Hacker News Reader

Comment by zokier | Hacker News Reader

To me chemistry suffers from being actually quite difficult field. Even by doing these sorts of experiments it is difficult to build up good intuition and actually start using chemistry to achieve some specific goals; contrast to e.g. electronics or programming where after some very basics you can start building your own things

View on HN

Mechanical engineering (I'm an ME) turned out to be very very hard to do as a hobby. Even with a complete machine shop, it takes a loooong time to make something simple, such as a steam engine. My engine finally worked, but the hours expended on it were extensive.

With software I can build amazing programs in short amounts of time.

Engines usually push (their contemporary) materials to their limits so it's not really a surprise that they're harder to make from scratch. The infrastructure to make ICEs is one of the largest in the world and they are only so cheap because of their scale and production steadily getting automated since the post war consumer boom.

Making that automation, however, has been easier than ever! DIY CNC designs are a dime dozen ranging from routers to pick and places to mills, there are at least half a dozen serious open source competitors for the firmware to drive them, and the welding equipment and skills necessary to make a CNC rigid enough to cut steel is more accessible than ever (designs exist even!). All the parts you need for all kinds of mechanical contraptions and automation buildouts are available from McMaster Carr with same/next day shipping in some regions and there are even low cost open source robotic arms with six degrees of freedom! Something as simple as aluminum extrusions have created a sea change in how mechanical structures are built and modified from cells in automated factories to 3d printers to automated labs. Hell, you can make 3d printed liquid cooled rocket engine nozzle in metal by uploading a CAD file and test it in Mojave a few weeks later without ever touching manufacturing equipment. /rant

Steam is basically the classical example of an ECE

That sounds pretty awesome. None of that was available when I was working on my engine. I had a lathe and a milling machine, and little skill with them.

The typical "model engineering" steam engine doesn't push materials to the limits. Cylinders are typically cast iron or bored out of brass rod. Tolerances aren't too tight.

It is a slow process because metal is kind of recalcitrant, and making one-off parts is a lot more time consuming than manufacturing in series. To be a model engineer you have to love the process and love spending time alone in the workshop. Another tedious/fascinating aspect is the jigs and fixtures that are often necessary to make even one example of a particular part. Model engineers have to devise ways of performing machining operations (often conceptually very simple operations) and it really has to be something you love. If you see the manufacturing process purely as a barrier to realizing what you designed on paper, it can be a real drag.

So, rather than a high-tech thing, pushing limits, I think the task here (making a small steam engine) is an example of craftsmanship (with lathe, file etc.) and of doing something that is intellectually/conceptually/materially very simple and undemanding, but doing by hand using inherently slow traditional processes. It's an aesthetic choice (and CNC, aluminum extrusions etc. would be ruled out for aesthetic reasons).

And yet I get way more satisfaction out of building some small wood contraption than I get out of building anything in software, big or small.

> chemistry suffers from being actually quite difficult field

I wonder how much of that is inherent, and how much is artifact of highly suboptimal education? Chemistry education research has been unusually scathing, characterizing chem ed content using adjectives like "incoherent". My favorite high-school state standard insisted students be taught both that atoms are conserved in chemical reactions, and that atoms are by definition electrically neutral. :P And the incentives around say teaching orgo, are more about serving as med school proxy, than using chemistry to achieve real goals. What might chemistry look like if taught well? How difficult would it still be? It seems a very open question?

Chemistry is fundamentally microscopic in a way that machinery is not, and chaotic in a way that electricity is not. (A big part of electronics is wrapping microscopics in friendly abstractions like "batteries" and "resistors".)

Molecules are small and diverse (it takes alkynes to make a world!) and the world is big.

Certainly chemistry also has abstractions that make it easier to handle. For instance, acids and bases are such; if you know something is an acid, you can trust it to behave in certain ways in certain situations regardless of the microscopic structure of the molecules.

For me it was only once I realized chemistry is just a specialization of physics that I really started to understand and care about it more. I learned way more about chemistry from a couple college physics classes than the chem class and lab I was required to take.

Out-of-school chem self-studier here. Can you elaborate?

Everything I learned in my first chem class was just bad rehashing of electron orbits and configurations from physics. I'm sure things get more interesting with organic chemistry, but the classical stuff/basic elements I thought were easier to understand just thinking about the physics of their atomic structure.

I would start the same way my high school chemistry teacher did when she said. "sometimes science isn't safe". My uncle's childhood chemistry set had radium in it. There are a lot more of us than there were back then, it is ok if some of us die in the pursuit of knowledge. It is a worthy goal, in that people like Curie paved a way to a better standard of living for the rest of us. If you can't demonstrate any cool experiments how are you going to attract future chemists? A different uncle who actually became a chemist had a crc reference that made your fingers itch just to turn the pages. I'm half convinced it would qualify as a superfund site and it was just a reference book. We worry too much about the wrong things. I'm sure all the big pharma in the water supply is fine, but chocolate has way too much cadmium!

A great man[0] once said, "Chemistry is easy, you just remember all the rules, and remember all the exceptions to the rules."

[0] https://www.youtube.com/c/ExplosionsFire2 I couldn't find the specific video, despite having a great time of rewatching a bunch of them. Amateur chemistry isn't quite dead yet. Ed: found it https://youtu.be/psUc_oBXE6c?t=18

I think (low confidence) this is due to basic chem being not-composable. And of course, possibly dangerous in a hard to detect way.

My thoughts exactly. When our little program crashes, even in C or assembly, your operating system cleans it up for you. Just redo-it better next time. Compare that with chemistry.

Maybe what is needed for chemistry is an "operating system" that safely handles and cleans up the experiment ran. Maybe this is what actually should be sold for basic chemistry. Like the kitchen robots but for nasty materials.

You would also be wrong.

Chemistry is very composable, if you are doing these experiments you are likely learning the theory and can quickly work out that variants exist, and enumerate those variants to find new things to test, expanding knowledge.

As for hazards: most of the hazards are very easy to detect or mitigate. Work at a small scale, wear PPE, etc.

The problem is that you need to be fairly far up the chain of mathematics and physics before you can build intuition about inorganic chemstry.

Oddly, organic chemistry is, to a certain extent, easier to build intutition about as the lower levels are a lot of "this reaction takes thingit A from the molecule and puts on thingit B with n% efficiency." This make organic very similar to solving jigsaw puzzles.

Come on, I was balancing equations at the age of 8. Knowing the multiplication table is enough.

It's very easy to think up balanced reactions that are completely nonsensical and the reason for that is hidden deep in quantum realm.

It is, but this will have to wait until late high school, when kids are old enough for matrices and complex numbers and Dirac equation and the origin of 2x^2 (2, 8, 18, 32 — the electron shells etc). You still can have meaningful and useful exercises meanwhile.

Given a reaction, you can balance it with fairly simple math.

However, if you want intuition as to how to create a new reaction or a novel molecule, simple math really isn't enough.

That is true, and something that I just realized by watching some Chem YTbers

Because it's like "oh look, this works like this, ok"? Then something similar doesn't work. Or works in a different way. "Oh just try the $guys_name reaction" ah yes because this guy discovered that you could do it like that (which is not a really intuitive way)

Oh look, this is actually simple to try at home. But if you try this other thing that might look similar it can kill you in a horrible way...

Reply on news.ycombinator.com