**** BEGIN LOGGING AT Thu Feb 28 02:59:57 2019 Feb 28 03:00:07 brb Feb 28 03:00:09 Sorry. Feb 28 03:09:15 generally you should use whatever kernel series is the default in current beagleboard.org images Feb 28 03:09:21 which right now is the 4.14-ti series Feb 28 03:25:28 Okay. Feb 28 03:43:11 zmatt: yes, was working from memory - the mcp23s17. Now that I see the documentation, this too makes understand the system much easier. Feb 28 04:30:40 So does the AM5729 do CAN-FD or what? :P Feb 28 04:45:55 no Feb 28 04:46:11 it has two DCAN instances, like the beaglebone Feb 28 04:47:59 there are more recent variants of the SoC, namely DRA7xxP, that have an MCAN instance which does support CAN-FD Feb 28 04:48:47 AM574x has it too Feb 28 05:40:21 I wish there was a PWM interface like the new userspace GPIO interface :\ Feb 28 05:40:52 you want the pwm interface to suck like the new gpio interface? why? Feb 28 05:42:04 interacting via FDs and well defined messages, locking the devices by opening them, and having a kernel-level assurance that the devices will reset when my process closes the FDs or is terminated Feb 28 05:45:36 I run heater mosfets via the PWMs, so it's nice to have some idea what will happen if my process suddenly stops existing :) Feb 28 05:46:03 most of these seem of dubious benefit, especially considering the loss of functionality (the ability to declare named gpios in DT with ability to constrain direction, applications don't have to know or care about gpio numbers, access control per individual gpio) Feb 28 05:46:03 the watchdog helps, but resetting the entire board is a bit of a brute force answer Feb 28 05:46:27 add an ExecStopPost action to your systemd service that resets the state Feb 28 05:46:43 then whenever your process crashes, it'll get cleaned up after it Feb 28 05:46:47 (or otherwise exits) Feb 28 05:46:47 TBH, I see the numbering change as a net benefit Feb 28 05:46:57 I really don't Feb 28 05:47:06 the numbers we use today were never actually stable anyway Feb 28 05:47:13 correct, I don't rely on those either Feb 28 05:47:24 and I believe you can still trace from the name in a DT to a pin/chip in sysfs Feb 28 05:48:17 the names are specific to the sysfs interface (they are provided when they're exported to sysfs by gpio-of-helper) Feb 28 05:48:53 and gpios are typically placed on "leftover" pins, and may be moved around if those pins turn out to be needed for a more specific function Feb 28 05:49:12 by naming them in DT applications don't need to know or worry about that Feb 28 05:50:01 more importantly, by defining in DT whether pins are input or output I ensure that unprivileged users can't change a pin's direction from input to output (which could easily damage hardware) Feb 28 05:50:21 actually I'm not sure whether that's more important, but I do care about it Feb 28 05:50:48 I believe querying the GPIOs via the new API can return the same names, no? Feb 28 05:51:08 if it can return names, I have no idea from where it obtains them Feb 28 05:51:20 check out `gpioinfo` Feb 28 05:51:30 and https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/tools/gpio/lsgpio.c?h=v5.0-rc8 Feb 28 05:51:54 -bash: gpioinfo: command not found Feb 28 05:51:55 :P Feb 28 05:52:07 ah, it's in RCN's Ubuntu 18.04 images Feb 28 05:52:17 but yea, it looks like the new API returns both "name" and "consumer" Feb 28 05:52:27 and consumer appears to map to the names in the DTB Feb 28 05:53:23 from where? the only time I've seen per-gpio nodes in DT is as part of gpio-of-helper, which is specific to the sysfs interface Feb 28 05:54:33 regardless, even if I could name gpios, userspace would still need to open every gpiochip to find the pins it needs, and therefore have access to all gpios Feb 28 05:54:37 that's not acceptable Feb 28 05:55:22 lol Feb 28 05:55:23 right now my application simply opens e.g. /dev/gpio/dsp-reset for the reset line to the dsp on our board Feb 28 05:55:27 that's about 20 lines of C Feb 28 05:56:07 I'm happy to have more well-defined behavior and locking Feb 28 05:56:12 I'm not saying it is hard, I'm saying it is a serious step back and a security problem Feb 28 05:56:26 what's not "well-defined" about the sysfs gpio interface? Feb 28 05:56:43 an exported GPIO is exported to anything in the system that cares to poke at it Feb 28 05:56:54 and no one knows what will happen if your process terminates Feb 28 05:57:21 I do, exactly nothing happens to the gpio Feb 28 05:57:24 and you also have to probe to figure out if the GPIO you want has already been exported or not because it doesn't un-export when your process dies Feb 28 05:57:48 if you define your gpios in DT they are automatically exported at boot and cannot be unexported Feb 28 05:57:49 see, I find exactly nothing to be a poor behavior Feb 28 05:57:57 I don't, it's usually what I want Feb 28 05:58:07 and if I want something else, that's not hard to arrange Feb 28 05:58:17 I'm quite happy with the idea that when my process isn't around anymore to ask that this GPIO be an