Interesting, I (obviously) wasn't aware of that.  Thanks.

--Terry

At 04:59 PM 3/23/2006 -0500, Charles M. Hannum wrote:
>On Thu, Mar 23, 2006 at 03:49:36PM -0500, Terry Moore wrote:
> > Yes, but because of lack of rotational latency sometimes people do
> > things substantially differently.  Also, non-local error properties
> > of NAND flash change things.  Erase blocks in NAND are some multiple
> > of sectors.  Writing a sector in an erase block can cause other
> > sectors in the erase block (that you've already written and verified)
> > to grow errors.  As far as I know, the error properties of CD-RWs and
> > DVDs are not as bad as this.
>
>Actually, they are.  Sectors are interleaved pretty aggressively to
>disperse multi-bit errors from scratches, so writing one sector can in
>fact damage other sectors -- possibly worse than in NAND flash.