- Jun 13, 2022
- 39
- 113
Here are some general tips:
The user interface (UI) is not great, I know, with 0.1 and all, but for something like this, usability and the user interface are very important. Hide the Patreon options and make the configuration editable in the options as well. If possible, only have the chat interface with all the icons in the system settings with the cog thing as well. Have a prompt in the beginning asking for your OpenAI API key. You've already hidden tons of things behind the Patreon thing. The stain on the character alone is enough for most to consider. But having an interface on the right where all but two buttons don't give me red text shouting out Patreon isn't needed at all.
Next would probably be stable diffusion integration. Match the set prompts already for the girl, and if you interface in a certain way, you'll get a stable diffusion-generated image. The prompts for the stable diffusion girl are already there. You have the personality, the looks, all you would need to add is the extra prompting of position and looks of the character based on what the user wants to do next. This is a more grandiose approach to the game as a whole, as it would probably take more time for the AI and us to adjust. But potentially, you could have her react in real-time to typing or voice commands. It might already be possible with text2video.
No, I haven't played the game yet to test the prompts and stuff like that, but I do plan in the near future when this becomes a little more developed. These are just my initial tips for it as I see the dev looking and responding in the thread.
The user interface (UI) is not great, I know, with 0.1 and all, but for something like this, usability and the user interface are very important. Hide the Patreon options and make the configuration editable in the options as well. If possible, only have the chat interface with all the icons in the system settings with the cog thing as well. Have a prompt in the beginning asking for your OpenAI API key. You've already hidden tons of things behind the Patreon thing. The stain on the character alone is enough for most to consider. But having an interface on the right where all but two buttons don't give me red text shouting out Patreon isn't needed at all.
Next would probably be stable diffusion integration. Match the set prompts already for the girl, and if you interface in a certain way, you'll get a stable diffusion-generated image. The prompts for the stable diffusion girl are already there. You have the personality, the looks, all you would need to add is the extra prompting of position and looks of the character based on what the user wants to do next. This is a more grandiose approach to the game as a whole, as it would probably take more time for the AI and us to adjust. But potentially, you could have her react in real-time to typing or voice commands. It might already be possible with text2video.
No, I haven't played the game yet to test the prompts and stuff like that, but I do plan in the near future when this becomes a little more developed. These are just my initial tips for it as I see the dev looking and responding in the thread.
Last edited: