Given that indexes could be empty, inclusive ranges should be able to represent the case where size is zero.
To do this, to=-1 is proposed to represent ranges that are empty.
Historically, ranges were designed only for segments where empty files were not possible. As recently empty indexes became a possibility when small segment files are created, then the initial design of BytesRange became an issue as it's designed as inclusive-inclusive.
Given that indexes could be empty, inclusive ranges should be able to represent the case where size is zero. To do this,
to=-1
is proposed to represent ranges that are empty.Historically, ranges were designed only for segments where empty files were not possible. As recently empty indexes became a possibility when small segment files are created, then the initial design of BytesRange became an issue as it's designed as inclusive-inclusive.