[Date Prev][Date Next] [Chronological] [Thread] [Top]

[Syntaxes] Missing GeneralizedTime/UTCTime restrictions



Other problems with GeneralizedTime and UTCTime:


Day numbers 29-31 should not be allowed in all months.  Or at least, the
implementation should be allowed to treat invalid dates as errors.

Maybe the wording should be a bit vague about that which dates are
valid, so an implementation can be allowed to either treat all years as
Gregorian or to properly treat pre-someting years as Julian.  (The
Julian calendar had leap years every 4. year including each 100. year.)
Also, there is the 'hole' in the calendar where some 10 dates are not
valid at all, since they were dropped realign the Gregorian calendar
with the seasons.


Also, maybe the implementation of GeneralizedTime should be allowed to
treat "false" leap seconds as errors.  Either by checking against a
table of leap seconds that have occurred so far, or with a rougher check
that the leap second is at 23:59:60 UTC.  (Leap seconds do not happen at
other times of the day.)

-- 
Hallvard