output held high, it goes back to being unconfigured and unused Feb 28 05:58:27 what does that mean, "unconfigured" ? Feb 28 05:58:51 reset to the default initialization state? Feb 28 05:58:56 which is? Feb 28 05:59:06 where it that defined? Feb 28 05:59:16 a safe default that the kernel driver presumably has built for me Feb 28 05:59:18 with gpio-of-helper I define the initialization state in DT Feb 28 06:00:09 for most of my gpio applications there's no "safe default", and the kernel better not fucking touch them unless I tell it to Feb 28 06:00:36 because specific sequencing needs to be done Feb 28 06:01:07 then you've of course written a kernel driver to ensure your hardware can be driven according to precise specifications, right? Feb 28 06:01:25 most definitely not, it would be completely insane to do that in the kernel Feb 28 06:02:17 well no, I guess it could be done, but I'm not remotely inclined to do things in kernelspace that can be done easily in userspace with no benefit to doing them in kernelspace Feb 28 06:04:50 I understand that it can be convenient to reset gpios to fixed values if your application exits for whatever reason, but that's easy to arrange in any case (like I said, with an ExecStopPost action that performs whatever actions are needed to reset things to a safe state), while the problems of the new interface are hard to work around Feb 28 06:06:00 locking seems like a weird thing to want (what on earth are you doing to have multiple processes access the same gpio??) but if you want to you can just lock the value attribute in sysfs Feb 28 06:06:54 also it sounds very annoying for debugging, I often access gpios manually via sysfs while they're "in use" by a service Feb 28 15:19:17 rcn-ee, you told me two days ago that you had luck with bcmdhd driver but not with brcmfmac one. Can you tell me if I should be looking at the driver from android kernel, please? I don't see this one in linux kernel. Feb 28 15:21:41 that driver is not publicly released, there's snapshots floating around, you need to bug the supplier of the module.. (it's really a shitty situation) Feb 28 15:22:02 I've tried basically each firmware file I could fine and each time I get "HT Avail timeout" error. It seems that driver is loading the firmware file and after that it isn't able to configure HT clock. I gathered all the data here: http://pastebin.com/raw/KTvX7BZY. Feb 28 15:22:20 Right, I understand, I'll try that. Many thanks Feb 28 15:23:00 td-kros, i see that same error on mainline with imx6 (wandboards) Feb 28 15:24:55 That's odd, I thought Cypress supports nxp well, at least all tutorials I found are for NXP Feb 28 15:25:54 I was trying to use CYW4343 but now flyleading Murata 1DX module Feb 28 15:26:32 if this would work we'll need to remake PCBs with Murata module Feb 28 15:27:35 if you need to remake, just use the wl18xx, while it's spendy, we know it works.. Feb 28 15:30:30 I'll check if we can fit it :-) (https://imgur.com/d03Apea) Feb 28 15:38:27 ah good point.. yeah nope.. Feb 28 15:40:10 Okay, many thanks for the hint regarding bcmdhd driver, still I have a few things to try. Feb 28 16:10:56 m Feb 28 18:29:17 Hello, has there been issues w/ Python3 and serial.Serial not accepting the fact that serial has a module named Serial on the 4.14.x kernel? Feb 28 18:29:35 Sorry. module = attribute Feb 28 18:31:07 no? that sounds like you simply didn't install pyserial ("sudo apt install python3-serial") Feb 28 18:31:16 Oh! Feb 28 18:31:24 I used pip instead. I got what happened now. Feb 28 18:31:32 pip3 install pyserial should work too yes Feb 28 18:31:42 Let me try w/ apt and see what happens. Feb 28 18:31:47 I already used pip. Feb 28 18:31:52 then you used it wrong Feb 28 18:31:54 sorry...pip3. Feb 28 18:32:25 This line gives errors: ser=serial.Serial("/dev/ttyO2", 9600) Feb 28 18:32:31 I will check w/ apt. Feb 28 18:32:53 wait, so you can "import serial" but not execute serial.Serial? Feb 28 18:33:11 Right. Feb 28 18:33:56 what is the _exactly_ error you get? Feb 28 18:34:00 *exact Feb 28 18:34:04 Okay. Please hold. Feb 28 18:34:46 if I do "pip3 install pyserial" and then in python3 "import serial" I definitely do have a class named serial.Serial Feb 28 18:35:07 File "GPS.py", line 7, in ser=serial.Serial("/dev/ttyO2", 9600) AttributeError: module 'serial' has not attribute 'Serial' Feb 28 18:35:08 ... Feb 28 18:35:22 Okay. Let me check that too. Feb 28 18:35:25 where does "serial" come from? Feb 28 18:35:33 what's the import you're using? Feb 28 18:35:38 import serial Feb 28 18:35:41 wtf Feb 28 18:35:44 Yep. Feb 28 18:36:19 python3 -c 'import serial; print(serial.__file__)' Feb 28 18:36:21 what does that print? Feb 28 18:36:24 I looked in the python.org pages online. I saw where serial.Serial was located. Feb 28 18:36:27 Okay. Let me check. Feb 28 18:36:29 set_: How did you name your script file? Feb 28 18:36:37 GPS.py Feb 28 18:36:50 it sounds like it's finding the wrong "serial" module somewhere Feb 28 18:37:14 print serial.