Luis Michaelis

> Are you sure? At least in place, where exception been hit, there is access: Yes, that's what I mean. If you try to set the value of a const...

Whoops. And that's why I need script tests. It was an incorrect switch fall-through. Fixed in dc0e3404a7f1e56e8866bac297c3805032156832.

> > You just repeat that until i == 0 > > Nit: other way around. `repeat(i,32)` is equivalent to `for(; i > So far naked call, with no stack...

You're right I totally forgot about `continue;` and `break;`. With those in might, I believe sub-classing the VM is actually a good idea. Here's my new take. ### Things considered...

> Another issue is class-declaration. Class ZCARRAY is defined by script and engine should not be aware of it. So need some solution to threat some classes as plain-structs, with...

> There would be a pretty simple way of achieving that which I'll implement later this week. It ended up being not quite as simple as I thought. I've pushed...

> strings - have no goof solution how to manage them, as we want to work with opaque strings, but Ikarus sometimes writes directly to .data and .size, also assumes...

> Specially I'm not sure, if it's possible to have MEM_Free, after setting instance. In that case - dangling reference inside instance. Hm but everything inside it is allocated using...

Yeah I've had that problem it my GitHub Workflow too. GCC and Clang both compile and run fine with sanitizers so I doubt there is anything wrong at all. I'm...

It is already possible to disable sanitizer flags in _phoenix_. You can just `set(PHOENIX_DISABLE_SANITIZERS ON)` and define your own sanitizers globally if you'd like.