I should have qualified that "PT0S" was a suggestion of mine for a common ailment in date typing across many different descriptive standards, and not a specific suggestion to make Jenn's EAD validate. For EAD validation, Charles' suggestion from ISAD(G) below is the best recommendation, it would seem.
PT0S, like I said, is certainly inelegant unless a schema (whether EAD or not) uses xs:duration. Dates are complex enough in archival description that it would be difficult to implement in EAD, I would imagine.
TEMPER is a step forward in terms of dealing with these issues for archival description. The tilde notation for uncertainty, and the ability to list multiple dates (e.g. 1970, 1974-1976, 1999) that are common in finding aids are nice aspects. I hope it can make it into some of the community's descriptive standards in the future.
>>> Charles Blair <[log in to unmask]> 12/06/07 11:32 AM >>>
Record as a single date or a range of dates as appropriate. A range
of dates should always be inclusive =>unless<= the unit of description
is a record-keeping system (or part thereof) in active use.