Ask HN: Developers, what is the last significant thing you learned?

I just (in the last month or so) learned how to expose the Docker daemon to code running inside a container, so you can create a "container controlling container." This works out really nicely for something I've been working on, which involves running smoke-tests inside a container. This way, the tests can programmatically request a container instance of the system under test, and then destroy it, without the need for any sort of coordinator code running one level higher.

I've also recently (last couple of months) learned just enough Common Lisp to feel like I am close to "turning the corner" and starting to grok Lisp. If I keep at it, I think I'll be at a place where I could write something meaningful in Lisp before much longer. This feels significant to me, as "learn Lisp" has been on my "to get around to one day" list for about 20 years now.

When writing unit/integration tests and they fail, consider possibility the test is wrong.

Yesterday I spent two hours researching UUID mappings in H2 and PostgreSQL and trying various combinations of annotations just to find out that I accidentally deleted last character of the hardcoded "expected" result.

In Haskell you can generate an infinite list of all integers starting from 1 with [1..]

This is both very succinct but for best use also requires a sufficient understanding of what goes on in the backend, which is true for many features of the language and isn’t always encouraging for uptake. Syntactical elegance and efficiency sometimes comes at a cost of high information/context overhead.

While developing https://classroom.qvault.io I've been learning that I forget how not obvious literally everything about coding is.

In the first course I had feedback like, "I didn't understand code starts at the top and goes down"

Some items I can recall:

- using CLoader for YAML loading in Python is x10 faster. (Last weekend)

- Chrome DEV Tools has an option to simulate different network speeds, really useful if you are developing for the developing work when cellular speed can be very low. (~1 year ago)

I've just recently committed myself to take Javascript seriously and more than a language that can work with frameworks.

The last useful thing I learnt is the company always wins and you are worth nothing.