BBQ Atlas started as a personal frustration: every "best BBQ" list was either a national magazine's top-ten or a Yelp dump. Neither answered the real question — where is the serious pit in this county? The kind of place that doesn't have a website, opens at 11, and sells out by 1.
The v1 database was a spreadsheet pretending to be a database. It worked well enough to prove the concept — geography as the primary lens, not cuisine category or star rating. The map became the interface, not a feature.
v2 is a proper PostGIS-backed store with a new API, a new map renderer, and a submission pipeline so other people can contribute pits. The goal is to document every serious pit in the American South — not the famous ones, the real ones.
"Geography is the ingredient that nobody talks about. The same recipe tastes different forty miles north because the wood is different, the hogs are different, the tradition is different."
The technical stack is deliberately boring: Django for the data layer, PostGIS for spatial queries, React for the map interface. The interesting engineering is in the data — how to represent a pit's style without flattening it, how to handle the fact that regional boundaries are fuzzy, how to let the geography tell the story.