__file__ Feb 28 18:37:19 that's what I just said Feb 28 18:37:21 or well with brackets.. Feb 28 18:37:24 also that's a syntax error Feb 28 18:37:24 yeah Feb 28 18:37:38 parentheses Feb 28 18:37:48 ... /usr/local/lib/python3.5/dist-packages/serial/__init__.py is the print out from that cmd, zmatt. Feb 28 18:37:49 missed that line (I hate py3 for that brackets) Feb 28 18:38:12 Humpelst1lzchen: why? every other function call in python uses parens Feb 28 18:38:20 Okay. Feb 28 18:38:29 set_: ok that looks fine Feb 28 18:38:33 Okay. Feb 28 18:38:45 python3 -c 'import serial; print(serial.Serial)' Feb 28 18:38:53 Okay. Feb 28 18:39:04 zmatt: but I quickly need to write a lot of prints when I have a problem Feb 28 18:39:28 Humpelst1lzchen: be glad you're not programming in LISP ;) Feb 28 18:39:39 File "", line 1, in Feb 28 18:39:39 AttributeError: module 'serial' has no attribute 'Serial' Feb 28 18:39:52 what the f is this serial package you have installed Feb 28 18:40:06 I installed pyserial. Feb 28 18:40:15 w/ pip3. Feb 28 18:40:23 you have such bizarre problems Feb 28 18:40:27 I know! Feb 28 18:40:40 Well. Onto another image I guess. Feb 28 18:40:49 what? Feb 28 18:40:59 this isn't a problem with the image Feb 28 18:41:29 I know but... Feb 28 18:41:46 I have two separate images on the eMMC and SD card. Feb 28 18:41:59 It could be that old issue w/ the bootloader. Feb 28 18:42:05 what? no Feb 28 18:42:08 Oh? Feb 28 18:42:13 Okay. Feb 28 18:42:16 this has nothing to do with a bootloader or a kernel or anything like that Feb 28 18:42:21 What do you think I should look for? Feb 28 18:42:23 this is something weird going on with a python module Feb 28 18:42:28 Oh. Feb 28 18:42:42 set_: can you pastebin /usr/local/lib/python3.5/dist-packages/serial/__init__.py Feb 28 18:42:53 I just put on a new image and put my old files on the machine BBGW. Feb 28 18:42:59 wait, usr local? that's not something you installed with pip3 Feb 28 18:43:03 pip3 can't install there Feb 28 18:43:07 Okay. Feb 28 18:43:14 so do "pip3 install pyserial" Feb 28 18:43:59 if necessary do "pip3 install -I pyserial" to force it Feb 28 18:44:18 Requirement already satisfied: pyserial in /usr/local/lib/python3.5/dist-packages Feb 28 18:44:20 (i.e. if "pip3 install pyserial" says it has nothing to do) Feb 28 18:44:23 yeah that Feb 28 18:44:31 Okay. Feb 28 18:44:34 the -I flag should fix that Feb 28 18:45:10 testing w/ -I flag now. Feb 28 18:45:26 just to be sure: never use "sudo pip3", always just "pip3" on your normal debian account Feb 28 18:45:35 Oh? Feb 28 18:45:50 I always thought pip3 would give permission issues when installing. Feb 28 18:45:59 no? Feb 28 18:46:06 I will make sure not to use sudo w/ pip3 now. Feb 28 18:46:19 and I've told you before that using "sudo" with "pip" or "pip3" is always a mistake Feb 28 18:46:24 I've in fact told you this very recently Feb 28 18:46:48 I got all kinds of errors now. Feb 28 18:47:24 I'm not sure I even want to try to debug them, you probably just made a mess of your system again Feb 28 18:47:34 Heh? Feb 28 18:47:35 I can just do "pip3 install pyserial" and it installs without a problem, and it works Feb 28 18:47:42 Okay. No issue. Feb 28 18:47:45 I will keep testing. Feb 28 18:47:46 Thank you. Feb 28 18:47:52 maybe just reflash again :P Feb 28 18:48:23 Yep. I might have to. Feb 28 18:48:32 This software worked before I flashed last night. Feb 28 18:51:02 lemme double-check on the stretch-iot image (which has an older python than what I'm normally using I think) Feb 28 18:53:24 Okay. Feb 28 18:53:53 yeah it installs just fine: https://pastebin.com/raw/2SHzgqZU Feb 28 18:54:03 Humpelst1lzchen: The __init__.py file has some extra info. in it. What are you looking for in that file? Feb 28 18:54:07 The first couple of lines? Feb 28 18:54:20 the entire file probably Feb 28 18:54:49 Okay. Feb 28 18:58:07 https://pastebin.com/bXDrPw7Y is the __init__.py file, Humpelst1lzchen. Feb 28 18:58:59 ok, that one looks good Feb 28 18:59:12 Okay. Feb 28 18:59:13 indeed Feb 28 18:59:17 Yea boy! Feb 28 18:59:46 except there's literally no way for there to be no Serial attribute Feb 28 19:00:18 That is on my side of things, I guess. Feb 28 19:00:20 if you look at lines 21-33, every possible path either leads to the import of a "Serial" from somewhere, or an exception is raised Feb 28 19:00:21 As usual. Feb 28 19:00:37 so congratulations, you're getting an error that's logically impossible to get :D Feb 28 19:00:50 I just added this image to the board. Feb 28 19:00:55 Last night. Feb 28 19:01:06 which image is this? Feb 28 19:01:28 Linux beaglebone 4.14.71-ti-r80 #1 SMP PREEMPT Fri Oct 5 23:50:11 UTC 2018 armv7l GNU/Linux Feb 28 19:01:36 that's not an image Feb 28 19:01:42 OH. Feb 28 19:01:45 Please hold. Feb 28 19:01:58 the kernel version is irrelevant for all this Feb 28 19:02:10 BeagleBoard.org Debian Image 2018-10-07 Feb 28 19:02:16 okay so that should be fine Feb 28 19:02:32 That is the IoT image from the main site, beagleboard.org. Feb 28 19:02:53 yeah, it's the right