**** BEGIN LOGGING AT Wed Sep 18 02:59:58 2019 Sep 18 06:59:17 I am using beaglebone black wireless, kernel 4.14, Debian 9.5. I want to know how to connect to wifi network programmatically, I meant using shell script. I am building an UI to accept SSID, Password and connecting it using script. Can anyone pls help me on how to do that. I tried few ways but nothing worked. Sep 18 08:41:09 Raj37: depends entirely on what network manager you use. the default images use connman but I don't really know anything about it. we use systemd-networkd which just leaves wifi-management entirely to wpa-supplicant (which you can configure via a config file or at runtime via dbus or a control socket) Sep 18 10:17:26 Thanks zmatt. I know in general but could not find the detailed information. Did you come across anything like that in the entire beaglebone life cycle. I am printing the links which I tried before for reference. Sep 18 10:17:28 https://superuser.com/questions/1453120/connect-to-wifi-on-boot-beagle-bone-black-ioctlsiocsiwencodeext-invalid-argum Sep 18 10:17:41 https://github.com/Doodle3D/connman-api Sep 18 11:06:30 mm Sep 18 11:56:31 kremlin 2 m's instead of the normal 1? Sep 18 13:07:11 oops Sep 18 13:34:43 rcn-ee[m]: looks like you u-boot patch for Wandboards is broken with uboot >=2019 .. fyi Sep 18 13:36:22 rcn-ee[m]: two hunks fail :) one is autoboot stuff, the other is include/config/wandboard.h Sep 18 13:56:36 veremitz: wand usually has lots of churn, i'm only shipping v2019.04 today.. thought i saw something where u-boot added a 3rd mmc name optoin.. Sep 18 14:01:09 rcn-ee[m]: :( Sep 18 14:01:32 rcn-ee[m]: you should rejoin us in #wandboard .. its quiet there anyway :D Sep 18 14:03:02 rcn-ee[m]: ah I see 2019.07 is available Sep 18 14:03:26 rcn-ee[m]: I could take a punt and see where the issue arose .. Sep 18 14:07:12 rcn-ee[m]: do you have that patch in git somewhere? Sep 18 14:08:03 here is v2019.07-rc4: https://github.com/eewiki/u-boot-patches/blob/master/v2019.07-rc4/0001-wandboard-uEnv.txt-bootz-n-fixes.patch Sep 18 14:08:13 oh there it is :D Sep 18 14:08:22 rcn-ee[m]: thanks :D Sep 18 14:09:54 rcn-ee[m]: I'll check that and ACK it for .07-release Sep 18 14:25:59 rcn-ee[m]: yeah NACK on .07 release.. I'll see if I can do a bisect Sep 18 14:38:43 rcn-ee[m]: apologies, bad patch my end . looks like -rc4 patch applies OK to 0.7-release, fyi :D Sep 18 14:39:21 rcn-ee[m]: a few lines offset, so should be easy to update :) Sep 18 14:40:12 rcn-ee[m]: also means you can bump the wiki :D Sep 18 15:56:56 any of you fine folks here "dual-booted" a system by chainloading grub2 from uboot ? Sep 18 15:58:51 i'd go with one of the grub-efi implementations; should generally work with recent upstream u-boot versions Sep 18 16:01:27 vagrantc: cool thanks Sep 18 16:01:48 with arbitrary caveats, of course :) Sep 18 16:04:12 ah there's a good paper from suse .. Sep 18 16:05:02 bit old at 2017 but still Sep 18 16:06:40 it's been very active; every u-boot release the EFI implementation improves Sep 18 16:06:57 https://forum.odroid.com/viewtopic.php?t=26894 looks good too Sep 18 17:06:57 It seems pastebin is down for now. Sep 18 17:07:26 refresh worked. Sep 18 17:07:28 Forget it. Sep 18 18:28:14 vagrantc: does u-boot efi auto load the *.dtb yet? last i looked at it (a year ago on bbb's) you had to do it yourself? Sep 18 18:29:17 rcn-ee[m]: you either use the device-tree that u-boot uses, or you have to expressly load it yourself, yeah Sep 18 18:33:31 efi... because u-boot isn't quite bloated enough yet Sep 18 18:36:32 vagrantc: i just thought it was odd... grub-efi on top of u-boot, so you can select a kernel version, but it doesn't auto pull in the matching *.dtb.. it's like they assume *.dtb is real firmware.. Sep 18 18:37:08 if it were we'd be using open firmware, not efi :P Sep 18 18:40:01 rcn-ee[m]: yes, exactly Sep 18 18:40:45 zmatt: the main attraction to distros is that it allows using grub on all the platforms Sep 18 18:41:33 rcn-ee[m]: there are a lot of people who try to force the issue thinking that device-trees should be treated as "real firmware" shipped on devices and not maintained in the kernel... Sep 18 18:41:38 oh right so there's no actual efi but the arrangement is that u-boot pretends to be just enough efi to be able to load grub-efi or something? Sep 18 18:42:49 vagrantc: ideally that would be the case, but that would also require device-tree bindings for *all* aspects of the hardware to be determined in advance Sep 18 18:43:00 we ain't there yet Sep 18 18:43:26 zmatt: yes, u-boot has a minimalist EFI implementation Sep 18 18:43:49 it's the SBBR specification, if i'm remembering correctly Sep 18 18:44:00 or EBBR? maybe EBBR Sep 18 18:44:13 one's basically for "embedded" and one for "servers" Sep 18 18:45:17 well, i think the argument is that if device-treee continues to be maintained in the kernel, we'll never get there... Sep 18 18:47:48 but to be honest this seems silly... grub exists because the capabilities of bios to select a startup device anad load the OS from it are too crappy. it is silly to have to have two stages like that Sep 18 18:48:41 I guess maybe if u-boot is trimmed down to do nothing but platform initialization and loading grub then it might make more sense Sep 18 18:50:23 vagrantc: dunno if I agree, until bindings are finalized the DT needs to be in sync with the drivers that consume it Sep 18 18:50:27 u-boot + grub-efi is more that u-boot's user interface is limited, grub is known.. if u-boot just understood grub syntax or had a similar interface.. Sep 18 18:50:37 so a big hack ;) Sep 18 18:54:26 heh Sep 18 18:55:22 the boot stack is built mostly out of hacks with a few more hacks to hold it together Sep 18 18:55:46 i still think it would worthwhile doing dtb overlays in grub-efi, imageing having a menu in grub allowing you select specific overlays.. and being able to modify it on every bootup.. Sep 18 18:56:06 sure Sep 18 18:58:00 when exactly would you want to do that instead of using a config file? Sep 18 18:58:24 especially since it means that if you reboot your change will be gone, whereas a reboot isn't very likely to cause a physical hardware change :P Sep 18 18:59:14 to test between three different overlays? Sep 18 18:59:29 plus you'd need to access the serial console to use it, which is already a lot less convenient than using a config file Sep 18 19:00:02 I guess it might be convenient while working on an overlay in case it fucks up boot Sep 18 19:00:35 exactly Sep 18 19:01:09 but is grub really going to make that easier? Sep 18 19:53:51 @zmatt: Will it be possible to use the N-Channel MOSFET for the transistor in the diagram? Sep 18 19:54:54 I also have a BJT! Sep 18 19:55:05 NPN! Sep 18 20:11:33 Up, up, and, away. bbl. Sep 18 21:37:38 set_ what diagram? :D if you use a BJT put a current limiting resistor on the base (1-2k) if you are