>we're facing Reddits shutting down left and right, losing over a decade of arcane and non-trivial information about various processors, microarchitectures, non-documented standards and other artifacts of knowledge that you will be hard-pressed to find elsewhere.
This sucks, but people should have known better than to entrust important information to a for-profit company that already has a bad track record. If you want anything to last the ages, it needs to be backed up in multiple places by different people.
That's a really poor argument. Now I entrust.... who exactly? How is this a sell for "6 months ephemeral data"?
Also, who is guaranteeing that? Who says it isn't going to be scraped? What's the point if not making historical data hard for the average user to find and easy for a malicious user to find?
You don't entrust anyone, it's simple. If you want data to last a long time, it has to be distributed, and not under the control of any single entity. Otherwise, I can virtually guarantee there will be a failure, whether it's because of maliciousness or just an accident (how many old Hollywood movies were lost because of one big fire?). (https://en.wikipedia.org/wiki/List_of_lost_films) (https://en.wikipedia.org/wiki/1965_MGM_vault_fire)
This sucks, but people should have known better than to entrust important information to a for-profit company that already has a bad track record. If you want anything to last the ages, it needs to be backed up in multiple places by different people.