[pdal] PDAL: vector<T> too long

[Howard Butler]

> On Apr 5, 2022, at 10:14 PM, Peter Lim <peter.lim at gpslands.com> wrote:
> 
> Nevertheless, would the team consider to enhance the program by internally splitting the area into smaller portions to process,
> if such limits is encountered.

No. Map and reduce are the responsibility of PDAL users, not PDAL itself. The reason for this is because you know your data, and many of the guesses PDAL might make in determining how to split up your data for a given task are going to be wrong. Some pipelines might take lots of memory. Some might take a long time. 

PDAL is not a workflow engine for processing point cloud data. It is a library *for making* point cloud processing workflow engines. 

Howard