so that we don't use values before discharge for the long-term delta
calculation.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@1899 a1c6a512-1295-4272-9138-f99709370657
samples taken while the disk is spinning. If no 'disk idle' samples are
available, use the ones we have anyway (yeah, maybe we could interpolate too).
Also tuned the deltas detection too, and fixed some comments. This seems to
work quite a bit better for me.
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@1682 a1c6a512-1295-4272-9138-f99709370657
