Hey there space fans!
I’ve been spending the past few weeks tending to our other obligations, but now we’re back in full swing!
Since we last spoke, I did a lot of work on overall usability. I re-wrote the camera rig from scratch for something much more smooth, responsive and more functionally appropriate given how we go about navigating and building the station. I also had a chance to solve a few performance issues here and there and solve a metric butt-ton of bugs.
I spent this past week revisiting our AI system. Now that we’ve had a chance to put it through its paces, there were some obvious weaknesses. Since few of our other systems currently depend on the AI, it’s as good a time as any to revamp that.
It took me pretty much the whole week to research appropriate solutions (and fill my whiteboard with the scribblings of a madman), as well as sanity check my findings with AI experts I know. I’m pretty confident I’ve got a solid, flexible solution that’ll hold up in a few key areas.
Specifically, I needed something that could deal with needs-based decision making (eating, sleeping, socializing), while also being able to react to environmental elements on the fly (talk to Bob as you walk past, respond to a reactor fire).
If there’s any interest, I’ll write a more comprehensive post explaining how the whole system comes together. It re-uses some parts I’ve spoken of before (name utilities), but also has a bunch new bells and whistles.
It’s been a while! Since last time we last caught up I’ve been very busy. Mostly in nailing down loose nails and tightening screws to get what we had to a more solid state.
Shaping up the state of the game allows us to better evaluate where we’re at and what is left to do. During most of game development you often break the game in many ways at any time in order to add more features or change things. This of course leads to a difficulty in assessing the current state so in order to do that you need to at some point put everything to a fully(ish) implemented state before adding more atop of what already exists.
Oh! I also got the opportunity to dig into some paperwork. Paperwork is a source for everlasting enjoyment, oh boy it’s fun!
Hey there! I’ve been working on trait re-balance. Basically, the traits that are in the system right now have been more-or-less placeholder.
First I worked out the optimal number of traits and the rulesets about their construction to have a perfectly balanced personality system. I then reduced the problem to a math/combinatorics problem where there is one correct solution out of about 16 million possible configurations.
If I can solve the math problem, I’ve balanced the system. I’ve been working on this for about two weeks — finding the correct configuration which satisfies all my balance rules is not just hard, it’s NP Hard (actually it’s NP complete but that joke didn’t work as well).