**** BEGIN LOGGING AT Wed Mar 28 03:00:01 2018 Mar 28 05:34:39 I just got a brand new bbb wireless, but without sd card. Do I need one? Mar 28 05:43:58 you mainly need some patience :P Mar 28 06:21:57 can anyone point me in the right direction to get a live webcam streaming server running on a BBBW? I've tried motion and ffmpeg/ffserver but even at 320x240 resolution and 2 frames/sec it is using 70% of the CPU. I need to be able to get a lot better quality Mar 28 06:22:32 lol Mar 28 06:22:37 lol Mar 28 06:22:57 yeah im still stuck on that Mar 28 06:23:18 find a webcam with built-in H.264 encoding to avoid having to spend cpu time on it? Mar 28 06:23:56 it'll also reduce the amount of cpu time used by dma-less usb transfer Mar 28 06:24:07 i think one of the ones I have has it, but the video from it comes out all messed up, and i'm not 100% sure it has h.264 built in Mar 28 06:24:45 also make sure the webcam gets decent power (to test that, try inserting a powered usb hub between bbb and cam) Mar 28 06:25:08 hmmm could that make the video all choppy/jumbled? Mar 28 06:25:22 it's a possibility Mar 28 06:25:46 also disable usb autosuspend, some cams react badly do it (though mainly if you try to use the cam for still shots) Mar 28 06:27:19 lol why does usbcore have a module parameter "blinkenlights" Mar 28 06:27:37 sry got disconnected... ill give that a try and see if it works. thanks for the help Mar 28 06:28:13 here's what you might have missed: https://pastebin.com/raw/eA2FhNgF Mar 28 06:29:03 awe you missed me saying 2 fps was almost a still shot Mar 28 06:29:13 hehe Mar 28 06:29:54 yeah ill dig out my powerd hub and give that other cam a try again and if still messed up ill probably get a cam I know has h.264 Mar 28 06:30:33 could drivers cause that or is it more likely if drivers were an issue it wouldn't even find the cam Mar 28 06:30:36 the issue with still images and autosuspend is that the cam is taken out of suspend just for a single image, but some cams need some time after exiting autosuspend (probably to stabilize internal power supplies) and the first image is complete garbage Mar 28 06:31:09 well usb cams should just use the generic usb video class driver Mar 28 06:31:26 ok Mar 28 06:31:54 though, before spending money on a cam, you may want to research what the current state of uvc1.5 support in linux is Mar 28 06:32:51 ok Mar 28 06:33:54 anabody is there Mar 28 06:34:04 anybody is there Mar 28 06:34:19 that is a statement, gramatically. and correct. Mar 28 06:34:53 i have one problem can you guide to solve that problem Mar 28 06:36:37 I can guarantee you that noone can help you with a problem if you don't actually first explain what problem you're having Mar 28 06:37:34 sir i want to use gpio output pins P8-3 to P8-6 Mar 28 06:38:06 thats why i am flashing the os in SD card not in EMMC Mar 28 06:38:16 you're the same guy as "Prakash" yesterday? Mar 28 06:38:22 yes Mar 28 06:38:31 now its working fine Mar 28 06:38:49 but how can i use EMMC as external sotarage Mar 28 06:38:52 storage Mar 28 06:39:31 ehm, you can't use eMMC if you're using P8.03 - P8.06 as gpio, I explained that. those two things conflict Mar 28 06:40:17 so its not possible to using emmc as external storage purpose Mar 28 06:40:28 because of its having 4GB Mar 28 06:40:33 eMMc Mar 28 06:42:41 the eMMC is connected to the same processor pins as P8.03-06, P8.20-25. so if any of those are used as gpios, no communication between processor and eMMC is possible Mar 28 06:44:18 kk tq Mar 28 06:48:27 zmatt: I'm trying