@LuisMendo Nice

Speaking of further changes, something to support having multiple test cases when the code uses G would be good

i.e., when the code uses G (or worse, nG), the usual trick of wrapping it in ` ... DT doesn't work properly because G doesn't change

If the input is a single value, we can assign each input to H (in the header) and use H instead of G in the multiple test cases code

but if we have 2 or 3 inputs per test case for eg., it's hard to replicate 1G or 2G in the inner code and have it be really equivalent

A neat way to manage this would be to clear G after one input set has been processed

In fact, an overall clear all command to reset H, K and everything to their original values would be good, for this use case

Zx seems thematically appropriate and sounds nice and final

@sundar In those cases you can use modular addressing in G. So 0G always gives the last input, -1G the second-last etc

Anyway, if you think clearing all clipboards would be useful we can add Zx as you say.

However, since G is an automatic clipboard, I'd have to think if clearing it could bring some complications.

Same for M

@LuisMendo But that would mean 1 more byte for the -, which means that's not going to be the main submission version. That's what it would be nice to avoid - having (slightly) different versions of the code for the single test case version and the multiple cases version.

And I'm not sure if there's any current way to neatly change this answer from single test case two-input version to multiple test case two-input-per-case version (,it0)'k'-tUw6*+qw3L)U*]>G for the Go rank qn)

Having a Zx-like command will mean we could treat every iteration as a fresh new program, which makes it easier to switch to multiple test cases in cases like that

(To be clear though, I'm just explaining it in long, not saying it's a big crucial feature or anything. Just a nicety that would make some things easier, assuming it's not a daft idea in the first place.)

@sundar Yes, that one would be tricky

I can include Zx (unless Dennis replies affirmatively to this). It's only a bit troublesome because the way program state is initiallized right now, but no big deal.

Hi