Maybe it is a good idea to switch the example for "very large data" to something with less boilerplate and naming it after the library/file type that is being used. This would make it easier to discover which example to look at and open the door for adding more examples.
Potential other streaming/read on demand/read only specified subset ideas:
Maybe it is a good idea to switch the example for "very large data" to something with less boilerplate and naming it after the library/file type that is being used. This would make it easier to discover which example to look at and open the door for adding more examples.
Potential other streaming/read on demand/read only specified subset ideas: