C++ - Reddit

ML Code Optimizations Curious who is working on this right now? What I am talking about is profile guided optimizations using ML approaches rather than traditional algorithmic statistically holistic approaches. For example, traditional POGO provides global feedback on estimates for branches taken etc, whereby ML approaches can provide context sensitive analysis that can generate different code paths as required based on probabilistic bifurcation. I think this line of research could be very beneficial, even beyond C++ but perhaps mainly for Jitters, especially Cloud oriented Jitters. A cursory look for scholarly articles resulted in little, so throwing this open. https://redd.it/n6i0e3 @r_cpp

implementing affine type in C++ using stateful metaprogramming https://godbolt.org/z/PnPPrnPjY for more type systems in C++, see: https://github.com/IFeelBloated/Type-System-Zoo https://redd.it/n6dxar @r_cpp

Detecting memory management bugs with GCC 11, Part 1: Understanding dynamic allocation https://developers.redhat.com/blog/2021/04/30/detecting-memory-management-bugs-with-gcc-11-part-1-understanding-dynamic-allocation/ https://redd.it/n6fpye @r_cpp

Qt Creator 4.15: What's new in C++ support? https://www.qt.io/blog/qt-creator-4.15-whats-new-in-c-support-0 https://redd.it/n6f3ze @r_cpp

szLogVar::variant. C++17/20. O(log n) recursive template instantiation depth. Now updated with C++20 features. https://github.com/SteveZhangSZ/variant https://redd.it/n6d56j @r_cpp

A Conan, CMake, Qt6 project template for C++ https://github.com/zethon/CppCCQTemplate https://redd.it/n6c2qw @r_cpp

Read and write pickle files from C++ Hello, everyone! I am working on a C++ program that needs to write and read .pkl files (in C++, without calling a Python script to write them). Pickle is a Python package. My program invokes a Python script that reads a .pkl and outputs another one. I am having a hard time to find a solution for this. It seems that PicklingTools is not updated. I need to use the latest pickle protocols (e.g., 4 and 5). It seems Boost.Python guide isn't updated and I couldn't finish building it. It isn't clear how I could use Pybind11 to achieve this. Can someone make a small example with Pybind11? Is there another possible solution that you know of? Thanks! https://redd.it/n678hs @r_cpp

fmtlog: fastest C++ logging library using fmtlib syntax fmtlog is a performant fmtlib-style logging library with latency in nanoseconds. https://redd.it/n67jhg @r_cpp

Generate compilation databases (compile_commands.json) online for GNU make C/C++ projects Hi there, I built a simple web app a few months ago to generate compile_commands.json for GNU make C/C++ project at https://texttoolkit.com/compilation-database-generator, for the sake of easy use. If you work with C/C++ projects and want to use VSCode/CLion/clang-tidy/Sourcetrail, it could help in just two steps: 1. Copy and paste the make output of, for example, `make clean && make -nw` 2. Download the compile_commands.json file, or copy it out. If you encounter any issues or have suggestions, welcome to post here or email me, I will try my best to improve it. https://redd.it/n5h20g @r_cpp

Announcing Meeting C++ 2021 online conference! https://meetingcpp.com/meetingcpp/news/items/Announcing-Meeting-Cpp-2021-online-conference-.html https://redd.it/n631rd @r_cpp

Memory error checking in C and C++: Comparing Sanitizers and Valgrind https://developers.redhat.com/blog/2021/05/05/memory-error-checking-in-c-and-c-comparing-sanitizers-and-valgrind/ https://redd.it/n61axk @r_cpp

