I have posted the slides of a lightning talk I gave at the PyGotham 2014 meeting on 20140816.
The abstract follows:
This presentation examines the behavior of IDs (memory locations) of “whole slices” of Python sequences, and particularly whether those IDs are the same or different for different copies of those slices.
In the standard implementation (CPython), whether or not the IDs are the same depends on which sequence is involved. The pattern for the standard implementation is itself different from the patterns for two other major implementations, PyPy and Jython. ID behavior of whole-sliced sequences is most consistent and predictable in Jython, where whole-slicing a sequence produces a new object in each and every case examined.