image. what errors did you get when trying to use pip3 install -I pyserial ? (without sudo) Feb 28 19:03:05 None. Feb 28 19:03:13 It installed. Feb 28 19:03:19 Again. Feb 28 19:03:30 I think I should erase and purge them. Feb 28 19:03:32 19:46 < set_> I got all kinds of errors now. Feb 28 19:03:41 Oh...that was w/ running the software. Feb 28 19:03:51 Not w/ installing via pip3 w/out sudo. Feb 28 19:03:51 python3 -c 'import serial; print(serial.Serial)' Feb 28 19:03:56 what does that say? Feb 28 19:03:58 I just rebooted. Please hold. Feb 28 19:05:00 Feb 28 19:05:09 okay so your Serial problem has been fixed indeed Feb 28 19:05:16 Yea boy! Feb 28 19:05:17 I still don't understand why you had it, but it's been fixed Feb 28 19:05:30 Hmmm. Something we may never know. Feb 28 19:05:46 and what did we learn? do not use "sudo pip3" Feb 28 19:06:04 not to use sudo w/ pip3 install ! Feb 28 19:06:28 normally you wouldn't need the -I flag either, that was just to work around the mess you created by previously using sudo pip3 Feb 28 19:06:39 My software I found online now is attrociously f'ed. Feb 28 19:06:48 assert not ('sudo' in cmd and 'pip' in cmd) Feb 28 19:06:51 Oh. Feb 28 19:07:10 I guess things need to change. Feb 28 19:08:15 function sudo () { [[ "$*" == *pip* ]] && { echo 'no !' >&2; return 1; } || command sudo "$@"; } Feb 28 19:10:57 I had to use apt install python-serial and things are working now. Feb 28 19:11:09 that's python2, not python3 Feb 28 19:11:18 don't use python2 Feb 28 19:11:32 python2 is obsolete and unmaintained Feb 28 19:11:57 Okay. Feb 28 19:12:37 I think this python-serial module is making my GPS.py file work for now. I will check again. Feb 28 19:12:46 that means you're using python2 Feb 28 19:13:24 I know. Feb 28 19:13:36 But...I cannot make the dang things work yet w/ Python3. Feb 28 19:13:36 any debian package that starts with python- is for python2, python3 packages start with python3- Feb 28 19:13:43 I understand. Feb 28 19:13:52 then just fix whatever minor issues it has when using python3 Feb 28 19:13:56 what errors are you getting? Feb 28 19:14:01 where did you get this GPS.py from? Feb 28 19:14:08 toptechboy.com. Feb 28 19:14:15 link to the file? Feb 28 19:14:17 Some guy that likes GPS. Feb 28 19:14:18 Okay. Feb 28 19:15:08 https://pastebin.com/0ZKdMjjN is the toptechboy.com file from that fellow. Feb 28 19:15:19 I changed some but it is mostly his stuff. Feb 28 19:16:59 I don't see a huge amount of obvious problems... apart from trying to write to the hardcoded path "/root/GPS_data/GPS.txt" Feb 28 19:17:22 https://pastebin.com/HmBS6VXh is the issue w/ python3. Feb 28 19:17:25 there might be a bytes vs str issue Feb 28 19:17:33 Yes...w/ python3. Feb 28 19:17:35 ah yeah Feb 28 19:17:52 it works w/ 2 but not w/ Python3 yet b/c of that issue. Feb 28 19:18:33 Should I change serialutil.py? Feb 28 19:18:42 so, every string that's written to the serial port (the "$PMTK..." ones), change them to use b"" instead of "" Feb 28 19:18:45 no Feb 28 19:19:04 the problem isn't in the serial module, the problem is being reported *by* the serial module Feb 28 19:19:20 Okay. So, I need to try b"" instead of just ""? Feb 28 19:19:41 yeah, so change e.g. "$PMTK220,10000*2F\r\n" to b"$PMTK220,10000*2F\r\n" Feb 28 19:19:55 Okay. B/c they need to be bytes? Feb 28 19:20:00 Not unicode? Feb 28 19:20:11 yes, a serial port expects a sequence of bytes Feb 28 19:20:20 not a unicode string Feb 28 19:20:22 Okay. Feb 28 19:20:23 Got it. Feb 28 19:21:07 the next problem is not being able to open the path "/root/GPS_data/GPS.txt", which you should just change to "GPS.txt" or whatever Feb 28 19:21:34 why the fuck does it open and close it twice o.O Feb 28 19:22:00 I do not know. I may have changed something to work b/c things were not working as is. Feb 28 19:22:03 for the same reason people hit ctrl-S five times maybe Feb 28 19:22:34 Well...it opens to write and read. Feb 28 19:22:35 nah it's actually writing it in a loop, but I don't understand why it's closing every time instead of just flushing it Feb 28 19:22:41 it never reads from it Feb 28 19:22:45 Oh. Feb 28 19:22:47 "a"? Feb 28 19:22:54 append Feb 28 19:22:54 Aw! Feb 28 19:22:55 Sheesh. Feb 28 19:23:41 oh god, what the fuck is this doing Feb 28 19:23:46 What? Feb 28 19:24:03 it's spin-waiting for data with 100% cpu load Feb 28 19:24:09 Yikes! Feb 28 19:24:20 Maybe it was meant to fail? Feb 28 19:24:39 it's just written horribly Feb 28 19:25:49 So, does everything w/ "" need to be w/ b"" now? Feb 28 19:26:31 So, like if NMEA1_array[0] == "GPRMC":? Feb 28 19:27:07 some stuff there will need to be fixed too indeed Feb 28 19:28:04 Yea. I see where I am receiving error codes on line 53 & 86. Feb 28 19:28:10 Off to it! Feb 28 19:28:12 jeez what is this hideous cargo-culting Feb 28 19:28:17 like calling flushInput twice and such Feb 28 19:28:43 I think it was for two ways of baud rates. Feb 28 19:28:46 I just one though. Feb 28 19:28:53 ? Feb 28 19:28:55 I just use one though. Feb 28 19:29:13 also, flushing