Re: [boost] Interest in a "Heterogenous Container" system?

Subject: Re: [boost] Interest in a "Heterogenous Container" system?
From: Mathias Gaunard (mathias.gaunard_at_[hidden])
Date: 2009-07-31 13:10:24

Stewart, Robert wrote:

> Yes, containers of shared_ptrs would work, but they add overhead that isn't always wanted.

Then use containers of unique_ptr, like you should (shared_ptr is meant 
for shared ownership or making it easy to fix broken code), or use the 
boost pointer containers if you can't (which move semantics mostly 
render obsolete).
You could alternatively use boost.any, adobe.poly, clone_ptr, or 
whatever is relevant.

> The heterogeneous containers, being implemented as class templates, need only be defined in terms of operations supported by its parameterizing types.

This merely looks like a container of Adobe.Poly.

As far as I can see, the only advantage of heterogeneous containers is 
when dealing with node-based containers.
A heterogeneous node-based container would have the opportunity to store 
the object directly into the node, while a container of boost.any or 
anything like that would only have a pointer to the object in the node.

I believe however a better approach to benefit from that would be to 
define a dynamically-sized object protocol and have containers extended 
to support it. (obviously, data structures that rely on the fact the 
size of all instances of the same type is the same couldn't do that)