driving it from IC. You can blow either end of the connection if you aren't careful (IE the BJT Or the IC output) Sep 18 22:02:30 larger value is probably fine too, like 10K Sep 18 22:03:08 this is the diagram he's referring to: https://photos.app.goo.gl/ESZvoHb6eZY1hwi77 Sep 18 22:03:31 (to invert a signal if you don't have an inverter around) Sep 18 22:05:17 I'd suggest 10K as value for both resistors but there's probably a fairly wide range that will work Sep 18 22:07:27 The classic BJT inverter. Sep 18 23:51:50 quick check if the live chat is working through the web interface Sep 18 23:52:08 :-D Sep 19 00:21:00 you'd think one could see it is working without typing that Sep 19 00:27:04 ERROR Sep 19 00:28:42 "Press the Any Key to continue..." Sep 19 00:29:08 Bad or missing keyboard. Press enter to continue. Sep 19 00:29:30 "Keyboard error or keyboard not present. Press F2 to continue..." ;) Sep 19 00:29:46 I always loved that one. Sep 19 00:47:49 "Classic BJT inverter" is a thing? Dang! Sep 19 00:48:41 100mA BJT, NPN, 45v as my source of this "classic BJT inverter." Sep 19 00:52:08 yeah probably any random NPN BJT will do unless you'd have a really weird one Sep 19 00:54:15 set_: yeah this is just how you make an inverter in resistor-transistor logic (which is the oldest type of digital logic) Sep 19 00:55:49 Dang! Sep 19 00:56:01 I just learned what my book on circuits did not teach me. Sep 19 00:56:27 I could have skipped it or not made it that far yet. WHo knows? Sep 19 00:57:43 @zmatt: I have a three pin BJT (BC547BTA). Sep 19 00:58:13 BJTs have three pins yes Sep 19 00:58:57 https://www.digikey.com/short/pffjc3 Sep 19 01:00:26 if I wanted to look it up I know how to use google set_ Sep 19 01:01:29 Oh! Sep 19 01:01:46 I just figured you would like a nice tiny link if you were interested. Sep 19 01:02:12 I see it takes in 5v on the emitter base voltage. Sep 19 01:02:58 That is max. Sep 19 01:03:56 I will try to make it work w/ 3.3v instead which it can. Sep 19 01:05:18 0.7 is usually BJT on-threshold .. Sep 19 01:05:43 anything a lot above that will just ensure the bjt saturates (which doesn't matter for a simple switch) Sep 19 01:05:46 set_: there's no "try to make it work", the schematic as given will work Sep 19 01:06:26 zmatt: "do, or do not; there is no 'try' .. " - Yoda. Sep 19 01:06:34 Oh? Yea boy! Sep 19 01:07:22 I will use the BJT as, alongside the resistors, it is described in the schematic. Sep 19 01:07:43 set_: these are very basic circuits, are well-tested :p Sep 19 01:07:55 otherwise, our chips would never work ;) Sep 19 01:08:07 Confuscious says, "Man w/ hand in pocket..." Sep 19 01:08:13 "Feels cocky all day." Sep 19 01:08:19 *groan* Sep 19 01:08:20 Sorry. Sep 19 01:08:24 I saw Yoda. Sep 19 01:08:37 Confuscious and Yoda would have been friends. Sep 19 01:09:03 I Suspect they would have met in the bar periodically .. Sep 19 01:09:28 or The Restaurant [at the end of the Universe] :p Sep 19 01:09:47 basic circuitry eludes me to this day. Oh and to that point, yes, they would have met in the "universal bar." Sep 19 01:10:16 Now, I have a circuit for making inverted signals on basic machinery. Yea boy! Sep 19 01:10:31 You guys. Come on. You guys. Sep 19 01:10:43 * veremitz -sigh- Sep 19 01:10:49 Ha! Sep 19 01:11:13 Okay. I am done w/ jokes for now. I need to collect the correct 10k resistors now. Sep 19 01:11:34 I have the BJT. I have some caps too if there was another method