Build 26499813

This patch makes the Experimental AI optional, restoring the Classic AI by default. While Experimental AI has come a long way since the initial Early Access release, it still has some disappointing results and consistently slow decision-making. Experimental AI will continue to be developed, but players may now freely choose which opponent AI to use in the Settings menu.

Support has also been added for adjusting animation speed scale, switching to windowed video mode, detecting user-provided localization/translation, and other minor improvements. Read below for the full details.

Features

• Adds the ability to set Animation Speed Scale in the settings, speeding up gameplay in some specific cases. Note that this won't speed up the Experimental AI's thinking time.

• Adds experimental support for a traditional windowed video mode. This may have a few minor bugs currently, such as windows not being exactly the right size.

• Adds experimental support for user-provided translations. In the game files, localization strings are found in resources/app/dist/src/resources/locales/en/index.json. If you place your own translated version of that file in a language-specific directory, using either its language code or locale code, the game will allow you to select that language in the Settings menu. For example, if you create the file locales/ko/index.json and then start the game, the Korean language will become selectable in the Settings menu.

Game Improvements

• AI opponents now correctly draft decks in the Gauntlet and Rift modes.

• Fixes an issue which caused the game to reload itself when the window size changed, causing it to swap assets in and out. The game now always uses high-quality assets.

• Reduces the HP of Generals in the Campaign's capital city encounters from 20 to 15.

• Improves campaign dialogue to better explain some encounters.

• Cleans up the profile creation menu, including more information about Historic Mode.

• Adds a Steam Deck keyboard hint to the profile creation menu.

• Adds a tooltip to Rift explaining that multiple concurrent runs are supported.

• Adds in-match dialogue to enemy Generals during capital city encounters.

Experimental AI Improvements

• Improves Experimental AI handling of Walls, Creep, Artifacts, Eggs, Mechaz0r, ranged units, triggered effects, and large threats like Worldcore Golem.

• Experimental AI now shows a "Confused" emote when it ran out of thinking time on its turn.

• Adds a hard cap of 30 seconds per turn, preventing situations where the Experimental AI becomes unresponsive.

• Fixes an error which could prevent Experimental AI from ending its turn.

• Fixes an error which could cause Experimental AI to discard multi-step plans.

Bug Fixes

• Fixes a database error which could occur when starting a Rift run.

• Fixes a typo in campaign battle description text.