[Benchmarking] Vector data

[Jeff McKenna]

More thoughts on this:

- is this pre-processing of the roads data done for other servers? 
(e.g. GeoServer)
- for MapServer I would normally leave the county data in each of their 
folders, generate a shptree index for each individual county roads file, 
and then create one tileindex.   This is best for performance.

Comments?

(also available on IRC)

-jeff


Jeff McKenna wrote:
> I'm getting started with the vector data, but I am surprised to see that 
> all of the TIGER roads for Texas have been merged into one file 
> "tiger_shp.shp".  Since US Census Bureau distributes the files by 
> county, and since we are doing 'real world use cases', shouldn't we also 
> have each county file for Texas (and then rely on indexing for display)?
> 



-- 
Jeff McKenna
FOSS4G Consulting and Training Services
http://www.gatewaygeomatics.com/