chore: consistent slice behavior with Python

[usamoi]

Current subslice behavior is consistent with PostgreSQL:

SELECT ('{0, 1, 2, 3, 4, 5, 6, 7}'::real[])[3:6];
  float4   
-----------
 {2,3,4,5}
(1 row)

SELECT ('[0, 1, 2, 3, 4, 5, 6, 7]'::vector)[3:6];
  vector   
-----------
 [2, 3, 4, 5]
(1 row)

Making it consistent with Python should be better.

>>> [0, 1, 2, 3, 4, 5, 6, 7][3:6]
[3, 4, 5]

SELECT ('[0, 1, 2, 3, 4, 5, 6, 7]'::vector)[3:6];
  vector   
-----------
 [3, 4, 5]
(1 row)

[VoVAllen]

haha. Programmers should count from 0, unless you are a mathematician.