Hacker News Reader

Comment by HarHarVeryFunny | Hacker News Reader

Those types of allocation technique were common back in the day for efficiency reasons, maybe still relevant for things like embedded programming where you need to be more careful about memory usage and timing, but I would say that nowadays for normal application usage you are better off using smart pointers.

It's not a matter of one being strictly better than the other, but rather about using the right tool for the job.

View on HN

Many soft-realtime systems make use of these techniques, specifically 3D graphics and game engines.

I disagree, as group lifetimes are conceptually and architecturally often easier and simpler than having lots of individual lifetimes managed by smart pointers. And sure, you can often slap shared_ptr around the place, or hopefully a less lazy smart ptr choice, but it makes the code harder to understand by obscuring ownership rather than eliminating it as a concern.

I don't understand why you see it that way... Whether using an arena allocator or smart pointer, in both cases you need to allocate the individual object (p = arena_allocate(arena, ...), or p = std::make_unique<T>()). In the case of a smart pointer that is all you have to do - deallocation will be automatic. In the case of the arena allocator you will also need to deallocate the entire arena at the appropriate time.

How are you conceptualizing this that the arena allocator is simpler?

How are you conceptualizing smart pointers as "obscuring" ownership, when the entire point of them is to make ownership explicit! The smart pointer IS the owner!

Reply on news.ycombinator.com