[sqlalchemy] Re: GIS support for SQLAlchemy

2009-08-06 Thread Sanjiv Singh
Awesome. Thanks for elaborating. Sanjiv On Thu, Aug 6, 2009 at 7:38 PM, Ants Aasma wrote: > > On Aug 6, 4:09 pm, Sanjiv Singh wrote: >> > Also the GeometryDDL and Geometry classes could use some refactoring >> > to support extensibility of adding new dialects. Probably via a >> > registry that

[sqlalchemy] Re: GIS support for SQLAlchemy

2009-08-06 Thread Ants Aasma
On Aug 6, 4:09 pm, Sanjiv Singh wrote: > > Also the GeometryDDL and Geometry classes could use some refactoring > > to support extensibility of adding new dialects. Probably via a > > registry that maps dialects to their geometry implementations. Also, > > the GeometryDDL doesn't account for the

[sqlalchemy] Re: GIS support for SQLAlchemy

2009-08-06 Thread Sanjiv Singh
Hi Ants, Thanks for the valuable feedback. On Thu, Aug 6, 2009 at 2:30 PM, Ants Aasma wrote: > > I don't have a lot of experience with geodatabases so I can't evaluate > this from a practical perspective, but I do have a couple of thoughts > about the general design. > > I'm not too keen on havi

[sqlalchemy] Re: GIS support for SQLAlchemy

2009-08-06 Thread Ants Aasma
I don't have a lot of experience with geodatabases so I can't evaluate this from a practical perspective, but I do have a couple of thoughts about the general design. I'm not too keen on having the geometry functions as methods on the comparator. Having them there creates a namespace collision ha