It's interesting to see the typo fixes and how long different stories last. At 2:11 a story relevant to us all. I wonder (with regards to his point about things disappearing forever) if a distributed effort among "tech people" to catalog a lot of websites would work, different people run nodes that are given tasks. Would be neat to keep a track of the top 1,000 article driven sites for a year (not just news, reddit, hn etc.)

> Due to an errant cron task that ran twice an hour

This has happened to me a few times. Recently I had accidentally collected over 100,000 copies of a website over a 3 month period after forgetting the thing was running.