[postgis-devel] ISO WKT/WKB

[Mark Cave-Ayland]

On 20/06/11 18:19, Paul Ramsey wrote:

> I'm ready to implement these. However, note that it's a scary time,
> more particularly I think for WKB than WKT. It's possible to build a
> WKB handler that can ingest both ISO-WKB and EWKB (in fact, we have
> one) but not every client software will have one of these already in
> place. Problems only occur with higher-dimensional geometries.
>
> ST_AsText(geography) and ST_AsBinary(geography) already return ISO.
>
> {{{
> postgis20=# select st_astext('POINT Z (0 0 0)'::geometry);
>   st_astext
> ------------
>   POINT(0 0)
> (1 row)
>
> postgis20=# select st_astext('POINT Z (0 0 0)'::geography);
>      st_astext
> -----------------
>   POINT Z (0 0 0)
> (1 row)
> }}}
>
> Anyhow, implementing them would bring us to a "new consistency" but
> would require any client that expects ST_AsBinary() and ST_AsText() to
> consistently return 2d to be upgraded. Folks already consuming 3d/4d
> will necessarily be using ST_AsEWKB and ST_AsEWKT so changing
> ST_AsText() and ST_AsBinary() won't effect them, nor will the
> differences in representation between ISO and EWKB/T.

I think if we're going to bite the bullet and do this then 2.0 is going 
to be the right time, as there are enough other breaking changes to make 
people want to do some serious testing with their applications.

Probably the one thing we have been guilty of in the past is not 
shouting loud enough to our companion projects when making changes like 
this. It would be good to start making noises on various lists, e.g. 
MapServer, Geoserver so that any related 2.0 fixes are included in at 
least 1 release before the full 2.0 release.


ATB,

Mark.

-- 
Mark Cave-Ayland - Senior Technical Architect
PostgreSQL - PostGIS
Sirius Corporation plc - control through freedom
http://www.siriusit.co.uk
t: +44 870 608 0063

Sirius Labs: http://www.siriusit.co.uk/labs