Hello, all!
It’s time to again share some new updates regarding game development!
This month I’m hard at work on developing and refining AI types and AI behaviours, as well as spending time on bug fixing and optimization for the upcoming free game demo that should hopefully be ready for release sometime in the next few weeks.
AI Behaviour & Difficulty Modes
In terms of AI behaviour, work has been underway for over a year on a ‘Smart NPC’ class that allows all AIs to have the same set of functional behaviours; all characters can be damaged or interacted with in similar ways. This means that characters in towns and cities don't just patrol a random loop, but instead have more compelling behaviour routines (storekeepers will stock shelves, roam their shop, wipe the tables, etc.) - and those same characters can also engage in combat, fleeing, etc.
Recent work has been heavily focused on adding dynamic difficulty scaling. Many games have rather uninteresting/disappointing systems for difficulty scaling (i.e., they simply scale up enemy HP/Damage). For Hermit, I wanted to ensure that NPCs feel harder without simply changing these parameters.
Harder difficulty modes should feel harder because of more intelligent AI behaviour rather than attributes. The new difficulty mode setting therefore changes some features of AI behaviour sequences to manage combat difficulty.
On harder difficulty modes:
AIs are more likely to block or dodge
AIs are more likely to be able to heal
AIs are more likely to fix bleeds and fractures during combat
AIs strafe and approach more quickly
AIs attack faster and with less delay between strikes
The result is a change in the way combat feels rather than simply longer fights versus enemies with more HP.
Since there are extremely tight time/budget constraints for Hermit (as a solo-developed project), these kinds of thoughtful design choices in game programming can make the world feel more alive without adding too much overhead.
Enemy NPC Types
One of the other areas of focus right now is on making sure that NPCs fit the world and behave in natural/interesting ways. All characters, including those that are not critical to the game's main story line (such as the Undead and other mobs), should clearly fit into the world in a natural way. Importantly, minor characters shouldn't just be 'generic filler NPCs' and instead should clearly belong to the places in the world where they are found.
As a result, many types of combat NPCs (even simple ones like Skeletons) are being redesigned to ensure that they feel like they belong with their surroundings:
Abandoned mines have undead miners
Ancient archives/libraries have undead librarians
Dungeons/catacombs have undead peasants and soldiers/guardsman
Dungeon kitchens have undead cooks/chefs
The hopeful result of this is that AI characters will more diegetically ‘fit’ with the world giving everything a more cohesive feel.
Settings, Optimization, Bug fixes
Work continues on optimization and bug fixes in advance of the upcoming Free Game Demo to be released soon on Steam. More than 50 bugs have been fixed over the past two weeks and the game demo is approaching a level of stability and optimization now where it is almost ready to begin close beta testing.
Other recent developments in this category:
Added separate world and music volume sliders
Added game difficulty settings to the settings menu
Added new settings for graphics scalability
Added new settings to enable/disable motion blur
Added new settings to enable/disable auto-hiding the hotkey bar
Known Issues
While significant work has been done on improving the game, there are still some areas where issues persist. Some of these are unlikely to be fixed before the initial Demo release, but are on the task list for the main game release.
The use of Torches and Bows is a bit cumbersome; while quick-throwing torches alleviates this somewhat, my intention for the full game release is that Bows, Torches, and Weapons will each have separate equipment slots on the Player Character (which should make swapping between them significantly smoother)
FSR3 is Implemented, but DLSS and XeSS are unlikely to be implemented in the Demo’s first release version
Some known bugs with quests, combat, and donkey companion behaviour still exist and are being looked into
Control remapping is currently still unavailable
Thank you for your continued support of the project!