Andreas Rossberg

#### Checking argument vector sizes > Done. Thanks. However, this is passing structs by-value now, avoiding which is an explicit requirement of the C API (since FFIs tend to not...

Given typical linking models, which rely on a more or less global name space for public symbols, I don't that can be done in common environments. This isn't a problem...

I see the problem, but it is a bit unfortunate that every vec allocation & deallocation now involves another function call. My understanding is that inline functions have static linkage,...

I see, good point. In that case, can we somehow make linking between client and lib with different DEBUG settings fail in all cases, such that users are forced to...

@jakobkummerow: > LSan is more thorough (and convenient) than handcrafted accounting anyway. Yes, but imposes much stricter requirements on the overall program, I think. In particular, will it work when...

@jakobkummerow, the use case would be embedding V8 with foreign language bindings where the host language runtime may not be LSan-ready, but the user (of the language) has no control...

Do you mean `std::span`? For C++ alone, that would be possible, but I'm afraid it will make it harder to implement the C API on top of it efficiently, which...

Note that the API does not yet support the threads proposal (e.g., no shared memories are expressible yet). So far I had assumed that staying within bounds is the responsibility...

Keep in mind that this is an embedder API, not an API that should be made accessible to untrusted client code. The embedder has to be part of TCB. Since...

A more efficient and direct call interface (in both directions) is something we have discussed. It clearly is desirable, but it is less clear how to provide it or implement...