Why Markdown Keeps Eating New Jobs

Markdown keeps absorbing new jobs — email, slides, dashboards, generative UI — not because it wins any single benchmark, but because it is plain text: the one substrate a new tool can be poured into without rewriting from scratch. The new tools are not built on Markdown the parser. They are built on plain text the substrate.

The pattern is visible in one week of community attention. "Quarkdown – Markdown with Superpowers" drew 358 points on Hacker News ¹. "Email.md – Markdown to responsive, email-safe HTML" drew 380 ². A Markdown superset for docs, dashboards, and slides drew 152 ³; a Markdown-as-protocol for generative UI drew 130 ⁴. And the loudest reply, "Why are we still using Markdown?", drew 233 ⁵. The receipts and the backlash arrived together. That is the argument worth having.

What most people believe Markdown is for

Most people file Markdown under "lightweight writing format" — README files, blog drafts, the thing you type in a chat box. That filing is fair and historically accurate. John Gruber described it in 2004 as "a text-to-HTML conversion tool for web writers" ⁶. For two decades that was the whole job. Write prose, get HTML.

The belief is not wrong. It is just out of date. The CommonMark spec records that Markdown "started to be used beyond the web, to author books, articles, slide shows, letters, and lecture notes" ⁷. The drift from "web writer's shortcut" to "everything-authoring format" was underway long before 2026. The 2026 wave is the same drift, accelerated.

The pivot — the new jobs skip the parser

The new tools do not adopt Markdown the parser — they adopt the plain-text base and extend it. Quarkdown was "born as an extension of CommonMark and GFM" and "brings functions to Markdown" ⁸. MDV calls itself "strict CommonMark plus four additions" ⁹. The shared move is one move: keep the readable syntax, bolt on what the new job needs.

Look at what those jobs are. Email.md turns Markdown into "responsive, email-safe HTML that renders perfectly across every client" ¹⁰, a notoriously hostile target rebuilt on a text file. MDV lets you "write reports, dashboards, and slide decks in a single Markdown-native workflow" ⁹.

One builder went further still. They turned Markdown "into a protocol for generative UI," using code fences as the transport an AI assistant uses to mount live components mid-conversation ⁴. Email, slides, dashboards, agent UI: four jobs Markdown was never designed for, all reaching for the same base layer.

Why plain text, specifically

The reason is not nostalgia — it is reach. Plain text is the lowest-common-denominator container: any program can read it, any program can write it, and nothing about it expires. That is the whole advantage, and for a tool that wants to outlive its own version number, it is enough.

Steph Ango, Obsidian's CEO, put the principle plainly: "File over app is a philosophy: if you want to create digital artifacts that last, they must be files you can control, in formats that are easy to retrieve and read" ¹¹. The corollary is the part that bites — "Apps are ephemeral, but your files have a chance to last" ¹¹.

A builder inventing a new tool faces a choice. Invent a fresh binary format and own the parser, the migration story, and the lock-in complaints. Or pour the new capability into plain text and inherit twenty years of tooling for free.

Gruber's original design goal explains why the second path keeps winning. A Markdown document "should be publishable as-is, as plain text, without looking like it's been marked up with tags or formatting instructions" ⁶. The substrate is legible to a human before any tool touches it. That legibility is the durable part, not the feature list any one tool wraps around it.

The honest limits — what the contrarian gets right

The "why are we still using this?" backlash is not a strawman, and it is mostly correct. The honest case against Markdown has three parts, and conceding them is what makes the substrate argument hold. The dialects are a mess. The base is not.

First, these new tools are supersets, not vanilla Markdown. Quarkdown is "an extension of CommonMark and GFM" ⁸; MDV is "strict CommonMark plus four additions" ⁹. Your Markdown is not my Markdown. The durable thing is the plain-text base underneath — not any one dialect bolted on top.

