- From: George T. Joseph <gtj@peakin.com>
- Date: Fri, 21 Jan 2000 13:27:57 -0500
- To: <www-xml-schema-comments@w3.org>
I'm writing the timeInstant and timeDuration validators for Xerces-Java and need some clarification. Is a negative timeInstant meant to represent time B.C.? I'm presuming "yes". What sign would you assign to the following timeDuration results: "CCYY-MM-DDThh:mm:ss.sss/PnYnMnDTnHnMnnS" presume positive. "CCYY-MM-DDThh:mm:ss.sss/-PnYnMnDTnHnMnnS" presume negative. "-CCYY-MM-DDThh:mm:ss.sss/PnYnMnDTnHnMnnS" presume positive. "-CCYY-MM-DDThh:mm:ss.sss/-PnYnMnDTnHnMnnS" presume negative. "PnYnMnDTnHnMnnS/CCYY-MM-DDThh:mm:ss.sss" presume negative. "PnYnMnDTnHnMnnS/-CCYY-MM-DDThh:mm:ss.sss" presume negative. "-PnYnMnDTnHnMnnS/CCYY-MM-DDThh:mm:ss.sss" presume positive. "-PnYnMnDTnHnMnnS/-CCYY-MM-DDThh:mm:ss.sss" presume positive. The assumptions are based on the idea that if the resultant time instant occurs after the base instant then the duration is positive and if a resultant time instant occurs before the base instant then the duration is negative. Please confirm or correct. I'm assuming that you meant to include right-truncated forms timeInstant and didn't mean for it to ALWAYS be expressed down to seconds to 3 decimal places. Thanks, George Joseph
Received on Friday, 21 January 2000 13:29:08 UTC