A problem with explicit interning is that libraries, when creating objects, cannot make an informed decision whether to offer some runtime performance in exchan

A problem with explicit interning is that libraries, when creating objects, cannot make an informed decision whether to offer some runtime performance in exchange for the expectation of a decrease in memory usage. And it is even less than an expectation. Many interning libraries never free objects that they created, so they can keep lots of objects that never are needed again around, and thus increase memory usage. I think the ideal API would somehow a) be simple and b) have the program communicate with the system about what they desire from the system. As a first step, the system has to know how much memory the program is willing to use, how much longer it expects to run, and for how long the program plans to retain data it is creating now. Simple examples: - if the program is shutting down, and there’s a megabyte of memory available, it’s likely detrimental to intern any new data. - in a “json grep” tool, interning json keys likely isn’t worth it, as most data will be discarded soon. That’s probably at least a dozen of ph.d’s of research, and likely not attainable, though.