver and 8MByte serial flash on a 17x7.5mm print, that had to fit into an 8mm hole :) Sep 17 11:48:35 KotH: not asking what hole that is Sep 17 11:48:40 totally not asking Sep 17 11:48:44 av500: a true hole! Sep 17 11:48:45 true holes. Sep 17 11:48:59 Russ: :) Sep 17 11:52:41 av500: did you know, that in freshmen psychology you learn about all the kind of measurment devices used... even those that fit into holes ? ;) Sep 17 11:53:19 of course Sep 17 11:53:24 gotta learn your tools Sep 17 11:59:44 helllo Sep 17 11:59:48 mrpackethead: https://plus.google.com/u/0/photos/112266164281670850856/albums/5924558535711349313 Sep 17 12:00:17 while building kernel for BBB....how can we install filesystem???? Sep 17 12:00:37 uImage is ready Sep 17 12:00:52 ? Sep 17 12:00:54 but cannot find filesystem Sep 17 12:00:58 av500: what is the cellphoe for? Sep 17 12:01:05 music Sep 17 12:01:27 av500: dont you have a record player? Sep 17 12:01:34 at home Sep 17 12:01:44 though i'll get my parents' one soon Sep 17 12:02:13 can you pls helpp me Sep 17 12:02:25 for what? Sep 17 12:02:29 what filesystem? Sep 17 12:02:53 bin,opt,usr,var... Sep 17 12:03:03 sigh Sep 17 12:03:22 your BBB came with a filesystem Sep 17 12:03:25 why not use it? Sep 17 12:04:01 actually i am rebuilding the kernel....so as a first step i have builded the image successfully... Sep 17 12:04:19 but now the filesystem.. Sep 17 12:04:19 good Sep 17 12:04:28 you want to rebuild the file system? Sep 17 12:04:35 yes Sep 17 12:04:50 http://www.angstrom-distribution.org/building-angstrom Sep 17 12:05:06 http://www.angstrom-distribution.org/ Sep 17 12:05:11 read and follow steps Sep 17 12:06:14 thanks...av500...lemme check Sep 17 13:07:21 Has anyone here booted a 3.10.x kernel on beaglebone? If so, could you please point me to the patch sets. Sep 17 13:09:44 https://github.com/beagleboard/kernel/ Sep 17 13:09:49 georgem: there is .11 Sep 17 13:09:50 and .9 Sep 17 13:10:05 so if you take the average of both..... Sep 17 13:10:32 av500: haha, I'm in the process of doing that Sep 17 13:10:42 greetings Sep 17 13:12:01 I was going to try pulling patches from both to see which applied and try to sort out the conflicts Sep 17 13:12:48 Figures koen or whoever would skip the long term support kernel, lol Sep 17 13:14:22 georgem: feel free to step up and take that over Sep 17 13:14:59 chasing one kernel is enugh Sep 17 13:15:02 +o Sep 17 13:17:32 yeah. I may do that. I'm sure others could benefit from a long term support kernel, especially if I also maintain a branch with RT-PREEMPT. Sep 17 13:20:59 I built 3.11, now to see if it actually boots. That would be a good start. Sep 17 13:32:15 What's the real use case for https://github.com/ungureanuvladvictor/BBBabb? Sep 17 13:32:29 Is it talking to ROM boot? Sep 17 13:37:19 Hi Sep 17 13:37:45 I'm trying to connect my bluetooth device via usb Sep 17 13:37:55 it shows in hci0 Sep 17 13:38:31 I'm getting "Error: unrecognized/unsupported machine ID (r1 = 0x00000e05)." after booting my kernel on Beaglebone Black. see http://pastebin.com/pXXdd7g1 for more details Sep 17 13:38:37 but I want to learn which virtual terminal correspinding to USB like dev/ttyACM* Sep 17 13:38:56 cfo215: you are missing the dtb file Sep 17 13:40:58 I'm using the Angstrom distribution. It seem to put all that stuff in the rootfs. it doesn't create a uEnv.txt file Sep 17 13:41:49 nor a dtb file Sep 17 13:43:50 Hi, I'm trying to connect Bluetooth dongle its connected to bbb USB port and its on hci0 but I want to learn its virtual serial port like /dev/tty* (Note: ls /dev/tty* controlled before and after connecting but didnt change) Sep 17 13:44:13 vvu|Log, none of my images have a dtb file. the emmc-flasher, cloud9 etc. Sep 17 13:44:35 vvu|Log, maybe cloud9 did. don't have one handy. Sep 17 13:45:38 the dtb is appended to the kernel Sep 17 13:45:59 vvu|Log, ah, thank you. Sep 17 13:46:44 Angstrom has the translation to Portuguese of Brazil? Sep 17 13:48:45 No one can answer my question? Sep 17 13:55:56 Looks like panto has a 3.10 beagle branch on github, lets try this out. Sep 17 13:58:18 Angstrom has the translation to Portuguese of Brazil? Sep 17 13:59:23 GeorgeHenrique: care to explain what angstrom *itself* should have in translations? either the pacakges have it, or nto. Sep 17 14:00:09 LetoThe2nd I want to translate the interface, the menu, the windows Sep 17 14:00:36 GeorgeHenrique: that has nothing to do with angstrom. Sep 17 14:00:58 GeorgeHenrique: that is gnome, or whatever desktop you are using. Sep 17 14:01:23 yeah Sep 17 14:01:25 I want to translate my BEAGLEBONE with linux already has it. Sep 17 14:02:15 GeorgeHenrique, beaglehueso? Sep 17 14:02:37 GeorgeHenrique: go read https://wiki.gnome.org/TranslationProject then. that has nothing to do with angstrom or the beaglebone Sep 17 14:03:56 GeorgeHenrique: after that, read http://www.student.montefiore.ulg.ac.be/~merciadri/angstrom/files/angstrom-manual.pdf Sep 17 14:04:00 GeorgeHenrique: pg24ff. Sep 17 14:14:07 how do i set my ubuntu (pc) to recognize BBB (using ubuntu on usd) via usb? Sep 17 14:15:20 beep-blip: have the BBB load whatever g_* module you want - storage, serial, etc. Sep 17 14:15:55 Hi, is it possible to build qt5 without OpenGLES so that I don't need all this SGX-bla-foo? Sep 17 14:17:07 letothe2nd: i am not sure i understand you...bbb not shown in lsusb Sep 17 14:17:47 beep-blip: well if you do not tell your ubuntu to provide something over usb, it will not show up. Sep 17 14:18:01 beep-blip: your google buzzwords are "linux usb gadget" Sep 17 14:19:50 letothe2nd: i had no problem with angstrom. hmm... Sep 17 14:20:31 beep-blip: yes, because the angstrom images are configured such that they provide gadget serial and gadget storage by default - which ubuntu probably does not. Sep 17 14:22:33 letothe2nd: i had similar problems with wifi