Maybe one day I'll feel ready (or need :) ) to take on part time or short term contracts. If so, I'll be back, but for now let me say how much I've enjoyed...
Abstraction is the idea that the class works in the way you want it to work. Encapsulation hides all the nasty translation of that abstraction into the bits you have to use to do it.
AFAIAA, that technique won't work with a template, since you can't have template friends, so each "finaliser" has to be specific to the class it's finalising.
Without getting into the question of whether NVI should be enforced or not, I still tend to stick to one of my basic dicta, which is: "you need a reason to make something non-private".
Because the implicit "this" pointer within foo::bar would be null and any attempt to access a member would imply dereferencing a null pointer, which is undefined behaviour.
Nope. Memberwise copies. For members of class type, it will call the appropriate copy constructor, for members of basic type (int, pointer, etc.) it's likely to be bitwise, but it's better to think...
Interesting - Comeau fails it with the same errors. Presumably implicit default ctors are not considered wrt const objects. I'll see if I can find the relevant section of the standard.
Emphasis on the may. And, yes, I could see it as a possibility - after all, they can unroll loops if it's more efficient. Don't forget, optimisers have access to compiler implementation details that...
...until a day or two after this assumption gets deeply embedded into the code, at which point the requirements will change in such a way as to make the assumption invalid.
Don't forget that the optimiser may well hoist the declaration out of the loop for you (and in a more efficient way than you could achieve manually), so the point could well be moot.
The only way you're going to get a definitive answer is to put together a test harness and time the two approaches. Never try to second-guess such fine differences.
