Re: boost.bind and g++ 2.95.3 woes

Peter Dimov wrote:

[snip]

A mistake, but not a stupid one. The boost::bind(bar, 0) subexpression is interpreted as a nested bind; that is, bind assumes that you want foo(bar(0)), but you actually want foo(bind(bar, 0)). You can "protect" the nested bind from being evaluated by using the helper function protect() defined in boost/bind/protect.hpp; see the last two paragraphs of

http://www.boost.org/libs/bind/bind.html#nested_binds

Hah, yes of course! Thanks a lot, Peter! Markus