On Tue, 8 Mar 2011 07:58:01 +0100, Karin Bredenberg wrote
> Sounds like the solution to me.
> >>> Ray Denenberg <[log in to unmask]> 2011-03-07 20:36 >>>
> I've changed the subject line (was RE: [DATETIME] Sv: [DATETIME] Last
Call for Use Cases)
> From: Karin Bredenberg
> Sent: Monday, March 07, 2011 10:27 AM
> > Cant find and may have missed, if something is undated and is stated to
> > be undated how to express that?
> There is nothing in the current draft spec that directly addresses this,
but here's one approach that I think would fit well with the current syntax.
> The letter 'u' is used to mean "unspecified", so 199u means "some
unspecified year in the 1990s" and 1999-uu means "some unspecifed month in
1999". I think it would be completely logical for uuuu-uu-uu to mean "date
> We could add this as an explicit example.
I think the predicates for uuu-uu-uu--- and for that matter any use of u---
should be opened. Unspecified means that something is perhaps known but not
expressed. Unknown means that the something is not known. Undated does not
mean unknown date nor unspeficied date but expresses that no date has been
There is, however, more to it..
This goes back to the dicussion I tried to open some months ago on
readability and precision. Does 1900-uu mean something different than 1900?
I have argued that they are not the same:
- 1900 means the year 1900 with a precision of year.
- 1900-uu means 1900 with a precision of month.
what does 19uu-uu mean?
I suggest known or determinable (or else we could not speak of precision) in
the 20th Astronomical century with a precision of month.
In this light the following:
are not the "same".
Edward C. Zimmermann, NONMONOTONIC LAB