Why don't you directly specify the number of bytes to reserve to store data in memory?

Answer 1

You don't directly specify the number of bytes to reserve in memory to store data because you don't necessarily know how many bytes that data requires.

Instead, you specify the type of data and the number of elements, and the compiler and run-time library figures out how many bytes to allocate.

Its a portability issue. You don't want to tie your code to a specific word size because the code would then break when the underlying platform changes. One very famous example of this is the int data type in MS Windows 3.x. That was a 16-bit integer. When Windows "grew up" and the int became a 32-bit integer, the int changed to 32-bit, and everyone's code broke, except those that planned ahead and wrote the code correctly.

Answer 2

We don't need to specify the number of bytes we require for storage because that's the job of the language compiler. The compiler knows the length of every data type we use, including user-defined types. As well as making it easier for programmers (the entire point of using a high-level language in the first place), it ensures consistency. That is; we do not have to explicitly check if two variables of the same type are also of the same length: that much is a given.

The compiler can also cater for differences between hardware implementations. For instance, on a 32-bit system a pointer variable may only be 4 bytes long but on a 64-bit system it may be 8 bytes long. If we need to know the actual length of a given type on a given architecture then we can interrogate the system (usually at compile time). But for general usage, the length of a type is usually redundant information; it's far more important that we use an appropriate type for the range of values we wish to store.