What is the proper way to return some extra information from a constructor that is really not the part of the constructed object? Consider the following scenario. I'm writing a font class that loads a font from a file in its constructor and provides facilities for text rendering. The constructor takes as arguments the font file and a collection of Unicode code points to be included in the font object. The problem: some of the requested code points may not be present in the font file, but this is not a failure condition. The object should still be created and then it is a separate question on what to do when attempting to render text including missing code points. However, it should still be possible to communicate which of the requested code points were not found. This might be useful information which we should be able to preserve (by opting in). Clearly, I cannot throw an exception since this is not an error and I want my object to be constructed. After some brainstorming I came up with some potential solutions. Here they are, starting from the worst and moving to slightly better ones. 1. Include the collection of missing code points in the font object. This is wasteful and just wrong. 2. Provide an extra optional parameter to the constructor, say a vector to be filled with the missing code points that can be inspected after the construction. This is clunky to use. It also feels wrong conceptually since I believe that a constructor should not have arguments that the constructed object does not depend on. 3. Provide an extra optional parameter to the constructor that is a callback that is called for each missing code point. Usability-wise this is better than the previous option, but suffers from the same conceptual problem. 4. Have a vector of missing code points as a static member of the font class that can be inspected after each constructor call. This is better than option 2, it is both easier to use and doesn't clutter the constructor calls with incidental arguments. 5. Store the callback as a static member of the font class. This is a variant of option 3, it is most flexible and easiest to use and also doesn't pollute constructor calls. Among the ideas I had, I like this one the most, but I'm not sure if it really is a good solution. Is there a standard solution to this problem? What do you think about my ideas? Do you have some other suggestions? https://redd.it/n5av23 @r_cpp

The Little Things: Testing with Catch2 https://codingnest.com/the-little-things-testing-with-catch-2/ https://redd.it/n5rne0 @r_cpp

Semantically ordered arguments should be lexically ordered too https://quuxplusone.github.io/blog/2021/05/05/its-clamping-time/ https://redd.it/n5mauy @r_cpp

[clang] Add -fc++-abi= flag for specifying which C++ ABI to use https://reviews.llvm.org/D85802 https://redd.it/n5hyrw @r_cpp

Video lecture series on advanced C++? Any resources. I want to learn advanced C++. But as of now I do not have an idea what would some advanced concepts be. Like socket programming, mutliprocressing, GUI etc. Can someone please help me with the list of advanced concepts and good video resources to learn them for C++? https://redd.it/n5ebfv @r_cpp

Portability is Reliability http://evan.nemerson.com/2021/05/04/portability-is-reliability/ https://redd.it/n5e0w9 @r_cpp

Any idea how to apply A* or Dijkstra here? Hello, I'm a University student in second semester of my Bachelor degree and I got this task to apply Dijkstra's algorithm for following assignment. We got a string of digits 0-9, the digits can repeat. And all you have to do is to find the shortest path to reach the end of the string. There are 3 allowed movesets: 1. One index forward, i + 1, if it's possible 2. One index backward, i - 1, if it's possible 3. Jump to the same digit anywhere in the string, if the index of your position isn't the same as the index of the digit you wanna jump to. Result should be number of jumps and indexes visited. Example input => p = 123415 Example output => 2 jumps, visited indexes: 0, 4, 5 ///////// Example input => p = 180557994366325331615166132279 Example output => 6 jumps, visited indexes: 0, 17, 16, 9, 8, 7, 29 ///////// Example input => p = 12345 Example output => 4 jumps, indexes visited: 0, 1, 2, 3, 4 I'm really desperate, because our teacher never told us anything about these algorithms and I don't know how to apply it even though I've been studying it since the assignment. https://redd.it/n5d6rn @r_cpp

vcpkg now (finally) has site with search functionality https://vcpkg.io/en/index.html https://redd.it/n547g0 @r_cpp

Standard Library <random> slow to load/save? I'm using std::mt19937 mes; std::uniform_int_distribution<int32_t> dist; as a random number generator in a game, and I need to save its state every frame. To save/load the state I'm doing this: std::string Random::save() const { std::stringstream s1; s1 << mes << dist; return s1.str(); } void Random::load(std::string s ) { std::stringstream restore(s); restore >> mes; restore >> dist; } It works, but it's slow. In fact, the save() function here is currently taking about 1.5ms on my machine - this is about 90% of the total execution time for one step of my game engine! This seems really slow to me, am I doing something obviously wrong? https://redd.it/n53g4v @r_cpp