Index: ogr/ogrsf_frmts/gml/drv_gml.html
===================================================================
--- ogr/ogrsf_frmts/gml/drv_gml.html (revisione 23630)
+++ ogr/ogrsf_frmts/gml/drv_gml.html (copia locale)
@@ -181,6 +181,41 @@
+
TopoSurface interpretation rules [polygons and internal holes]
+
+Starting since OGR 1.9.0 the GML driver is able to recognize two different interpretation
+rules for TopoSurface when some polygon does actually contains any internal hole:
+
+ - the previously supported interpretation rule assumed that:
+ - each TopoSurface may be represented as a collection of many Faces
+ - positive Faces [i.e. declaring orientation="+"] are assumed to
+ represent the Exterior Ring of some Polygon.
+ - negative Faces [i.e. declaring orientation="-"] are assumed to
+ represent an Interior Ring (aka hole) belonging to the latest declared
+ Exterior Ring.
+ - ordering any Edge used to represent each Ring is important: each Edge is expected
+ to be exactly adjacent to the next one.
+
+ the new interpretation rule now assumes that:
+ - each TopoSurface may be represented as a collection of many Faces
+ - the declared orientation for any Face has nothing to deal with Exterior/Interior Rings
+ - each Face is now intended to represent a complete Polygon, eventually including any possible Interior
+ Ring (holes)
+ - the relative ordering of any Edge composing the same Face is completely not relevant
+
+
+
+The newest interpretation seems to fully match GML 3 standard recommendations; so this latest
+is now assumed to be the default interpretation supported by OGR.
+Using the GML_FACE_HOLE_NEGATIVE configuration option you can anyway select the
+actual interpretation to be applied when parsing GML 3 Topologies:
+
+ - setting GML_FACE_HOLE_NEGATIVE NO (default option) will activate
+ the newest interpretation rule
+ - but explicitly setting GML_FACE_HOLE_NEGATIVE YES still enables to activate
+ the old interpretation rule
+
+
Encoding issues
Expat library supports reading the following built-in encodings :