Jake Lishman

The design means that there's no overload of the constructor that allows a single-statement reconstruction, which is what's needed for `eval`, while looking like a single object. You have to...

I've tagged this "good first issue", though we wouldn't want one PR to add `repr`s to every class - that would be too much work, and a good `repr` needs...

Given that it's an alternate class method constructor, I'm still not sure it would be appropriate for a `repr` (and `Instruction`, `Qubit` and `Clbit` don't general have a `eval`able reprs...

I would tend to agree on removing `Qubit` - there's not really anything sensible we can provide for it at the moment. I do _also_ agree that its repr doesn't...

It's not just `seed_transpiler`, I think, it's all the arguments, but the docstring does say this: https://github.com/Qiskit/qiskit-terra/blob/a296ca009bf440b5a2cb00f61b8221c6ce9aa044/qiskit/compiler/transpiler.py#L73-L76 That said, if we're going to have type hints, they should at least...

@aniken04: the current actionable part of this issue is to update all the type hints in the manner I'd described in my comment above. The other discussion here is more...

Thanks, I've assigned you, and let us know if you've got questions!

Hi @Eriken79, sorry for the delay: for simple things like `int`, it's better to directly write `Union[None, int, List[int]]` in the `transpile` docstring, because then a user can quickly read...

The docstring looks fine as it is to me - the comment at the top addresses that all options are also allowed to be a list of values, and there's...

Hi @Eriken79, sorry, I didn't see this at the time. - `basis_gates` is a collection of strings in the singleton case - for `initial_layout`, that's the sort of thing where...