Not detracting from this post, but has anyone else noticed there's a front page post about Quaternions or Kalman filters on about a monthly cadence? Wonder why that is?
I have noticed too that HN has an enduring fascination with quaternions, and they seem to reach front page surprisingly frequently (considering that there's not a huge amount of discussion of other geometry topics). I'm certainly not complaining though - I love quaternions too!
HN's going to need a windscreen cleaner and a bucket if 3blue1brown and Terence Tao ever collaborate on using adaptive Kalman filters for optimal paths in Quaternion spaces.
I never noticed it until I read Pynchon's Against the Day, I assume I just ignored quaternion posts before then. Great book, enjoyed the way he used quaternions and vectors towards literary ends.
Edit: just noticed we also have a thread on bifurcation, another big topic in Against the Day.
At least this is a topic that the average HN reader is likely to be able to understand and which is closely related to software. Several years ago there was a period where there were periodic posts about things like homotopy type theory and research-level algebraic geometry which inevitably spark only inane misunderstandings and uninformed speculation in the comments. I can only attribute it to some kind of fetish for the frontiers of pure math.
AG maybe isn't so connected to software, but HoTT connects to the "functional programming" crowd. https://youtu.be/MVtlD22Y8SQ is me doing some examples.
Octonions are cool! The one application I’m familiar with is in crystallography - you can represent the interface between two crystals with a unit octonion https://doi.org/10.1016/j.actamat.2018.12.034
They’re both quite basic building blocks in state estimation and as SV’s focus shifts from web apps to drone warfare it will only increase in frequency.
In my experience in a university ECE program, you'd start with understanding the high level properties of transistors, then combining transistors to make AND and OR gates, then XOR and other gates, then MUXes and half/full adders, then flip-flops and eventually into synchronous (clocked) logic.
The lab component of such coursework did start with TTL chips but the timing of the coursework was such that you'd have most of the asynchronous logic theory taught by the time the chips came out.
Was it an Electrical Engineering, Electronics Engineering or Electrical Engineering Technology Program? My digital course skipped over transistor level and spent that time on basic FPGA's instead.
Not OP but I did Electrical and Electronic Engineering undergrad and we started with diodes at the materials level, then BJT and FET transistors, then logic gates, flip flops, timers, ALUs and eventually working up to build a Motorola 68K micro controller from mid level components. There was some VHDL and FPGA in the later stages as well from memory.
I would comment the difference is I have GUIs to set things up, for non-technical users. I did a survey of various projects and found most of them had a configuration step such as "now SSH into the box and input your API key". My mum is never going to do that I'm afraid.
Edit: not to poop on this mode of setting things up, it's good for personal projects but I'm trying to take another step here.
Not sure you can eliminate it, but you can certainly reduce the impact and scale by fighting anti-competitive behavior . A major reason this sort of mass manipulation is so lucrative and effective is because you only have operate on a couple of platforms to reach a majority of eyeballs.
Reminded me of the Amazon show called Upload. Sort of a lighter, more humorous take than Black Mirror, but the darker philosophical complexities still surface regularly.
Supporting EINK on mine could potentially help make it more visually compelling like yours, but I've done other EINK display projects in the past (see https://github.com/Mrjohns42/DoggieClock) and screen burn-in was definitely an issue.
Burn-in is very possible. It is ink particles inside a solvent. Drive it the wrong way and you'll damage the ink or the solvent. Just like LCD where if you drive it the wrong way, the liquid crystal gets damaged. The difference is LCD voltage waveforms are built into hardware and are much simpler since there's no moving components.
reply