Nikita Kniazev

I am afraid of renaming the next macros: ``` PHOENIX_LIMIT PHOENIX_CONSTRUCT_LIMIT SPIRIT_VERSION SPIRIT_STATIC_LEXER_VERSION SPIRIT_DECLARE_ARG SPIRIT_USING_ARGUMENT SPIRIT_DECLARE_ARG SPIRIT_USING_ARGUMENT SPIRIT_X3_VERSION SPIRIT_VERSION ```

@ned14 What is your interest in this? Do you want to automatically rename macros to be able to use multiple Boost versions in the same codebase? ```cpp #define BOOST_SPIRIT_VERSION 123456...

> The current suggestion on boost-dev for how to handle these is that for the next three to six releases of Boost, a deprecation warning be given to code which...

@saki7 please provide a [minimal reproducible example](https://stackoverflow.com/help/minimal-reproducible-example) when you open a bug report. > stackoverflow.com/questions/45282293/spirit-not-able-to-use-x3skipskippersome-recursive-rule-in-its-rule-defin @cppljevans I do not think #237 is a right fix, it should build a stack...

@cppljevans I was concerned that after #237 the following code will print `42 123 123` instead of `42 123 42`, but it is not a case. I am still did...

> What is needed is some type-erasure around the iterator and attribute types. There is [`any_parser`](https://github.com/boostorg/spirit/blob/master/test/x3/any_parser.cpp) that does it.

There is no Phoenix integration in X3, it would need some other interface and since it is much easier to create customs parsers with X3 I am not sure all...

> Now, I'm not saying exactly this should be implemented in Spirit, but please let me know what you think about this approach. I think that you can turn `FileParser`/`DirectoryParser`...

I do not know how `hold` was introduced into `Qi` and why that solution had been chosen. I feel like even `Qi` must lose `hold`, because it (even on C++11...

Since I did not even open a PR you may have misunderstood me. Or maybe you have a counterexample?