Template: Markdown post

May 1, 2026 May 4, 2026 · templaterendering

This is a template post. It exists so you can see how every element renders. Delete it (or replace its body) before publishing anything real.

A brief preamble

This paragraph exists so you can confirm that body copy in the chosen serif reads pleasantly at a 66ch measure with generous line height. Adjust --measure in global.css if your taste differs. Setting type is one of the few decisions that’s worth getting right early¹.

Smaller heading

Subsection-level headings should feel related to their section, not shouted. Run a few of them in a row to confirm vertical rhythm.

Lists

Unordered:

• An item.
• Another item with a longer line, long enough to wrap so you can confirm that the second line aligns the way you expect against the first.
• A third item.

Ordered:

1. First.
2. Second.
3. Third.

Code

Inline code: array.flatMap(x => x.children) should sit comfortably inside a sentence without disturbing the line height.

A TypeScript fence:

type Post = {
  title: string;
  pubDate: Date;
  tags: string[];
};

function recent<T extends { pubDate: Date }>(items: T[], n = 5): T[] {
  return [...items]
    .sort((a, b) => b.pubDate.valueOf() - a.pubDate.valueOf())
    .slice(0, n);
}

A Python fence — different language to confirm Shiki picks up the grammar and the dual-theme swap renders both cleanly:

import torch
import torch.nn as nn

class TinyMLP(nn.Module):
    def __init__(self, d_in: int, d_hidden: int, d_out: int) -> None:
        super().__init__()
        self.net = nn.Sequential(
            nn.Linear(d_in, d_hidden),
            nn.GELU(),
            nn.Linear(d_hidden, d_out),
        )

    def forward(self, x: torch.Tensor) -> torch.Tensor:
        return self.net(x)

A blockquote

The truth is rarely pure and never simple.

— Oscar Wilde, paraphrased

A captioned image

A table

Footnotes

The footnote reference earlier should jump down here, and a back-link should return you to the spot you came from².