Why did we choose [to do the 4.0 preview]? What did we learn from 4.0? 4.0 was in testing for three months before it went to preview, and then we left it on for another month and a half. How did this process go?

Well, so during the development of 4.0, there was a lot of new technology coming on, right? So it's a major patch that changes a lot of the fundamentals. Obviously throughout the year, we've been developing server meshing, we've done tech previews on server meshing. And so getting to the end of that process, basically exiting the tech preview streams, coming back with the 4.0 content, it was a major endeavor to synchronize all that, make sure the new content that was produced fit with the technology and that it was working. Even the results that we got with the 3.24 release, for example, were completely obscured by the new content. At the end of the year, we took a decision to split the two releases.

So leave 3.24 up, put 4.0 in a preview.

Correct. Because as we were getting closer to the end of the year, and time is running out, what we knew from the integrations is that the patch was not quite ready because we needed more concurrency on it to discover more issues. And we were kind of at the maximum of what our PTU concurrency could actually deliver. But we knew it wasn't ready for a live release because we want to be fair with our assessment of where the patch was at. We knew there were some egregious issues that we had yet to discover. I'll talk a bit later about the fog of war of crashes. And so we decided to go with the lesser of the two known problems, which was leave 3.24 live for players to enjoy in a stable release, right? But at the same time, give them access to a 4.0 preview. Now, since we had taken the decision months ago that the 4.0 patch was going to be a full wipe, that gave us the ability to basically preview the patch but allowing you guys to keep your persistent progress in the preview as we move up to at some point bring 4.0 on the live environment. So overall I think that was a very difficult decision to do this option. But I think it was the best for the project because it allowed us to de-risk a little bit the .0 initial release, which had a lot of issues even during the preview. A lot of the skeleton crew teams worked throughout the break to actually document, hotfix some of those issues as well and help players that got into weird situations. And start the ball rolling for the year with already important fixes to the underlying technology.