Vintage Store Map Deprecated
Mapped Los Angeles vintage stores by category and vibe, to spot where a new one fits. Retired, no longer updated.
Deprecated. The scheduled monthly refresh was switched off and the map is no longer updated, so its page was removed rather than keep showing stores that may have closed.It's no longer updated, so the map was taken down rather than keep showing stores that may have closed.
Searched as a Box, Counted as Districts
A box is how you find everything without missing the edges; a district is the only unit in which 'there is nowhere to buy this near here' means anything. 2,277 stores found, 2,174 read, 453 on the map, the rest read and correctly ruled out. 28 of 35 districts now hold evidence.
The search area is a rectangle, but the answers are reported by neighbourhood, which is the only unit in which 'there is nowhere near here to buy this' means anything. 2,277 shops found, 2,174 looked at, 453 on the map. Most of the rest were not secondhand.
A Closed Vocabulary, and Low Confidence Allowed
A fixed list rather than free text, because a map you can filter needs categories that mean the same thing twice. Confidence is recorded and allowed to be low: a shop the read was unsure about is marked unsure rather than filed confidently in the nearest bucket.
Every shop is sorted into one of five kinds, from a fixed list, so the map can be filtered. If it is not sure about a shop, it says so instead of guessing.
Politely, and Slowly, On Purpose
One request at a time with a two-second pause, behind a User-Agent that says what it is. Nothing about the job needs to be fast, and a crawler that hammers a small shop's site to save an hour is a bad trade made on someone else's behalf.
It reads one site at a time with a pause between, and identifies itself honestly. Nothing here needs to be quick, and hurrying would cost the shops rather than me.
The Gaps Are the Output
An absence is the finding, so it is stated rather than left as blank space. The report is deliberately unranked. Naming a 'best' gap would be a recommendation the data cannot support.
A re-run leads with what moved, and it now runs monthly on its own, the pace shops open and close at.
What is missing is the point, so it says what is missing rather than leaving a blank. It does not rank the gaps, because that would be a recommendation it cannot back up.
A re-run leads with what changed. It now runs on its own, once a month, the pace shops open and close at.