So, this is the problem of persistence of URL's always referencing the original content, regardless of where it is hosted, in an authoritative way. It's an oka

So, this is the problem of persistence of URL's always referencing the original content, regardless of where it is hosted, in an authoritative way. It's an okay idea to link to WB, because (a) it's de facto assumed to be authoritative by the wider global community and (b) as an archive it provides a promise that it's URL's will keep pointing to the archived content come what may. Though, such promises are just that: promises. Over a long period of time, no one can truly guarantee the persistence of a relationship between an URI and the resource it references to. That's not something technology itself solves. The "original" URI still does carry the most authority, as that's the domain on which the content was first published. Moreover, the author can explicitly point to the original URI as the "canonical" URI in the HTML head of the document. Moreover, when you link to the WB machine, what do you link to? A specific archived version? Or the overview page with many different archived versions? Which of those versions is currently endorsed by the original publisher, and which are deprecated? How do you know this? Part of ensuring persistence is the responsibility of original publisher. That's where solutions such as URL resolving come into play. In the academic world, DOI or handle.net are trying to solve this problem. Protocols such as ORE or Memento further try to cater to this issue. It's a rabbit hole, really, when you start to think about this.