usb device before and gave up--using lan cable now ;p Sep 17 14:24:01 beep-blip: totally unrelated. Sep 17 14:25:17 vvu|log, do you know what the correct id should be? Sep 17 14:26:44 pfff dunno Sep 17 14:26:57 you need some custom distro ? Sep 17 14:27:17 i can help you flash the eMMC with my tool *need some testers* Sep 17 14:29:02 letothe2nd: would udhcpc work? i am not sure yet...still searching Sep 17 14:30:29 beep-blip: what "would udhcpc work"? i have no clue what you mean, or waht you want it to do. Sep 17 14:48:44 any idea where i can find BeagleBoard pngs images with the icons ? Sep 17 14:59:08 Anyone out there using a distro other than Angstrom on a BB Black with LCD7 Cape and Qt4? Console based Qt. Not X11/Qt Sep 17 15:03:41 Hi, I'm trying to connect Bluetooth dongle its connected to bbb USB port and its on hci0 but I want to learn its virtual serial port like /dev/tty* (Note: ls /dev/tty* controlled before and after connecting but didnt change). How can I learn or configure? Sep 17 15:16:55 Hi, I'm trying to connect Bluetooth dongle its connected to bbb USB port and its on hci0 but I want to learn its virtual serial port like /dev/tty* (Note: ls /dev/tty* controlled before and after connecting but didnt change). How can I learn or configure? Sep 17 15:30:15 acs, good question I may be way off but you might be able to config that to always be the same tty device. Sep 17 15:30:38 I would start by looking at the USB hardware identifiers Sep 17 15:31:41 I am new at bbb and linux I search evrywhere but nothing found in 1 week Sep 17 15:33:14 I've been there and done that, I am by no means highly experienced. yet I do recall small bits here and there of how things work Sep 17 15:33:31 Also in ubuntu 13.04 cant find /var/lib/connman/settings file Sep 17 15:33:53 http://www.michaelhleonard.com/enable-bluetooth-on-beaglebone-black/ Sep 17 15:33:57 I reverted back to 12.04 for other reasons Sep 17 15:34:32 i mean angstrom and ubuntu relases are different Sep 17 15:35:07 I know I have 12.04 LTS installed and installed LXDE last night Sep 17 15:36:00 N2TOH please keep in touch mail: cagdaseckin@gmail.com Sep 17 15:36:39 is there any link about that? Sep 17 15:46:39 acs, no sorry no direct link. finding anything on the internet is about hitting the correct search therms that are close enough to yield the answers you need Sep 17 15:48:17 N2TOH: correct search therms? Sep 17 15:51:42 acs, your looking for a way to coralate your USB serial adapter to it's location in the ttys locations? Sep 17 15:53:46 N2TOH thanks. Sep 17 15:54:51 have you tried "how to find tty address for USB adapter" Sep 17 15:55:13 and then maybe append the name of the OS after that Sep 17 15:55:41 man Sep 17 15:55:57 someone just got murdered at one of the indian buffets i sometimes go to for lunch Sep 17 15:56:07 who the hell stabs someone to death at an indian buffet Sep 17 15:56:24 to clarify "correct results" are the one/s that are useful to you. Sep 17 15:56:30 i'm normally so happy when I am eating indian food. It never occurred to me to stab someone. Sep 17 15:57:02 where to buy in europe, cheapest ttl 3.3.V (ftdi) usb cable, and + 5V DC (2A ?) adaptor ?... Sep 17 15:58:19 what are ur favorite electronical shop Sep 17 16:00:27 RypRap: you mean http://en.wikipedia.org/wiki/Ur? Sep 17 16:00:39 no idea if they have electronics shops, sorry. Sep 17 16:09:54 this is not yet in europe Sep 17 16:11:00 Hello, any suggestion for rechangeable bettery for beaglebone, I need to charge it with solar panel. Thanks. Sep 17 16:12:19 you may ask your solar cells seller Sep 17 16:13:12 Thanks. how pepple use temprature sensor, with i2c ? Sep 17 16:13:41 and 5 volt bettery should work ?? Sep 17 16:42:30 I reimaged my sd card with Ubuntu Linux, I connected the beaglebone black to the host computer, it says to log in with an Id and password. But I never set one when I installed it, Nothing told me to set up a password or id. Sep 17 16:42:35 SilicaGel: Maybe someone ate all the Naan? Sep 17 17:51:48 Hello i have a problem with the i2c-0 port. How can i activte this network at pin 17 and 18? Sep 17 17:54:35 anyone here ever use emmc-prepare.sh from meta-beagleboard/contrib/bone-flash-tool? Sep 17 17:58:46 what's the general consensus for using terms on the mailing list that some may read as name calling but is an accurate description per the dictionary? Sep 17 18:10:10 ds2, please don't refer to linus as "floozy" Sep 17 18:13:34 well... something with this 3.10 patchset is borked. Anyone seen this before "omap_hsmmc 48060000.mmc: unable to obtain RX DMA engine channel 3". Looks like its using the dma settings for mmc2 instead of mmc1. Sep 17 18:15:38 Russ: not that list :P Sep 17 18:15:48 the beaglelist specifically Sep 17 18:57:59 <_vicash_> Hi. I am trying to write a couple of bytes to the EEPROM on the baseboard of a BBB and it doesn't work. i am using "dd if=/home/root/eeprom of=/sys/bus/i2c/devices/0-0050/eeprom bs=1 count=256" to do so. Sep 17 18:59:03 <_av500_> "it does not work" sounds so harsh Sep 17 18:59:03 <_vicash_> I have taken the original EEPROM using dd as well and then changed a few bytes at position 0x50 in the EEPROM file and then am trying to rewrite that EEPROM back to the device. Sep 17 18:59:42 <_vicash_> however, the dd succeeds but when I read it back it doesn't show the updated EEPROM but gives me the earlier version Sep 17 19:10:16 eh can you really write eeproms with dd like that? :O Sep 17 19:10:34 sorry I'm a bit of a n00b on things embedded linux .. :) Sep 17 19:11:19 i used cat to dump my whole image to my eeproms .. just fine Sep 17 19:11:51 reading I understand, that's easy to do in the driver and completely unintrusive.. but writing.. Sep 17 19:18:29 i was using it to write .. just fine Sep 17 19:19:02 hello all Sep 17 19:19:49 i have some questions about the BBB. 