Guide to the C++ Filter View in C++ Programming - Nicolai Josuttis - CppCon 2024

[@LeDabe]

Great talk! This reminded me of the C++ Should Be C++ talk by David Sankel. Something weird has happened in the std committee to let such feature get standardized even though it is clearly broken.

[@FalyPr1nc3]

"We do not want that" (c) that's for sure. To be honest I was shocked while watching the presentation. As a C++17 user I was looking forward to using C++20 views. It turns out they have introduced an implicit state into undoubtedly stateless api as begin() and empty(). And this all could be avoided with just an explicit 'cache' call. Such a disappointment.

[@pyajudeme9245]

By far the best (and funniest) C++ teacher!

[@PaulMetalhero]

The design alternatives should be implemented as policies

[@robmorgan1214]

Wow. Jump through millions of hoops to make indecipherable typesafe code that still doesn't do what you want and still isn't memorysafe threadsafe or typesafe just to avoid raw void pointers... rustification of c++ has a few downsides.... despite the fact that any data on disk or from the network has no type info until we pour it into our objects... so we tolerate c cast of raw void pointers all over the place, but just pretend we don't.