scale_name property should be removed from Scale objects

[wch]

One issue with simply removing this property is that the Scale objects will no longer have labels that indicate their subtype. For example, the brewer and grey scales are subtypes of color scales.

[thomasp85]

Is there a real issue or is it just a nicety? I cannot find anywhere in the code where this property is queried so there should be no problem in just removing it. On the other hand there doesn't seem to be an urgent need for the removal either as it is just a label...

[hadley]

Mostly just a nicety as part of continuing to clean up the API so it's easier to understand

[hadley]

It should really become the ggproto() class name, not a separate attribute.

[hadley]

But this doesn't seem super important, so lets not worry about for this release

[hadley]

Needs a little thought as ScaleContinous$transform() currently does:

type <- if (self$scale_name == "position_c") "continuous" else "discrete"

[paleolimbot]

Scale$scale_name should probably be set to the function that the user used to create the scale (e.g., "scale_x_continuous") so that it can generate the appropriate error message (I can't think of a better way to generate an informative error message given the other information available at transformation time). The current values of Scale$scale_name() only generate good error messages for x and y scales (which most commonly generate errors).

I do, however, feel that Scale$call should be removed, since it is never used.