1) one byte is required for the ” – ” sign
that makes sense, so if I set INT unsigned it will default to 11 and if not then default to length 10
2) I practically always make an autoincrement PK a BIGINT UNSIGNED .. but probably 'overkilll' in most situations!
correct me if I am wrong, but it is possible to change (increase specifically) the size later on (as data needs require) without much difficulty right? Also, in theory, better performance is obtained from a smaller number, all other things being equal?