search

opengeospatial / ogc-feat-geo-json

Apache License 2.0
32 stars 14 forks source link

Avoid the use of start instant + duration #32

Closed chris-little closed 2 years ago

chris-little commented 2 years ago

In section 7.2 there is a highlighted Note: ISO 8601 also supports intervals by a duration (a start instant and the duration or the duration and an end instant). Should this also be supported or does that make parsing just more complex for clients? I propose not to do this because IETF RFC 3339 very carefully does not. The RFC only establishes the monotonic ordering of timestamps in a restricted ISO8601 notation. The RFC avoids any hint of differencing two instants to get a duration. The subtraction is not deterministic, at the precision of a second, or perhaps a day, if the calendar algorithms and intercalations (leap seconds, leap days, etc) are not well understood. The ordering of the timestamp instants is rigorous providing only one clock is used, thus comparisons (before, after, equal) give proveably correct results.

cportele commented 2 years ago

Thank you for highlighting this important aspect. I agree, it would be safer to not support durations.

cportele commented 2 years ago

I am reopening this, because we still need a decision in the SWG what to do with the note in the document.

cportele commented 2 years ago

Meeting 2022-02-21: We agreed that adding support for duration adds complication that we don't need or want in Part 1. @cportele will remove the highlighted text and then close.