[postgis-devel] [PostGIS] #1586: [raster] st_dumpaspolygons makes not valid polygons

[PostGIS]

#1586: [raster] st_dumpaspolygons makes not valid polygons
----------------------+-----------------------------------------------------
 Reporter:  jomarlla  |       Owner:  pracine      
     Type:  defect    |      Status:  new          
 Priority:  high      |   Milestone:  PostGIS 2.0.0
Component:  raster    |     Version:  trunk        
 Keywords:            |  
----------------------+-----------------------------------------------------

Comment(by pracine):

 Replying to [comment:4 jomarlla]:
 > ST_Makevalid is making st_dumpaspolygons much slower. Dont know if it is
 worth it to do it at the source. I was wondering if the algorithm can be
 changed for not making invalid polygons instead of using st_makevalid
 without a lot of work.

 I guess the amount of work for producing a valid geometry from what GDAL
 is returning us would be equivalent to ST_MakeValid but I might be wrong.

 > I just dont like invalid polygons because they just bother me about any
 future spatial analysis about unpredictable results.
 > As you said users can do it easily but in that case maybe the docs
 should point out that st_dumpaspolygons can return invalid geometries,
 then the user can take care about that.

 If there is no simple low cost way to make them valid we will add a note.
 It's the first time someone is complaining about that so I guess the type
 of invalidity does not affect most analyses. Would be great if you could
 find an example of invalid post st_dumpaspolygon analysis.

-- 
Ticket URL: <http://trac.osgeo.org/postgis/ticket/1586#comment:5>
PostGIS <http://trac.osgeo.org/postgis/>
The PostGIS Trac is used for bug, enhancement & task tracking, a user and developer wiki, and a view into the subversion code repository of PostGIS project.