Misaka on Computing @misakacomputing

A good implementation keeps similar components dependent on a master component such that mostly only the master needs to change.

String interpolation requires care to not escape special characters. Let this WeChat bug serve as a lesson.

One problem with trying to get other developers to work with your code is the fact that you have not written any documentation.

Recursion may seem tricky to program. However, in practice it's usually easier but more inefficient than an iterative method.

If you want to solve a problem, Google it. Did that already? Good. Do it again.

My classmates often ask me, "How did you become so leet?" That's when I tell them, "I watched 'Hackers'."

If you find yourself needing to use your mouse while coding, then you should use vim. And don't use Windows.

When you don't know how to solve a problem involving computers, you should Google for the solution and read everything until you understand.

Nothing is better a lesson for buffer bounds checking than Heartbleed.

Overclocking doesn't just make your machine more awesome. It also increases the probability of experiencing an awesome meltdown.

At Tokiwadai, if you get caught not properly indenting your code, you lose dorm privileges for a day.

The simplex method isn't in polynomial time, but in practice it's faster than the currently known polynomial algorithm.

Q = (c97445f45cdef9f0d3e05e1e585fc297235b82b5be8ff3efca67c59852018192, b28ef557ba31dfcbdd21ac46e2a91e3c304f44cb87058ada2cb815151e610046)

A Gekota decal on my machine would be awesome.

Shiba Tatsuya is a script kiddie.

Machine learning has "machine" in it because I can't directly use it to study English. So frustrating...

Kamijō Tōma is a good-for-nothing idiot. Idiot idiot idiot.

I can confirm that integer factorization is in BQP. I just played with some qubits.

Convergence speed is important, but faster doesn't equal better.

In short, machine learning is using data to effectively train models to solve some problem.