MS RFC 21: MapServer Raster Color Correction

[Frank Warmerdam]

Howard Butler wrote:
> +1.  Cool and useful feature.
> 
> I'm still anti-PROCESSING option because I think it ends up being an 
> alternative and free-form mapfile syntax.  Someday maybe it would be 
> useful and worthwhile to formalize PROCESSING and FORMATOPTION in some 
> way.  I do not expect these complaints to be addressed by this RFC, 
> however.

Howard,

In fact, my original intent was that processing options would be used
to "shake out" the more useful features which would then be formalized
with their own mapfile syntax.

We could consider that, but it does add a bunch of complication.  Not
only do we need to upgrade the mapfile parser, but this also generally
means there are specific new fields in layerObj, mapscript needs to
be updated with methods to set/get etc.

If the only concern is that the mapfile format is less reasonable this
way, then we could potentially provide specific mapfile syntax for setting
these things, but then have the parser actually just insert the corresponding
entries into the PROCESSING option list in the layerObj.  That way, the
majority of the code stays the same, and mapscript can just use generic
processing option setting.  The specific keywords become "syntactic sugar"
at the front end.

I *do* think that MapServer 5.0 would be the time to deal with this if
we are going to.

I think the situation is different with FORMATOPTION.  It is hard to
formalize these since they vary by driver and for output drivers in
GDAL (which it is mostly used for) there is no way for mapserver to know
in advance all the creation options that could apply.  We essentially need
a pass through.

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