COLORITEM & kin

[Attila Csipa]

On Thursday 10 May 2007 16:29, Frank Warmerdam wrote:
> > Is there interest (as in would you support) development of a COLORITEM
> > (and related BACKGROUNDCOLOR, etc items) functionality ? This would work
> > as LABELITEM and similar attribute related tags. The rationale behind
> > this is that I have maps with too many colours to squeeze them into
> > separate classes, and worse yet, these should (dis)appear dynamically. Or

> If we don't have a way of reading the color from the feature then I think
> this would be a desirable addition in some fashion.
> Is this intended to be for continuous ranges like RFC 6, or where the color
> value to be used is read from the feature?

The original intent is the latter, but these might actually overlap. From a 
consistency standpoint COLORITEM should interpret values, since that's what 
all [prefix]ITEM keywords do. Note that you cannot replace this functionality 
with gradients as they move on only one axis, while the value interpretation 
holds all of the (in a common case) 3 dimensional colour data.