Coffee Diary - May 2026

Hey Exanimates, The complete update is currently in testing and at this point we're just looking for issues that need to be urgently fixed before release and tuning various things. The latest test includes the changes we discussed in the last diary, many improvements and additions to the Gardens including a sub-level, the new shaper and their thaumaturgy domain, another new creature, new items and various other things. Some of the changes and additions are quite significant and involve major engine level changes, so they certainly need some testing, but we've already solved a good portion of important issues in a very brief period of testing. Of course the other new creature was a lot of work, but after having just done a big overhaul of character physics and general improvements for the last one, we didn't run into much of the same issues again. This other creature does have some unique elements that required new features and this time we wanted to create its rig entirely from scratch using our new tools. This meant some significant additions to the tools and we made many improvements along the way. It definitely took longer developing the tools first, but with them in place we are able to do what would normally take several days in about an hour, and do it better too. We will be adding more creatures and redesigning some creatures later, and after the 1.0 release we will work on "human 2.0", which will bring many improvements and an overhaul of core assets for humans. This will require updating many things including hundreds of items, and the new tools will be critical in dealing with the workload. In general this is a new class of tools for our engine that will have many applications besides and we've done a lot of the ground work to support them and the new workflows they enable. The new creature did still highlight some issues with character physics, AI and combat mechanics, much smaller in scale than last time, but they still led to some general improvements. We fixed some old issues with foot-ground interaction which result in smoother and more consistent footwork behaviour and some horizontal swing targeting improvements which all applies to players too. We also made more combat AI improvements which are very noticeable, though at this point it feels like we need to handicap low skill AI more aggressively. The new thaumaturgy domain brought some interesting challenges with power mechanics and also teaching the AI to use them. We always develop powers as player usable first, then teach the AI to use them after. Some of the new powers require new general features and kind of broke all the rules we had in place for when and how powers can be used, what interrupts them, how you control their duration etc. so we had to expand on those systems and it got quite complicated, but useful for more powers to come. For the AI we also had to come up with a better tactics system that involves not just evaluating each power individually, but also considering tactical information returned by other available powers. The AI works just like players here, they don't have scripted combat behaviours or patterns, but simply a list of available powers and the same restrictions as the player, so they need to be able to make decisions and plan ahead without a predefined overall strategy. A notable addition is the ability to acquire powers from defeated shapers. This allows you to learn basic powers from other domains, including the one we