At CppCon 2016, Jon Kalb gave a very interesting Lightning talk, entitled. “unsigned: A Guideline for Better Code", in which he argued against the use of unsigned types (except for bit-masks). Indeed it appears that majority of C++ expert opinion agrees with him. Yet, I haven't yet heard arguments that seem to be adequately convincing. So with all due humility and with blessing from Jon (and James McNellis), I want to argue the opposite case (that unsigned types have their place) if only to move the public conversation forward so that everyone clearly understands why I'm wrong. In this talk, I will argue that (1) shunning unsigned types is the wrong solution to the core problem that Jon identifies and that (2) with the correct solution in place, the benefits of unsigned types will often outweigh their costs. I intend to make a video of these arguments soon so I'm eager to hear the feedback and discussion in this London meetup so I can improve that video.
The class template polymorphic_value is proposed for addition to the C++ Standard Library.
The class template, polymorphic_value, confers value-like semantics on a free-store allocated object. A polymorphic_value may hold a an object of a class publicly derived from T, and copying the polymorphic_value will copy the object of the derived type.
I will present the design and reference implementation, all of which is available here https://github.com/jbcoe/polymorphic_value