to disable autosuspend using [ sudo echo on > /sys/bus/usb/devices/1-1/power/control ] but getting permission denied. Mar 28 06:49:00 that's because you're redirecting the output of [ sudo echo on ] Mar 28 06:49:07 but the redirection is still done as you Mar 28 06:49:22 one solution is: echo on | sudo tee /sys/bus/usb/devices/1-1/power/control Mar 28 06:51:17 ok i think that worked Mar 28 06:51:54 if you want to disable usb autosuspend globally, you can add this kernel parameter (to 'cmdline' in /boot/uEnv.txt): usbcore.autosuspend=-1 Mar 28 06:54:02 ok, ill probably do that Mar 28 07:03:31 zmatt: well the video is still all messed up. If i send u a link to the stream you think if you see how its messed up you would know if it is power related or not? Mar 28 07:04:28 I'm willing to take a look, but I doubt I'd be able to tell anything useful from it, but who knows Mar 28 07:04:47 can i send you a PM? Mar 28 07:04:52 ok Mar 28 07:07:03 looks more like some weird kind of pixel format error? Mar 28 07:07:34 yeah its got swirls and stuff Mar 28 07:07:47 im using this to launch ffserver Mar 28 07:07:49 ffserver -f ~/ffserver.conf & ffmpeg -v 2 -r 5 -s 640x480 -f video4linux2 -i /dev/video0 http://localhost:8090/webcam.ffm Mar 28 07:08:29 im not sure what the ffmpeg commands are doing Mar 28 07:08:38 -v and -r Mar 28 07:09:12 and this is what the config file looks like https://www.moreno.marzolla.name/software/linux-webcam-server/ffserver.conf Mar 28 07:10:00 the very bottom has the stream that is going to the webserver Mar 28 07:10:06 I think -s, used as input option here, might be fucking things up Mar 28 07:10:13 but I'm not sure Mar 28 07:10:18 hm, maybe not Mar 28 07:10:45 so drop the -s 640x480 or just -s Mar 28 07:11:17 no nm I guess it's probably understood by the video2linux2 module to select the resolution Mar 28 07:11:49 oh i also made a few tweaks to the config file.. but ion that section I only changed videosize from qvga 320x240 Mar 28 07:11:50 you should probably use -framerate instead of -r Mar 28 07:13:22 based on what the manpage is saying Mar 28 07:13:48 see also https://trac.ffmpeg.org/wiki/Capture/Webcam#Linux Mar 28 07:15:39 ok ill research ffmpeg and its options Mar 28 07:17:43 how does built in h.264 work? Would ffmpeg tell the built in encoder to do the encoding or would ffmpeg just recive an already encoded stream from the camera? Mar 28 07:18:07 I have no idea, that's why I recommended you'd research the current state of linux support for it first :) Mar 28 07:18:19 IIRC it's all part of the v4l2 magic Mar 28 07:18:29 lol, yeah i'll do that Mar 28 07:18:38 but I guess it would be similar to support for mjpeg encoding by the cam which is much older Mar 28 07:18:40 there are tools to inspect v4l2 devices Mar 28 07:18:45 most cams support it I think? Mar 28 07:19:03 yeah, I think mjpeg is sort of default stream output Mar 28 07:19:08 (see the page I just linked for an example of how to check) Mar 28 07:19:35 ok ill read that page and do a bit more research Mar 28 07:20:43 btw, it might also help to tell ffmpeg not to reencode "-c copy" Mar 28 07:20:58 hmmm Mar 28 07:21:08 i think thats what i want Mar 28 07:27:22 tbr: so would I make the ffmpeg command like this ffmpeg -c copy video4linux2 -i /dev/video0 http://localhost:8090/webcam.ffm Mar 28 07:27:40 I think after the input definition Mar 28 07:28:30 like this? ffmpeg video4linux2 -i /dev/video0 -c copy http://localhost:8090/webcam.ffm Mar 28 07:29:30 yes Mar 28 07:36:14 tbr: its giving me https://pastebin.com/rbDw4SiF