r.slope.aspect converts to meters

[annakrat]

I am getting here to an old problem when r.slope.aspect converts automatically horizontal distance to meters. Details are described in this old discussion (only Helena and Glynn involved):

​http://lists.osgeo.org/pipermail/grass-dev/2004-January/013630.html

As a result of this behavior, users with location in feet, have to use zscale parameter 0.3048 to get correct result. Most users have their DEMs with the same horizontal and vertical units and they would assume no vertical scaling is needed, and they end up with incorrect results.

This behavior is very unintuitive and is source of many errors.

I attached a diff. I added a warning when running it in location with units other than meters and zscale different than one (that's the case I was mentioning above). I know some old scripts will start to give wrong results, but we can avoid future problems of new users.

[annakrat]

r.slope.aspect diff

[annakrat]

If no objections, I will commit it tomorrow.

[annakrat]

Replying to annakrat:

If no objections, I will commit it tomorrow.

Done in r64177 and backported.