Hacker News new | ask | show | jobs
by k__ 1825 days ago
Reminds me of learning what electricity is.

I learned it multiple times. In middle school, in high school, in university, on youtube explained by a quantum physicist.

Everytime I understood less of it.

2 comments

> I learned multiple times. Everytime I understood less of it.

Isn’t that how to spot seniority? The junior says “I know ReactJS and SpringBoot!” The senior says: “I don’t know much…”

Unrelated, but that reminds me how my Masters Degree teachers touted the importance of their subject in their introduction course, all explaining the Ariane V explosion from a completely different angle:

- The measurements professor: “Ariane V crashed because engineers tripped themselves into different imperial/metric units, this is why Measurements & Precision is the most important topic!”

- The programming teacher: “They fit a long inside an integer and it looped to negative, which inverted the trajectory of Ariane V and triggered its destruction, this is why learning C properly is the core of your teaching this year.”

- The quality assurance teacher: “They didn’t check the contract of the component! This is why QA is the most important when creating big systems!”

- The management teacher: “It’s the story of two teams who designed two components with different assumptions, one team worked in imperial units and they didn’t communicate clearly about assumptions, that’s why management is the one topic you should really work on.”

They were all right. Or rather, they were all wrong: Everyone knows Ariane V exploded because the officer pushed a red button ;)

Actually the first flight of Ariane 5 exploded automatically when it started to fall apart, because the flight computer read an Ada exception as flight data and from this, it decided to turn as fast as possible.

They reused the launch code of Ariane 4 in Ariane 5, but Ariane 5 was much faster to take off. It was an overflow on the acceleration and bad testing because reading an Ada exception as flight data is not great.

We learn that at school in France many years later.

That's why proper engineering is to address all the causes, even if fixing just one of them would have prevented the accident.

You'll see this regularly in the series Aviation Disasters on TV. It has lessons for all engineering projects. I watch every episode :-)

>They were all right. Or rather, they were all wrong: Everyone knows Ariane V exploded because the officer pushed a red button ;)

...and this is why learning the value of drawing the boundaries and selecting stop points in the analysis of complex topics, and the employment of humor is a powerful rhetorical tool. This is why you composition is the most important topic this semester!

Sorry... Couldn't resist. <Queue the follow up psychology is the most important topic you'll learn this semester, followed by Biology, Social Psychology, Anthropology, all getting stucktrying to get in the door.

Also, what school teaches QA These days?

> Also, what school teaches QA These days?

Interesting question! INSA Lyon in France, but that was in 2005, you could mock that the Old Continent does a lot of V-Cycle waterfall projects and had missed the Agile turn of 2001.

BUT learning how processes help is, instead, a very important step to judge what exactly we give up with Agile.

The irony is I went on creating software for requirements, and I can testify that all of the hardware industry does QA more diligently than ever!

That's because QA and Statistical Process Control were born out of manufacturing due to the high stakes with processes and tooling not being able to change on a dime like software does. I'm not surprised at all there.

Software Quality Assurance is much more... Spongey.

small point: the imperial vs metric is not an Arianne V thing but a Mars Climate Orbiter one ;)
I once read a book called "There Are No Electrons". I'm not sure I'd recommend it, but its approach was interesting: the author reasons that unless you're in grad school studying physics (and perhaps even then), everything you've been taught about electricity is a lie anyway, so the author attempts to present a framework of easier to understand lies intended to make the reader able to better reason about and predict the behavior of electrical systems, than if the reader had only the lies that are usually taught.
For everyone other than the aforementioned physics grad student, it's a lie. For the physics grad student, it's a mystery.
Turns out, the world is really really complicated.

So it's better to say that our models are simplified, not lies.

> So it's better to say that our models are simplified, not lies.

To be fair, for most purposes (atoms, molecules, metals) there usually (very-) technically aren't any electrons, just configurations of the relevant quantum fields (eg in the form of electron orbitals) whose asociated conserved quantities would allow them to convert into a certain number of free-flying electron particles if you dumped in enough energy to make up the difference.

You see this a bit more obviously with ('virtual'[0]) photons, where some non-particulate field configurations simply can't be thought of as particles at all (eg attactive electromagnetic forces).

0: https://profmattstrassler.com/articles-and-posts/particle-ph...

More generally: we have stories that say that "things behave as if they are made of ..." but too many people misread or mishear them and think that the form is "things are made of ..."

There are no quantum fields either, just as there are no atoms. Things (of the right size) behave as if there they were composed of quantum fields (or electrons or atoms), and the less we have to say "... except that ...", the more comfortable we are with the story.

All models are wrong, but some are useful. (George Box)
The hydraulic analogy is at least honest!
Solid state physics was my favorite class in college (and it was a senior course that people regularly flunked so not really a beginner-friendly tutorial). It was also very hard, and one of only 2 courses I actually attended while studying because I couldn't just show up for the test and ace it. It was fascinating to finally understand the physics behind circuits I'd been building for years (I'd understood RLC circuits since high school, but transistors, op-amps, diodes, and all that stuff I knew how to use but the "how the fuck does this amplify current?" mystery remained).

Unfortunately I don't remember what book I used. But yeah, this definitely opened my eyes to how electricity works.

If this is the book with "Greenies" in it, i've read that, and it was interesting to read but i don't know that it gave me any better idea of how to build a circuit. I lost all ability to design any circuit when it was explained that transistors work because the places for the charges to go moved, not the charges themselves.

I look at "quantum computer" components and go "what does a grid of wires have to do with 'computing'? And then you realize the big secret - There's regular computers that take the 'output' of the qubits/QC stuff and 'decide' what the results are, since it's all just a blob of probability anyhow...