🚀 go-pugleaf

In <4fu7tj$9a1@solutions.solon.com> als@fl.net.au (Andrew Snow) writes:

>Well, when I learnt C, the following applied:
>
>If you explicetly wanted a 16 bit data variable, you would use a
>"short int".  If you explicetly wanted a 32 bit variable, you would
>use "long int".

Then, you learned C when there was only one implementation available:
the original Ritchie compiler for PDP-11.  Soon after that, C was
implemented on a platform with 9/18/36-bit integer types.  This is
mentioned in K&R1, with the purpose to show that assumptions on the sizes
of various types (and on the signedness of char) are unsafe.

>If you didn't mind, and were prepared to use either
>16-bit or 32-bit, whichever the particular operating system you are
>compiling for decided was best, then you just use "int".

Operating systems don't make such decisions.  Compiler implementors do.
Both 16 and 32-bit compilers are available for MSDOS/Windows.

>To make your code completely portable, you should specify long or
>short.  In situations where it really doesn't matter, you can just
>use "int".

To make your code completely portable you shouldn't rely on more than
16 bits for short and int and on more than 32 bits for long.  If your
code breaks because a short is 64 bits (such implementations do exist)
it's entirely your fault.

Dan
--
Dan Pop
CERN, CN Division
Email: danpop@mail.cern.ch 
Mail:  CERN - PPE, Bat. 31 R-004, CH-1211 Geneve 23, Switzerland