[postgis-devel] liblwgeom.so

[Paragon Corporation]

> > > For the reasons mentioned in the email, but in summary:
> > >
> > > 1) Smaller binaries
> > 
>>  Don't particularly care about this.
> 

I don't particularly care about this.  a 885 kb vs. 500 kb or whatever seems
so miniscule to care about
and certainly for regress testing I fail to see the point.

Can we please get back the speed I enjoyed before we went down this messy
road
that made PostGIS 2.0 twice as slow as 1.5.  (May not be related I'm just
venting).

 
>>  I feel like a better approach for the overall foss4g 
>>  ecosystem, and for our own internal work would be to have 
>>  those aspects of the API in OGR, which is already a system 
>>  level utility library.

+++++20 for this.
Guys the way I see it, the next big change we will have is taking full
advantage of the foreign data wrappers
functionality introduced in 9.1.  This will be big for both raster and
postgis.  OGR/GDAL seems like the most likely avenue
for taking advantage of that.  That said -- our dependency on OGR/GDAL is
going to go from 50% to 100%

It's already used and demonstrated its flexibility and use in other
projects.  Let's make it stronger and if we are going to be breaking our
functionality lets put our efforts there.
We've already got Paul, Jorge, and I think Bborie already contributing work
to GDAL.  We are already a one big happy family.
The marriage is already half there.