**** BEGIN LOGGING AT Thu Jan 28 03:00:03 2021 Jan 28 03:00:07 Oh. Jan 28 03:00:33 Hmm...so I have to figure out w/ what motion makes it stable enough to move? Jan 28 03:00:46 This should be a good learning experience. Jan 28 03:01:18 Do you think the top base is larger than the ground base or vice versa is better? Jan 28 03:01:23 If that makes sense. Jan 28 03:01:42 umm perhaps you should read up on it instead? There are tons of papers written between the 1970's and 2012. Jan 28 03:01:49 Okay. Jan 28 03:01:55 No issue. Jan 28 03:02:06 I finally looked up ideas tonight. I will read more. Jan 28 03:02:09 I suggest 6 legs so that you have 3 legs on the ground at all times, tripod. Jan 28 03:02:28 Oh. I thought tripod was three. Jan 28 03:02:41 right I said that read slower. Jan 28 03:02:52 Oh. I got you. Jan 28 03:03:08 Double the tripod. Tripod * 2! Jan 28 03:03:28 Too late. The MotorCape only comes w/ up to four motors attached. Jan 28 03:03:46 You probably only need 2 actually. Jan 28 03:03:52 Balance. Jan 28 03:04:00 one for each set of legs. Jan 28 03:04:05 Aw. Jan 28 03:04:26 if you want it to look nicer 2 for each set of 3 legs. Jan 28 03:04:50 So, quadruped it is. I am not going against you but this will be fun. I cannot wait to crash and burn into victory! Jan 28 03:05:24 Some dents here, dings there but in the end, primordial slippage. Jan 28 03:07:00 Ok do read as I said it is actually the most difficult to move. Despite biologists thinking it was easier. Jan 28 03:07:27 you know smart people do DUMB things. I've met a lot of dumb things happening. Jan 28 03:10:20 Ha. Jan 28 03:10:25 Okay... Jan 28 03:10:36 I appreciate the ideas and facts. Jan 28 03:11:39 I just want to try this now b/c of the fact that flying has taken place, four wheels "be" rollin', and other bots are becoming not as complicated. I need to complicate things to supply my needs. Jan 28 03:11:43 I am needy. Jan 28 03:13:10 An elevator bot that knows how to move into forward and reverse and left to right? That does sound complicated. Jan 28 03:16:28 It seems like one would need two motors per leg. Yikes. Jan 28 03:17:53 well one problem at a time I guess. I suppose I should finish a robot sometime. Jan 28 03:18:59 Yes! Jan 28 03:19:04 Finish. Jan 28 03:19:32 This one will be done and ready to be deployed soon. It is for learning and not conquering mankind (for now). Jan 28 03:20:09 you need to memorize Pinky and the Brain quotes. Jan 28 03:20:49 they would be pretty hilarious to quote as you are making your videos Jan 28 03:22:31 Okay. Jan 28 03:22:54 I am more of a Bugs Bunny, dude. But okay. Marvin the Matian? Jan 28 03:24:03 well I suppose they are from similar studios. They made animaniacs from the original studios that made bugs and buddies. Jan 28 03:25:51 https://journals.sagepub.com/doi/full/10.1177/1729881419839596 <<< look here under Preception. Jan 28 03:26:30 They have some commonly used sensors. SOme of which I never heard of...of course. Jan 28 03:26:53 force-torque sensors? Jan 28 03:26:55 Hmm. Jan 28 03:27:54 force sensor measures physical pressure against something IE force Jan 28 03:28:15 torque sensors force in a rotational moment / movement Jan 28 03:28:37 Oh. Jan 28 03:29:01 sense force no sensors force (face palms) Jan 28 03:29:19 So, like if I was to get pushed, I can measure it. Or if smashed by something, I can measure that too. Jan 28 03:29:21 I got you. Jan 28 03:29:58 torque sensors would make me understand rpm. Jan 28 03:30:29 here is a force sensor from alps https://tech.alpsalpine.com/prod/e/html/sensor/piezo/hsfpar/hsfpar003a.html Jan 28 03:30:56 torque sensor shows how much strain is being rotated but not tell you how fast things are rotating. Jan 28 03:32:38 Oh. Jan 28 03:32:43 that force sensor is a little bigger than a 0805 resistor. Jan 28 03:33:10 Okay. SO, if there is friction that the rpm are hitting in time, the sensor accounts for that idea. Jan 28 03:33:38 Stator...or whatever. I guess the stator is the mechanical aspect. Jan 28 03:34:16 stator is part of a rotating motor. Jan 28 03:34:21 So, when the sensor is aware of the mechanical issue of friction, the torque sensor accounts for this friction. Jan 28 03:34:22 Oh. Jan 