[PostGIS] #5893: regress/core/geos39 failing on geos 3.14.0dev

[PostGIS]

#5893: regress/core/geos39 failing on geos 3.14.0dev
----------------------+--------------------------
  Reporter:  robe     |      Owner:  robe
      Type:  defect   |     Status:  new
  Priority:  blocker  |  Milestone:  PostGIS GEOS
 Component:  postgis  |    Version:  3.4.x
Resolution:           |   Keywords:
----------------------+--------------------------
Changes (by robe):

 * owner:  pramsey => robe


Old description:

> {{{
> +++
> /projects/postgis/tmp/3.6.0dev_pg17_geos3.14_gdal3.9.2w64/test_119_out
> 2025-05-02 22:03:29.465152000 -0400
> @@ -1,10 +1,10 @@
> -mic-box|POINT(49.5117 49.5117)|POINT(50.5111 98.9796)|49.4779
> +mic-box|POINT(49.495 49.495)|POINT(50.4947 98.9799)|49.4950
>  mic-empty|POINT EMPTY|POINT EMPTY|0.0000
>  mic-null|||
>  mic-line|POINT(49.5117 49.5117)|POINT(50.5111 98.9796)|49.4779
>  mic-mpoint|POINT(48.584 49.9512)|POINT(99 98)|69.6453
>  mic-point|POINT(0 0)|POINT(0 0)|0.0000
> -mic-cte|POINT(49.512 49.512)
> +mic-cte|POINT(49.495 49.495)
>  rp-1|POINT(1.4 19.3)
>  rp-2|POINT(1 19)
>  rp-3|POINT(0 20)
> }}}
>
> Grrh seeing this failure on winnie and debbie testing geos-main

New description:

 {{{
 +++ /projects/postgis/tmp/3.6.0dev_pg17_geos3.14_gdal3.9.2w64/test_119_out
 2025-05-02 22:03:29.465152000 -0400
 @@ -1,10 +1,10 @@
 -mic-box|POINT(49.5117 49.5117)|POINT(50.5111 98.9796)|49.4779
 +mic-box|POINT(49.495 49.495)|POINT(50.4947 98.9799)|49.4950
  mic-empty|POINT EMPTY|POINT EMPTY|0.0000
  mic-null|||
  mic-line|POINT(49.5117 49.5117)|POINT(50.5111 98.9796)|49.4779
  mic-mpoint|POINT(48.584 49.9512)|POINT(99 98)|69.6453
  mic-point|POINT(0 0)|POINT(0 0)|0.0000
 -mic-cte|POINT(49.512 49.512)
 +mic-cte|POINT(49.495 49.495)
  rp-1|POINT(1.4 19.3)
  rp-2|POINT(1 19)
  rp-3|POINT(0 20)
 }}}

 Grrh seeing this failure on winnie and debbie testing geos-main

--
Comment:

 As Martin and Paul mentioned in matrix/irc -- GEOS was changed recently

 "MIC has been enhanced to compute exact results for small polygons
 (triangles and quadrilaterals).  So this could change the result value -
 but only to within the tolerance value provided by the user.  Perhaps the
 tests are not using that same tolerance?"

 So sounds like we need to revise the regress.  I'll take a stab at making
 it work for newer and older geos.
-- 
Ticket URL: <https://trac.osgeo.org/postgis/ticket/5893#comment:1>
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.