Ingvar Stepanyan

@netvl I'm surprised Cargo itself didn't strip the suffix for you as it does nowadays for any new project.

Actually, I see https://webassembly.github.io/threads/binary/types.html#binary-limits actually has slightly different types/notations here. I guess Overview is simply out of date?

@rossberg Thanks. That notation (`lim,0:limits`, `lim,1:limits`) was also a bit confusing - is it documented somewhere? I don't think I've seen this one in particular before, usually it's either literals...

> It may be a bit too smart in this case. Yeah I have a feeling that it would be easier to grasp if it was written down as two...

> Seems more natural in this case. Why? Or, more generally, are the rules on what should be checked at parsing time vs validation time codified (written down) anywhere? To...

> But to some extend it's a choice. Right, I guess that's the blurry line that sometimes makes me uneasy. It makes sense for what you said about shared tables,...

@rossberg To get back to the original issue - the resolution here is that someone will update the spec as per > I tend to adjusting the spec and make...

@binji I see you renamed the issue to "Make shared tables a validation failure", but I think the resolution was to also have the maximum on memory limits be a...

> so we shouldn't automatically expose them They're already explicitly marked with `(export ...)` so it seems rather odd and backwards incompatible not to expose them when engine supports interface...

> To be clear, you're just proposing a change to the boundary between an invalid and a malformed module, right? Yep.