Re: [boost] more thoughts on c2_functions