After the spill of Invisible Ink, Part B in House Colvan after the imprisonment of their illegitimate heir leading to a lack of maintenance in their alchemical facilities is known, a instant ecological problem emerged from the reality that Invisible Ink is highly toxic to aquatic lifeforms as the Mors substrate is binded to the receptors where such lifeform receives Air. The Wise Alchemist must maintain the balances of the ecosphere like the cardinal elements, and careful operation at 7 parts per 16 rate is necessary to ensure perfect elimination of the compound.

------------------------

We think that the metric for ??? is (10*(rate - 16/7))^2, and 16/7 rate = 0 metric score. To achieve this; we made differential solves; used omsim plus a custom python wrapper to punch out all the metric value that is possible (and then learnt that such work was already done by predecessors and we're reinventing the wheel); then used pysr plus another script to do symbolic regression on each of the solves until the regressor spit out square(22.857042 - (rate * 9.999995)).

Especially with how easy this workflow is; I think there is more than 30 people able to find out the optimal rate; and making such a solve is probably very easy. Making it cheap tho; is really hard. We originally used dual inputs for our rate solve but we knew one reagent solve is doable; and then found out one reagent solve is cheaper than two reagents (since; again you don't need to make two sets of debonders). 

Unless you're somehow able to remove a debonder by reusing it *while* keeping the rate up; I don't think it is possible to get cost any lower. - Alchemist Clarissa

--------

Clarissa did some good work figuring out the metric for this one. I was a little disappointed there was only one metric involved but the actual solve was pretty interesting. I think we are pretty near min cost for one reagent but I'm not getting my hopes up. We theorised combining the two single arms into a hex arm but couldn't find a layout that fed into it. I'm 90% sure its geometrically impossible but they're could be a layout we missed. maybe there's just a cheaper way to do it in general....

That aside we didnt explore 2 reagent solutions much. It would require reusing debonders to get under 180 cost. It doesn't seem very possible but I look forwarding to see how rebix did it. - Alchemist Alice