It is not related to squashing or not. One of the main rule when using git (or any other VCS I guess) is that each commit should be atomic.

IMO squashing commits is something that you should do locally, not remotely. For example I'm currently debugging a fairly large C application on an embedded system. Each bug has it's own branch. When debugging / testing / trying to fix it I tend to do a lot of commits. When the bug is fixed I do an interactive rebase to pick / drop / squash commits in order to have only a clean one at the end (and then push it).