If you can spot scraping, maybe have a list of randomly selected ‘plausible’ txts that you silently start serving instead of the real ones once the scraper threshold is tripped?
(Maybe throw in things like half the fortune database, the I-ching, the text of the Kama Sutra for good measure - plausible stuff, but wrong.)
The scraper won’t know they’re not getting the real thing, but any ‘archive’ or Google index-y redirect trickery would just be completely and utterly incorrect.
I’d be happy to contribute false-but-plausible messages to add to a random file for that purpose.
