And i can imagine, when they just doubled the amount of possible characters, they just correlate these tweet buckets now. Like, maybe each tweet already had spa

And i can imagine, when they just doubled the amount of possible characters, they just correlate these tweet buckets now. Like, maybe each tweet already had space for attributes/opcodes, so one can express things like "deleted". Maybe there is another attribute now: "has successors". And in the API one just return many tweets as one. In that way, one does not have to rewrite history. But who knows!