I have also received quite a bit of private email begging me to keep Outcome free of a hard Boost dependency if accepted. [snip] Actually there is overwhelming evidence that Boost **users** do not want a dependency on Boost. But I've been saying that for years now.
It seems like these users are best satisfied with just github.com/ned14/outcome. i.e. They wouldn't even need Outcome to be part of the Boost collection of libraries, if they desire Outcome to be standalone and independent of Boost.
That's not what I've been told. They want Boost libraries without all the hassle that comes with Boost libraries. I keep saying this for years, but no one here on boost-dev appears to listen.
Meanwhile, those Boost users who do have some expectation or desire that libraries included in the Boost distribution benefit from all the work done in other libraries (e.g. compiler workarunds, compiler feature detection), or simply not duplicate code that another Boost library has an optimal implementation of, would have different expectations of a "Boost" Outcome.
The customisation points are available for any end user to change a minimum default implementation provided by Outcome to any other implementation, including a Boost implementation. I think that a perfectly fine situation if it avoids a hard Boost dependency, it's not like anybody is losing out here. In numbers terms, far more people have asked for Outcome to remain free of a Boost dependency than the relatively few here on boost-dev. As with all the other changes I've agreed to, I'm following the numbers, if a clear majority want something, I try to provide it if I can. Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/