Trying to translate over as much medi data as possible is computationaly expensive, right? Because what we have to do is hat to generate a dump, and then the dump hasa different dat to model. So there's a reason why it takes so long. We also don't want to have it happen during we're probably going to be run at a low like process priority,. Which also means we want to do it, like, on a friday, in the afternoon, so that it's over the week end and not when kidhum gets hit hardest.