On 2 Jul 2015 at 13:47, Gottlob Frege wrote:
Changing something like boost::optional to have conversion to tribool instead of bool, or removing operator*() and forcing users to use get(), etc. Those are big breaks.
I'm getting the sense that the biggest head wreck for people here is the idea that optional and future are the same thing. As in, literally the same implementation code with the same API.
I have no idea how to get users to actually be guinea pigs for those experiments. Should we have 3 different optionals to choose from? (Should we randomly select which optional you get in your boost distribution so as to get good A/B testing :-)
I've tried to not deviate from source compatibility unless there was a good reason. Either from optional, or expected. My hope is that with a bit of template aliasing the implementations could be swapped. Regarding your comments that WG21 is a lousy place to design libraries, I agree in the strongest possible terms. Boost once was the place to playpen exciting new C++ libraries - if you read this mailing list back in 1999-2001, there was a palpable excitement here as people tried new things and showed one other what could be possible. Rather like in the Rust mailing list nowadays.
From 2008-2009 onwards things haven't been as good here, and especially since Dave left. It doesn't help when members of the steering committee conspicuously fail to perform their duties, and specifically disavow taking any leadership position except to intentionally prevent and inhibit change.
I strongly support your earlier comments Tony, and your rationale, and it's long overdue for an Boost.Experimental library distro. If Boost can't and won't do it, someone else should. (For the record, it won't be me, I'm likely going to be unemployed next week and my plate this year is too full. But I'll support anyone who does best I can). Niall -- ned Productions Limited Consulting http://www.nedproductions.biz/ http://ie.linkedin.com/in/nialldouglas/