Mar 28 07:41:26 you should probably 1. verify the cam supports mjpeg 2. select an mjpeg format somehow Mar 28 07:42:16 yeah, that looks like it doesn't get an compressed datastream Mar 28 07:45:43 looks like i need to do some more research. thanks for the help Mar 28 08:07:45 Hi: Mar 28 08:08:24 I just got my BBB Wireless, I am unable to mount it with my Window 7 pc Mar 28 08:13:57 do the LEDs light up and how do they blink? Mar 28 08:15:24 Yes. 2 blue LED at 5v connector Mar 28 08:17:34 the USR0 LED is blinking Mar 28 08:18:29 also, i am using the mini usb cable provided Mar 28 08:19:36 then you should by now probably see a usb drive Mar 28 08:19:53 also device manager should show some devices missing drivers Mar 28 08:22:43 not detected. under device manager-other devices---> unknown device Mar 28 08:30:12 then try to reboot it and if that doesn't help flash a SD card and boot from that (by holding down the user button) Mar 28 08:32:08 OK. I have to go grab a sd card and try again. But anyway I installed the 64bit USB driver from BBB org, restarted the BBB Mar 28 08:32:55 there is an error message--failed to installed usb driver Mar 28 08:34:31 you should see the storage device regardless Mar 28 08:35:07 No. None. Mar 28 08:35:17 then try the SD card Mar 28 08:36:18 ok. I think that is the last resort Mar 28 08:37:25 by the way, any difference between the mini and usb a port? Mar 28 08:37:57 yes, one is a host the other is a device port Mar 28 08:38:17 both could be reconfigured in software but would need hardware changes Mar 28 08:38:35 can i use the host to mount ? Mar 28 08:41:54 both USB ports require the OS running on the device to configure them and expose functionality. If the OS doesn't do that then you can't force it to. Mar 28 08:45:12 ok. ssh what a bad start up Mar 28 08:46:17 yeah, although generally it's often a good idea to reflash a new device anyway since you never know how old the software on it is anyway Mar 28 08:46:29 ew, double 'anyway', ugly sentence Mar 28 08:47:39 I try with another win7 pc, same thing, ... arrr Mar 28 08:47:48 SD card Mar 28 08:48:24 yeah I'd recommend grabbing the stretch-iot image and reflashing eMMC Mar 28 08:48:39 hopefully that will fix problems Mar 28 08:53:29 any specific steps i can follow once i have the sd card? Mar 28 08:54:14 download the "Stretch IoT" image near the top of https://beagleboard.org/latest-images Mar 28 08:54:33 flash it onto sd card, and boot the beaglebone from it Mar 28 08:55:26 ok Mar 28 08:55:38 assuming that all works okay and you can successfully log in on it, you can easily flash it to eMMC by uncommenting a single line in /boot/uEnv.txt on the card (see the paragraph below the download links) and rebooting Mar 28 09:00:40 this step can be fine at this web site? Mar 28 09:02:09 I have to Rasp PI 3, i have no issue with it. (any way it came with SD and Preloaded) Mar 28 09:02:27 hi. Does anyone have a ccs7 target configuration file for bbb? Mar 28 09:03:09 RPi is more like a toy, so gets more handholding Mar 28 09:07:56 lol Mar 28 09:08:10 he's not wrong though, it ought to work out of the box Mar 28 09:09:12 yes Mar 28 09:20:59 Sorry, I fell asleep Mar 28 09:21:17 And I don't know where ALSA is reading the input from Mar 28 10:26:53 Is beaglebone equipped with a built-in JTAG emulator providing connectivity over the same USB cable that powers the board? Mar 28 10:27:44 only the original beaglebone (aka the beaglebone white) had that, later models removed it to reduce cost and because pretty much noone cared