input from a serial port is a great way to lose sync Feb 28 19:29:29 I wonder what the data received actually looks like Feb 28 19:29:44 9600...this used to be 57600 to start and then it would get changed somewhere along the software path. Feb 28 19:29:55 On line 16 or something. Feb 28 19:30:12 I will show you. Feb 28 19:30:14 Please hold. Feb 28 19:33:43 set_: could you try https://pastebin.com/LL435Va9 Feb 28 19:33:50 just to see what the GPS module is producing Feb 28 19:35:30 Okay. Please hold. Here is what I would put in a kml wrapper: -xx.xxxxxxxxx,xx.xxxxxxxx,x.x and etc. Feb 28 19:35:51 ? Feb 28 19:37:34 Yea. It is a bunch of numerial values for a kml wrapper to put into Google Earth. Feb 28 19:37:56 So, I was using this software in Python2 to track myself on bike, by foot, and so on. Feb 28 19:38:04 I don't understand what you're saying to me or why, or how it related to anything I said Feb 28 19:38:40 Okay. Feb 28 19:38:55 and as usual, it makes trying to help you a rather energ-draining activity Feb 28 19:39:16 I thought you wanted to see what those sentences looked like on my end. Feb 28 19:39:51 "I wonder what the data received actually looks like" Feb 28 19:40:00 zmatt: Okay. Feb 28 19:40:05 I did, but instead of providing the output of the script I gave, you starting talking about what you use this data for Feb 28 19:40:08 I am receiving sentences on my end. Feb 28 19:40:10 *started Feb 28 19:40:22 I got you. Feb 28 19:40:24 Sorry. Feb 28 19:40:52 I also wonder why you're trying to do this with custom python code rather than use existing software like gpsd Feb 28 19:41:02 zmatt: I tried out that example. Feb 28 19:41:08 It works. gpsd? Feb 28 19:41:10 What is that? Feb 28 19:41:44 a service that parses the raw data and makes it available for any application that cares Feb 28 19:41:48 with python bindings available Feb 28 19:41:50 Oh. Feb 28 19:41:53 Oh! Feb 28 19:41:59 I will read about it later. Feb 28 19:42:20 basically it's what every linux system that cares about gps data (including every android phone) uses Feb 28 19:42:30 Oh. I jotted it down. Feb 28 19:43:08 So, w/ that example you gave me, I have, what I guess, NMEA sentences spewing on my terminal. Feb 28 19:43:16 Is that it? Feb 28 19:43:28 Or, should I keep trying to look up gpsd? Feb 28 19:43:42 yes, but now that I think about it, I don't actually want to waste time on parsing that when you should actually just be using gpsd Feb 28 19:44:00 I mean...thank you sir but I guess I will listen to whatever whenever. Feb 28 19:44:06 gpsd! Feb 28 19:44:24 you can probably just install it with sudo apt install gpsd Feb 28 19:44:32 it probably needs some configuration Feb 28 19:44:40 zmatt: Serious? A new task. Cool! GPSD. Feb 28 19:44:43 Let me try. Feb 28 19:45:44 Yep. That was it. Feb 28 19:46:38 Should I search debian.org for gpsd? Feb 28 19:46:58 Or another site. Ubuntu jumped up on my search. Feb 28 19:47:12 I'm installing gpsd myself to see if it has an obvious config file or something Feb 28 19:47:25 zmatt has a new task. cool! Feb 28 19:47:42 zmatt: Okay. I will stay patient. Feb 28 19:48:28 in /etc/default/gpsd change DEVICES="" to DEVICES="/dev/ttyO2" Feb 28 19:48:39 Okay. Feb 28 19:48:41 Please hold. Feb 28 19:48:50 that would be my guess anyway Feb 28 19:48:58 my node client is updating right now. Please hold. Feb 28 19:50:58 Well. What should I do now? Feb 28 19:51:05 and then probably sudo systemctl start gpsd Feb 28 19:51:11 Oh! Feb 28 19:51:36 oh crap Feb 28 19:51:42 uhh, pin config Feb 28 19:51:43 Oh? Feb 28 19:51:49 or is this on the blue/ Feb 28 19:51:56 Oh yea. I already did that. BBGW. Feb 28 19:52:07 oh. I may have rebooted. No! Feb 28 19:52:09 BBGW, so pin config will be needed to make this work from boot Feb 28 19:52:25 best would be to enable the overlay for uart1 Feb 28 19:52:32 to ensure those pins are configured right from boot Feb 28 19:52:38 I have to use uart2 anyway. Feb 28 19:52:42 It is a Grove GPS. Feb 28 19:52:43 uart2 sorry Feb 28 19:53:10 What is wrong w/ UART2? Feb 28 19:53:21 ? Feb 28 19:53:28 I can type up a service for UART2? Feb 28 19:53:37 Oh. I just got what you were typing. Feb 28 19:53:37 nothing is wrong with it? Feb 28 19:53:43 My bad, sir. Feb 28 19:53:57 Okay. Feb 28 19:54:11 Where would I look for the files? Feb 28 19:54:19 I suggest loading /lib/firmware/BB-UART2-00A0.dtbo via /boot/uEnv.txt Feb 28 19:54:26 Okay. Feb 28 19:54:42 i.e. uncommend one of the uboot_overlay_addr* variables and change its path to the BB-UART2-00A0.dtbo Feb 28 19:54:45 *uncomment Feb 28 19:54:48 I usually just start a service w/ a .sh script. Feb 28 19:54:49 Okay. Feb 28 19:54:58 I will do what you are saying insteaad. Feb 28 19:55:01 yeah but in this case it's an existing service Feb 28 19:55:08 so it would be messy to hack in config-pin stuff Feb 28 19:55:13 Oh yea. That makes sense. Feb 28 19:56:57 I most likely need to reboot, heh? Feb 28 19:57:02 yes Feb 28 19:57:31 So, since when is this gpsd thing happened to be? Feb 28 19:57:41 Bad