Matthew Dillon wrote:
Most of the arguments quoted are incorrect. The one about the ECC
length is correct, but the inter-sector gap argument doesn't apply
to most modern drives because they already do full-track reads and
writes, without gaps between sectors.
Yeah, I'm a bit puzzled by this, too. I think they meant that drivers
and firmware will need to deal with 1/8th of the number of blocks. I
could argue either way whether this is that big of a deal -- on the one
hand, silicon is still many times faster than drives; on the other,
linear bit density increases faster than silicon.
