Re: [boost] Guidelines to implement Boost library evolution policy (was Boost 2.0)

On Sun, Jun 8, 2014 at 5:45 PM, Edward Diener wrote:

On 6/8/2014 11:54 AM, Beman Dawes wrote:

...

On Sun, Jun 8, 2014 at 9:51 AM, Edward Diener wrote:

Rob Stewart wrote:

...

...

Also, I urge you not to think in terms of language standards. Think in terms of compiler versions and their features.

Realistically this requires a knowledge of particular compilers and their versions that it is nearly impossible for any particular developer to have. Am I really expected to use or not use a C++ language feature in a particular release of my theoretical library because Compiler X, version Y does or does not support some C++11/C++14 feature ? I do not believe such thinking is conducive to expert programming.

A developer does not have to track compiler/library releases. Boost.Config macros take care of tracking, backed up by the regression testers.

I agree this is the best way to work with Boost and the C++ language standard. But when I do this I am not thinking "in terms of compiler versions and their features" as Stephen Kelly suggests but only in terms of C++ features as they are supported by Boost in Config.

Good point. --Beman