Query timestamps lead to empty patch

[Modexus]

Given the warping of tiles (#548) and the handling of the query mint and maxt values even tile based datasets (NAIP) return empty patches sometimes.

(mint, maxt) are based on the hit but different tiles have different capture dates which lead to queries that the original hit does not contain (because of warping) but the next tile is not returned in the intersect as it has a non-intersecting (mint, maxt) with original hit.

It may make sense to set (mint, maxt) always to the values in the roi instead of intersecting with the hit.

[adamjstewart]

It may make sense to set (mint, maxt) always to the values in the roi instead of intersecting with the hit.

I think this depends strongly on the particular application you are working on. For example, if you are working with time series data (#640), then this solution wouldn't be useful. In general I think we need more builtin samplers to handle these kinds of situations, one of which could very well include the suggestion you describe. I'm going to move this to a discussion since it's not exactly a bug (it's by design).