It's not as large as Chromium by far (~55k lines + ~27k lines support library), but I quite like the Thrill codebase. It's a research prototype (but not what yo

It's not as large as Chromium by far (~55k lines + ~27k lines support library), but I quite like the Thrill codebase. It's a research prototype (but not what you would typically associate with "research code") of a big data platform like Spark but in C++. It makes nice use of lambdas with return type deduction, some template metaprogramming in the internals, smart pointers & RAII, and generally uses lots of modern C++ features. https://project-thrill.org/ https://project-thrill.org/ and https://github.com/thrill/thrill https://github.com/thrill/thrill (disclosure: I know the authors and have worked on/with it for a bit).