PL history research, Sep 2026: bundle index
17 Sep 2026. This page is the 13 Sep synthesis of an external research run. The Mo it describes — brace syntax, effect rows, a Rust implementation, a package registry, an RFC process — was never Mo’s design. Mo’s decisions are the spec chapters under
spec/design-v0/and the decision-log. Read this page as landscape only.
Headline
A 2026-09 research bundle covering ninety years of programming-language history, eight design camps, an implementation engineering menu, and a synthesis for Mo. Every wiki summary in this section links back to the raw material under raw/plang-history-2026-09/.
What’s in the bundle
The bundle is organized as five nested layers, from surveys to Mo-specific decisions:
History (three eras) — a source-cited walk through the field:
- plang-history-lambda-to-1970s — Lambda calculus through the paradigm explosion of the 1970s
- plang-history-1980s-to-2000s — Objects, functional purity, the internet, and virtual machines
- plang-history-2010-to-2026 — Systems renaissance, verification, and the AI era
Design camps — the ideological landscape:
- plang-design-camps — Eight camps: paradigm, types, memory, concurrency, syntax, compilation, philosophy, ecosystem
Implementation menu — the engineering choices:
- plang-implementation-menu — Parsing, IRs, back ends, GCs, package managers, tooling
Mo synthesis — the point of the exercise:
- plang-mo-synthesis — What history says to Mo
- plang-decision-matrix — 18 axes, each with a Mo direction and rationale
Language-by-language deep-dives
Fifteen numbered deep-dives sit under raw/plang-history-2026-09/deep-dives/; thirteen of them are summarized in research/languages/, the other two (rust, elixir) in research/comparisons/:
- c · lisp · smalltalk · ml · haskell · erlang · cpp · java · python · javascript · go-history · zig · dependent-types
- (rust — already covered under
research/comparisons/rust.md) - Elixir has both a language page (elixir under comparisons) and is treated in the history summaries.
Three findings that shape everything
- Annotation economics are inverted. When agents write code, an annotation that costs the language designer thirty minutes and buys machine-verifiable structure is nearly free — the human review it saves compounds forever. Ceremony that was expensive to humans is cheap to agents.
- Supply-chain security is the highest-leverage decision. Every modern package ecosystem (npm, PyPI, crates.io) has produced live incidents (typosquatting, dependency confusion, post-install scripts, slopsquatting). A language whose module and dependency system is validated at edit time avoids an entire class of production breach. See q17-package-management-and-supply-chain and d30-supply-chain-security.
- Effects + capabilities + regions are complementary, not competing. The 2010s treated them as rival research programs; the 2020s (Austral, Hylo, Koka, Roc, OCaml 5) show they compose well. Mo’s d15-effects-via-capabilities takes this seriously.
How to use the wiki pages
- The history pages are for orientation — read them once, refer back for citations.
- The camp and implementation pages are for design decisions — read them when you’re about to choose a direction.
- The Mo synthesis and decision matrix are the working outputs — this is where “what should Mo do?” gets answered with links to specific research.
- The language deep-dives are for depth on a single tradition. Read them when a specific language is being cited as a model or anti-model for Mo.
Cross-references
- Existing overview: research-summary-2026-09
- Existing shortlists: steal-list, fork-in-the-road
- Related concepts already in the wiki: language-landscape, case-against-new-languages, capability-module-lineage, supply-chain-defenses, comparison-synthesis-draft