[fdo-internals] Question about FDOOGRProvider

[Gabriele Monfardini]

Hi Jackie,



> Is it worth caching this extent in the provider for future calls to other
> spatial context readers from this provider?
>

the information is indeed cached until "cached schema information" is
purged.


> An extent of a Spatial Context isn't exactly guaranteed to be an accurate
> barometer of the true bounds of a Feature Class (eg. SQL Server's spatial
> context bounds are mostly just inferred from the bounds of whatever spatial
> indexes have been set for their geometry columns). It seems to be just a
> bounds and not *the* bounds of the Feature Class.
>

Also in PostGIS there is a ST_ExtimatedExtent function which is quick and
based on table statistics.
Unfortunately the expensive ST_Extent is used instead.



> The brute force EXTENT() function in MapGuide is a last-resort method for
> obtaining the bounds of a Feature Class. This would only be used if there
> was no SpatialExtent() function support in the FDO provider (which sadly
> OGR
> is one of). I know the MapGuide Schema Previewer is one such place that
> uses
> such logic.
>

my question is "how are these bounds used in MapGuide?".
If giving MapGuide clearly wrong bounds do not cause any apparent problem,
why FeatureClass bounds are needed AT ALL?

Regards,

Gabriele Monfardini
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/fdo-internals/attachments/20140327/cd8d44cf/attachment.html>