[Proj] datum matters

[Richard Greenwood]

2009/3/28 strebe <strebe at aol.com>:

> This notion of having to think in terms of coordinate systems, not just projections, is only true within some domains. Large-scale maps? Yes. Medium and especially small-scale maps? No. Analysis of projection distortion? No. Believe it or not, a LOT of medium and small-scale maps are made. People tend to view the world through their own myopic interests.

Datum has nothing to do with scale. Even the smallest scale maps can
benefit from being related to a datum. Look at the very interesting
work Jan Hartman is doing bring maps that are many hundreds of years
old onto a modern coordinate system.

> Gerald Evenden and I do not agree on many things, as the years have demonstrated, and I certainly do not understand flouncing off in a huff over something so obscure. Still, I do think it is more reasonable to keep projections and coordinate systems distinct than it is to conflate them. They ARE distinct. Perhaps we save some people from their own ignorance by insisting on an integration of the two, but then we impose a computational and conceptual burden upon those who don't need them, as well as encourage those who do need them to confuse themselves into believing their own interests are the norm.

The "computational burden" of comparing the input and output datum is
no more intensive than converting from radians to degrees. Hopefully
you will not argue that we limit ourselves to radians. The "conceptual
burden" is more subjective, but closer to my point: That to a growing
number of users, datum does matter.

-- 
Richard Greenwood
richard.greenwood at gmail.com
www.greenwoodmap.com