grammar. Sorry. Feb 28 19:57:59 Is gpsd new? Feb 28 19:58:00 ? Feb 28 19:58:32 new? no, not remotely Feb 28 19:58:46 Oh. Feb 28 19:59:19 alright. So now w/ the u-boot overlay, what should I do? I rebooted. Feb 28 19:59:31 I am signed in and ready to look around. Feb 28 19:59:41 is gpsd running? journalctl -u gpsd Feb 28 20:00:13 systemctl status gpsd Feb 28 20:00:17 -- No entries -- Feb 28 20:00:20 Please hold. Feb 28 20:00:26 maybe it needs to be enabled Feb 28 20:00:38 try sudo systemctl start gpsd Feb 28 20:01:06 Okay. That is most likely what took place. I did not enable it after starting it. Feb 28 20:01:09 if that works, and journalctl -u gpsd shows no problems, do sudo systemctl enable gpsd to make it start at boot Feb 28 20:01:21 Okay. Feb 28 20:01:27 of course having it start at boot might not be strictly needed Feb 28 20:01:40 it would autostart anyway when any program tries to use it Feb 28 20:01:51 so I guess you don't need to do systemctl enable Feb 28 20:02:08 Oh. just start it each time I sign into the Debian Distro? Feb 28 20:02:17 w/ systemctl? Feb 28 20:02:28 it should start automatically whenever any program tries to connect to it Feb 28 20:02:57 Okay. So, I have started the gpsd daemon w/ systectl. Feb 28 20:03:04 systemctl. Feb 28 20:03:17 I checked the status. it is running. Feb 28 20:03:18 anything interesting show up in journal? Feb 28 20:03:21 Good. Feb 28 20:03:25 Oh. Let me see. Feb 28 20:04:23 -- Logs begin at Thu 2019-02-28 19:57:32 UTC, end at Thu 2019-02-28 20:03:41 UTC. -- Feb 28 20:04:32 It ran for five minutes? Feb 28 20:04:39 huh, does it not use that journal tag? Feb 28 20:04:57 I tried journalctl -u gpsd Feb 28 20:05:12 maybe it just doesn't produce any log output Feb 28 20:05:17 Oh. Feb 28 20:05:31 So, someone knows right now where this GPS is positioned but I do not? Feb 28 20:05:37 No! Feb 28 20:06:01 annoyingly, it seems the package for its client utilities has dependencies on x11 :( Feb 28 20:06:08 maybe just try the python module: pip3 install gps3 Feb 28 20:06:41 Okay. I just checked to see if u-blox is supported. Yep! Feb 28 20:07:27 gps3 or gpsd? Feb 28 20:08:19 gps3 Feb 28 20:08:26 Okay. That worked. Feb 28 20:08:35 Look at me not using sudo! Feb 28 20:09:07 https://pypi.org/project/gps3/ has some examples Feb 28 20:09:57 Okay. I will goof w/ those until I can configure things. Feb 28 20:10:08 I've tried the first two, both of them work for me... except of course I get Altitude = n/a and Latitude = n/a since I don't have any gps receiver Feb 28 20:10:40 Okay. I will test real quickly. Feb 28 20:10:57 the API looks a bit inelegant to me, but if it works it works Feb 28 20:12:37 no module named 'gps3' Feb 28 20:12:42 That is my first error. Feb 28 20:12:44 No! Feb 28 20:13:04 zmatt: Do not worry about it. Feb 28 20:13:18 I will test later and again and again. I am hungry now. i need to eat. Feb 28 20:14:49 https://pastebin.com/qfc1m284 Feb 28 20:14:56 this will show all the info in that data record Feb 28 20:15:28 Okay. I will test that out right after this bunch of movements at my place. I have been sitting too long. I am going ape nuts. Feb 28 20:15:38 Thank you, sir. Mar 01 01:30:08 zmatt: Check this out! http://manpages.ubuntu.com/manpages/bionic/man8/gpsd.8.html. Nice, heh? Mar 01 01:35:34 ? it's the manpage, what about it? Mar 01 01:38:20 Nothing really. This daemon seems very, um, testing still. Mar 01 01:38:21 ... Mar 01 01:38:56 I mean, -b is for some stuff and it can create the GPS receiver to become inaccessible totally. Mar 01 01:39:07 I am officially scared. Mar 01 01:39:22 I tried the instructions. Mar 01 01:39:49 gpsd /dev/ttyO2 and nothing happened. I was going to put in a -b but then I read what it can do. Mar 01 01:40:29 Should I use -N? Mar 01 01:42:14 what? why are you doing any of this? I thought you already had it running Mar 01 01:42:25 you can't start it twice, the serial port will already be in use Mar 01 01:42:32 Oh. Mar 01 01:42:34 don't mess with any of this Mar 01 01:42:47 Fine. Why did you bring it up to me earlier in the day? Mar 01 01:42:54 Just curious! Mar 01 01:43:06 ?? I helped you to install it Mar 01 01:43:12 So, I should stick to programming the device instead of using the daemon? Mar 01 01:43:24 and it sounded like it should have been working Mar 01 01:43:24 Oh. Mar 01 01:43:24 and I even gave a python examplre Mar 01 01:43:27 I saw it. Mar 01 01:43:29 Thank you. Mar 01 01:43:33 did you try it? did it work? Mar 01 01:43:36 No. Mar 01 01:43:40 what happened? Mar 01 01:43:41 It did not work. Mar 01 01:43:45 Errors. Mar 01 01:44:11 stop saying shit like "errors". that is of absolutely zero informative value to anyone Mar 01 01:44:17 give the actual errors Mar 01 01:44:17 Okay. Mar 01 01:44:18 Fine. Mar 01 01:44:37 Please hold. Oh and I cannot config-pin my P9.22 uart pin. Mar 01 01:44:59 didn't you enable an overlay for it? Mar 01 01:45:04 OH yea! Mar 01 01:45:06 to ensure it would be configured right Mar 01 01:45:07 I forgot. Mar 01 01:45:21 Way to go, zmatt. You have