[bwhmather]

Commit history serves two largely separate purposes:

1) To provide a code-level record of what has changed in individual files and why.

2) To provide a high-level record of what features were introduced and what bugs were fixed over a period of time.

Often people will forget about one of them when arguing for a particular approach.

Squashing can make 2 easier but annihilates 1. Rebasing gives you 1 but makes 2 difficult, or requires that you track high level changes in an external system. In theory, approaches using merge commits can give you both but they are often difficult to apply in practice.