w/ caps. Sep 19 01:13:52 there is not Sep 19 01:14:35 Okay. Just having high hopes. Sep 19 01:14:43 notions. Sheesh. Sep 19 01:15:44 Three pin cap? Jk. Sep 19 01:17:04 I found them. Three pin capacitors. ! Sep 19 01:18:01 caps do not have three pins Sep 19 01:18:34 Please hold. Sep 19 01:18:43 I will show you. Sep 19 01:20:07 https://mou.sr/2V7DzuT Sep 19 01:23:35 I think one makes a stator while one makes a another section for starting motors. Sep 19 01:23:37 I mean, you can add an extra pin if you really want to (for mechanical stability most likely) but I've never seen them in real life Sep 19 01:23:41 And one is GND> Sep 19 01:24:02 I have not seen them either. Just online. Sep 19 01:24:33 rotar and stator legs and then gnd. Sep 19 01:24:47 sorry you're now just using random words Sep 19 01:24:54 Please hold. Sep 19 01:25:23 a caoacitor has two terminals, these things just have an additional pin for mechanical stability (which is either connected to one of the other two pins or simply unused) Sep 19 01:25:56 that's all there's to it Sep 19 01:26:05 it's just to make it less wobbly when mounted on a pcb Sep 19 01:26:11 Oh. Sep 19 01:26:47 I found something about them used for starting motors or HVAC stuff too. Oh well. Maybe I have terrible info. Probably! Sep 19 01:27:25 I'll assume you just understood whatever it was you read Sep 19 01:27:31 *misunderstood Sep 19 01:30:21 Trimmer capacitors are what they are called. Sep 19 01:32:05 I think they are used for calibration. Sep 19 01:33:20 those are not what you linked to, and those still have only two terminals (if they have a third pin it's again just for mechanical stability) Sep 19 01:33:30 Okay. Sep 19 01:33:39 I moved on and kept searching. Sep 19 01:33:41 Sorry. Sep 19 01:33:50 they're just variable capacitors Sep 19 01:33:54 Right. Sep 19 01:34:06 Oh well. More nonsense from me again. Sep 19 01:34:10 No issue. Sep 19 01:56:37 BJT's don't care about the BE voltage except for polarity and being greater then Vf Sep 19 01:57:29 the base resistor has to be small enough so Ibe * Beta.min is >> your desired Ice Sep 19 01:57:37 otherwise, it may not work as a switch Sep 19 02:07:46 Aw. Sep 19 02:07:53 https://www.mouser.com/ProductDetail/BeagleBoard/BBONE-AI?qs=%252B6g0mu59x7IfEw1Zb81%252B%252BQ%3D%3D Sep 19 02:07:58 I found the BBB AI. Sep 19 02:08:01 Odd days. Sep 19 02:08:14 $125.00! Sep 19 02:10:26 I want to use those other cores. M4. Baby, baby. Sep 19 02:16:11 @zmatt: That dang schematic is not easy on a breadboard but doable. Here goes nothing. Sep 19 02:17:25 Okay. I have the breadboard set up (I think). Power is on! Testing software and EZ2! Sep 19 02:22:12 https://pastebin.com/nfHRiCQR is the error for this source at https://pastebin.com/HW2LFCb5. Sep 19 02:22:24 Do you see where I may be making a mistake? Sep 19 02:25:44 set_: what about the code I suggested? Sep 19 02:26:02 https://pastebin.com/RbDjLFjM Sep 19 02:26:14 it should give clearer messages Sep 19 02:26:50 and of course you should carefully verify everything is hooked up right, verify that P9.22 is configured to uart mode, etc Sep 19 02:27:55 Oh. Sep 19 02:28:08 Okay. Please hold. Sep 19 02:28:49 What is re in that software? Sep 19 02:28:58 Looking it up. Sep 19 02:29:49 module for regular expressions (which I use to parse the output from the module) Sep 19 02:29:56 Aw. Sep 19 02:29:58 Okay. Sep 19 02:30:13 I am seeing this src for the first time. I must have missed it