1. How do i restore it to factory default? Everytime i try to do an opkg upgrade and install, it fails Sep 17 19:20:16 secondly, why dont many of the Unix commands work Sep 17 19:21:02 dhclient doesnt work Sep 17 19:21:10 sudo restart doesnt work Sep 17 19:23:11 rmedia It usually doesn't have sudo installed by default. You typically login as root with a new device. So you should be able to do just 'restart' Sep 17 19:24:32 rmedia for the latest images see: http://beagleboard.org/latest-images Sep 17 19:25:18 rmedfia (these are for flashing your SD card or eMMC. Typically, with the BBB, the entire image resides on the eMMC and no SD card is used.) Sep 17 19:28:40 o.O Sep 17 19:29:36 awesome Sep 17 19:29:42 <_vicash_> rmedia: try reboot instead of restart Sep 17 19:30:25 perfect Sep 17 19:30:51 lastly, how do access the /etc/network/interfaces? Sep 17 19:31:21 thats where i normally setup the ethernet connections however it doesnt work here Sep 17 19:31:23 _vicash_, what's the difference between reboot and restart on the BBB? Sep 17 19:32:14 I have been using reboot and it keeps booting from the uSD card, I would like to know how to get it to boot from the onboard eMMC with the uSD card still installed. Sep 17 19:36:43 <_vicash_> N2TOH: i don't know. to boot from eMMC instead of uSD with the uSD still installed you may have to remove the /boot/uImage file from there... not sure... do you plan to keep the Linux on the uSD ? Sep 17 19:37:43 <_vicash_> rmedia: there is a /etc/network/interfaces in my version of Angstrom. you should have it on there too. you do not need dhclient.. it automatically connects to ethernet and uses DHCP without you having to do anything manually. just insert an ethernet cable into the port and try running ifconfig Sep 17 19:40:20 _vicash_, yes as it's a useful tool to be able to use the factory default OS to config other/new installations without extra hardware. and the safety net of not having to worry about bricking the board Sep 17 19:40:36 when i type nano /etc/network/interfaces theres nothing there Sep 17 19:40:54 The network settings shoudl be there Sep 17 19:41:24 <_vicash_> rob_w_: which version of Angstrom were you using then ? it doesnt work on the latest update version. is there a flag that needs to be set as the EEPROM is write protected. I have tried writing from u-boot and from Angstrom Sep 17 19:41:46 i am rolling my own... Sep 17 19:42:31 for reference it's possible to reflash the eMMC with new code, but you must boot from the uSD card first before you can do so. Sep 17 19:42:34 <_vicash_> rob_w_: i am doing that too and the EEPROM writing succeeds but when I read it back it doesnt show the updates Sep 17 19:43:09 is it a own baseboard ? Sep 17 19:43:17 or the a known cape ? Sep 17 19:43:54 <_vicash_> N2TOH: i dont know the answer to your question. u can flash the eMMC from the uSD but if your uSD is installed u-boot will pick that. u can go to the u-boot command prompt and change your boot device Sep 17 19:44:07 <_vicash_> rob_w_: baseboard. not using capes. Sep 17 19:44:29 N2TOH, you can put a uboot config on the sd which rolls the onboard mmc boot Sep 17 19:45:21 did you try cat yet ? Sep 17 19:45:35 not that it makes a difference but for the stats Sep 17 19:47:14 <_vicash_> rob_w_: yes i tried cat and dd ... Sep 17 19:47:34 ah thanks, the part that was confusing me was the need to push the button near the uSD card to get it to boot the first time. then after that I was like what the hell? Sep 17 19:47:43 has anyone installed Beaglenmt? I want a simple beagle server Sep 17 19:48:11 <_vicash_> _vicash_: here is what i did. using dd i read the EEPROM and save to file. then i edit 2 bytes on the EEPROM and using dd/cat write it back to the i2c-eeprom file on sysfs.. then i read it back again but no changes have happened. no failures happen either. Sep 17 19:48:30 I suppose a script could be installed on both the eMMC, and uSD media to ask every time during boot up what media to use. Sep 17 19:48:41 <_vicash_> rob_w_: here is what i did. using dd i read the EEPROM and save to file. then i edit 2 bytes on the EEPROM and using dd/cat write it back to the i2c-eeprom file on sysfs.. then i read it back again but no changes have happened. no failures happen either. Sep 17 19:48:42 like LILO and the other PC boot loaders Sep 17 19:48:51 <_vicash_> N2TOH: it is called u-boot Sep 17 19:49:01 <_vicash_> N2TOH: and it is already there.. Sep 17 19:49:14 ah, I will have to go RTFM on that Sep 17 19:49:29 _vicash_, ur sure that your file edit worked ? Sep 17 19:50:01 <_vicash_> rob_w_: yes the file i edited can be hexdumped and u can see the changes and even diff with the original EEPROM dump. Sep 17 19:50:10 well actually i never messed with the bbb eeeprom .. only with our capes additional ones Sep 17 19:50:26 <_vicash_> rob_w_: well there you go.. i am not talking about capes.. Sep 17 19:50:35 sorry Sep 17 19:51:05 did the readout of the eeprom bring a valid data ? mean the real data you would exspect Sep 17 19:51:10 <_vicash_> rob_w_: no problem. Sep 17 19:51:47 then i dont know sorry .. Sep 17 19:51:51 <_vicash_> rob_w_: yes. that is valid.. as per the BBB_SRM.pdf the EEPROM is write protected on the base board.. that's the problem.. i need to find how to un-protect it Sep 17 19:51:54 in such cases i pull out my scopes Sep 17 19:52:06 ah Sep 17 19:52:20 i was not sure about write protection so i didnt suggested it Sep 17 20:02:41 <_vicash_> The BBB_SRM manual section 5.3.2 says the EEPROM has a test point to allow the EEPROM to be programmed. Does anyone know where this test point is on the board ? Thanks. Sep 17 20:04:28 if you pull up the datasheet for the eeprom you can probably figure out which pin is the write lock, reference the BBB schematic and you should be able to find it Sep 17 20:05:06 <_vicash_> georgem: thanks. Sep 