28 03:34:26 Armature. Jan 28 03:34:27 Sorry. Jan 28 03:35:13 I had to look up armature after typing it. I get all words confused if you have not noticed. Jan 28 03:36:30 So say, my 75% motor duty is at 100% true, then my friction to that 75% duty is what the torque sensor picks up? Jan 28 03:36:49 Ncie. Jan 28 03:36:50 nice. Jan 28 03:39:25 Dijkstra’s algorithm is something you told me to look up a while back. I just got to it. Jan 28 03:39:33 hmm say you are setting the PWM to get the motor to rotate in your actuator. The torque is the rotational force needed to overcome the static friction of the actuator. Jan 28 03:39:34 Interesting. Jan 28 03:39:52 Hmm. Jan 28 03:42:30 I am still on the same paper. I am reading about swarm intelligence techniques in robotics. i remember seeing a "nice" representation of swarms on mobile bots that hoard each other like bees to honey. Jan 28 03:44:41 Robotics is a larger field than I thought w/ all these new algorithms. Jan 28 03:54:08 yes it's massive it's been being studied since the 1920's when the term robot was coined. Jan 28 04:05:23 20's! Gee-galaroree. Ha. Jan 28 04:05:41 Dang it. Past 10:00. Snip on my sweet sleep? Jan 28 04:05:48 Sorry. Jan 28 04:05:49 BBB! Jan 28 04:07:34 It seems that analog circuits and digital circuits need to be located on a pcb at separate locations. Jan 28 04:08:12 most times but they also need other care as well. further digital circuits are just simplified analogue circuits. Jan 28 04:08:18 For instance...all of the digital elect. should be on a piece and then another analog section should be at another location. Jan 28 04:08:23 Anyway, sleep ideals. Jan 28 04:10:54 GenTooMan: Are you still awake? Jan 28 04:10:56 Nice. Jan 28 04:12:01 I am neck deep in motors, some other persons ideas on robotics, and a general consensus on mechanical torque from electrical characteristics. Do not forget current! Jan 28 04:35:49 right... I've mentioned that before torque is proportional to current speed is proportional to voltage. Jan 28 04:36:11 Aw. Jan 28 04:37:52 I was just reading about what you were saying. So, Fluke says if PWM is ON for 0.05 seconds in a 0.1-second cycle, the duty cycle equals 50%. Jan 28 04:38:43 B/c 0.05 is half of 0.10. Jan 28 04:39:26 So, that makes sense. Jan 28 04:39:34 I am starting to understand. Jan 28 04:40:38 So, I can create my own timing scheduling, right? Then, I would use that start or ON pwm to account for my timing cycle. Jan 28 04:42:00 But. I need to record this ON PWM ON period for when my seconds cycle is being recorded to get my duty_cycle. Jan 28 04:42:17 I get it, sort of. Jan 28 04:43:58 So, say I record the ON time for 10 seconds at 100%. my timing should be 10 period? Jan 28 04:44:59 No matter how it is figured out, that is what happens usually w/ period, duty cycle, and timing. Okay. Jan 28 04:45:13 I never once said it while meaning it. Jan 28 04:45:15 Thank you. Jan 28 04:55:38 Dang it. Jan 28 15:40:04 aargh, I do _not_ want to look in drivers/tty/serial/ Jan 28 15:40:53 why can't someone decide on _one_ driver for omap/am33 uarts and kill the other one? Jan 28 15:43:40 honestly I still think a separate driver is the better approach Jan 28 15:43:58 sure Jan 28 15:44:04 but right now it seems to be broken Jan 28 15:46:59 "The Linux kernel has switched to 8250-omap driver for the UART in TI processors a few years ago, and the omap-serial driver hasn't been used and tested in Processor SDK Linux releases since then." Jan 28 15:47:28 so why hasn't it been purged? Jan 28 15:48:50 and the hell does the 8250 driver create device nodes for ports that don't exist? Jan 28 15:49:11 yeah no idea wtf that is all about Jan 28 15:49:29 the number of 8250 ports is configured in kernel config with optional override via kernel cmdline Jan 28 15:49:35 really bizarre Jan 28 15:49:45 well, mine are not consecutive Jan 28 15:50:24 yep so you'll have non-functional ttyS* devices filling the gaps Jan 28 15:50:29 really dumb Jan 28 15:50:34 really annoying Jan 28 17:21:06 hah, patched it Jan 28 17:27:15 https://lore.kernel.org/lkml/20210128172244.22859-1-mans@mansr.com/T/ Jan 28 17:28:35 I'm sure greg will find a creative way to tell me I'm stupid for