an awesome memory. Mar 01 01:45:33 Let me check that it is still there. Mar 01 01:45:34 my memory is shit, but I guess yours is even worse Mar 01 01:45:43 My is "shittier." Mar 01 01:45:53 Excuse my language. Mar 01 01:46:31 Hey. Let me go and check that software you gave out. Mar 01 01:46:36 Be right back! Mar 01 01:48:18 ImportError: No module named 'gps3' Mar 01 01:48:28 this line: from gps3 import gps3 Mar 01 01:48:56 pip3 install gps3 Mar 01 01:49:28 I tried this already. I will try it again. Mar 01 01:49:51 like I said, it worked right away when I tested it Mar 01 01:50:07 Installing collected packages: gps3 Mar 01 01:50:26 Successfully installed gps3-0.33.3 Mar 01 01:50:32 Same error as before. Mar 01 01:50:41 The import error. Mar 01 01:51:35 are you accidently using python again instead of python3 ? Mar 01 01:51:43 No. I am using python3 Mar 01 01:51:47 3! Mar 01 01:52:03 python3 myGPS.py Mar 01 01:52:21 python3 -c 'from gps3 import gps3' Mar 01 01:52:26 that gives the same error? Mar 01 01:52:39 Please hold. Mar 01 01:53:04 no info and no error. Mar 01 01:53:32 so that import line simply works. you must have a typo or something in your code, or maybe a weird invisible character, I dunon Mar 01 01:53:34 I run the cmd, python3 -c 'from gps3 import gps3', and nothing happens. Mar 01 01:53:34 *dunno Mar 01 01:53:46 yeah, that mean it works fine Mar 01 01:53:49 Okay. Mar 01 01:54:17 huh what Mar 01 01:54:34 I'm getting the same error if I download my paste Mar 01 01:54:36 wtf Mar 01 01:54:41 See sir. Mar 01 01:54:54 oh never mind, I'm an idiot Mar 01 01:55:01 Oh? Mar 01 01:55:06 nah, it works Mar 01 01:55:07 No, you are smart. Mar 01 01:55:14 Help a brother help you help me. Mar 01 01:55:17 I was a moron and saved the paste as "gps3.py" Mar 01 01:55:20 Oh. Mar 01 01:55:30 which means it found itself when trying to import the module "gps3" Mar 01 01:55:34 I saved it as GPSIII.py Mar 01 01:55:35 Oh. Mar 01 01:55:39 Do you think that is my issue? Mar 01 01:55:56 no idea, but if I simply save https://pastebin.com/raw/qfc1m284 as foo.py and run it, it works here Mar 01 01:56:49 I get the same error. Mar 01 01:56:58 no module named 'gps3' Mar 01 01:57:29 Oh well. Mar 01 01:57:30 try putting this line before your "from gps3 import gps3": Mar 01 01:57:36 import gps3; print( gps3.__file__ ) Mar 01 01:57:38 Okay. Mar 01 01:58:49 same thing. no module named gps3 Mar 01 02:00:32 I mean, I don't know what to say... you confirmed that the gps3 module does in fact exist and can be found, since you ran python3 -c 'from gps3 import gps3' Mar 01 02:00:45 which is the same as putting that single line into a python script and running it with python3 Mar 01 02:00:52 There is something wrong w/ that BB-UART2-00A0.dtbs file. Mar 01 02:01:00 that has absolutely nothing to do with python Mar 01 02:01:22 you always say such completely random things Mar 01 02:01:47 I tried another file that used to work. Mar 01 02:02:11 No longer does this file work b/c of this issue: Set pin mode failed for uart channel Mar 01 02:02:42 yes? that's expected Mar 01 02:02:51 Oh. Okay. Mar 01 02:02:52 also, if gpsd is running you shouldn't try to use the uart manually Mar 01 02:03:06 I just... I don't understand what on earth you are doing Mar 01 02:03:27 you confirmed you can import gps3 just fine, and then you somehow can't Mar 01 02:03:28 What do you mean? I am trying to run some GPS files w/out issue. Mar 01 02:03:46 Oh. Yea. I have no ideas on why this may work on your end and not on mine. Mar 01 02:03:50 I am listening to you. Mar 01 02:04:00 it's not even about "on my end" vs "your end" Mar 01 02:04:07 you literally confirmed you can import gps3 Mar 01 02:04:24 I can but when the file gets run, I end up not being able to import it. Mar 01 02:04:28 I cannot explain why as of now. Mar 01 02:04:48 how exactly are you running the file? can you pastebin all relevant lines from your terminal? Mar 01 02:04:56 Sure. Mar 01 02:04:58 you must be doing *something* weird as hell Mar 01 02:04:58 Please hold. Mar 01 02:06:06 https://pastebin.com/6JmLGfQc is the input and output of my terminal when I run this software in a .py file format. Mar 01 02:06:27 why the FUCK ARE YOU USING SUDO Mar 01 02:06:48 Oops. Mar 01 02:06:55 Should I not? Mar 01 02:06:57 DO NOT RANDOMLY USE SUDO Mar 01 02:07:03 Okay. Mar 01 02:07:04 @#^&*$ Mar 01 02:07:07 Hhahahaha. Mar 01 02:07:14 I'm done with you for today, good night. Mar 01 02:07:20 Sorry man. Sir, I always thought I needed to run sudo w/ python files. Mar 01 02:07:30 * zmatt slaps set_ with a trout Mar 01 02:07:44 Yummy, Yummy Fish Heads? Mar 01 02:07:47 sudo is for limited system-administrative tasks like installing packages Mar 01 02:07:52 Okay. Mar 01 02:07:56 Now, I know. Mar 01 02:08:19 any time you use sudo, there must be a clear reason for it Mar 01 02:08:25 never use sudo because "I dunno" Mar 01 02:08:31 Okay. It works! Mar 01 02:08:41 So, let me just type this out. Mar 01 02:08:58 "sudo" is for adminitrative tasks only. Mar 01 02:09:07 add an s