Post History

# [BQN](https://mlochbaum.github.io/BQN/), 19 bytes<sup>[SBCS](https://github.com/mlochbaum/BQN/blob/master/commentary/sbcs.bqn)</sup>
```
' '+32⊸<⊸×+⌜˜⌽⊸⌊↕59
```
[Run online!](https://mlochbaum.github.io/BQN/try.html#code=JyAnKzMy4oq4POKKuMOXK+KMnMuc4oy94oq44oyK4oaVNTk=)

This uses the whitespace allowance, padding by four spaces on all sides to make the arithmetic work better. It can be adjusted by replacing `↕59` with `4+↕51` at the cost of two characters.

BQN has no built-in idea of the alphabet, but it has a nice system for Unicode [character arithmetic](https://mlochbaum.github.io/BQN/tutorial/expression.html#character-arithmetic). This solution depends on the length 26 of the alphabet, and the distance 33 (`'A'-' '`) from space to the start. Fortunately these are close together. We will build a matrix of numbers that goes from 0, for `' '`, at the corners to `33+26-1`, or 58, at the center, then bring all numbers below 33 to 0. At the middle of each edge we get 29, leaving 4 spaces before `'A'`. [Before](https://mlochbaum.github.io/BQN/tutorial/combinator.html#before-and-after) plays a big role here.

```
' '+32⊸<⊸×+⌜˜⌽⊸⌊↕59
                ↕59   # Range            0...58
             ⌽        # Reverse:         58...0
              ⊸⌊      #  Before Minimum: 0…29…0
          +⌜          # Addition table
            ˜         #  with itself
    32⊸<              # More than 32
        ⊸×            #  times the original
' '+                  # Add to space
```

Another 21-character "exact" solution is `(+⌜˜⊏-∘≠↑'A'+⍷)⌽⊸⌊↕51`. Now `'A'` is used as the basis and spaces come from fill elements in Take (`↑`), so it will work for any starting character.