Panu Lahtinen

If the area reduction in https://github.com/pytroll/satpy/blob/master/satpy/scene.py#L897-L906 is commented out, the results are as expected.

Seems to be a bug in `pyresample.spherical_geometry.intersection_polygon()`:  In green: EPSG:4326 boundary, in blue: GOES-15 boundary, in red: intersection. The minimum longitude of the intersection is -187.376... degrees, and should...

I know @mraspaud started something, but I guess he got distracted on something else. I'll check with couple different scenes tomorrow morning to check if there are differences in resampling...

@ameraner good point! I'll see if I can come up with something. @djhoese thanks for the fixes.

Maybe this is better regarding the other decompression possibilities?

I downloaded the CI logs, and the new tests seem to be failing with: ``` 2022-03-04T13:27:02.2647707Z > super(NetCDF4FileHandler, self).__init__( 2022-03-04T13:27:02.2648022Z filename, filename_info, filetype_info) 2022-03-04T13:27:02.2648393Z E TypeError: super(type, obj): obj must...

Also, in Python 3.10 the `hdf5plugin` needs to be import before `Scene` creation, otherwise the NetCDF4 handling fails with `RuntimeError: NetCDF: Filter error: unimplemented filter encountered`. So either the import...

Unfortunately I haven't had time to dig into this. Maybe after Easter holidays.

I'm now thinking that maybe we should just indicate the different decompression possibilities for this reader and not include the import in Satpy in the first place.

The rebase was a bit more complicated than it should've been. Hopefully git history isn't completely effed-up. I'll re-check the documentation and error message @mraspaud commented on after lunch.