WCS scaling extension: scalefactor is handled wrongly

[Dimitar Misev]

The WCS standard OGC 12-039 states:

Requirement 5 getCoverage-scale-by-factor-positive:
The Scal::scaleByFactor parameter in a GetCoverage request, if present, shall have a
positive float value.
Note A value of 1.0 leaves the coverage unscaled, a value between 0 and 1 scales down (reduces target domain), a value greater than 1 scales up (enlarges target domain).

In petascope however, a scalefactor > 1 scales down, which seems wrong.

[Bang Pham Huu]

yes, scalceByFactor=4 will return ¼ image (scale down) and scaleByFactor=0.25 will return an image (size x 4),are not intuitive for a very long time, but that is what the scaling extension test of CITE expected ​https://github.com/opengeospatial/ets-wcs20/blob/master/src/main/scripts/ctl/ext_get-scal.xml. That means, to fix this problem, it will need to fix in Petascope first and ask CITE to update script later or the scale extension test will not pass.

[Dimitar Misev]

we should ask them to fix the test right away in my opinion.

[Peter Baumann]

Reason behind is that common scale factors should be integers; otherwise, tricky rounding issues might occur.

[Dimitar Misev]

@Peter should we contact Jinsongdi about fixing the test or someone else?

[Bang Pham Huu]

CITE is acknowledged with ​https://github.com/opengeospatial/ets-wcs20/issues/43