🚀 go-pugleaf

In article <41vlgv$q33@druid.borland.com> pete@borland.com (Pete Becker) writes:

> >:       Or, to turn the question around the other way, why separate them?
> >
> >Obviously when using an array where each object is initialized to a
> >different value.  Especially a resizable array.
> 
> That's not the example you posted, which allocated a single
> object. So, let me ask again: in the code you posted, what benefit
> is there in separating allocation from initialization?

There occasionally are benefits; that's why placement operator new and
explicit destructor calls exist.  A lot of container classes use those
features, for example.  (Not that Pete needs lectures from anyone else
about container class libraries.)  

Needless to say, though, I'm happiest when low-level stuff like this
is nicely encapsulated away deep in the guts of some class where I'll
never see it.  Storage allocation and initialization are naturally two
parts of creating a new object, and it's best to keep them together.
In fact, even operator new is a bit too low level for my tastes, and I
prefer to encapsulate it, too, whenever possible.
-- 
  Matt Austern                             He showed his lower teeth.  "We 
  matt@physics.berkeley.edu                all have flaws," he said, "and 
  http://dogbert.lbl.gov/~matt             mine is being wicked."