You have discovered Wally's playground. Wally is one of the guys from the Dilbert comic strip.

Basically, many incentives are "misaligned" in such a way, that micromanaging demise is (locally) considered more valuable than actually bringing in cash from customers by actually giving them something.

As long as the company's owners don't pick up on this and are effective at doing something about it, "all bets are off."

Epigram 102:

102. One can't proceed from the informal to the formal by formal means.

— Alan J. Perlis, Yale

That’s a great way to put it.

What I keep running into is that teams know the work is informal at the edge, but systems are designed as if formality can be enforced at capture time. In practice, that just pushes the work elsewhere.

In your experience, where does that mismatch hurt the most: audits, safety reviews, customer disputes, or something else?

Everywhere.

Computing and formal mathematics rely on exhaustive case analysis and binary logic. And even with The Excluded Middle, there are unprovable statements.

There are at least two incompatible ternary truth tables (hello there SQL NULL) in formal logic. Then there is fuzzy logic, but that is also a formalization.

(NP-complete problems and uncomputable problems in binary logic are another sore point.)

And for informal systems the best people have found so far is hypothesis testing, which is to say that only the rejection of hypothesis based on measurements works, but not confirming a hypothesis.

Turtles and blech all the way down.

Agree with your pov, especially the idea that you can only reject hypotheses, not confirm them.

What I keep seeing operationally is that teams are forced to act before hypotheses can be tested or falsified. For example, inspections completed, assets redeployed, and customers responded to. Only later are they asked to prove correctness.

When that gap shows up for you, what was the concrete trigger? An audit, a customer dispute, a safety review, or something else?

There is also the Vernier scale on certain styles of mechanical calipers.

I love the Vernier scales! But I honestly hadn't known about the sense in which slide rules have a Venier scale until your comment, which led me to [1]. In another class on laser-cutting I had students make a kerf-meter, which became more accurate with a Vernier scale.

[1] https://dram.page/p/slide-rule-vernier/

Oh my God that is so clever. Thanks for sharing!

Slide rules are awesome.

Many irrelevant difference between programming languages are now exposed for what they are.

Thinking clearly is just as relevant or encumbering as it always was.

Wait until you hear about the pixel size restrictions on safari canvases.

WebGL punts to WebGPU for decent compute shaders.

That was also my research group's approach.

How else is the data going to make it to your phone?

And with a JS-friendly tool you can also test your plots on a tablet and a phone in your local wifi.