Hackaday editors Elliot Williams and Mike Szczys peruse the great hardware hacks of the past week. There’s a robot walker platform that wirelessly offloads motor control planning to a computer. We take a look at automating your fishing boat with a trolling motor upgrade, building the Hoover dam in your back yard, and playing Holst’s Planets on an army of Arduini. Make sure you stick around until the end as we stroll through distant memories of Gopher, and peek inside the parking garages of the sea.
Take a look at the links below if you want to follow along, and as always, tell us what you think about this episode in the comments!
Direct download (60 MB or so.)
Episode 138 Show Notes:
What’s that Sound?
- That sound was Tesla coils playing the Super Mario Bros. Theme (Artist: ArcAttack!)
- [cmbalancio] was randomly drawn from 66 correct responses and wins the shirt!
New This Week:
- A Coolant Leak The Likely Culprit For Aussie Tesla Battery Bank Fire
- It’s Wildcard Time, Your Last Chance To Enter The Hackaday Prize!
Interesting Hacks of the Week:
- 3D Printed Research Robotics Platform Runs Remotely
- A Promising Start For The Doritos Space Program
- Arduino Orchestra Plays The Planets Suite
- Open Source Autopilot For Cheap Trolling Motors
- Bluetooth Vulnerability: Arbitrary Code Execution On The ESP32, Among Others
- Reinforced Concrete: Versatile At Any Size?
Quick Hacks:
- Elliot’s Picks
- Mike’s Picks:
“an army of Arduini”. I will use this.
I remember people referring to the class of DEC VAX computers as “Vaxxen”, so why no Arduini? And while we’re at it, how about Raspberry Pies? I sure as heck don’t want to ever call them “Raspberry Pis”.
I’m pretty sure that Data General’s minicomputers were commonly called Novae or Supernovae, and of course a pile of two-button pointing devices are meeces. That’s all I’ve got, right now.
Raspberry Pii?
B^)
It turns out that you cannot turn a larger uniform random number distribution into a smaller uniform distribution using the modulus operator unless the size of the final range evenly divides the source range. This is an incredibly common mistake.
True! But since 2^128 is kinda big relative to 66…
It’s still better for a lot of purposes not to do this, if only because you’re perpetuating the problem in public; people don’t necessarily understand why it’s flawed. There are versions of this issue all over code; things like bad hash functions that produce low hash table performance even though they “shouldn’t”, bad random selection and permutation algorithms, etc. (BTW, for hash tables, there’s a whole host of other folk myths people believe when implementing them, like the idea that they’ll perform better if the table width is prime.)