the other day. Sep 19 02:30:19 I will check it out. Sep 19 02:30:25 oh maybe you weren't around when I wrote it Sep 19 02:30:26 Please hold if you are willing/have time. Sep 19 02:36:34 https://pastebin.com/BXDPw7Yu seems to explain something. Sep 19 02:36:51 hmm, did I mess up Sep 19 02:37:07 I have not a clue as how you came up w/ that so quickly. Sep 19 02:37:16 I have no clue. Sep 19 02:37:43 weird, this should be correct based on the pyserial docs Sep 19 02:37:46 lemme check its source code Sep 19 02:37:55 Okay. Sep 19 02:38:30 it definitely does Sep 19 02:38:49 Hey. You have port=port. Should I type in my uart2 port as /dev/ttyO2? Sep 19 02:38:51 it should be correct Sep 19 02:39:03 no, the port is specified in the example Sep 19 02:39:07 Okay. Sep 19 02:39:12 as ttyS2, which is correct Sep 19 02:39:20 ttyS2? Sep 19 02:39:31 I always thought it was ttyO2. Sep 19 02:39:43 Oh. That is for Adafruit_BBIO. Sep 19 02:39:44 Okay. Sep 19 02:39:46 Got it. Sep 19 02:39:46 it hasn't been for many years, although it still works for backward compatibility Sep 19 02:39:56 Aw. Odd days. I have not read that once. Sep 19 02:41:03 how did you install pyserial? using apt or pip3 ? Sep 19 02:41:14 I cannot remember. I think w/ pip3. Sep 19 02:41:58 I am installing python3-serial now. Sep 19 02:42:28 you can check what you've installed with pip3 using its list command, i.e. "pip3 list" Sep 19 02:43:50 oh the fuckers changed the name of the parameter Sep 19 02:43:56 I must have only used venv to install the pyserial. Sep 19 02:44:02 w/ pip3. Sep 19 02:44:32 I have pyserial 3.4 installed via pip3. Sep 19 02:45:35 okay I've updated my source code on pastebin Sep 19 02:45:37 wait Sep 19 02:45:41 Okay. Sep 19 02:45:50 there, now it's updated Sep 19 02:47:08 I just refreshed and saw nothing thus far that was changed. It is the same link? Sep 19 02:47:16 same link yes Sep 19 02:47:21 Okay. Sep 19 02:47:45 I changed self.read_until( expected=b'\r', size=5 ) to self.read_until( b'\r', size=5 ) on lines 13 and 17 Sep 19 02:47:57 I see. Sep 19 02:50:31 RuntimeError: Timeout while waiting for data Sep 19 02:50:58 well then it's not receiving any data Sep 19 02:51:07 Okay. Sep 19 02:51:17 I guess I will need to do something. Sep 19 02:51:23 I will test it. Sep 19 02:51:40 if P9.22 is configured right, that probably means you didn't hook up the hardware right... either that or something else is wrong :P Sep 19 02:51:49 Right. Sep 19 02:52:00 It might be a shady solder job. Sep 19 02:54:13 it would be a good idea to test the inverter separatly... configure P9.22 to gpio, disconnect the input of the inverted (from the TX output of your ultrasonic sensor) and confirm (e.g. using show-pins) that P9.22 is high. connect the input of the inverter to 3.3v and confirm that P9.22 is low Sep 19 02:55:28 Okay. Sep 19 02:58:39 @zmatt: The emitter is the pin that goes to GND, right? Sep 19 02:58:49 yes Sep 19 02:58:53 Okay. Just making sure. Sep 19 02:59:08 Aw. Sep 19 02:59:22 Too late to cut my loses but bedtime is near. Blah. Sep 19 02:59:26 Until another day! Sep 19 02:59:32 base is input of inverter (via resistor), collector connects to the output of the inverter Sep 19 03:00:03 My set up is shot. Sep 19 03:00:06 I just got it. Sep 19 03:01:00 I can show you a photo of this terrible set up I have but your brain might go into frenzy. Sep 19 03:01:07 poweroff Sep 19 03:01:08 Dang it. **** ENDING LOGGING AT Thu Sep 19 03:01:42 2019