Hey Exanimates, We've finally started insider testing of the update! We've done one patch and we've got another big patch on the way. The testing release already feels very solid, with only a couple of very specific issues that we want to do some more targeted testing on. The new map is massive, varied and looks very different from anything so far. We've been trying some more complex things with the environment design and they have been well received and are working well. We're still adding a bit of visual polish to some assets and things like a few more small objects here and there, but these are small touches. We're going to make one significant addition in the test build still, the new thaumaturge / shaper encounter, which needs a little revision after many changes to character physics and AI, but that's it. Leading up to the test release it's mostly been finalising content and just transitioning from a still a bit prototypey dev build to a release build. Much of this was migrating all our creature types to the new race system and adjusting their motions and AI after all the changes to those systems too. A few of the older creatures got a visual upgrade too. All the changes to muscle function really messed with existing motions, and this meant having to redo or at least modify almost every attack motion and many others, not just for creatures, but also humans with all the different weapon types. There's literally hundreds of these, so it was quite the slog, but between the improvements to character physics and the fresh look at motion design things look and feel great. There was really a lot of testing new creature physics, new motions and the new combat AI which is still a work in progress. This led to a lot of improvements beyond the initial scope as we noticed various little issues, including some notable combat mechanics problems that had been reported in the past. In the last stages of development of an update we're much more hands on with the game instead of deep in the engine code and artwork creation. This, and especially when we do these rounds of insider testing, leads to uncovering and fixing many little old issues beyond the scope of the latest build and features. Maybe not game breaking stuff, but it all makes for a more polished and enjoyable experience. Our list of things we'd like to improve is many pages long and always growing, with many things that often seem small, but can take a lot of work to do. We do gradually complete the items on the list, which usually include something that someone in our community will be particularly happy to see. One thing that is a more serious issue that we've been giving particular attention to since the test release is pathfinding, particularly the tendency for NPCs to fall into pits as well as take some unlikely paths. This is a very difficult problem. On the one hand we don't want NPCs to take dangerous paths that could lead them to falling somewhere bad, on the other we do want NPCs to be able to take the same dangerous paths the player can, such as crossing a pit via a plank placed over it, or walk along a thin ledge between a wall and a drop. We also want NPCs to avoid walking against walls, except for cases like the narrow ledge, where it's better than falling. To be able to drop off a small ledge, but still prefer taking the stairs in most cases and so on. This requires complex pathfinding logic and nuanced decisions, things like this may seem straightforward to a human, but for a computer program that doesn't understand what these things are it's a whole different story. Our pathfinding isn't based on anything hand designed or designated walkable areas, it's just "polygon soup", characters can walk on anything including dynamic objects which change and create potential paths. We've been iterating on our pathfinding a lot, and while it keeps improving, these problems remained. It's one of those difficult open ended problems that we can't be sure even has a solution. Recently however we've had some new ideas on how to better represent things and how AI decides if and when to take dangerous paths. We've implemented these changes and our problematic test cases seem to be solved and we're also seeing generally improved pathfinding behaviour, with NPCs taking more sensible paths. We're also considering further additions, like perhaps companions voicing their reluctance to cross a particularly dodgy path, but being willing if encouraged by the player. This is not something we were confident we could solve, and it's been bothering us for a long time, so it's great and a relief to see such an improvement. It's been a hectic time working on a huge amount of different things to get everything ready, but it's been exciting and uplifting to finally see it all come together, especially after a stressful few months of unexpected and sometimes seemingly insurmountable problems. We can't say for sure how long we'll keep things in testing, so we won't, but things are looking good. Best, Bare Mettle