But the paywall, i.e. the metered paywall where some articles are behind a paywall by author choice and an unsusbcribed reader can read three of those a month, is up there. And I'm just coming at it from the opposite side. Many people here don't want to see a paywall. That's fine. But I'd like to see their reaction if the quality of what was behind it changed. A no might still be a no, but I think the tenor would change. So that's how my original take on fix quality first fits into the anti-paywall headline of other people. This model works well enough for us in the worst possible implementation that I believe it will work all the way if we have a better implementation, especially if there is more of a sense of that there's at least something valuable behind it.