Abyss-W4tcher

If we use a ridiculously big value, then the `context.object("array")` will still instantiate a very large array in memory even if 99.9% of the time it will iterate over no...

I know it might sound overkill when we could indeed just put one comment and keep the old way, but if we can achieve greater precision and coherency with the...

Yeah so originally as the output was the same I didn't realize it wasn't the correct way of iterating. When we iterate over arrays, we know their start and end...

Oh, I think there was a misunderstanding as I pushed this helper and other commits this afternoon. Please check the latest diff and tell me if it fits the framework...

Based on that, as I said earlier even in the new helper I don't think we can rely on existing vol3 array mechanisms for unbound arrays? The current ones always...

Yeah so using `utility.array_of_pointers` could work, because I realized I was maybe wrongfully thinking it was returning a list and not an iterator. In the second case then yes it...

Thanks for the details, I think it basically is the same as an iterator in the end. I initially thought that a thousand Pointer objects would be created but if...

So in the end I think you can merge #1801 and I'll handle potential conflicts here (there shouldn't be any) as well as a new design to fix the circular...

Really cool work, it makes me wonder if every line in Vol3 should be enclosed in a try/except given that smear can occur pretty much everywhere 😄. A wrong bit...

> A single capture of an `InvalidAddressException` is kinda what we have (see the CLI exception handling code), we just can't recover from it without knowing that it was ok,...