That seems kind of dumb, but with Moore's Law still mostly working for storage, it's also cheap -- certainly cheaper than building actual libraries. There's also currently no way to mark URLs as temporary, so we can't distinguish the permanent from the ephemeral. Finally, it's far cheaper than having everyone who wants to save URL content archive it on their own personal box (plus however many backups), which is what I do now if I really want to use something again.
No comments yet.