[lcuff]

Fun article, but the example of Hilbert Curve usefulness with hard drives is quickly aging out. SSDs, which are increasingly prevalent, have more-or-less constant access time. There is no equivalent of the seek time or rotational latency that slows down HHD access when successive reads are on different tracks, at different rotational angles.

[jandrewrogers]

It wasn’t even really a thing for HDD. There was a broad consensus that it was inefficient for such purposes by the 1990s. If you had random access at all, you could do better.

The killer app was tape drives.