> >>> Notice that in the following piece of code shmget() returns a value
> >>> incremented by (64 *1024) on each invocation.  Is this expected?
> >> Is there some reason to think it's wrong?
> > Because the return value from shmget() seems to be very different
> > from other OS'.
> 
> So what?  Isn't a shmid an opaque cookie anyway?  Then why are you
> concerned over NetBSD using a different way of generating those cookies
> than FooOS for some foo?
> 

Easy....   This was just a naive question from one who noticed a rather
striking difference.