17 20:39:40 do I need special incantations to add a route on beagleboard? Sep 17 20:40:00 this doesn't work: ip route add 10.9.8.0/24 via 131.159.74.53 dev eth0 Sep 17 20:40:06 all I get is RTNETLINK answers: No such process Sep 17 20:40:49 does eth0 exist? Sep 17 20:41:24 yup: http://paste.ubuntu.com/6121112/ Sep 17 20:42:10 is there a route to 131.159.74.53? Sep 17 20:42:28 has anyone installed Beaglenmt (bbb server)? Sep 17 20:42:42 mru: http://paste.ubuntu.com/6121119/ Sep 17 20:42:50 (yes) Sep 17 20:43:07 <_vicash_> Has anyone use the Test Points TP1-8 on the BBB ? Sep 17 20:43:25 honestly: and what does that route look like? Sep 17 20:43:42 "ip route get 131.159.74.53" Sep 17 20:44:28 _vicash_: ground TP4 to remove write protect Sep 17 20:44:28 http://paste.ubuntu.com/6121128/ Sep 17 20:44:33 from EEPROM Sep 17 20:44:59 honestly: so it's already an indirect route Sep 17 20:45:06 mru: yeah... Sep 17 20:45:18 so how would that work? Sep 17 20:45:24 heh. Sep 17 20:45:35 oh my. Sep 17 20:46:04 whelp, that is problematic. Sep 17 20:47:26 let's hope I can solve it by just adding that host into the vpn. Sep 17 20:47:56 what does this have to do with beagle* ? Sep 17 20:48:02 nothing Sep 17 20:48:08 just checking! Sep 17 20:48:12 yeah, nothing, it turns out (: Sep 17 20:48:12 like most questions here Sep 17 20:48:28 the networking stack on ubuntu beagle boards works just fine Sep 17 20:48:29 well _vicash_ had a reaaaaal beagle-question Sep 17 20:48:35 well, it doesn't, but that's not the problem here (: Sep 17 20:48:42 err... Sep 17 20:49:49 _vicash_: oh and if you fry your beagle, it's not my fault :) Sep 17 20:49:59 forgot the disclaimer Sep 17 20:51:58 <_vicash_> Vaizki: you are right. that's what the SRM says.. i have to ground TP4. just wanted to know if someone had done it and any tips or cautionary tales ? Sep 17 20:53:06 no and no from me Sep 17 20:53:52 but you can see on the BBB assembly video that they plug into TP4 during the test bench phase where I suspect they also write the eeprom from uboot via i2c write commands Sep 17 20:54:43 <_vicash_> Vaizki: do you have a link to that video ? Sep 17 20:55:06 let me find it Sep 17 20:55:20 <_vicash_> Vaizki: i am not an electronics engineer so i appreciate any help. Sep 17 20:55:26 http://www.youtube.com/watch?v=FcqQvH41OR4 Sep 17 20:55:32 I am not an EE either Sep 17 20:55:45 3:29 onwards Sep 17 20:56:03 <_vicash_> Vaizki: Awesome. Thank you. Sep 17 20:56:04 plugs in all the ports + TP4 Sep 17 20:56:42 but I just peeked in the SRM to find out it's TP4 really Sep 17 20:57:11 and you can see there that it's normally pulled high via a 10k resistor so to change the WP pin state you ground TP4 Sep 17 20:57:22 if you want to double check, read the data sheet for the eeprom Sep 17 20:57:42 <_vicash_> Vaizki: yes i looked at the schematic and the SRM, have the data sheet and it also says the same.. Sep 17 20:59:04 yea DS is clear on that, low on WP pin = not protected Sep 17 21:02:08 hmmh that eeprom is a 32Kb one.. I thought the BBB only uses like 2Kb of it? Sep 17 21:03:22 <_vicash_> Vaizki: the video doesnt show using the TP4 .. it is more generic testing using expensive equipment. Sep 17 21:03:58 _vicash_: you see that one wire sticking out straight from the back of the BBB? that's in TP4 through hole Sep 17 21:04:19 <_vicash_> Vaizki: that 32K is for the whole I2C EEPROM.. only 256 bytes are used for the baseboard. the rest are for the other capes Sep 17 21:04:39 what? Sep 17 21:04:42 <_vicash_> Vaizki: ah ! yes that's the black wire.. right. Sep 17 21:04:46 each cape has its own eeprom Sep 17 21:05:29 <_vicash_> well as far as i can tell the eeprom has 256 bytes. after which the 256 bytes repeat if u access it from Linux Sep 17 21:05:53 <_vicash_> so maybe it just burns the same 256 bytes multiple times Sep 17 21:06:05 EEPROM-RAID! Sep 17 21:08:00 <_vicash_> Vaizki: if i can write to the EEPROM then I can see if i can write to a position > 256 bytes to see if it works. Sep 17 21:08:48 I think you have to rewrite the whole eeprom at once if you do Sep 17 21:09:58 <_vicash_> Vaizki: yes that's what i mean but with data in the > 256 byte position. anyway. i will try it out tomorrow as i dont want to end the day with a fried BBB Sep 17 21:11:29 well seems like it has a 32byte page or so Sep 17 21:12:47 but apparently you can write single bytes no problem, it will just internally erase and rewrite the 32byte page Sep 17 21:13:30 this shit is just too easy with modern chips Sep 17 21:13:54 <_vicash_> yea Sep 17 21:14:20 btw why are you modifying the eeprom? Sep 17 21:14:27 changing board id? Sep 17 21:18:06 <_vicash_> no. just trying to see if i can store an RSA certificate in the EEPROM or a SHA sum if there is no space for an RSA cert. this can then be used later in special applications for certain security applications Sep 17 21:18:57 <_vicash_> i am using u-boot from the mainline git repo and they have RSA implemented although only for the Chromebook. i might try to make it work for the BBB and if i succeed in doing all this then great.. i might send a patch forward. Sep 17 21:19:23 so that u-boot will only boot signed kernels? Sep 17 21:19:52 <_vicash_> the key is that the certificate be in the EEPROM so that one cannot modify it unless they really want to. this can be used for custom embedded devices that use sensors and want to transmit data encrypted. Sep 17 21:20:11 <_vicash_> Vaizki: sure yes that as well. to boot signed kernels. Sep 17 21:20:41 I'm all for a signed bootloader and kernel, as long as I control the keys Sep 17 21:20:44 <_vicash_> if RSA doesnt work then a simple SHA sum check will suffice and i can store the SHA-sums in the EEPROM to verify with. Sep 17 21:20:50 Signed binaries would be even better Sep 17 21:21:16 <_vicash_> spikebike: wel this is just a project for a custom application. not