BeatlesAnswers.org has a narrow mission by design: the UK Beatles canon — archival, recording, and session history — treated with the specificity it deserves. Not another lyrics site, not a news feed. The questions we want to answer are the ones that send people to the back half of a Mark Lewisohn book: which take, which studio, which console, which mix.
The first architectural decision was the least fashionable one available: static HTML. No CMS, no client-side framework, no database. Every page on this site is a self-contained HTML file generated ahead of time and served as-is. For a reference work whose facts change on the timescale of official reissues, that trade is nearly all upside — speed, security, zero runtime dependencies, and pages that can be audited byte-for-byte.
The generator chain
The site is produced by a Python generator chain from structured source data: a canon spreadsheet covering the released catalogue, and an extracted text corpus of Lewisohn’s The Complete Beatles Recording Sessions (1988) used for cited quotes. The generator emits a complete versioned snapshot of the site into its own deployment folder — deployments/v1/, v2/, and so on — while all working material (scripts, corpora, caches, audit logs) stays in separate build directories that are never deployed.
That separation earned its keep immediately. We went through eight versions in the first week — comparison audits of v1 through v3, then successive quality passes — and because every version is a clean drop-in deploy, rolling forward or back is a folder swap, not an archaeology project.
Shipping
Version 8 closed the week: a multi-pass quality build, zipped and deployed to Cloudflare Pages behind the beatlesanswers.org domain. The footer on every page reads “Built 2026-05-11” — the day the site first came together as a coherent whole.
What is deliberately not here yet: monetisation beyond the scaffolding (an AdSense account and affiliate details are prepared but nothing is submitted), and the deep per-song recording detail that is the whole point of the project. The image library alone runs to roughly 2 GB of source material awaiting curation. The plan from here is enrichment — and the discipline to version every step of it.