[raster]: ST_HillShade why did we switch to degrees in 2.1

[robe]

​http://postgis.net/docs/manual-dev/RT_ST_HillShade.html

I propose we switch back before release. I'm flagging this for 2 reasons

1) It's a breaking change from 2.0 2) We dinged pramsey for trying to sneak in use of degrees in his ST_Project function.

Our general convention is to use radians unless if there is another over-aching standard that requires degrees and just demonstrate in the docs how to flip between the two using the built in PostgreSQL degrees and radians functions as demonstrated in:

​http://postgis.net/docs/manual-dev/ST_Project.html

and ​http://postgis.net/docs/manual-dev/ST_Azimuth.html

[dustymugs]

If you look at the docs for ST_Slope, ST_Azimuth and ST_Hillshade in 2.0, there was absolutely no units described. So, when I was going through the code to make sure the results matched that of ArcGIS (I assume most folks are going to compare to that application), I noticed that ArcGIS uses degrees by default. And since our functions where not documented, I went with degrees…

​http://postgis.net/docs/manual-2.0/RT_ST_Aspect.html

​http://postgis.net/docs/manual-2.0/RT_ST_HillShade.html

​http://postgis.net/docs/manual-2.0/RT_ST_Slope.html

Personally, I like degrees…

[robe]

Okay I'm fine with that. Hopefully pramsey isn't listening cause then he'll want his degrees back.

[dustymugs]

Also, gdaldem and GRASS use degrees…

​http://www.gdal.org/gdaldem.html

​http://grass.osgeo.org/grass64/manuals/r.shaded.relief.html