It is usual for programmers to declare a variable and then dereference it as in:
This very much looks like object oriented programming in C++, Java, JavaScript, etc. where o is considered the target.
In Ada, you write that the other way around as in:
Note, however, that Ada 2005 introduced the concept in the language. If the first parameter is of the same type as o, then the syntax in (1) is now valid. If you are still using an Ada 95 compiler, then you will get an error.