Second, fragmentation is real. CommonMark exists precisely because "there is no unambiguous spec," so "implementations have diverged considerably," and Gruber's canonical description "does not specify the syntax unambiguously" ⁷. The critic behind the 233-point counter-thread sharpens it: "The only reason feature creep exists is because of unclear specifications," and inline HTML means "recurring XSS vulnerabilities across major Markdown implementations" ⁵. These are genuine costs, not nitpicks.

Third, Markdown does not replace HTML, and it is not the best format for every job. Gruber said so when he called it a conversion tool, not a destination ⁶. For where each format actually wins, the layer comparison is its own argument ¹².

Here is the turn. Asked what to use instead, the same critic answers: "the answer is plain text" ⁵. The most thorough case against Markdown lands on the exact substrate this essay defends. The dialects fragment. The base does not.

What to do about it

Keep your own work in the substrate, not in any one tool's dialect. The move is small, and it compounds. Author in plain text, treat every superset as an exit rather than a home, and judge tools by how easily your sentences leave them.

• Author in plain text. A .md file opens in every editor, today and in a decade.
• Treat supersets as exits, not homes. Use Quarkdown for a deck or MDV for a dashboard, but keep the source readable as plain text so leaving costs nothing.
• Judge tools by how easily your sentences leave them, not by the feature list on the landing page.
• Remember that upvotes measure demand, not durability. A 380-point Show HN ² proves people want the job done, not that this tool outlives the next one.

This is the build-on companion to a separate argument about reading: the web is increasingly converted to Markdown so machines can read it efficiently ¹³. Both point the same way.

Whether you are building new tools on plain text or feeding it to an AI, the format that keeps absorbing jobs is the one that asks nothing of the reader. The most durable version of "your notes" is the plainest — stored locally on your own device, in open Markdown you can open in any editor ¹⁴.

Frequently Asked Questions

Why are we still using Markdown in 2026?

Because the new jobs are built on plain text, not on Markdown the parser. Builders keep reaching for a readable text base, then extend it, because plain text is the one container every tool can read and write without lock-in. The honest costs are real: fragmentation and ambiguity, conceded above ^{7 5}.

Why do new tools keep building on Markdown instead of replacing it?

Inventing a new format means owning a parser, a migration story, and lock-in complaints. Pouring the capability into plain text inherits twenty years of tooling for free. Quarkdown is "an extension of CommonMark and GFM" ⁸ and MDV is "strict CommonMark plus four additions" ⁹. Both extend; neither replaces.

Why is Markdown used for email, slides, and dashboards now?

Each is a different output target generated from one plain-text source. Email.md compiles Markdown to "email-safe HTML that renders perfectly across every client" ¹⁰; MDV produces "reports, dashboards, and slide decks in a single Markdown-native workflow" ⁹. The source stays legible while the tool handles the hostile target.

Is Markdown a real standard, or does every tool do it differently?

Both, honestly. There is "no unambiguous spec," so "implementations have diverged considerably," and CommonMark exists to fix exactly that ⁷. Treat any one tool's flavor as a dialect of a shared plain-text base. The base is the durable part; the dialect is not.

Who made Markdown, and when?

Markdown is "a plain text format for writing structured documents," developed by John Gruber with help from Aaron Swartz and released in 2004 ⁷. Gruber's design goal was that a document should be "publishable as-is, as plain text" ⁶. That legibility is what the 2026 tools are still building on.

Are these new Markdown tools actually Markdown?

Mostly they are supersets: Markdown-shaped syntax on a plain-text base, then extended. Quarkdown "brings functions to Markdown" ⁸; MDV adds front-matter, data blocks, and containers to strict CommonMark ⁹. The vanilla-Markdown part is the durable substrate; the additions are each tool's bet on a specific job.

The format that keeps eating new jobs is the one you will still be able to read when those jobs, and the tools that did them, are gone.

This argument owes its honest center to the critics who keep asking "why are we still using this?" and answering, against their own thesis, "plain text" ⁵. If you want your own notes to keep that property, MNMNOTE keeps them as plain Markdown on your own device, readable in any editor — mnmnote.com.