It's frequently a loose requirement; trouble is to know how much information is necessary to replicate, you typically have to do the actual replication (and ite

It's frequently a loose requirement; trouble is to know how much information is necessary to replicate, you typically have to do the actual replication (and iterate with authors over details). Not to mention that journals tend to dislike overly long papers.