Paul Fultz II
Paul Fultz II
> If you think the committee chose to use a simpler method for delivering the intent (which was described by Dave Prosser's pseudo code), how can we explain the item...
The best way is to use [cget](https://github.com/pfultz2/cget) which will avoid the rebuilds and can cache downloads. In cmake-get, we could store in a file which dependencies were installed and not...
It sounds like it should be called an Alias Class.
> MPL uses unpack_args for something slightly different, namely an adaptor that lifts Metafunction Classes to operate on Sequences. This is actually similar to the Fit library where [`fit::unpack`](http://fit.readthedocs.io/en/latest/include/fit/unpack.html) is...
> The real problem at hand really boils down to changing the meaning of apply to match that of std::invoke Well that's backwards. C++ has had a long history of...
> We currently do not apply these findings to our code and if we are not able to dogfood it we cannot expect others to use it. We should. I...
> That could just be the ugly std::transform ones though Whats the problem with `std::transform`? > std::string::find_first_of(): - https://trac.cppcheck.net/ticket/10777 We dont suggest this. > std::copy() - https://trac.cppcheck.net/ticket/10706 We suggest `std::copy`...
> In the past I had a slow function which was in the red hot path which was using that function several times (we modernized it from memcpy to std::copy...
> It would be more readable if the replacement was more like: > > return any_of(suffixes, p, isPrefixStringCharLiteral(str,q,p)); I dont that would be more readable as it not clear that...
> But this PR is still about flagging iterators, which also make for unreadable code. This true, using the algorithm in this case is not anymore verbose than the original...