Alex Beregszaszi

The alternative is following up on #39 and and having a "bytecode version" field in front of the code in every single case, but that is still affected by problem...

> Unless main looks like main(requestOrResponse) and is supposed to have handlers for both types. I was actually thinking about this and somehow concluded the above, with reference to the...

Wouldn't you want to propose this PR in conjunction with changing `selfDestruct` at the same time?

The case when the argument is zero, could be easily solved by embedding a call to a function `check_for_zero`, which traps if the argument is zero, at run time. If...

@pepyakin good catch, all these four are. Perhaps the way forward is to specify a deterministic version of these four instructions, which are either: a) provided by the "ewasm vm"...

How are they deterministic with division by 0?

pp45: > Some operators are partial, because they are not defined on certain inputs. Technically, an empty set of results is returned for these inputs. pp63: > Where the underlying...

I think the spec could be improved, because one needs to jump around to unfold the entire story. I did found some missing test cases though which I'll try to...

Somebody needs to review it? 😉

For context, this was discussed between the ewasm and research team as well as it was one of the options before the launch of original Eth 1.0.