What's not in the course

[bernhardmgruber]

This is intended as a meta issue to collect and track C++ constructs that are not yet part of the course. If we decide to add them, we will create new issues for them with more details and remove them here. In case we decided to not add them, we provide a reason here. Please edit the table below to add new content instead of commenting.

C++ feature	What is this about?	Reason if we don't want it
freestanding	A mode of C++ for non-hosted programs (i.e. embedded systems)
the C++ memory model	The theoretical model on which memory and concurrency is built
incomplete types	Restrictions and use cases for declared-only types.
alignment	alignas, alignof, overalignment, aligned allocation with new and align_val_t
unicode	char8_t/char16_t/char32_t and their literals, u8string, wchar_t and why char is sometimes enough
conversion operator	operator T() on a class
thread_local	variables global for a thread
anonymous unions	Union members inside structs which have no names
volatile	Disables optimizations around e.g. a variable.
pragmas	Discuss some important pragmas (e.g. unroll, pack, etc.)
[[attributes]]	We have [[fallthrough]], but there are a lot more
package managers	how to get external libraries via conan, vcpkg, spack, ...
typeid	It is mentioned once around std::any, but not explained. Include std::typeindex here.
namespace aliases	namespace stdx = std::experimental
ref qualifiers	struct S { void f() && { ... }};
member pointers	pointers to member functions and non-static data members
function pointers	Pointers to functions and how to define their type
full template specialization	we only have 1 example for partial specialization
function template specialization vs. overloading	Why overloading is preferable to specialization. How SFINAE and requires help here.
template meta programming	Type lists, type functions, template template parameters, algorithms on type lists, ...
bit fields	struct S { int a : 4; int b : 28; };
type traits and named requirements	trivially copyable, trivial, copy constructible, POD, standard layout, aggregate, etc.
std::bitset
std::deque
file IO	probably std::fstream
all of <filesystem>	how to handle paths, iterate directories, copy/delete files, etc.
IO manipulators	std::endl, std::left, std::hex, etc.
std::stringstream
Lists	std::list and std::forward_list. Why they are slow for many applications and when they are useful
Locals	Changing language, formatting decimal points, etc. Which functions are affected? Why you cannot read back a text file on your en_US notebook, written on a de_AT system.
Associative containers	std::set, std::multiset, std::unordered_set, std::map, std::multimap, std::unordered_map, std::flat_set, std::flat_map, ...
std::hash
std::queue
std::stack
std::pair and std::tuple
std::integer_sequence	This connects to std::tuple and meta programming
Optimization utilities	std::unreachable, [[likely]], etc.
std::valarray
std::array
<chrono>	Basically all the header: clocks, dates, times, timezones, calendars, weekdays, etc.
std::regex
<system_error>	All about error codes and handling platform specific errors
Allocators	Discuss that many STL constructs can be augmented with allocators to change their allocation behavior
Memory resources	All from std::pmr and <memory_resource>
Barriers and latches	From <barrier> and <latch>
Bit and byte manipulation	From <bit>, e.g. bit_cast, bit_ceil, rotl, endian, byteswap, ...
Math. constants from <numbers>	We can use std::numbers::pi now instead of non-standard M_PI
<semaphore>
std::source_location	As a replacement for __FILE__, __LINE__, etc.
std::expected	Alternative to errorcode/exception based error handling. We mention it briefly around exceptions.
stack traces	use std::stracktrace to get a stack trace. And you can print it.
floating points	There are a lot of myths around floats and also IEEE 754. We could discuss a bit how the format works, INF, NAN, rounding modes, precision, comparison, fast-math, ...
iterators	we mention them around the STL, but we never discuss their requirements, the iterator hierarchy, and some concepts behind them
memory order	Atomic operations have a memory order and we don't talk about its semantics. Especially, that relaxed is faster than the default seq_consistent, and when we need acquire and release semantic.
deducing this	Member functions with explicit object parameters
value categories	We only discuss rvalue and lvalue, but not xvalue, glvalue and prvalue

[stale]

This issue or pull request has been automatically marked as stale because it has not had recent activity. Please manually close it, if it is no longer relevant, or ask for help or support to help getting it unstuck. Let me bring this to the attention of @klieret @wdconinc @michmx for now.