[postgis-devel] Current regression test status of jdbc driver update

[Markus Schaber]

Hi,

This morning, I just retested my java code against a server running
PostGIS 0.9 CVS, PostGIS 0.8.1 and PostGIS 1.0 CVS.

TestParser results: 1.0 CVS passes all tests, while 0.9 equally and
0.8.1 fail on 22 tests. There are three categories why those tests fail:

- Deeply nested Geometries not accepted by hwgeom implementation
  (e. G. a Geometry Collection containing two MultiPoints or both
  Points and MultiPoints), 8 tests fail due to this.
  "Server side error: [...] couldn't parse Objects in Geometry"

- Deeply nested Geometries flattened by hwgeom (e. G. a Geometry
  Collection containing Multi* geometries of distict types gets
  flattened to a Collection directly the base types), this causes
  8 Tests to fail.
  "[reparsed] Geometries after SQL are not equal!"

- Empty Geometry Representations not understood by old PostGIS (but
  OpenGIS compliant). This only hits MULTI* geometries, the simple
  Geometries and GeometryCollection work. Causes 6 more failures.
  "Server side error [...]"

As you can see, all those tests are no JDBC java code problems, but
either located on the server side or just invalid test cases for this
server version, depending on your way of seeing it :-)

I do not think it is worth the effort to correct those issues in 0.9
branch (except someone pays for it :-), we should concentrate on getting
1.0 out and stable.

But maybe we should document those test failures in the readme (btw, the
readme should be heavily updated, maybe I'll find enough time to do this
today).

HTH,
Markus
--
markus schaber | dipl. informatiker
logi-track ag | rennweg 14-16 | ch 8001 zürich
phone +41-43-888 62 52 | fax +41-43-888 62 53
mailto:schabios at logi-track.com | www.logi-track.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 256 bytes
Desc: OpenPGP digital signature
URL: <http://lists.osgeo.org/pipermail/postgis-devel/attachments/20050120/d1e54e03/attachment.sig>