Primary material
Manuals, data sheets, service manuals, original source releases, contemporary developer notes, and scans of period magazines or books.
Sources & standards
Code Like It's 198x is part curriculum, part reference shelf, and part public workshop. That only works if the site is clear about evidence: what was run, what was read, what was inferred, and what still needs checking.
The source ladder
stronger evidence firstManuals, data sheets, service manuals, original source releases, contemporary developer notes, and scans of period magazines or books.
A lesson program that builds from the repository, runs in the named emulator or on hardware, and produces the screenshot or behaviour shown on the page.
Photos, capture cards, logic analyser traces, disk or tape flux captures, and other evidence taken from real machines when a claim needs that bar.
Modern retrospectives, interviews, emulator documentation, preservation projects, and community research used with attribution and checked against stronger evidence where possible.
What each section owes you
site contractThe code must build, run, and show the thing the prose says it shows. If a ROM is needed, the page says so without redistributing it.
A pattern is reusable only after it exists as working code, with the system, language, and limits named clearly.
Historical claims need a source trail. Primary sources win; oral history and modern writing are useful, but they are labelled for what they are.
Images should be generated from our builds, our emulator captures, real hardware captures, or software we have actually booted. Captions should say enough to understand the origin.
What counts as valid
source rulesEvidence labels
plain English markersThe source was built from the repository.
The program was run in the named emulator or on named hardware.
The screenshot, image, audio, or video came from that run.
The factual claim has a source trail.
The prose is drawing a conclusion from sources or observed behaviour, not quoting a source directly.
The page is honest that the claim is not fully verified yet.
AI is not a source
AI can help rough out prose and code, but it cannot stand in for a manual, a trace, a build log, a screenshot, or someone who was there. If an AI draft introduces a fact, that fact still needs an ordinary source before it belongs on the site.
Screenshots, copyright and ROMs
The site can show screenshots and short clips from software that has been booted for testing, review, or documentation. It does not distribute commercial ROMs, disks, tapes, Kickstarts, kernels, or copyrighted assets as downloadable course material.
Emulator versus hardware
Emulator evidence is valid for lessons, screenshots, normal program output, and the behaviour learners are expected to reproduce. Hardware evidence matters when a page claims exact timing, analogue display behaviour, physical media behaviour, or anything where the emulator itself is part of the question.