Ross Tate

Unfortunately low-level structural heap types tend to be quite large. If you think of a language like Java, a class's structural type includes not only all of its fields but...

@tebbi Unfortunately there are a number of assumptions in your suggestion that start to get why this is such a hard problem despite seeming straightforward. > For writing, add something...

@aardappel @Horcrux7 The last two comments seem to be on a separate topic. I think understanding casting performance is one of many items that will help inform a discussion on...

That's not how a Java cast to a (non-final) class works nor how an `rtt` cast is typically implemented. (Sorry to be abrupt.) They both use the same cast mechanism,...

That's true, which is why the microbenchmark only ever casts to classes.

I suspect a C benchmark would perform the same or worse. It would not incorporate the decades of engineering the JVM has benefited from (remember that every write to an...

Not sure. It's possible the JVM is already employing such a technique. But also, my understanding is that WebAssembly is not supposed to rely on inline caching for good performance.

Thanks for all the responses above! @jakobkummerow What are you doing currently? (I understand that it's temporary; I'm just curious what it is at the moment and how the teams...

Thanks for the update! I'm glad to hear y'all are looking into the wrapper-free approach, given that the research on efficient interop suggests wrappers can be quite problematic for performance....

Ah, good point on immutability of the prototype being useful for parallelism/concurrency! In prior discussions, similar concerns have been raised about supporting multithread-safe runtimes of Java and the like in...