Hi there,

That is a great discussion and I applause Evens efforts and all the 
contributions, so but don't get me wrong by asking that:

What do we want to accomplish with that class hierarchy?

Will GDAL 2.0 be able to do thinks like:

$ gdalinfo <dataset>

And the proposed classes will be able to figure out what kind of dataset is 
that?

It could be:

- raster file
- vector file
- cvs file or dbf file
- some kind of server-based connection string for raster or vector
- some kind of service connection string for raster or vector
- multi-dataset raster file like HDF, NetCDF

Will the GDAL Open mechanism probe for all those possibilities?

My best regards,

Ivan

> From: [email protected]
> Date: Thu, 27 Mar 2014 14:05:29 +0100
> To: [email protected]
> CC: [email protected]
> Subject: Re: [gdal-dev] Draft GDAL/OGR class hierarchy for GDAL 2.0
> 
> On 27 March 2014 13:40, Etienne Tourigny <[email protected]> wrote:
> > How about multiple inheritance for mixed class?  Say you have a class for
> > rasters and a class for vectors, and a dataset that would support both types
> > would inherit from these two base classes.  Is that what you mean by
> > composite?
> 
> I meant separate base class for mixed dataset, then use "has a"
> relationship, like in
> 
> class MixedDataset
> {
> public:
> //
> private:
> RasterDataset* pDtsR
> VectorDataset* pDtsV;
> }
> 
> 
> Best regards,
> -- 
> Mateusz  Łoskot, http://mateusz.loskot.net
> _______________________________________________
> gdal-dev mailing list
> [email protected]
> http://lists.osgeo.org/mailman/listinfo/gdal-dev
                                          
_______________________________________________
gdal-dev mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/gdal-dev

Reply via email to