"Eric Niebler" <eric@boost-consulting.com> writes:
David Abrahams wrote:
"Eric Niebler" <eric@boost-consulting.com> writes:
A different approach to achieve the same ends might be to define overloads of the C string APIs that worked with std::string,
Gee that smells like a familiar idea ;-)
so the migration of legacy code can actually eliminate bugs instead of making them less severe. Has this approach been considered?
I think that's my point, Eric :)
:-P That's what I get for only skimming the thread -- I didn't see you suggest this approach. Sorry for the noise, and consider this an endorsement of Dave's suggestion.
To be fair, I didn't suggest that specifically until forced into it by Gennadiy. My fundamental position is that you should probably *not* be using a fixed-sized string for this sort of transition. The details of how you use a variable length string are less important to me. If you want to keep a <cstring>-like interface, that's fine, but ideally all that code should be revisited anyway and you could easily end up with better code by translating it to a more natural and expressive interface. Regards, Dave -- Dave Abrahams Boost Consulting www.boost-consulting.com