There should be an option to ignore charset differences

[kormoc]

Charset differences should be able to be ignored when generating schema differences. This leads to a lot of false positive changes as developers run in latin-1 and production runs in utf-8.