DE wrote:
For a simple demonstration of the basic mechanism (untested, so not sure if it will compile, but it should give the basic idea) [code here] oh now i get it! i didn't realize there are operator rv<T>& and its const counterpart hidden behind the macros
on 26.02.2010 at 21:27 Jeffrey Hellrung wrote : <snip> thanks for your effort explaining this
No problem. I have suggested to Ion that a section of the Boost.Move documentation should be devoted to explaining how the mechanism works. <snip>
so summing it all up i say that this solution is far too tricky to be useful
I think the same could be said for a lot of "tricks" Boost libraries employ ;) I'd have to put Boost.EnableIf and Boost.Preprocessor in the category of code that uses language features in non-obvious ways.
no offence it seems to me too restrictive and too incomplete the infrastructure is too complicated
Regarding "too restrictive and too incomplete", you still can't get perfect forwarding. But otherwise, you pretty much get identical behavior as with C++0x rvalue references. So I would disagree with this claim. Regarding "too complicated", the basic mechanism (and most of the nastiness) can be encapsulated behind a macro, and as long as you adopt certain conventions, writing move-aware code otherwise is very easy. From the user's point-of-view, there is some complication and extra consideration, but not really any more than you'd have with true rvalue references. However, everyone's entitled to their own opinion ;) <snip>
another question arises: who is the target audience (aka users) of this lib?
I use it, and I think Chris stated the motivation for using it very well. I believe it can be targeted to anyone wishing to write move-aware code. - Jeff