Sure, dating is a mixed bag no matter your body size, but I always say that dating as a plus-size girl can be…different.Oftentimes, it’s downright frustrating, especially when I compare my dating life to that of my thinner friends (read more about that here). Find me on github to see what I'm working on this week. There's really so much: compile times, better support for Generic Programming, reflection, ranges, Unicode support, etc, etc. As a long-time member of the Committee, I probably have a different perspective on this than most. But on the whole, I think the committee does a good job for what it is: a group of volunteers. Do you think that if there were one person or someone esle who would be making all of the decisions — like W3C — it would be better, or is the Committee a good compromise? I've occasionally wondered if C would be better with a BDFL like Python has in Guido. Do you always understand multi-page errors that you get when using Boost? Does it worries you that using Boost makes compiling manyfold slower for some projects? As a library developer, I don't often interact directly with the outside world.
Parts of C have already been declared deprecated (throw specifiers) and some have actually been removed (export templates and auto as a storage specifier). Some are from standardizing features we didn't have enough experience with (allocators), and some are just bloated and poorly designed (iostreams, locales, string).
I've never seen a compelling use case for virtual inheritance (OK, I use it in my own code for simulating Concepts, but I look forward to a time when it's not needed). I would de-emphasize that feature and provide a clean, simple way to achieve dynamic polymorphism with value semantics (see: Sean Parent, «Inheritance is the Base Class of Evil»). Other languages have awesome tooling support that is hard or impossible in C because of the preprocessor.
In it's place, I would prefer an AST-based, hygenic macro processor.
Let me manipulate my program's AST at compile time. What is more important for you: aesthetics (to make language or ibrary more beautiful, elegant) or practice (to make it more useful for daily tasks)?
The Committee was ready to say yes to just about any coherent Ranges proposal.
There was an almost audible sigh of relief when I first presented my work.Deprecate the bad stuff (see: throw specifiers, auto_ptr).Write tools that help people program in the modern subset (see: clang modernize).When I look back on all the ugly hacks in Boost to work around the lack of decltype, I am deeply grateful that we don't have to live like that anymore.The same question, but about the most controversial changes.How do you think the decision should be made about what language to use in a large projects? Do you think that it’s bad for C that it’s rarely used on mobile devices? I'm dumbfounded that C isn't used more heavily on mobile. The prospect of maintaining parallel codebases in different languages is just too awful to consider. (Cue the angry trolls telling me what a Luddite I am for disparaging Emacs/vim/whatever.) How did you convince the Committee that the Ranges library is important?