Mar 28 10:28:30 hmm..how can i debug code made for the PRUs? Mar 28 10:29:00 there are pru debuggers you can run under linux Mar 28 10:29:12 the PRU debug interface is via memory-mapped registers, not via jtag Mar 28 10:29:32 oh! that's nice... Can I use ccs for that? Mar 28 10:29:38 no Mar 28 10:29:45 at least, I doubt it Mar 28 10:30:14 google 'prudebug' Mar 28 10:30:48 thanks Mar 28 10:31:11 maybe I'll also some day add some debugging functionality to my py-uio project Mar 28 10:31:36 it already has a low-level interface for core control including single-stepping Mar 28 10:33:19 ok. So without extra components all PRU programming/debugging is done using assembly Mar 28 10:33:21 ? Mar 28 10:33:38 ? Mar 28 10:33:57 assembly language Mar 28 10:34:10 assembly is most commonly used for pru, more recently TI also made a mediocre C compiler for it Mar 28 10:34:48 C and PRU are not a great match however, the instruction set was designed for assembly programming, not for being targeted by a C compiler Mar 28 10:35:19 I see. thanks very much for the info Mar 28 14:16:45 hi there channel :-) Mar 28 14:17:32 fortnight, are you available? Mar 28 15:47:05 myself, are you available, there? Mar 28 15:47:55 gesss... any ideas for rev.a5c. Mar 28 15:50:02 why do you keep bringing up the board revision when you've been told countless times already that the board revision is irrelevant? Mar 28 15:50:36 eMMMC /dev/mmcblk1p1 1.8G 1.3G 368M 78% / Mar 28 15:51:05 yes it has less eMMC space? so what? Mar 28 15:51:12 irrelevant, what you mean zmatt? Mar 28 15:51:26 2GB should be fine for most purposes, and if it isn't, boot from sd card instead Mar 28 15:51:35 * tbr agrees Mar 28 15:52:16 but you keep including the board revision in your questions, as if it matters Mar 28 15:52:21 it doesn't matter Mar 28 15:53:04 maximun amount sd-card supported Mar 28 15:53:39 1.21 jiggabites! Mar 28 15:53:53 no max on sd card, at least no limits that matters Mar 28 15:54:49 2 TB is fine, if such a card existed yet Mar 28 15:55:11 you kidding Mar 28 15:55:17 no we are not Mar 28 15:55:31 why would there be a limit? Mar 28 15:55:44 catch it, (y) Mar 28 15:55:58 (there is only a marketing/licensing bla limit, but there is no technical limit) Mar 28 15:56:41 tbr, you got A5C one? Mar 28 15:56:43 2 TB is the technical limit Mar 28 15:57:02 (2^32 sectors * 512 bytes/sector) Mar 28 15:57:32 dury: possibly, I'd need to check. it's somewhere, collecting dust. Mar 28 15:59:01 just want to build a simply fly project or project with wheels Mar 28 15:59:25 so why does it matter if tbr has an A5C ? Mar 28 16:02:06 see how did he install the OS, I guess there are different ways to install OS in 2GB eMMC using at the same time an sd-card Mar 28 16:02:37 an = a sorry Mar 28 16:03:28 I haven't used my BBB in about a year or more Mar 28 16:03:38 you can either boot from sd card and ignore eMMC entirely, or reflash eMMC and don't bother with an sd card Mar 28 16:04:01 what he says Mar 28 16:04:13 (using both eMMC and sd card is possible but is probably more headache than it's worth except in special situations) Mar 28 16:05:48 booting from eMMC is faster than sd-card Mar 28 16:06:08 or depends on the sd-card Mar 28 16:06:48 it's much more dependent on optimization of the distribution Mar 28 16:07:19 if you are at a point where IO performance matters, you're probably done booting in less than 2 seconds Mar 28 16:11:01 have to go guys... catch all in other time cheeerss Mar 28 16:13:00 anyone recommend a channel for microcontrollers? Mar 28 16:13:41 