Background reading on the reference counting vs tracing garbage collection debate
Eight years ago I answered a question on Stack Overflow about the suitability of OCaml and Haskell for soft real-time work like visualization: " for real-time applications you will also want low pause times from the garbage collector. OCaml has a nice incremental collector that results in few pauses above 30ms but, IIRC, GHC has a stop-the-world collector that incurs arbitrarily-long pauses " My personal experience has always been that RAII in C++ incurs long pauses when using non-trivial data (i.e. nested, structured, collections of collections of collections, trees, graphs and so on), non-deferred reference counting has the same problem for the same reason, tracing garbage collectors like OCaml work beautifully but there are many notoriously bad tools like Java that have given tracing garbage collection a bad name. Now that I am revisiting this issue I am surprised to find many individuals and organisations repeating exactly the same experimental tests that I did and coming...