The volatilty of digital archives is vexing in a lot of ways.
I read a while back that, in the 1980s, the Canadian Broadcasting Corporation committed a large portion of its video archive to a custom-developed laser disk format. Years later, the disks remained but no one had thought to preserve the playback machines or to transfer the material to more current formats. I think, in the end, they had to rebuild one of the machines in order to retrieve the stuff.
Web-based text is odd in a different way. It doesn't just disappear, but it can change over time, and not necessarily for very good reasons. Stories on sites like BBC News, NYT, etc. are constantly being revised, sometimes , both on the day of publication and subsequently, often as a result of political or corporate pressure. A writer might reference a particular article which said one thing at the time of her/his writing, only for it to tell a different stroy when a reader follows the link. I always take a PDF snapshot of anything I'm referencing for that reason. Outside of morning and afternoon editions of a newspaper, that sort of disappearance of text doesn't occur in print. The storyline might change in subsequent editions but the original document persists.