On 29 May 2015 at 12:31, Stefan Seefeld wrote:
Rather, I would like to know whether people here (and Boost.Python users in particular) have any particular thoughts about this proposal. Any strong reasons not to move forward with this ? Things to consider while doing it ? Etc.
This is exactly what proposed Boost.APIBind is designed for: allow a legacy Boost library to both exist as part of monolithic Boost, and as a standalone library with no Boost dependencies (with the C++ 11 STL swapped in instead). Identical code base can do both targets. Have a look at https://svn.boost.org/trac/boost/wiki/BestPracticeHandbook#a16.COUPLIN G:Considerallowingyourlibraryuserstodependencyinjectyourdependencieson otherlibraries. If that whets your appetite, have a look at my C++ Now Presentation on APIBind: https://github.com/boostcon/cppnow_presentations_2015/raw/master/files /A-review-of-Cxx-11-14-only-Boost-libraries-Fiber-AFIO-DI-and-APIBind. pdf The video of that talk is probably also very useful to watch, but I don't think they are posted yet. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/