[postgis-tickets] [PostGIS] #2634: geography regression failure ST_DWithin/ST_Distance between 2.0.4 and 2.1.2 for dimensional mix of geographies

[PostGIS]

#2634: geography regression failure ST_DWithin/ST_Distance between 2.0.4 and 2.1.2
for dimensional mix of geographies
---------------------+------------------------------------------------------
 Reporter:  robe     |       Owner:  pramsey      
     Type:  defect   |      Status:  new          
 Priority:  medium   |   Milestone:  PostGIS 2.1.2
Component:  postgis  |     Version:  2.1.x        
 Keywords:           |  
---------------------+------------------------------------------------------
Changes (by pramsey):

  * priority:  high => medium


Comment:

 I'm not seeing this one much, the differences look to be spheroid-vs-not,
 basically not respecting the sphere/spheroid flag correctly throughout the
 logic. Still odder case
 {{{
 with geogp(gid, geog) AS (VALUES (1, 'POINT(-11.1111111 40)'::geography),
 (1, 'POINT(-11.1111111 40)'::geography))
  , geoglm (gid, geog) AS (VALUES
       (3, 'LINESTRING M (-10 50 1,50 -10 1)'::geography),
      (4, 'LINESTRING M (-10 50 2,50 -10 2)'::geography) )
 SELECT geogp.gid As p_gid, geoglm.gid As lm_gid,  ST_Distance(geogp.geog,
 geoglm.geog) As dist, _ST_DistanceTree(geogp.geog, geoglm.geog)
  , _ST_DistanceUncached(geogp.geog, geoglm.geog) As distunc
  FROM geogp CROSS JOIN geoglm ;
 }}}
 tree distance is different from the other two! I'd expect it to be one
 (sphere) or the other (spheroid) but not a whole new number.

-- 
Ticket URL: <http://trac.osgeo.org/postgis/ticket/2634#comment:8>
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.