Org mode

The oldest continuously used plain-text system in this corpus's tools-for-thought set, and the one it did not hold. From the Org manual (T1, first-party): "Org mode is an authoring tool and a TODO list manager for GNU Emacs."

That double description is the whole point. obsidian, logseq, roam-research and tana each pick a centre — notes, blocks, graph, database — and grow outward. Org grew from an outliner (it sits on top of Emacs's Outline mode) and then absorbed everything a working researcher does around an outline: TODO states, scheduled and deadline dates, "dynamic agenda views" computed across files, tables with a spreadsheet, executable source blocks, and export to LaTeX, HTML, and the rest.

The format is the durability argument. Org files are "strictly plain text files, the most portable and future-proof file format" — readable in any editor, though Emacs is where the behaviour lives. Set that beside roam-research's hosted database and the local-first turn the corpus has been tracking: Org made that choice in 2003 and never had to reverse it.

The manual describes the design stance as a toolbox: the basics stay simple, the rest is there when wanted, and nothing forces the user through the full system. That is a different answer to complexity than the one tiddlywiki5 or logseq give — not a smaller feature set, but an optional one.

Its distinctive contribution: code in the document

Org Babel — executable source blocks in many languages inside a single file — is the piece nothing else in this corpus matches, and it has a peer-reviewed account of its own (org-mode-literate-programming). It puts Org in two of the spoke's threads at once: tools-for-thought, and the executable-markdown / literate-computing lineage that runs from augmenting-human-intellect through today's notebooks.

From that paper, four details the manual does not frame this way:

• Org does both literate-programming and reproducible-research in full, which its authors claim no tool but Scribble had done, and it is the only one of the ten they survey that passes data between blocks written in different languages.
• The block-reference syntax for reordering code on tangle is noweb's, adopted unchanged — Org did not invent a notation there, it joined an existing one.
• Adding a language is a small contract: tangling, editing and export are language-agnostic, and only evaluation and live sessions need per-language Emacs Lisp functions. 30+ languages by 2012, and by then users rather than maintainers were writing most of the new ones.
• Running a block asks permission by default (org-confirm-babel-evaluate), argued for on the grounds that a reader cannot audit code in a language they do not know.

The paper also dates the reach: as of Emacs 24, all of this ships in the Emacs core rather than as an add-on.

But Babel is not why most people are here. In emacs-user-survey-2020, among the 4,980 respondents who described what they use Org for, only 10.2% mention Babel, literate programming or source blocks at all; notes and journalling draw 61.6% and agenda/TODO/GTD 38.9%. The feature this corpus singles out as Org's unmatched contribution is a minority practice inside Org's own user base. Both things are true — Babel is what nothing else does, and note-taking is what nearly everyone came for — and the gap between them is filed in synthesis.

Uptake

Measured at last, in org-mode-adoption. The headline: 79.7% of Emacs survey respondents use Org and 51.2% use it daily or exclusively; org-mode is 20.9% of all questions on emacs.stackexchange.com and 0.0088% of questions on Stack Overflow. Deep saturation of a small population, near-invisibility outside it — and no denominator anywhere, since the number of Emacs users has never been measured, so Org cannot be placed on an adoption curve at all.

Refined 2026-08-30 (claim drain). "Never been measured" is too strong. What does not exist is a census; two published extrapolations do exist and they disagree by more than an order of magnitude. The 2020 Emacs User Survey (~7,000 respondents, run by Adrien Brochard) supports roughly 90,000 users worldwide when scaled through the ratio of self-reported MELPA maintainers to total MELPA packages; Debian popularity-contest data points instead at a few million. That gap is itself the finding: the disagreement is about what counts as a user of a free editor, not about Org. The denominator problem stands, and now its reason is named rather than asserted.

Still not read here: how Org has fared against the newer generation head to head. The comparison with obsidian, logseq and roam-research would need a user count for each, and Org's does not exist.