Simon Krajewski

While I'm saying that it could work, I have some doubts about the cost/benefit here. I imagine that a hybrid cache would work like this: 1. Keep any module that...

I agree. All this can be summarized as the difference between "doing something" and "doing nothing" that I mentioned in the hxb PR. Our profiling suggested that we spent by...

Another aspect that is often overlooked is the amount of persisted memory. While we tend to treat memory as this big cheap thing that everyone has plenty of, a large...

The module to binary ratio is about 5:1 in the codebases we tested. Will be interesting to compare this once #11511 has been implemented.

This change makes a lot of sense, I don't think "something else might be faster anyway" is a good reason to not make it. ... but I can't even reopen...

I'd like to know what's going on with cppia there because this looks quite harmless. Will try to isolate the problem, but it's not very important for the next release....

Could you update the branch? I'd like to see where we're at.

So cppia still fails: ``` Command: /home/runner/work/haxe/haxe/tests/unit/bin/cppia/Host-debug [bin/unit.cppia,-jit] src/unit/TestMainNow.hx:6: Generated at: 2024-02-08 14:35:21 src/unit/TestMain.hx:31: START src/unit/issues/Issue10831.hx:16: null Command exited with 134 in 1s: /home/runner/work/haxe/haxe/tests/unit/bin/cppia/Host-debug [bin/unit.cppia,-jit] terminate called after throwing an...

No, if anything those could go into hx4compat. I don't know how many types there are that work on the JVM target, IIRC quite a few of them have `__java__`...

Let's separate this into the map and EReg implementations. Also a question: Does the Map.size implementation have linear cost on any target?