in there somewhere. Mar 01 02:09:13 or other select use cases that absolutely require root privileges Mar 01 02:09:19 Okay. Mar 01 02:09:21 using it for random shit is always a bad idea Mar 01 02:09:53 Okay. I got used to it for some reason. I forget why now but things were not allowing me to have permission to do specific tasks. Mar 01 02:10:01 I just started using it all the time. Mar 01 02:10:14 don't Mar 01 02:10:21 that generally just creates a mess Mar 01 02:10:28 Now, I know it was a terrible idea to not learn why I was using it instead of learning why I should be using it. Mar 01 02:10:38 including making files with root privileges, causing later stuff to break without root privileges and thus getting you deeper into the mess Mar 01 02:10:41 zmatt: You are right. I know already. Mar 01 02:11:39 I have used sudo until I was blue in the face. Now, I must learn my lesson. Off to read about sudo. Mar 01 02:13:04 I am 7.4 ft. above sea level now. Mar 01 02:13:07 Few. Mar 01 02:13:14 I mean. Phew. Mar 01 02:13:28 No more sudo. I am sorry. Come back to me! Mar 01 02:14:06 zmatt: I cannot tell you how hard it is to break that motion and action in life. I have been using sudo for everything. Mar 01 02:16:25 Are you telling me that all that work I have done, that did not work, may not have worked b/c of me using sudo for all my files and cmds? Mar 01 02:17:41 Damn mothers. I blame me first off. Mar 01 02:18:20 Hey. Let me just say this idea. Out of all those books I have read, not once was it mentioned on when and why to use sudo. Mar 01 02:18:52 Not sparingly, not for administrative tasks, and not for random use. Nothing. Mar 01 02:19:01 Okay. Done now. Mar 01 02:20:20 using sudo indiscriminantly is like swinging around a huge sword blindfolded when you're trying to slice a piece of cheese or bread or maybe chop up a carrot Mar 01 02:20:36 Okay. Got it. Mar 01 02:20:39 you're bound to cut *something* Mar 01 02:20:46 what about w/ touch or nano? Mar 01 02:20:47 but maybe not what you had intended Mar 01 02:20:52 Got it. Mar 01 02:21:02 or mkdir Mar 01 02:21:02 ? Mar 01 02:21:23 so, editing files in /etc, for example, is going to require sudo Mar 01 02:21:28 Aw. Mar 01 02:21:37 or some other privledge-escalation Mar 01 02:21:40 (usually) Mar 01 02:22:03 what about /home/debian/ <---- Here? Mar 01 02:22:08 but editing files in your user's home directory, or making directories in your home directory, you should be able to use your user's regular privledges Mar 01 02:22:16 Okay. Mar 01 02:22:29 unless you've been using sudo indiscriminantly, then you might have some cleanup to do Mar 01 02:22:42 Damn. Did this just change or change in 2015? Mar 01 02:22:55 * vagrantc blinks Mar 01 02:22:59 I have been using sudo like a sword chucker. Mar 01 02:23:08 no...okay. Mar 01 02:23:25 how should I clean up sudo issues? Mar 01 02:23:26 it's been this way since before sudo even existed :) Mar 01 02:23:31 Oh! Mar 01 02:23:43 it's just a basic "unix permissions" issue Mar 01 02:23:58 Now that you mention it, I think I may have read about portions of sudo text before. Mar 01 02:24:04 Oh well. Mar 01 02:24:09 I know now. Mar 01 02:24:11 sudo gives you privledges of a different user, usually privledged. Mar 01 02:24:16 Right. Mar 01 02:24:18 That is what I read. Mar 01 02:24:29 But, see, on my system, it is just me for now. Mar 01 02:24:42 e.g. root is generally the user that can look at, edit, or mangle anything on the system Mar 01 02:24:51 Right. Mar 01 02:24:59 Got that down. Mar 01 02:25:21 and when you run arbitrary things as root (e.g. with sudo) they can break anything on your system Mar 01 02:26:16 vagrantc: Thank you for the info. I have not been aware of things breaking b/c of sudo use but now I know it may be due to using sudo w/ a anxious finger. Mar 01 02:26:37 Dang sudo when it is not needed can ruin my system. Mar 01 02:26:41 Aw! Mar 01 02:27:13 https://www.xkcd.com/149/ Mar 01 02:27:38 So, just for kicks, using sudo on my system could break a file content of a particular file even w/out changing and saving the file from w/in? Mar 01 02:28:28 Get this. I always thought someone was changing my files. I would have to re"do" them and program them when I had not changed them. Mar 01 02:28:28 ... Mar 01 02:28:32 Odd days, mate. Mar 01 02:29:09 Just think. All those files that failed when they actually worked. And this was going on behind the scenes w/out my knowledge. Mar 01 02:29:19 if you create a file with sudo in your homedir, try to edit it as a user, you may not be able to save your changes, sure Mar 01 02:29:22 Or b/c I would forget. Mar 01 02:29:34 Oh. Okay. Mar 01 02:30:11 That has happened to me on more than one occasion. Mar 01 02:30:12 ok, well... heading off to other things, but hope that helped :) Mar 01 02:30:20 It did. Mar 01 02:30:28 Thank you. Mar 01 02:31:00 Sometimes all it takes is some person to shove sudo ideas down my throat to make me understand the complications. Mar 01 02:31:03 Yes! **** ENDING LOGGING AT Fri Mar 01 02:59:57 2019