[Proj] Datum shifts in PROJ's EPSG table

[Frank Warmerdam]

Martin Vermeer wrote:
> Both of these mechanisms, while better than their absence, seem
> arbitrary. 

Martin,

True

> Has there been thinking on how to support more than one datum shift if
> available? What would that require? Could the EPSG labels be modified
> inside Proj.4, e.g. <4123:1>, <4123:2> etc? Or is this issue bigger than
> Proj.4?

For any given datum EPSG keeps a list of ways of transforming to other
datums - usually to wgs84, and often with detailed notes on where they
are most applicable.

The idea would be for us to offer all of these in some fashion and let
the user select.  The proposed approach (names like <4123:1>) would work,
but gives the user little clue about which is more appropriate.  It is also
annoying that we would need per-datum-shift-value options for every
projected coordinate system based on the datum.

> (If we are talking about slightly different +towgs84 for different parts
> of one datum target area, the use of a shift grid could be an
> alternative solution)

I don't see this as a direct analog though it is a possibility.

I will say that I have no plans for a systematic approach to dealing with
this whole many-datum-shift-options issue.

Best regards,
-- 
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | President OSGeo, http://osgeo.org