wanting to do that Jan 28 20:25:53 Hello and thanks in advance, does anyone know if its possible to advertise a default route from the beagle's usb subnet? Jan 28 20:26:35 in essence you would not need to configure a static gateway on the device connecting to the beagles usb subnet? Jan 28 20:32:16 i think that's one of the attacks badusb does to intercept traffic. Jan 28 20:38:19 you could configure the dhcp server to advertise a default route, yes Jan 28 20:51:01 dhcp server from the usb interface as in the default gateway would be 192.168.7.2? Jan 28 20:51:53 and when you say DCHP server, do you mean built into the linux on the beagle or do you mean installing a dhcp server program Jan 28 20:53:01 my problem is I'm forward traffic through the beagle from the usb to the ethernet, but unless i statically tell the device connected through the usb port what the default gateway is it will not work Jan 28 21:04:49 that's because the dhcp server on the beagle doesn't advertise a default route Jan 28 21:04:55 since most people wouldn't want that Jan 28 21:05:09 you can enable it in some config file Jan 28 21:07:37 The peeps at Node-RED say that beagleboard.org packages node and node-red themselves AND they have not received any doc updates.  Who can I talk to about updating the this page to reference the stable version of Debian Buster (10.3) rather than "Alpha Testing"? Jan 28 21:07:48 https://nodered.org/docs/getting-started/beaglebone Jan 28 21:08:32 To expand on my question: I'm trying to upgrade to Node v12 and did so by updating my apt lists but I still get v10 when running "node --version" even though  node v12 was successfully installed. Jan 28 21:09:58 I also rebuilt my npm packages but node-red still says it's using V10 in the journal output.  Node-RED peeps say I'm better off uninstalling and re-installing per the beagleboard docs page which hasn't been updated in quite some time... Jan 28 22:41:26 I'm not familiar with IRC.  Am I stuck between getting a response within my timeout or not getting a response at all? Jan 28 22:41:48 not getting a response at all Jan 28 22:42:26 That's ok, I already emailed Jason. Jan 28 22:43:18 it can take time to get a response since most people only glance at irc occasionally... and if those people don't know (e.g. I have no experience with node-red) you won't get a response Jan 28 22:43:36 (hence the "please be patient" in channel topic, not that it actually helps :P ) Jan 28 22:44:09 btw, my beagles are happily running 5.10 now Jan 28 22:44:21 my devicetrees needed some fixes Jan 28 22:44:29 HA!  I shall expand my patience.   (Not that I'm in a hurry but a little excited) Jan 28 22:45:44 engelhardtw: what exactly did you do to try to update to node 12 ? since the default apt package repositories in buster do not have that, you'd need to add e.g. the nodesource one for v12 Jan 28 22:46:38 mru: fixes not covered by using the newer am33xx.dtsi and other includes? Jan 28 22:47:09 whoever wrote them hadn't used the AM33XX_PADCONF macro Jan 28 22:47:13 That's what I did.  I got the commands from another vendor for another device  (I know..bad)  and I was doing that to test a node-red package specific error. (node-red-contrib-state-machine) Jan 28 22:47:17 and it was changed to produce 3 values instead of 2 Jan 28 22:47:38 so all the pinmux settings were wrong or invalid Jan 28 22:47:41 engelhardtw: what does "which node" say? Jan 28 22:48:04 mru: wtf, they broke backwards compatibility of pinctrl,single ? Jan 28 22:48:47 not quite Jan 28 22:49:21 they changed #pinctrl-cells from 2 to 3 Jan 28 22:49:34 i.e. they broke backwards compatibility of pinctrl,single Jan 28 22:49:36 so the new dtsi is incompatible with old dts Jan 28 22:49:42 the driver handles both Jan 28 22:49:48 oh Jan 28 22:49:54 ah Jan 28 22:49:59 what's the new additional argument? Jan 28 22:50:21 they split pull-up/down and mux mode into separate values Jan 28 22:50:28 instead of being | together Jan 28 22:50:50 how does that make sense? pinctrl,single normally doesn't need to know or care what's what Jan 28 22:50:56 @zmatt Booting.  My issue was "node --version" gives v10.23.1 but if I do the same command from /opt/node-red/bin  then I get the recently installed v12.20.1.  