It is usually assumed that it's implied by using the term “array” and accepting a size argument. One could perhaps argue that the standard does not state explicitly how the size argument and the array length are related.
Previous upstream discussions for other *n* string function strongly went in favor of the array length interpretation, even in cases where existing programming practice is clearly different (although I believe this is just a documentation issue, the implementations work as programmers expect).
Anyway, it's probably a good idea to bring up the snprintf case on libc-alpha.
It is usually assumed that it's implied by using the term “array” and accepting a size argument. One could perhaps argue that the standard does not state explicitly how the size argument and the array length are related.
Previous upstream discussions for other *n* string function strongly went in favor of the array length interpretation, even in cases where existing programming practice is clearly different (although I believe this is just a documentation issue, the implementations work as programmers expect).
Anyway, it's probably a good idea to bring up the snprintf case on libc-alpha.