As for the JSON library, if I'm correct there is already something in Boost.PropertyTree that works pretty well ? See: http://www.boost.org/doc/libs/1_55_0/doc/html/boost_propertytree/parsers.htm... However, maybe generalizing the idea of parsing tag based text files could be a great addition to Boost. A little bit like Boost.Proto is a library for creating DSL in C++, you could try to make a library to parse tag based language and as a proof of concept, show that it can parse XML, JSON and a subset of LaTeX. By the way, this gentleman would like to be mentored for GSOC 2014. I personally recommend him. Best, David On Fri, Feb 28, 2014 at 7:11 AM, Ganesh Prasad <sir.gnsp@gmail.com> wrote:

I already have done some work over such a library, but this library by me has limited features and need to be improvised. And hence, I am thinking about doing this work during this GSoC 14. But I have certain questions about it.

1) Who will be taking the design decisions for the development of the library ? Whether the student or the mentor will be responsible for the decisions ?

2) To which formats should the library parse the JSON ? The current library by me parses JSON to Maps and Vectors. I am intending to provide flexibility, but if some specifications will be available it'll be helpful.

3) Is the library to be kept as simple as possible in terms of User Interface ? Or, should we use factory patterns to provide more flexibility to the advanced users ?

Have a nice day, everybody. Best wishes Ganesh Prasad

_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost