Web scraping breaks first at the network edge. Before parsers, selectors, or pipelines get a chance, the connection either completes or it doesn't. That edge has changed: almost half of web traffic is ...