I think there is something like ##electronics or ##microcontrollers, but I don't remember exactly Mar 28 16:13:46 I'm testing the endurance of some 'emulated eeprom' on an Atmel samd21j Mar 28 16:14:17 the datasheet gives different endurance figures for flash and 'emulated eeprom' but as far as I can tell it's the same memory Mar 28 16:14:25 was interested in what makes the difference Mar 28 16:14:35 tbr: thanks - I'll have a look Mar 28 16:15:48 eh, so it's emulating an eeprom on top of some sort of flash? Mar 28 16:16:13 yeah - this ARM chip doesn't have EEPROM Mar 28 16:16:33 but you can reserve some of it's flash for emulating EEPROM Mar 28 16:18:59 https://imagebin.ca/v/3wRt8MQLPiSl Mar 28 16:19:41 factor of 4 difference for cycling endurance Mar 28 16:20:26 so the emulated eeprom gets much better numbers. do I see this right? Mar 28 16:21:09 yeah Mar 28 16:21:24 I'm up to 1M cycles atm Mar 28 16:21:29 does AT03265 cover this? maybe they add redundancy and wear leveling? Mar 28 16:21:47 good point Mar 28 16:22:06 yup Mar 28 16:23:38 eeprom page size is half that of the flash page size, that's causing a factor two Mar 28 16:23:43 and there's a factor two redundancy Mar 28 16:25:12 at least that's what I picked up from a quick skim Mar 28 16:25:59 why would page size effect this? Mar 28 16:26:43 most emulated eeprom doesn't erase a page on each write, so you get more endurance because you are erase less than you are writing, as long as the size of you data is smaller than a page you are going to erase once, then write, write, write multiple times until you filled the page and then you will do another erase Mar 28 16:27:15 your data Mar 28 16:27:22 mattvenn: because it means writing two eeprom pages results in writing only a single flash page Mar 28 16:28:14 flash endurance is determined by how many erases you do, not how many small writes Mar 28 16:30:46 the eeprom emulator will automatically erase a flash page when needed obviously Mar 28 16:32:00 so I might think I'm writing to the same address in emulated eeprom, but as I write over the same byte again and again Mar 28 16:32:21 the controller will be writing over different parts of the flash Mar 28 16:32:31 and only erase when the page is full? Mar 28 16:32:33 yes, that is usually how it works Mar 28 16:32:49 I guess I could just read this application note! Mar 28 16:32:56 thanks though, super interesting Mar 28 16:33:35 of course it all depends on how it is implemented, but what you just described is typical Mar 28 16:34:53 it actually looks like you should 6x the endurance if you keep writing the same page over and over Mar 28 16:34:57 same eeprom page Mar 28 16:35:29 my trivial test is writing 80 bytes consecutively starting always at the same address Mar 28 16:35:46 80 bytes spans 2 eeprom pages doesn't it? Mar 28 16:35:54 it seems here the eeprom page size is 60 bytes Mar 28 16:36:19 it's saying here that each flash row can store 2 eeprom pages Mar 28 16:36:39 yes Mar 28 16:37:38 80 bytes seems kind of large for an eeprom value, what are you storing? Mar 28 16:38:05 keyboard layouts for a programmable keyboard Mar 28 16:38:41 do people typically change all the keys or just one or two? Mar 28 16:39:15 I don't really have much of an idea. My guess is people will do a lot of messing around to start, getting things how they like it Mar 28 16:39:21 then the changes will happen later Mar 28 16:39:45 I'm testing to make sure i don't need an external eeprom Mar 28 16:39:49 but it seems fine. Mar 28 16:40:05 originally I thought it would