On 13/11/2013 08:34, Quoth Niall Douglas:
Boost community feedback is requested for the formal peer review of the TypeIndex library by Antony Polukhin. Feedback requested includes: [...] 2. Any conditions which should be attached to acceptance into Boost e.g. fixes, additional testing, changes to documentation. Please be as specific as possible here (bullet points are good!)
There are several instances where code inside the boost namespace (and often within the class in question) refers to the same class with implicit namespace (eg. "type_info", eg. type_info.hpp:147), and elsewhere with explicit namespace (eg. "boost::type_info", eg. type_info.hpp:130). I don't know if there's an official policy on which of these two is preferred (though I think I've most commonly seen the namespace omitted), but it seems wrong to be inconsistent about it. Especially given that in type_info.hpp:73 there is, within namespace boost::detail, an implicit reference to *global* type_info. And of course boost::type_info itself potentially shadows global type_info and/or std::type_info, which may negatively impact code that does "using namespace boost". (Not that I think that should be encouraged, and this is not unique to this library now that eg. shared_ptr is in std with C++11.) I don't have a problem with this personally (and it fits the "drop in replacement" goal) but I'm not sure where Boost policies stand on this.