[GRASS5] Database linking question

[Eric G . Miller]

On Tue, Jan 02, 2001 at 08:01:29PM -0800, Rich Shepard wrote:
>   Ah! Now I understand. Well, in my naivete, I believe you've hit on the
> reason raster GIS are easy and inexpensive to build and vector GIS cost so
> bloody much. ESRI has either a patent or some other very tight lock on their
> vector data structures. AFAIK, everyone who's developed a vector-GIS has
> rolled their own data structures (based on quadtrees or something similar),
> but written their own dbms.

Yea, vector stuff is fairly complicated, especially if you want to
support things like "routes", "regions", "networks", "z-ordering".
Non-topological systems like Shapefiles are easier in some ways, but
make some of these other things pretty much impossible.

My understanding is (in the U.S. at least) file formats can't be
patented, but certainly ESRI (or any other vendor) isn't going to freely
share the data structures of their cash-cows.  My understanding of the
Arc/Info Workspace is something like a segmented ISAM database with
"arcs" being spatially sorted and attributes being keyed in by id's on
the given type of spatial object (with a separate "table" for each
type).  There's got to be some kind of hierarchical tree to move between
levels of spatial "objects" (regions->polys->arcs->nodes->vertices). 

Wish I could spend more time on it (and perhaps get more schooling...).

-- 
Eric G. Miller <egm2 at jps.net>

---------------------------------------- 
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo at geog.uni-hannover.de with
subject 'unsubscribe grass5'