On Tue, 2011-05-31 at 10:14 -0300, Stefan Monnier wrote: > > This is assuming the theoretical story is correct in practice. In > > practice, I've seen flashes (especially NAND, but also NOR) die a > lot > > sooner then that.... especially in cheap consumer hardware :( > > But these cases aren't avoided by writing less.
They actually are, as their # of cycles are basically just lower then "average". So less writing means less failure means longer life. They are often used as the manufacturer makes sure that the software *they* deliver do very little writing. I know the OpenWRT situation is a little different, with much more frequent flashing often, and much more functionality, and therefore writes, but that is still not a good reason to waste write cycles. Ithamar. _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
