Sunday 10 October 2021

27th Anniversary of LBA 1

Finally arrived LBA 1 release anniversary. This year it is a pretty weird one. On one side it is a happy one because we got the announcement that a new entry to the universe will happen. On the other hand a little bit sad because our celebration stream was canceled. I initially planned to replace my stream from it with another dev update, but then life happened. 

I got super busy, partially upon my own request. Specifically, I work on a presentation for the GiC conference where I will be talking about a custom resource management system. I think that this will be an interesting lecture, mostly because I will be talking about the engine that powers the L.B.A. Tribute. If any of you will be there somewhere around and see me don't hesitate to come by and say: hello. 

Now after this long intro, let's focus on the project. From the last update, nothing really changed. The plan is to work on this nice small demo containing only Citadel Island. I will release it when I will be done with it (Sadly there is still no release date). 

Currently, I am focusing on AI. I was not really happy with my previous solution and decided to improve upon it. A few AI-related presentations later I felt like I'm still lost. I understood what they talking about but still did not sorted stuff in my head to make a final decision regards my approach to it. This came later after few beverages (alcoholic and non) and discussions with friends. The last one probably helped here the most.

I decided to use behaviors graphs. How they work for those who don't know: when designing AI you building different states and defining conditions that need to be met to transfer between them. A simple example can be seen below.

Little did I knew, how this decision will backfire on me. The next weeks were full of UI work. Not a lot of AI but mostly thinking about how to make the creation of graphs possible in the toolkit. This was important for me because I would spend a lot of time using this tool later.

Right now I think I'm done with most of this work. It took forever, lead to some interesting discoveries, performance fixes, bug fixes, UI improvements, changes in the workflows, and generally a lot of boring topics that I will spare you. What is important, now I plan to continue work on AI with the new tools I designed. How this will go we should see soon.