[Ietf-calsify] Resolutions of issues 8, 68 and 69: DST discontinuities

[Jeffrey Harris]

Hi Jonathan,

> The third bullet point here is the potentially controversial one, and would
> be an incompatible change with 2445.  Does anyone currently generate
> DURATION values of the form "P1DT8H"?

Chandler doesn't currently serialize durations, but internally we store
durations and I've been trying to move things over to serializing
DURATION instead of DTEND.

I'm not following why we'd disallow "P1DT8H".  What's wrong with this
heuristic for interpreting such a value: Add the nominal time first,
then add the accurate time delta?

Sincerely,
Jeffrey