[postgis-devel] release status

[strk at refractions.net]

The mapserver connector bug seems to be fixed, the only issue left
is the spatial_ref_sys entries.

The reported *bogus* entry was one of Paris ones  (27582).
We have more then a single entries for Paris, sharing most of
the parameters that requires correction, so what I don't know
is wheter all of them are really *wrong* or if different PROJ4
versions handle different parameters.

Frank, can you tell us something about this ?

Finally we have a minor problem with FLEX and YACC variables
being somehow overidden by postgresql Makefiles, did anyone here
ever compiled postgis using different lexical analyzer / parser
generators ? Should we directly invoke 'flex' and 'yacc' instead
of using variables ?

--strk;


On Mon, Feb 14, 2005 at 10:43:46AM +0100, strk at refractions.net wrote:
> It's almost one month since RC2 is out, and we have
> a bounce of changes logged:
> 
>         - transform(): missing SRID bugfix, better error handling
>         - jdbc2: small patches, box2d/3d tests, revised docs and license.
>         - jdbc2: bug fix and testcase in for pgjdbc 8.0 type autoregistration
>         - few changes in autoconf
>         - reduced precision of box2d output.
>         - some functions made IMMUTABLE from STABLE,
>           for performance improvement
>         - new performance tips chapter in manual
>         - BUILDDATE extraction made more portable
>         - fixed bogus 72 index bindings
>         - prefixed DEBUG macros with PGIS_ to avoid clash with pgsql one
>         - fixes in probe_geometry_columns() to work with PG72 and support
>           multiple geometry columns in a single table
>         - fixed spatial_ref_sys.sql to avoid vacuuming the whole database.
>         - jdbc2: Removed use of jdk1.4 only features to enable build with
>           older jdk releases.
>         - jdbc2: Added support for building against pg72jdbc2.jar
>         - documentation updates: pgsql72 requirement, lwpostgis.sql
>         - memory alignment handling
> 
> The "segfaulting points" issue seems to be fixed (memory alignment
> handling), but the following issues are still open:
> 
> 	- mapserver connector weidness
> 	- spatial_ref_sys entries (compatibility ?)
> 
> I think the special handling of the SRID=-1 should wait until 1.1.0, 
> as getting 1.0.0 out is already taking too much time, so I removed
> it from the list of "open issues" (it's a feature request).
> 
> If anyone has reports about mapserver connectivity or hints about
> new proj4 entries please send them out. I'd tag and boundle RC3
> this week. 
> 
> --strk;
> 
> 
> On Tue, Feb 08, 2005 at 09:11:29AM +0100, strk at refractions.net wrote:
> > Update changes:
> > 
> >         - transform(): missing SRID bugfix, better error handling
> >         - jdbc2: small patches, box2d/3d tests, revised docs and license.
> >         - jdbc2: bug fix and testcase in for pgjdbc 8.0 type autoregistration
> >         - few changes in autoconf
> >         - reduced precision of box2d output.
> >         - some functions made IMMUTABLE from STABLE,
> >           for performance improvement
> >         - new performance tips chapter in manual
> >         - BUILDDATE extraction made more portable
> >         - fixed bogus 72 index bindings
> >         - prefixed DEBUG macros with PGIS_ to avoid clash with pgsql one
> >         - fixes in probe_geometry_columns() to work with PG72 and support
> >           multiple geometry columns in a single table
> > 
> > Updated open issues:
> > 
> >         - segfaulting points
> >         - mapserver connector weirdnesses
> > 	- spatial_ref_sys entries (incompatible between PROJ versions?)
> >         - SRID=-1 special handling
> > 
> > --strk;
> > 
> > On Mon, Feb 07, 2005 at 01:25:41PM +0100, strk at refractions.net wrote:
> > > 12 days has gone since RC2.
> > > 
> > > This are changes so far:
> > > 
> > >         - transform(): missing SRID bugfix, better error handling
> > >         - jdbc2: small patches, box2d/3d tests, revised docs and license.
> > >         - few changes in autoconf
> > >         - reduced precision of box2d output.
> > >         - some functions made IMMUTABLE from STABLE,
> > >           for performance improvement
> > >         - new performance tips chapter in manual
> > > 
> > > This are the open issues in order of importance:
> > > 
> > > 	- segfaulting points
> > > 	- mapserver connector weirdnesses
> > > 	- SRID=-1 special handling
> > > 
> > > It would be useful to have most informations about the bugs handy.
> > > Paul, what about linking the bug tracker on the postgis web site ?
> > > 
> > > Is anyone experiencing the points bug on this list ?
> > > 
> > > --strk;
> > > _______________________________________________
> > > postgis-devel mailing list
> > > postgis-devel at postgis.refractions.net
> > > http://postgis.refractions.net/mailman/listinfo/postgis-devel
> > _______________________________________________
> > postgis-devel mailing list
> > postgis-devel at postgis.refractions.net
> > http://postgis.refractions.net/mailman/listinfo/postgis-devel
> _______________________________________________
> postgis-devel mailing list
> postgis-devel at postgis.refractions.net
> http://postgis.refractions.net/mailman/listinfo/postgis-devel