RAM Tracker

Newegg & Amazon product pages — prices come from live retailer product pages (Newegg /p/N82E168…, Amazon /dp/…), not search results or third-party feeds.

Raw-first archive — every page fetch is stored in full before parsing, so prices are derived from the original HTML we captured, not a stale cache.

RAM-only catalog — mini PCs, laptops, full systems, and other non-kit listings are filtered using title, spec, and structured-data checks before anything enters the catalog.

In-stock only — we skip unavailable listings. A page must show a buyable price before it becomes a product or price observation.

Spec validation — a listing only counts when we can extract desktop RAM details (DDR generation, capacity, and speed) from the page.

Pool cleanup — queued URLs that never produced a price, or unfetched links with no RAM context in stored pages, are removed so the catalog does not fill with junk.

Verified links — the top bar shows how many Newegg and Amazon URLs have at least one confirmed price observation each.

Sale context — when a retailer shows list vs. sale pricing, we capture both plus savings — not just the headline number.

Polite collection — one product page at a time, long pauses between requests, no search spam. Related kits are discovered from pages we already fetch.