only handle 10k writes so that seemed a bit short Mar 28 16:40:12 you wrote the eeprom emulator? or you are using something from atmel? Mar 28 16:40:26 http://ww1.microchip.com/downloads/en/AppNotes/Atmel-42125-SAM-EEPROM-Emulator-Service-EEPROM_ApplicationNote_AT03265.pdf Mar 28 16:46:43 mattvenn: so my understanding of that doc is that if you keep writing the same eeprom page over and over, there are two flash rows affected and each is erased only once per 6 eeprom page writes: https://pastebin.com/raw/5eQ7GW6G Mar 28 16:46:58 seems like the samd have a very small page size Mar 28 16:47:58 mattvenn: while writing one (even-odd) pair of eeprom pages results in each of the two rows being erased once per 4 eeprom page-pair writes: https://pastebin.com/raw/UQB5DWqn Mar 28 16:48:00 most of the arm chips i've played with are usually at least 1k Mar 28 16:48:51 how much flash are you dedicating to this emulated eeprom? Mar 28 16:49:36 ATM 1k, probablly up to 8k later Mar 28 16:49:40 lots of flash after all Mar 28 16:50:14 so i guess you have to compute the worst case scenario and then allocate the proper amount Mar 28 16:50:40 yeah Mar 28 16:51:01 to be honest, I think the worst case is like 100k writes, so I'm not going to have any problems Mar 28 16:51:12 but I was interested in the difference between the 2 endurance numbers Mar 28 16:51:20 didn't realise eeprom emulation was so clever Mar 28 16:51:25 it's like an SD card Mar 28 16:52:14 zmatt: thanks for the pastes Mar 28 16:58:18 well, no, it doesn't have proper wear-leveling Mar 28 16:59:13 if it did, the total number of writes you could do to the same page would scale with the total number of eeprom pages Mar 28 16:59:59 I think you could probably still trigger this with a bit of cleverness Mar 28 17:03:39 e.g. include a write counter in the page-pair you keep updating, and once every N writes update some other page-pair twice (to force it to rewrite a row), cycling through all eeprom page-pairs over time Mar 28 17:04:54 what's your hourly rate? ;) Mar 28 17:24:00 mattvenn: example with 8 eeprom pages (5 flash rows) and N=6: https://pastebin.com/raw/PMKMj0dr Mar 28 17:24:07 N=6 is way too low of course Mar 28 17:25:08 but even so, this lets you rewrite one eeprom page-pair 30 times while each flash page is only erased 4 times during this period, which yields an endurance of 7.5 times the flash endurance Mar 28 17:27:24 actually even N=2 would already yield benefit: 10 writes / 2 erases = 5x endurance Mar 28 19:57:55 Hey can anyone help me with my BeagleBone Green, im trying to connect to my schools wifi but I cant connect to it using my smartphone because it only allows me to enter a password when my school wifi requires a username and password to connect Mar 28 19:59:37 connmanctl Mar 28 19:59:57 use that command Mar 28 20:00:07 type connmanctl in your terminal. Mar 28 20:00:10 then... Mar 28 20:00:29 services Mar 28 20:00:31 agent on Mar 28 20:00:37 and then... Mar 28 20:01:07 copy your wifi signal on your terminal and type... Mar 28 20:01:21 connect "your wifi signal" Mar 28 20:01:28 then...enter your password. Mar 28 20:01:56 ... Mar 28 20:02:01 Boom! Connection! Mar 28 20:04:00 when I tried to enter the wifi signal it put out an error saying invalid argument Mar 28 20:04:22 did you type connect? Mar 28 20:04:58 so... Mar 28 20:05:04 Yes the signal ends in "...managed_ieeee8021x" Mar 28 20:05:13 but it wont connect Mar 28 20:05:56 Copy the entire part, i.e. wifi_Blah_Blah_Blah. Then, type connect wifi_Blah_Blah_Blah. Mar 28 20:06:18 