It’s both horrifying and fascinating. And there’s no way to learn except by shopping. Bugs = buying the wrong thing.
Conversation
I spent an hour this morning with calipers first figuring out, then buying... nuts and bolts. And there’s a chance I got it wrong.
1
19
The polarity of build vs buy decision-making has flipped in the last 30 years. There’s really no such thing as pure building anymore. It’s all degrees of buying. A “from-scratch” design really means “from commodity COTS” — the baseline is not unrefined raw materials but China.
1
4
40
I didn’t realize how utterly bad I was at this part till I tried doing non-trivial amounts of shopping in the last few months.
2
17
Example: a typical early mechanical engineering class is called kinematics of machinery. It’s applied geometry. You work a lot of problems about how one thing will move if another moves. The things are connected by pins, sliders etc. In the textbook, the pin is just a point.
2
1
18
In practice... buying a pin is a whole production. Took me hours of googling and Twitter questions to even shortlist the types of pins to choose from, and finding a cheap source of a sufficiently varied assortment box so I have room to experiment.
2
1
24
And now my whole initial design is junk because I thought pin selection would be trivial. Now it is the driving constraint around which the rest of my design is developing.
2
2
21
In fact I suspect “purchasing inertia” drives a lot of design. Selecting, characterizing, calibrating, and testing a catalog part is so much work that you will reuse that part in future designs until you’ve amortized the purchasing decision overhead.
5
2
41
I suspect, if you’re a part supplier, making it brain-dead easy to choose your part for a large, cheap, low-end application class buys you very loyal future customers for higher-end versions.
8
2
33
Oh yeah forgot a big aspect. One reason they don’t teach shopping is that actual shopping is expensive. Window shopping doesn’t cut it, since the learning is in the trial-and-error build-buy loop. You can fill up shopping carts as exercises but that’s like 25% of the learning.
Quote Tweet
Engineering competitions are a great way to get to do this twitter.com/vgr/status/137…
3
25
Exhausting afternoon of technical shopping. Spent around $270 on Amazon on 27 items. Some of the hardest thinking ever. Mostly electronics/electrical parts and tools.
Replying to
Wouldn't buy technical parts off Amazon without complete part numbers, and assume 1/10 won't be what you ordered.
1
Replying to
It’s mostly either brand name stuff or commodity consumable stuff that doesn’t have to be too exact. I too buy more precise things on Adafruit or McMaster-Carr etc.
1
1
Show replies



