Joaquin, thanks for your feedback. Joaquin M Lopez Munoz wrote:
John Torjo
writes: Hi all,
Today starts the formal Fast-Track review of the Boost.Functional/Factory library. [...]
Hello, this is not a review, but only a compilation of some doubts and suggestions for improvement:
1. I think value_factory requires that T be copyable, even if copying is ellided through RVO. This requirement is not stated in the docs.
Yes.
2. Maybe it's a good idea to have BOOST_UTILITY_VALUE_FACTORY_MAX_ARITY be automatically greater than or equal to BOOST_FUNCTIONAL_FORWARD_ADAPTER_MAX_ARITY, so as to make it simpler to forward adapt value_factory. Same for plain factory.
I'm not sure I like to couple these components. It's not /that/ likely to forward adapt factories, after all: I suspect that the arguments will be most commonly fed either by a) a central facility that might as well use lvalues or b) Boost.Function (no rvalue-ness inside).
3. Just as planned for the forward library, it'd be good to let the lib use variadic templates and rvalue refs when available.
Yes.
Just my 2 cents,
Taking care of the pence for the pounds to take care of themselves :-). Regards, Tobias