> I propose that 3 digits be used for a decade. [...] Thus '196' would have the same meaning as '196x'. [...] Comments?
What about years bigger than 9999? How would we represent the decade 10000 to 10009? We would need different ways to represent decades. Using the x seems more consistent, I think, because it would work for any year.
Likewise, I think that two-digit-centuries should be discouraged, because they cannot cover a "century" like the one from 10000 to 10099. Having completely different ways to represent centuries (and decades) before vs. after the 9999-10000-boundary is a bit more complex than needed.
> If ‘19’ means the same thing as ‘Y19e2p2’ (and same as 1900P2)
Let's clarify this one! Does "P" applies to "the future"? What would "1950P2" have as boundaries: (1900 and 1999) or (1950 and 2050)?
> if one of the characters '?', '~', '%' is included to the left of one (or more) of the date components, then that character should not occur to the right of any component.
> Use ‘?’ (or ‘~’ or ‘%’) to the right of a component rather than to the left of that component as well as all components to the left.
How should we meet an agreement between those two rules? Should we write "?2015-02-?31" or "2015?-02-?31" while none of them follows both rules mentioned above?
> (2) Drop the "P" notation, and only support two mechanisms for precision:
> (a) For scientific notation, anything to the right of the decimal is considered imprecise; AND,
> (b) For non-scientific notation, use "X" as a placeholder for any number of right-hand digits (as we do now, but extending the concept).
Maybe, this rule (2) could be made even simpler. What about:
Drop the "P" notation, and only support ONE mechanism for precision:
For scientific notation, anything to the right of the decimal is considered imprecise; AND,
For non-scientific notation, then use ... scientific notation instead of non-scientific notation.