Conceptually, common_type is a type_trait. We should not be so quick to
worship at the altar of removing dependencies that we compromise logical
design or introduce code duplication. I don't think a dependency-free graph
is necessarily desirable -- the whole reason we have Boost is to create
reusable components.
On Sun, May 17, 2015 at 1:51 AM, John Maddock
On 16/05/2015 18:44, Vicente J. Botet Escriba wrote:
Le 16/05/15 14:03, Peter Dimov a écrit :
Ion Gaztañaga wrote:
Is there anything that we could make to make to take away common_type
dependencies?
Let's just make it its own module and be done with it.
I agreed. John could you extract it for "version 2"? Could I do something?
I have to be honest, I don't like splitting stuff up, but OK I'll go with the flow. Let's split it.
I've just made a bug fix to develop - when the tests have cycled on that I'll merge to release and then merge the new type_traits to develop. I guess immediately after that would be a good time to spin off common_type - not sure what the process is?
John.
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost