It turns out that link rot and content drift are endemic to the web, which is both unsurprising and shockingly risky for a library that has “billions of books and no central filing system.”