by steve-myers » Mon Jul 07, 2014 9:42 pm
Mr. Prins is correct. What do you mean by "big inconsistencies on dataset level?" Are you asking about a data set name? Or are you asking about the contents of a data set?
A data set name is relatively simple. It has 1 to 44 characters; it is usually divided into several "index levels," separated by a period (.) character, like SYS1.LINKLIB. Each "index level" in turn, consists of 1 to 8 roman characters, "national" characters, or numeric characters. The first character must be a roman character or a "national" character. "National" characters vary; in the United States they are usually considered to be $, # or @.