It will ask for a passphrase. Mar 28 20:06:59 That is odd. Oh well. Mar 28 20:10:54 Hey fortnight you still here? Mar 28 20:10:59 I disconnected Mar 28 20:11:46 Yea. Mar 28 20:11:51 Here and willing. Mar 28 20:12:06 What do you need? Mar 28 20:12:17 If you are, when I type connect "my wifi signal" it says Error 'connect': invalid argument Mar 28 20:12:34 so...you typed connmanctl first? Mar 28 20:12:38 yes Mar 28 20:12:41 then services Mar 28 20:12:44 then agent on Mar 28 20:12:48 Hey. Mar 28 20:12:58 What did it say once you typed agent on? Mar 28 20:13:12 Agent registered Mar 28 20:13:17 Okay. Mar 28 20:13:44 This is the wifi signal " wifi_2cf7f1060384_4169724f72616e676558_managed_ieee8021x" Mar 28 20:14:17 Damn mate. Mar 28 20:14:31 Okay. What image are you using? Mar 28 20:14:37 What year? Mar 28 20:14:43 cat /etc/dogtag? Mar 28 20:14:59 ok one sec Mar 28 20:15:01 Okay. Mar 28 20:15:29 Debian 8.10 2018-02-01 4GB SD SeeedStudio IoT Mar 28 20:15:46 this is the image i donwloaded from beagle bone Mar 28 20:15:57 Try connmanctl and then wifi enable and then wifi scan. Then, try the rest of the commands mentioned. Mar 28 20:16:49 Hey... Mar 28 20:17:10 sorry about that when i type wifi enable it says no such command Mar 28 20:17:12 Yep...did you get that last part? Mar 28 20:17:25 Okay. Try the wifi scan too. Mar 28 20:17:38 says unknown command as well Mar 28 20:17:52 Hmmm. What is the name of the kernel you use? Mar 28 20:18:02 Wheezy or Stretch or something else? Mar 28 20:18:17 the commands needed to be reveresed haha Mar 28 20:18:21 Oh. Mar 28 20:18:30 Odd days! Mar 28 20:18:36 scan is complete and wifi is enabled Mar 28 20:18:41 Awesome! Mar 28 20:18:48 I am glad you figured it out. Mar 28 20:19:05 im still not connected to my wifi though haha Mar 28 20:19:09 Oh? Mar 28 20:19:25 i meant those command were completed, my fault haha Mar 28 20:19:28 You still need to complete the rest of the commands. Mar 28 20:19:29 Oh. Mar 28 20:19:39 I got you. Mar 28 20:20:22 so. connmanctl, agent on, services, copy the Wifi signal, connect Wifi signal, and add the passphrase. Mar 28 20:20:49 I would update to stretch if I were you. Mar 28 20:21:13 Right so we get to the wifi connect part Mar 28 20:21:28 JoFitz: Update to Stretch. Mar 28 20:21:28 but it still throughs the same errot Mar 28 20:21:38 Hmmm. It is the image you use. Mar 28 20:22:10 We are currently using 9.3 and higher, right? Mar 28 20:22:59 JoFitz: Use the 9.3 kernel if you can. Mar 28 20:23:06 Stretch... Mar 28 20:23:20 We are using Linux beaglebone 4.4.113-ti-r145 Mar 28 20:23:29 I remember the older kernels and images. They had given me a bad set up w/ connmanctl. Mar 28 20:23:33 Update your kernel. Mar 28 20:23:38 Oh... Mar 28 20:23:52 Just update the image. I forgot you cannot connect to WiFi. Mar 28 20:24:16 So, 4.9.x works well. Mar 28 20:24:49 Does it have to do with the type of security that my university is using? Mar 28 20:24:54 Stretch IoT 9.3 should work like a charm. Mar 28 20:25:15 For a beaglebone green wireless? Mar 28 20:25:19 Yea. Mar 28 20:25:23 Yes sir. Mar 28 20:25:45 You can update to whatever you like but first we need to get you working w/ WiFi. Mar 28 20:26:25 So, use IoT if you need it or you can just use the LXQT. Mar 28 20:27:24 JoFitz: That damn complication w/ the sign in menu is a pain. stick w/ connmanctl as a command to sign into WiFi. Mar 28 20:27:38 I have a green and green wireless. Mar 28 20:28:14 That green wireless had that damn sign in menu for "coolness" or something. It is a pain. I used