Move up to C++17

[SteveBronder]

Description

Jw, what are the blockers that stop us from moving to C++17?

There's a couple features that I think would be nice. From here

• Nested namespace definitions, e.g., namespace X::Y { … } instead of namespace X { namespace Y { … } }
• Fold expressions, for variadic templates (ala std::is_arithmetic<Args> &&...)
• A compile-time static if with the form if constexpr(expression) (No more #ifdefs for the OpenCL code!)
• copy-initialization and direct-initialization of objects of type T from prvalue expressions of type T (ignoring top-level cv-qualifiers) shall result in no copy or move constructors from the prvalue expression.
• Additional mathematical special functions, including elliptic integrals and Bessel functions see here

From that same wiki page it looks like we would have to bump up the supported compiler versions?

GCC since version 7 has complete support for C++17. GCC C++ standards status Clang 5 and later implement all the features of C++17. clang C++ standards status

Current Version:

v2.20.0

[SteveBronder]

For if constexpr in particular we could use BOOST_IF_CONSTEXPR but personally I'd really like to avoid having macros everywhere (you can see in the example branch we would have to use this everywhere).

But there's a lot of stuff that we can't macro out. There's a list of C++17 niceties here here of new C++17 features (full list here). In particular things like structured bindings, Template argument deduction for class templates, folding expressions, etc. Having these features would clean up a lot of current stuff that's gross we need to do that is simple with 17

[SteveBronder]

@bgoodri does rstan still need to test windows on older gcc versions or is the minimum version of gcc now 8+?

[bgoodri]

We will probably require R >= 4.x in April, in which case RTools based on gcc 4.9 would no longer be supported and the C++17 flag is an option. I don't think 100% of C++17 was in g++-8 but it is close.

On Tue, Jan 12, 2021 at 4:48 PM Steve Bronder notifications@github.com wrote:

@bgoodri https://github.com/bgoodri does rstan still need to test windows on older gcc versions or is the minimum version of gcc now 8+?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/stan-dev/math/issues/1300#issuecomment-759053369, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAZ2XKV5XTGSSHGAOXSVRZ3SZS7TXANCNFSM4IJJSR7A .

[SteveBronder]

Looking at the support sheet it seems like gcc8 is the version that finalizes 17

https://gcc.gnu.org/projects/cxx-status.html#cxx17

[bgoodri]

For some operationalization of final. The page it links to

https://gcc.gnu.org/onlinedocs/libstdc++/manual/status.html#status.iso.2017

lists the known exceptions, but it was pretty close in g++-8.

On Tue, Jan 12, 2021 at 11:24 PM Steve Bronder notifications@github.com wrote:

Looking at the support sheet it seems like gcc8 is the version that finalizes 17

https://gcc.gnu.org/projects/cxx-status.html#cxx17

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/stan-dev/math/issues/1300#issuecomment-759196417, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAZ2XKVJ5KYDJYOMUSM4E7DSZUN75ANCNFSM4IJJSR7A .

[SteveBronder]

ah icic, looking over that it mostly seems like the stuff that didn't get into 8 is related to working with chars, strings, file system, the parallism TS, and the memory_resource header. I think thats fine?