for something mass market. Sep 17 21:21:19 _vicash_: well sha sum needs updated once per kernel and makes it easy to brick Sep 17 21:21:30 _vicash_: but someone can replace the u-boot as well and boot an unsigned kernel that way Sep 17 21:21:30 RSA allows easy kernel updates without changing anything Sep 17 21:21:55 <_vicash_> spikebike: yes that is true. Sep 17 21:22:53 <_vicash_> Vaizki: yes that is true if the hardware is opened and available. if BBB is embedded in a larger piece of hardware which has a warranty then if the user wants to change the keys or change the bootloader they lose hteir warranty. if they want to play with their own custom stuff that's fine by us Sep 17 21:23:44 <_vicash_> we are using open hardware for that reason. so that those who are curious can do what they want but we wont support their warranty. but they have the freedom to do so and dont need to reverse engineer anything Sep 17 21:24:28 <_vicash_> this allows us to satisfy the hackers as well as our customers and corporate management. Sep 17 21:25:51 _vicash_: ya, sensible. Use our bit-identical image/env with support or have at it... with no support Sep 17 21:25:52 ok.. I don't quite get why you want the RSA/SHA stuff on the eeprom if people can replace u-boot or the kernel easily.. Sep 17 21:26:25 Vaizki: well the general idea if you have firmware check the bootloader signature, then the bootloader checks the kernel signature. Sep 17 21:26:47 although that's kinda silly if you don't require signed modules since there's no limit to what a module can replace Sep 17 21:26:50 <_vicash_> Vaizki: they cannot replace it without breaking the warranty. and right now this is still R&D .. we could add or remove features based on the RSA key. Sep 17 21:27:06 sure but the soc firmware can't be changed on BBB, but the MLO and uboot are on the FAT16 partition so very easy to change Sep 17 21:27:14 <_vicash_> so yea it is still all up in the air. the question is really : can it be done. Sep 17 21:27:33 <_vicash_> Vaizki: that's fine by us. most of our customers will not want to change it Sep 17 21:28:11 <_vicash_> those who want to change it we can cancel their support. that way they can own their own device however they want. we dont care. once support is cancelled the device will not communicate with our systems anyway Sep 17 21:28:14 so you are basically putting a piece of string on the door instead of lock so you can say, warranty void if string removed? :) Sep 17 21:28:18 Hello there ! Sep 17 21:28:45 I just got a BBB and i have noticed the perf tool is so different Sep 17 21:29:03 anybody have a link for me to install the perf tool on my BBB Sep 17 21:29:17 not a bbb-related question Sep 17 21:29:39 <_vicash_> Vaizki: yes. just like how Microsoft does with theri UEFI implementation. and sure MLO can change but if u want to change it then u r free to do so. it doesnt affect us. u have already paid for the device. Sep 17 21:29:51 ok Sep 17 21:30:14 <_vicash_> so unlike the phone companies and such we respect hte hackers and give them the freedom to fuck with their device. they paid for it so they can do what they want. Sep 17 21:30:42 centos added signed modules Sep 17 21:30:52 er RHEL Sep 17 21:30:57 yes, and they dont need to buy bbb from you neither Sep 17 21:31:10 soley so they can tell if you have unsupported modules when you call support Sep 17 21:31:20 <_vicash_> spikebike: i have to look into how Angstrom can do that. Sep 17 21:31:24 if you use their kernel Sep 17 21:31:38 _vicash_: it's pretty simple/direct Sep 17 21:31:55 cat /proc/whatever has a flag to show if the kernel + modules are signed or not Sep 17 21:32:36 and if you compile your own kernel you can fake that ;) Sep 17 21:33:00 <_vicash_> RypRap: we are not selling BBBs. we are using BBBs to build a custom device that does something and it is a part of the device. if someone is really inclined to change MLO on that, fine by us. the number of folks who do that maybe ten or so. we dont even plan to sell more than 1000 such devices. Sep 17 21:33:29 <_vicash_> spikebike: interesting. i will look into it. Sep 17 21:33:35 http://www.linuxjournal.com/article/7130 Sep 17 21:34:36 <_vicash_> spikebike: thanks. Sep 17 21:35:14 I guess your goal is to have people use your whole software load or none of it Sep 17 21:35:27 heh, no Sep 17 21:35:42 I mean _vicash_, not RHEL Sep 17 21:35:44 Sounds like they don't want to support folks tweaking it. Sep 17 21:36:04 don't replace the kernel and then spend hours on the phone with me trying to track why obscure things don't owkr Sep 17 21:36:22 <_vicash_> Vaizki: yes we are planning to do a software + hardware combination sale. right now the device isnt even ready. we are still experimenting. so what we do as policy with tweaking is something not in my domain. that is all management stuff Sep 17 21:37:37 <_vicash_> spikebike: yes. exactly. we dont want to support folks who replace the kernel. if too many folks want to do it then we can come up with guidelines but until then we are not putting the cart before the horse Sep 17 21:37:59 sensible Sep 17 21:38:04 I wish phone makers did the same Sep 17 21:38:23 <_vicash_> spikebike: i wish phonemakers made phones which did not drop calls Sep 17 21:38:37 heh.. phones.. 0.1% of users want to hack their phone. Sep 17 21:38:44 who cares about 0.1% Sep 17 21:38:46 Sure the default is 3d animated facebook photos flying in formation around my screen... just let me opt out to default android. I don't mind losing support. Sep 17 21:39:13 Heh, I don't want to "hack" I just don't want tons of crapware clogging up my phone Sep 17 21:39:21 <_vicash_> agreed Sep 17 21:39:41 I don't want to know when a NFL game is available for streaming, or then Nth nag about joining some random photo service. Sep 17 21:39:46 <_vicash_> that's the one thing good about the Blackberry. even though it is only an official phone for me they dont have useless apps on it.. Sep 17 21:40:01 well that crapware is a big part of driving down the price of the phone Sep 17 21:40:13 I know, been in the business.... Sep 17 21:40:21 <_vicash_> anyway. i think i am digressing from the point at hand which was to write to the EEPROM which all of you have helped me with. Sep 17 21:40:42 Vaizki: I'll happily pay for not having ads Sep 17 21:40:48 <_vicash_> so thank you for all your pointers. Sep 17 21:41:08 spikebike: yea me too. but we are statistical anomalies I'd say. Sep 17 21:41:32 * spikebike pats his nexus-4 Sep 17 21:41:38 http://blog.flurry.com/bid/99013/The-History-of-App-Pricing-And-Why-Most-Apps-Are-Free Sep 17 21:42:35 I was amused that angry birds on android made more $$ with ads than angry birds on IOS did as a paid app Sep 17 21:44:12 because selling the sizzle is easier than selling the sausage? Sep 17 21:44:48 yea it was a bit surprising.. but they didn't have a choice really, when they launched on android most of the world couldn't really PAY for android apps very easily Sep 17 21:45:00 get free ones, yes sure Sep 17 21:45:38 sure, similar situation here Sep 17 21:46:04 google sure as hell is not getting my credit card data. Sep 17 21:46:28 maybe they'll ask the NSA for it if they need it Sep 17 21:46:59 <_vicash_> kblin: if u have ever used Google Checkout they have your credit card info Sep 17 21:47:09 _vicash_: I haven't Sep 17 21:47:34 <_vicash_> in any case with all the receipts being emailed back to gmail they at least have your last 4 digits and maybe even the expiration dates. Sep 17 21:48:11 <_vicash_> kblin: but it is possible that all those javascript APIs that shopping carts load some might be sending identifying info to google via the API. i dont know just speculating Sep 17 21:49:05 <_vicash_> kblin: if your bank sends out your credit card statements to you email then google might be scanning those PDFs. they are unscrupulous that way Sep 17 21:49:06 what counts is plausible deniability. Sep 17 21:49:20 and no, my bank doesn't do emails Sep 17 21:49:48 <_vicash_> that's good. Sep 17 21:51:48 ah, better article on signed modules Sep 17 21:51:51 https://lkml.org/lkml/2007/2/14/169 Sep 17 21:52:41 Hello everyone! I am new to BeagleBone and need some quick help with FLASHING my board Sep 17 21:53:01 <_vicash_> spikebike: nice. that's worth bookmarking Sep 17 21:54:30 I am currently flashing my BeagleBone Black with the newest image downloaded from the official website through the micro-SD card. But it has been over 45 minutes and the LEDs have not stayed ON (signaling that the flashing has ended). Sep 17 21:55:20 so it's been 47 minutes already? Sep 17 21:56:35 yes. definitely Sep 17 21:57:12 well I don't know how long it takes but it may depend on your sd card as well Sep 17 21:57:30 Could it take MORE than one hour? Sep 17 21:57:38 BTW, thanks for answering Sep 17 21:59:23 yes Sep 17 21:59:47 give it about 2hrs to be safe... the uSD card you are using could be painfully slow Sep 17 22:00:28 they really should make the flasher image do some sensible output with the leds.. Sep 17 22:00:54 like an initial signal verifying that it's booted ok and starting flashing Sep 17 22:01:13 <_vicash_> Vaizki: one can use the UART cable and see it happening Sep 17 22:01:45 <_vicash_> or you can connect an ethernet cable to it as well. then you can ssh into the system and watch the emmc.sh script flash the eMMC using the "ps" command. Sep 17 22:01:49 instead of they, why don't you submit patches to implement it. the source is available! Sep 17 22:01:52 well my empirical research shows that maybe 2.4% of people coming to this channel have a usb ttl serial cable Sep 17 22:02:15 ds2, because I have never used ångström and never will :) Sep 17 22:02:31 never say never but I'm not inclined to start using it currently Sep 17 22:02:35 <_vicash_> ethernet works. when the flashing is happening the system boots normal into the Angstrom distro and then runs an emmc.sh script to flash. in the meantime you can do whateveer u wnat such as log in adn see Sep 17 22:03:12 right.. but they could periodically USR1 signal dd and read the stats, then change the blink rate of the led Sep 17 22:03:39 * _vicash_ thinks that sounds like parsing morse code Sep 17 22:04:06 well my camera battery charger works like that, flashes more when it's fuller Sep 17 22:04:36 the less "dark" there is, the fuller the battery Sep 17 22:04:55 of course it only has 5 steps Sep 17 22:06:01 and now.. this may be a completely idiotic ångström noob question but why do you even have to reflash your bbb? Sep 17 22:06:13 can't you just update the whole distro from the net? Sep 17 22:18:03 [kernel] RobertCNelson pushed 1 new commit to 3.12: http://git.io/_cvlDw Sep 17 22:18:04 kernel/3.12 84ef502 Robert Nelson: 3.12: update to v3.12-rc1... Sep 17 22:18:30 hey hoes! Sep 17 22:19:35 @Vaizki I need to flash it for a class im taking; they require the newest image Sep 17 22:20:59 hey, leds-pwm: http://permalink.gmane.org/gmane.linux.kernel.commits.head/369873 Sep 17 22:21:04 nice. Sep 17 22:24:16 why aren't the pru pins ever in the expansion header table in the srm? why no pru love? Sep 17 22:38:50 I thought there were no dedicated Pru pins Sep 17 22:39:49 I have another question, How can I SSH through the ethernet port? Sep 17 22:43:56 FIUstudent Sep 17 22:44:02 you get the ip Sep 17 22:44:04 and then you type Sep 17 22:44:10 ssh user@ip.. Sep 17 22:45:13 from where? Sep 17 22:45:57 try ifconfig from a root prompt.. using serial if you have to.. ssh to that address Sep 17 22:46:16 <_vicash_> or use a uSB cable to the BBB and ssh root@192.168.7.2 Sep 17 22:46:58 I got the IP address using ipconfign (on Windows, sorry) but when I use PuTTY to SSH into it i get an error. Sep 17 22:47:21 ...and by the way, I am using an ethernet cable to connect, Sep 17 22:47:24 <_vicash_> Vaizki: I am not an EE, so this might be a dumb question. How does one ground the TP4 test point. Can i connect it to the TP8 test point which says "System Ground" using a jumper wire ? Sep 17 22:47:32 the error i get is Connection Refused Sep 17 22:47:36 umm Sep 17 22:47:40 maybe you got the wrong ip Sep 17 22:47:49 ipconfig on windows would give you the windows ip address Sep 17 22:47:54 windows is not going to know what IP address the rest of your BB has. do you have it connected to a switch or anything? or just BB to your PC? Sep 17 22:47:54 get the ip from your router Sep 17 22:47:59 or just play around Sep 17 22:48:06 incrementally Sep 17 22:48:07 what you are doing is attempting to SSH into your windows machine Sep 17 22:48:15 ping ip's Sep 17 22:48:22 nmap works really well too Sep 17 22:48:41 Ok, I understand the issue with the iP address. I have it connected directly from BB to computer's ethernet Sep 17 22:48:52 :/ Sep 17 22:49:11 find a dhcp thingy and connect it there Sep 17 22:49:28 if you have serial connected, you can set the address that way Sep 17 22:50:39 emocakes , what do you mean by a dhcp thingy? Sep 17 22:51:49 you need to run a dhcp server on your windows box Sep 17 22:52:33 <_vicash_> FIUstudent: u can just use the USB cable that came with the BBB and ssh using the root@192.168.7.2 ip address. it uses USBnet and u shud be able to do it easily Sep 17 22:55:13 guess he either got it to work or bluescreened his windows box lol Sep 17 22:55:35 lol Sep 17 22:55:36 FIUstudent: the beaglebone gets the ip from ethernet using the dns. the ethernet-over-usb uses a static ip of 192.168.7.2. Sep 17 22:55:44 dont diss windows TehCaptain Sep 17 22:55:47 OS: Windows 8 Server Datacenter (full installation) (Version 6.2 Build 9200), CPU: 12 x Intel(R) Xeon(R) CPU E5-2630 0 @ 2.30GHz @ 2294 MHz 1536 KB Cache, MEM: 73693 MB, 90% (66240 MB) free, DISKS: total 19.0 Tb, 17.2 Tb free, GFX: NVIDIA GeForce 9800 GTX/9800 GTX+ 512 MB, SCREEN: Generic PnP Monitor, 1900 x 1140 @ 32 bit, AUDIO: Sound Blaster Z, UPTIME: 0 d, 21 h, 32 m Sep 17 22:55:48 :p Sep 17 22:56:04 emocakes: lol. 21 hours! WOW. Sep 17 22:56:08 although it does suck for any dev work :p Sep 17 22:56:12 why's that? Sep 17 22:56:25 i dunno, im more of a *nix person Sep 17 22:56:37 need some different software, just compile it Sep 17 22:56:42 so. vm it. it'll still run crazy fast. Sep 17 22:56:51 virtualbox, set it to all 12 cpus. Sep 17 22:56:55 :/ Sep 17 22:57:02 im avoiding virtualbox Sep 17 22:57:05 why? Sep 17 22:57:06 that is a nice machine.. would still run debian on it though Sep 17 22:57:07 i have vm's though Sep 17 22:57:16 virtualbox is shitty Sep 17 22:57:19 the performance difference will be unnoticably different. Sep 17 22:57:20 compared to hyper-v Sep 17 22:57:22 or vmware Sep 17 22:57:23 :p Sep 17 22:57:34 i have os x running in a vm Sep 17 22:57:36 pfft.. whatever. any vm will be absolutely acceptable. Sep 17 22:57:41 i know Sep 17 22:57:46 with 72gb of ram Sep 17 22:57:48 12x. daaang. Sep 17 22:57:49 seriously. Sep 17 22:57:50 lol Sep 17 22:57:53 ramdisk that biaaatch Sep 17 22:58:17 im on a mac at the moment, remote desktopping into the computer, its at my table and im too lazy to get up Sep 17 22:58:25 makes me want to upgrade my phenom 1090t... Sep 17 22:58:32 TehCaptain, its an amd Sep 17 22:58:37 what are you doing in the first place? Sep 17 22:58:54 best part about the computer is i can fit another cpu in it :D Sep 17 22:59:02 two sockets, holds up to 768gb of ram Sep 17 22:59:31 you could compile the whole world with -O3 -flto! :O Sep 17 23:00:32 i would install gentoo over and over Sep 17 23:00:34 just for fun Sep 17 23:00:37 hahaha Sep 17 23:06:08 * ds2 throws scooby snacks around Sep 17 23:07:15 ruh-ro Sep 17 23:09:31 so now i need a use for my bbb Sep 17 23:09:39 it has just been sitting there for a few months Sep 17 23:09:41 :( Sep 17 23:09:47 paperweight? Sep 17 23:09:51 doorstop? Sep 17 23:09:54 NTP Sep 17 23:09:57 art? Sep 17 23:11:23 cat toy? Sep 17 23:11:34 urinal cake ;p Sep 17 23:12:08 paperweight Sep 17 23:12:13 i wanted to get erlang running on it Sep 17 23:12:22 but havent really had much of a chance to sit down with it Sep 17 23:12:35 2 in 1, just plug it in it should attract house pets that will want to sleep on it. Sep 17 23:17:01 why would a house pet want to sleep on a urinal cake? Sep 17 23:17:16 hy not Sep 17 23:19:14 smell Sep 17 23:20:32 I've been pouring over the TRM for the beagle bone black's processor for a while now trying to understand how to set up and read the analog inputs with the PRU. I do have previous experience with configuring microcontrollers, I just need to be pointed in the right direction. Sep 17 23:21:29 rocky|beagle: the PRU is blackmagic Sep 17 23:21:56 Yeah, I've figured out that much. But even magic has some rules to it, right? Sep 17 23:22:08 you'll need a old TRM Sep 17 23:22:19 because the new TRM has the PRU seciton removed Sep 17 23:22:25 thanks TI! Sep 17 23:22:26 spru73c Sep 17 23:22:38 Really now? That's annoying *look of disapproval* Sep 17 23:22:41 spruh73c Sep 17 23:23:02 is that the document's id? Sep 17 23:23:03 ^^ just google that Sep 17 23:23:07 thanks Sep 17 23:24:22 thanks troll mru, i was just looking for that Sep 17 23:24:38 rocky|beagle: you get the pru working? Sep 17 23:24:57 rocky|beagle: the trolls in #beagle are freindly most of the time, if you dont' ask dumb questions Sep 17 23:25:13 yeah I did Sep 17 23:25:45 I had to modify the device tree to enable the pru Sep 17 23:25:52 19:23 < mru> spruh73c Sep 17 23:25:52 19:23 < rocky|beagle> is that the document's id? Sep 17 23:25:52 19:23 < mru> ^^ just google that Sep 17 23:25:52 19:23 < rocky|beagle> thanks Sep 17 23:25:52 19:24 < mrpacket_> thanks troll mru, i was