31 May
2007
31 May
'07
10:06 p.m.
I agree with you that boost::noninheritable is probably stronger than boost::nonderivable (even if the proposed exploit is a very subtle one). I read boost::noninheritable documentation and I found a typo: indeed, in the section "When to use", you refer to your library as "boost::noncopyable". Also, I suggest you to: 1) list every known supported compiler, and especially you should compare it with boost officially supported platforms 2) add to the documentation a section where you explain how to solve the problem about "value semantics" (copy constructor and assignment operator) Best regards, Manuel Fiorelli www.fioreltech.net