connmanctl and never looked back. Mar 28 20:28:51 Ok we're going to try that haha Mar 28 20:30:02 Sure. So, go to bbb.io/latest-images and get the top one. 9.3 IoT or the second one: 9.3 LXQT will work definitely. Mar 28 20:31:04 JoFitz: You have etcher to write the image and all that mess right? Mar 28 20:31:09 and then once we have that just run the same commands as before and we should then be able to connect to the wifi correct? Mar 28 20:31:19 yes i have etcher Mar 28 20:31:28 Good and 7-zip? Mar 28 20:31:35 yup all that good stuff Mar 28 20:31:38 Or whatever compression tool. Mar 28 20:31:41 Okay. Mar 28 20:32:26 Yea. Just use that image and use the services, agent on, connect, passphrase. do not forget to copy your wifi connection entirely. Mar 28 20:33:41 one more thing that we think is an issue is that in order to connect to our university wifi we also need to enter our usernames not just the password Mar 28 20:33:50 Sir...if you have to mess w/ wifi enable and wifi scan, you may. Mar 28 20:33:52 Oh. Mar 28 20:34:19 Hmmm. I am not sure how your system is set up at the University. Mar 28 20:34:31 Do they have user ids or user names per person? Mar 28 20:34:49 its usernames and passwords for everyone Mar 28 20:35:15 So? If you want to get online via your lap top, you enter a user name and password? Mar 28 20:35:26 Or just password? Mar 28 20:35:33 both user and password Mar 28 20:35:47 its WPA2-Enterprise Mar 28 20:35:53 the wifi that is Mar 28 20:35:57 Hmmm. This is a complication. Let me look up something real quickly. Okay. Mar 28 20:37:09 https://01.org/connman/documentation is the site about connman. There are other sites but I do not have an enterprise set up w/ multiple points/servers. Mar 28 20:37:15 ... Mar 28 20:37:33 Let me read on what to do. That site will bring you to an intel place w/ connman as the subject. Mar 28 20:39:15 https://git.kernel.org/pub/scm/network/connman/connman.git/tree/README Mar 28 20:39:21 Read that and I will too. Mar 28 20:40:52 connman.net too. It says it works w/ enterprise. Mar 28 20:42:50 I read the readme. The system works w/ enterprise. I a trying to find where it works and how but I came up short. Mar 28 20:44:19 Haha yea its all good, my friend and I are going to keep doing some digging but I really appreciate the help Mar 28 20:44:57 Hey...you may have to compile it. Mar 28 20:45:11 I do not know why it is not working out of the box. Mar 28 20:45:23 compile what? Mar 28 20:45:45 https://git.kernel.org/pub/scm/network/connman/connman.git/about/ is where you can read about connman (connection manager). Mar 28 20:45:55 and how to compile it. Mar 28 20:46:23 Damn. Sorry. I thought I could help more. Mar 28 20:46:52 So. Did you get the updated kernel and image? Mar 28 20:48:18 It seems to be more complicated because we've connected to our university wifi using a raspberry pi, the weird thing is that our beaglebone green wireless images have the same file structure bc its debian but it dosent have the same config files for wpa_supplicant Mar 28 20:48:38 Okay. Mar 28 20:48:57 You are connected w/ a usb to usb/computer right? Mar 28 20:48:59 We're not to sure if we're going to change images right now just because it takes a decent amount of time to do ~25 mins or so for us Mar 28 20:49:06 yea usb to computer Mar 28 20:49:10 Okay. Mar 28 20:49:12 No issue. Mar 28 20:49:34 If you have time, try the updated image 4.9.x and then use those commands. Mar 28 20:49:42 If not, Godspeed. Mar 28 20:50:01 I had updated and I have not looked back. Mar 28 20:50:38