Subject: Re: [boost] [c++11]
From: Niall Douglas (ndouglas_at_[hidden])
Date: 2013-06-19 16:32:29
> Sorry, I didn't mean to offend you. English isn't my native language and i
> sometimes use inappropriate words when I am excited about something :P
For the list's reference we sorted out misunderstandings off list. I should
have chosen my own words more carefully as well, and I do apologise to the
list for my earlier tone.
> Furthermore i
> believe that pushing that feature is also a good idea. However, that has
> implications and everyone involved needs to be aware of those. I will
> you in your efforts for going for a C++11 based solution under the condition
> boost will move on and deprecate all libraries that will be unusable with a
> compiler. Maybe Boost.AFIO should be the library that advocates and pushes
> this breaking change.
I'll paraphrase from that email discussion between myself and Thomas off-list
here: in essence, we discussed the striking difference in cleanliness and
tidiness of C++ code that can appear when it is written knowing that the main
C++11 features are guaranteed to hand as against code knowing that C++03
compatibility is a must. We agreed that the difference can be in orders of
magnitude, especially so for corner case scenarios such as the one I mentioned
earlier where twenty lines of variadic templates in C++11 can require pages
and pages of preprocessor generated specialisations and overloads in C++03. I
think we may even have concluded that such can be the difference that in some
ways, one could even consider that C++11 _is a different language_ entirely to
C++03 rather than merely a different dialect. Sure, it depends on perspective
and framing, and where you're coming from and going to. But I think it's safe
to say that anyone who has pushed C++11 hard wouldn't choose to go back to
C++03 willingly, and I think that emotional response is very good indicator as
to what comes next for Boost.
Anyway, as for proposed AFIO, my thanks to the list for such detailed feedback
on what will be expected for peer review. I think myself and Paul now have a
very good handle and what needs to be done between now and end of GSoC. We'll
get to it so, and I'll remind the list that progress in four of the seven GSoC
projects this year can be watched at https://github.com/BoostGSOC.
--- Opinions expressed here are my own and do not necessarily represent those of BlackBerry Inc.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk