Yak trails 2021-06-13 :: The Yak Collective

Herewith a roundup of what the Yak Collective is up to this week.

Yaks — jump on the server to join in.
If you’re not yet a member, consider joining here.

@now

⌗Internal Learnings. “[…] bot-based automation is one of the abundance variables we have that I think we should use to cover for the scarcities, in this case of coordination labor. To handle differences over bot UX, perhaps it is worth defining a new ‘yak style bot automation’ project to think them through and arrive at a ‘rough consensus and running code’ shared design philosophy.” Discussion on weekly ⌗Infrastructure call Wednesdays 10am ET.

⌗New Old Home and Country “is making a call for interesting case studies in the emergent roles & projects spun up or faltered during the course of the pandemic. This would begin as a slide presentation just like last year’s The New Old Home and could go for there to multimedia formats. DM Scott Garlinger or email him at scottgarlinger@gmail.com for info for next steps.”

⌗Online Governance Studies. A weekly call for discussing ideas about governance of online communities, collectives, networks, etc. We take a somewhat scholarly approach to discussions. Fridays 9am PT.

⌗Philosophy. “We finished a book last week and took this week off and the next. Our next book will be George Eliot’s translation of Spinoza’s Ethics. Currently reevaluating times.”

⌗Weak Signals. A sounding board for things that feel significant but you can’t articulate why. Then finding a consensus on what to measure and how to measure it. async

⌗General Discussion. “I’d like to throw up a discussion prompt… I don’t think we are nailing the ‘doing consulting’ part.… the question is, do we broaden the scope (while remaining action-biased)? … I’d like us to figure out what that emergent evolutionary path is, and either own it, or if it needs a course correction, attempt to make it.”

⌗yak rover updates

Nature is Murder. Printed chassis for a test rig for the board/motors, a lot of shopping. Reprint the chassis to hold the battery I got, which doesn’t fit in the space I though it would. To decouple a system into subsystems with a stub, the complexity of the stub (like my test rig chassis) is a function of the complexity of the systems being decoupled. In this case, the test rig has to hold 8 motors and the board and batteries and have 4 drive wheels just like the actual robot design, otherwise it won’t do much by way of useful decoupling.

Stubborn. Open Sourced stubborn project. Started working on a component for managing configuration vars in non-volatile memory. This should be really helpful for fine-tuning. Got a RaspberryPi booted up just to see how it felt. Finish up configuration system and focus on some movement tuning. Writing to EEPROM immediately crashed my rover. Got garbage on the serial and an immediate reboot. I was really stumped and thought I was really screwed because it’s not like I have a debugger or a core. After sleeping on it I realized I just messed up my pointer arithmetic and was probably overwriting my stack pointer. C is hard.

Wonderful Wandering Growth. investigated Twitter control and tried to recycle LCD display into sunglasses. try to solve the physical and smoothness problems of the sunglasses, build Twitter control. figuring out how to cut optically usable small pieces of plastic backed glass. drone obstacle detection algorithm may be able to provide data for automatic characterization of surroundings.

Infinity and Beyond. Print the solar panels for rover. Setup PlatformIO for RISC-V. Setup the YouTube channel. Go through one tutorial of OnShape and Descript for video editing.

Go and See. Reviewed where the Go and See build is at… I am switching gears to designing the custom build I thought I would start in October, now. I”m not able explore many of the features and concepts that I hope to be part of my longer term design using a cheap kit robot, and want to start working on a robot that can be used outdoors in a more complex environment. Prepare to present on the custom design — motivation, strategy, specific use cases, design considerations, and technologies. Then, do so! As often occurs, being intrigued by some aspects of the conversation in the Yak Rover meeting — this time the discussion of the natural locations of computer vision processing and component redundancy in the context of safety critical applications (cars) vs. less human centric uses (Mars rovers) was particularly interesting.