scrape-based processes feels like 'tech debt as a service' and yet

we're building more technology and supporting less of it. the future will include a lot of obsolescence (planned + unplanned), and 'approximate compatibility' services like this one may have a bigger role than we want