The real issue is that Node-RED journal output says it's using v10.  I don't even know for sure that I need v12 but I wanted to try it. Jan 28 22:51:17 engelhardtw: why does node-red have a custom nodejs installed in its own dit? Jan 28 22:51:23 *dir Jan 28 22:52:02 so you didn't update the system nodejs package to v12, you just installed a nodejs v12 in a weird place outside package management? Jan 28 22:52:37 ah, I knew I have a paste for it: https://pastebin.com/fAGYJRTp Jan 28 22:52:56 @zmatt I'm working on a Beaglebone Black with stable Deb 10.3  which comes with Node-RED v1.2.7 and nodejs 10.23.1 by default.  I got the image from Beagleboards site. Jan 28 22:53:15 yes I understood that part Jan 28 22:53:25 You might be right Jan 28 22:53:38 oh weird, that has its own nodejs on the default install too Jan 28 22:53:54 I didn't use beagleboard's  instructions because they reference Alpha testing.  The page needs to be updated. Jan 28 22:54:33 Node-RED (The organization) needs updated documentation from beagleboard.org (The organization) Jan 28 22:54:49 so again I don't know anything about node-red, but the link I just gave is how you update the system nodejs to v12 ... of course that will almost certainly break anything on the system that uses nodejs :P Jan 28 22:55:13 Unless I rebuild...which I did.... Jan 28 22:55:23 unless you rebuild indeed Jan 28 22:55:31 isn't broken the default state of nodejs things? Jan 28 22:55:32 right, but you haven't actually updated your system nodejs evidently Jan 28 22:55:37 So why is Node-RED  journal output still saying :28 Jan 19:33:39 - [info] Node-RED version: v1.2.7 Jan 28 22:55:38 28 Jan 19:33:39 - [info] Node.js  version: v10.23.1 Jan 28 22:55:38 28 Jan 19:33:39 - [info] Linux 4.19.94-ti-r42 arm LE Jan 28 22:55:39 28 Jan 19:33:42 - [info] Loading palette nodes Jan 28 22:56:03 Sorry I didn't format that nicely Jan 28 22:56:15 because that's the version of nodejs you have installed, as confirmed by "node --version" gives v10.23.1" Jan 28 22:57:10 not sure why node-red is installing a custom nodejs in /opt/node-red nor why it isn't using it... both of those things sound inappropriate to me :P Jan 28 22:57:55 But  when I navigate to the dir where node exists (/opt/node-red/bin/) and THEN issue "node --version" then  I get v12.20.1 Jan 28 22:58:38 Well peeps in Node-RED Slack channel are pointing the finger at Beagleboard so I'm getting frustrated. Jan 28 22:59:17 the nodered.service on the default image does seem to use /opt/node-red/bin/node Jan 28 23:02:13 yeah on my beaglebone (running a mostly-unmodified 2020-04-06 buster IoT image) node-red uses /opt/node-red/bin/node instead of the system nodejs Jan 28 23:04:01 node btw that node-red on the default image is provided by (or at least installed by) a package: bb-node-red-installer Jan 28 23:04:15 the upstream version of that package is v1.2.7 Jan 28 23:04:27 so all you need to do presumably is update that package Jan 28 23:05:18 since I don't know what you've done to your system, you may want to start with a fresh image and then just sudo apt update && sudo apt install bb-node-red-installer Jan 28 23:10:39 @zmatt I follow and I will try that.  Thank you for the guidance. Jan 28 23:15:14 engelhardtw: if it works, don't forget to also inform the node-red slack folks so they know the correct solution for future reference :) Jan 28 23:28:01 @zmatt Either way, y'all need to send them updated docs to AT LEAST eliminate the reference to "Alpha Testing" of Debian Buster.  Right? Jan 28 23:28:27 outdated docs are a neverending problem Jan 28 23:28:48 Then you have work to do. Jan 28 23:29:41 the doc looks fine though.. it's saying exactly the same thing I was: update the bb-node-red-installer package Jan 28 23:30:00 for some reason they also update the system nodejs but that's evidently pointless Jan 28 23:31:40 Ignore me.  I'm being weird today.  Signing off. Jan 28 23:31:48 only the first section is outdated Jan 28 23:31:58 which happens very rapidly with anything that uses the word "Currently" Jan 28 23:31:59 :P Jan 28 23:32:39 the workaround is to say "as of this writing" **** ENDING LOGGING AT Fri Jan 29 03:02:12 2021