> ... Should we print the "days" +/-
> sign only when it is not the same as the "year" sign?  Seems like the
> way to go.

It isn't quite that easy, but you have the right idea. My new code omits
signs unless there is a negative field value, then all subsequent fields
are signed.

That seems to be less ambiguous. fwiw, the traditional "Postgres format"
never was right in this regard (though internal calculations were done
correctly).

                    - Thomas

Reply via email to