[postgis-devel] PostGIS release

[Mark Leslie]

Looks like we're making some progress.  I'll apply the patch and get on
these failures as soon as I have a few minutes.
We'll get this release out yet.
Thanks,
Mark Leslie

Michael Fuhr wrote:
> On Tue, Jul 24, 2007 at 09:03:05AM -0700, Mark Leslie wrote:
>> I've put in a fix for this bug.  Please try the new candidate at
>> http://postgis.refractions.net/postgis-1.3.0RC3.tar.gz when you have the
>> chance.
> 
> Thanks; the regress_ogc tests are now succeeding on all of my
> platforms.  Some of the sql-mm tests are still failing due to
> differences of precision, with more failures on Solaris 9/sparc
> than on x86.  Here are the Solaris tests; I've attached the diffs:
> 
> sql-mm-serialize. ok
> sql-mm-circularstring. failed (diff expected obtained: /tmp/pgis_reg_4601/test_28_diff)
> sql-mm-compoundcurve. failed (diff expected obtained: /tmp/pgis_reg_4601/test_29_diff)
> sql-mm-curvepoly. failed (diff expected obtained: /tmp/pgis_reg_4601/test_30_diff)
> sql-mm-general. ok
> sql-mm-multicurve. failed (diff expected obtained: /tmp/pgis_reg_4601/test_32_diff)
> sql-mm-multisurface. failed (diff expected obtained: /tmp/pgis_reg_4601/test_33_diff)
> 
> Regression tests now fail entirely on PostgreSQL 8.3devel with the
> following error in regress_log:
> 
> psql:lwpostgis.sql:6189: ERROR:  syntax error at or near "query"
> LINE 5:  query text;
>          ^
> 
> This appears to be due to the following commit that makes QUERY a
> keyword in PL/pgSQL:
> 
> http://archives.postgresql.org/pgsql-committers/2007-07/msg00237.php
> 
> I've attached a patch that quotes "query" in the necessary places;
> another possibility would be to change "query" to something else.
> While making these fixes I found a couple of variables that aren't
> used; the patch deletes those lines.
> 
>