just looking for that Sep 17 23:25:55 19:25 < nomel> rocky|beagle: you get the pru working? Sep 17 23:25:57 19:25 < mrpacket_> rocky|beagle: the trolls in #beagle are freindly most of the time, if you dont' ask dumb questions Sep 17 23:26:00 19:25 < rocky|beagle> yeah I did Sep 17 23:26:03 19:26 < rocky|beagle> I had to modify the device tree to enable the pru Sep 17 23:26:08 eh what? Sep 17 23:26:15 why you cut and past Sep 17 23:26:25 Sorry, I'm using putty, and highlighting and right clicking (annoyingly easy to do) does that Sep 17 23:27:20 changing my settings, won't happen again. Sep 17 23:27:25 rocky|beagle: you don't have to modify device tree, you can use an overlay. Sep 17 23:27:54 although, i suppose that fits with "modify the device tree" Sep 17 23:27:59 My prof covered that today. I didn't know about it a couple days ago when I did the device tree method. Sep 17 23:28:07 oh, you're learning this in school? Sep 17 23:28:18 yeah, ECE497 Sep 17 23:28:21 what do you plan on doing with the pru? Sep 17 23:29:01 My senior design team is considering using it to handle the flight control of a quadcopter Sep 17 23:29:23 why not the main cpu? Sep 17 23:29:39 I want to ensure everything is real time. Sep 17 23:30:20 you mean with deterministic timing ;) Sep 17 23:30:33 It's a safety thing. If the kernel were to hang or too much cpu time is going to other processes, I want the copter to remain stable. Sep 17 23:30:47 that's pretty cool. Sep 17 23:31:12 there are a few debuggers in the works, last i checked. Sep 17 23:31:19 that'd be very nice :) Sep 17 23:31:21 (although, they seemed pretty lame) Sep 17 23:31:38 one is being used by a beaglebone opencnc project Sep 17 23:31:47 and i saw a few started on github. not sure they went anywhere. Sep 17 23:32:17 My plans right now are to port some other flight control software for the msp430 to the PRU. Or at least the algorithms. Sep 17 23:33:09 the macros can take parameters, and have default arguments Sep 17 23:33:17 so you can make the pru asm pretty clean looking Sep 17 23:33:56 Yeah, I did see some high level stuff (for assembly) in the wikis and whatnot. Sep 17 23:34:15 asm is all you get Sep 17 23:34:20 I know. Sep 17 23:38:45 isn't the PRU supplement checked into the github page with the assembler? Sep 17 23:45:57 yeah, it is. Sep 17 23:46:21 with all of it's errors and contradicting statements! :D Sep 17 23:46:45 there are errors in the truth Sep 17 23:46:46 and signal names that are used in single sentences, and never again. Sep 17 23:47:44 oh I love that about the latest TRM. An introduction sentence for the PRU is all it gets, not even a mention of where to look for more docs. Sep 17 23:47:55 it's technically unsupported by ti. Sep 17 23:50:18 When beaglebone black was received I was able to ssh over USB, now I am unable to updating to the latest black firmware. Is that normal? Sep 17 23:50:40 Note, I am booting from sdcard Sep 17 23:51:26 EvilSeed: did you read the instructions on how to update? Sep 17 23:51:45 yes, I see it booted and see the USB drive when connected Sep 17 23:52:07 download the compressed file, extracted with 7zip and used the disk image software to put it on SDCARD Sep 17 23:52:09 if you're booting from the sd card, you'll need the sd card image, and you'll have to modify the uEnv.txt to disable the emmc boot. Sep 17 23:52:20 booting from the sd card is not usual. Sep 17 23:52:26 Ahhhhh Sep 17 23:52:30 updating from the sd card is. Sep 17 23:52:33 which is described in the instructions. Sep 17 23:52:47 Forgive me for being a clod there. Thank you for your input. Sep 17 23:52:53 Appreciate it very much Sep 17 23:52:54 :P Sep 17 23:54:08 I don't think booting off of the sd card is all that unusual. We're doing kernel development as part of our class, and he's suggested that we have many sdcards with angstrom loaded to decrease downtime if we make the system unbootable. Sep 17 23:59:25 unusual in the majority sense. :P Sep 17 23:59:35 sure sure Sep 17 23:59:54 i boot from sd for that reason. Sep 18 00:00:25 if the card goes bad, i can pop it out and load a pre-loaded one. Sep 18 00:00:31 if you are doing kernel dev, why not tftp in the kernel? Sep 18 00:00:37 or did someone break that? Sep 18 00:01:32 didn't know about that, looks useful Sep 18 00:03:46 some people netboot everything, which is pretty rad. Sep 18 00:05:56 netboot is a bit more involved Sep 18 00:08:52 i spent 3 hours trying to get a bone to netboot Sep 18 00:09:08 most of which i spent in the tftp server code trying to figure out why the bone didnt like the packets it was sending Sep 18 00:09:30 i "fixed" it by moving it from a 100m full duplex port to a 10 half duplex port Sep 18 00:09:40 made me angry Sep 18 00:12:03 i'm impressed you have something with a 10 half duplex port. Sep 18 00:12:24 it's either expensive and configurable, or from the long long ago. Sep 18 00:13:00 can you netboot over wifi? Sep 18 00:13:25 i suppose you could. Sep 18 00:13:26 configurable Sep 18 00:13:43 and i "borrowed" it Sep 18 00:13:59 In uboot, is a linux kernel loaded? or is it just a standalone program? Sep 18 00:14:32 without a linux kernel, it'd be very difficult to get the wifi going I'd think Sep 18 00:16:20 heh, ya Sep 18 00:20:17 sounds like the code to reprogram the PHY is not in uboot Sep 18 00:21:21 you'd think default would involve autoneg. Sep 18 00:24:38 the phy on there does not power on with autoneg Sep 18 00:24:53 it needs to be put into autoneg mode otherwise it is in 10BT-HD Sep 18 00:36:09 why? wouldn't autoneg select that if it were needed? Sep 18 00:37:18 that's if autoneg was enabled Sep 18 00:37:27 question is why didn't the other end negotiate down Sep 18 00:38:21 true that. Sep 18 01:20:20 Does anyone here know of good resources to get started with learning DSP? **** ENDING LOGGING AT Wed Sep 18 03:00:00 2013