I thought a monthly review might be a good idea for a post, that way even If I am busy IRL, there would still be at least one blog post a month. I thought of this idea for several reasons but the main one I will explain below.
Last night, I felt myself asking the question
“It’s been another month already and I still haven’t done finished X,Y,Z”
I started to get a little upset with myself. Have I really wasted another month? How come I didn’t do as much as I thought? Then I remembered something, and I feel that if you are in a similar situation – then this might help you out too.
I keep a diary.
Anything thats going on in my life, I put into my little digital diary. Things that annoy me at work, family events and yes even Underhell development makes it into there. Having this information was super helpful to reflect on because I could see daily entries where I was pleased about progress I had made, or how I finally fixed a bug etc. It really helped me turn things around and realise that I am actually making progress. This can be something very easy to lose sight of when working on an indie project. So I encourage you, if you are working on a project like this, keep a diary! Yes I do have these blog posts, but these are not daily and do not reflect my personal life. It was definitely useful.
Aaaanyway, enough about my feelings!
Over this last month I really got down and dirty with A.I for the game. It’s not going to be anything amazing right out of the gate, but It’s gotten a lot more capable now. It can navigate through doors and react to threats (visually, audibly and through damage responses) and has had many improvements to melee & firearm weapons. It’s been a real learning process for me as ,unfortunately, I don’t know everything.
I thought that the AI was working well with doors, until I started to layer on other behaviour and realised it broke it, this caused me to rework it, lifting and shifting it over to a new system, but now its better than ever. It seems that a lot of the time, the issue is more about stacking these things together and getting them to play nice as opposed to being stuck with questions “How do I do this?” and is more of a “Where do I do this?” kind of approach.
My most recent addition has been improving the AI reacting to your presence and keeping up with you as you move around the world. It was far too easy to lose the AI before, and this was rather disappointing to play. I’ve improved the system (and still working on it) where the AI will use your last known location and randomly generate a few locations nearby where it ‘thinks’ you might have gone to. It will check these locations out before declaring that you’ve actually been ‘lost’. I think it works out pretty well, so we will have to see how it plays out.
The problem I face is finding out where the line is. Where do I leave the A.I and move onto implementing other systems that help get me to a stable build? I still don’t know the answer but I feel with more and more experimenting and development I’ll be able to answer these questions more confidently soon.
A lot of the A.I seen so far is scripted, meaning they move where I tell them to, play animations when told and speak when allowed to speak. These are fun to setup and I even setup almost all of the first Hermit meeting as a test of these systems which includes the 5 card quest and the reward for completing the quest!
The only downside is that the vents don’t work yet, so simply clicking on the vent instantly destroys it once you have the wrench – but you can use your imagination for now I think.
The actual A.I is still very much in chunks, a mix between old and new code that is kinda working. I feel that next month will still be focused on tidying this up. Every day typically brings progress, so that’s a good thing right? I’ll share more and maybe even a video once I have more progress on that.
I think that’s all for now. Happy Halloween!