This is the first of many to come weekly Dev-log posts for my work on 310 Games current project, The Krilling. This post is from a few weeks ago and is being posted late due to my recent recovery from being ill. This week I was given the following tasks:

· Create AI locations for the concert area.

· Create AI locations for the check-in desk

· Create AI locations for the bedrooms

· Randomize the visual models of the AI

· Test game optimization

For the AI locations I simply created new locations in the map called Places of Interest. The AI goes to these locations to complete given tasks, for more info on the AI system see my other post on the creation of the AI system. For each of these, I then create a new task outline to add to the POI.

For the bedroom locations, the original plan was to have the AI spawn from the bedrooms and leave via the bedrooms, but I decided it was simpler and achieved the same effect to just have the AI go out of vision into the bedrooms and return later.

To randomize the victim models I created a new script on our world state object that handles the victim models. It contains serialized lists to hold the prefabs of the models. This is to make it easier for the artists to create new models and then just add them to the list and they will be used.

This script then has 2 static methods for getting a random model prefab, one for the guests and one for the workers.

The victims then at the start of the game call the relevant function to get and instantiate a random model for themself.

For now I just threw in some random incomplete models, but in the future these lists will be populated by our artists with the correct models.

The bulk of my work this week was spent looking into game optimization. Our team is not aware of any optimization issues at the moment, but we were looking into it to make sure that all kinds of devices will be able to run our game. I ran the Unity profiler and nothing seemed out of the ordinary, our usage was mostly on rendering which makes sense, as well as the editor loop which is usage caused by the Unity editor and won’t appear in the actual game. [Profiler Image] The original plan was to then run the game on a virtual machine where we could limit the CPU and memory available to see how it would run on slower machines. This plan did not work out because I discovered that the virtual machine that I use, and from research most other virtual machines, do not work well with graphics cards. This results in, at least in my case, a maximum video memory of 128 MB. This is an incredibly low amount and not enough to run even the least demanding 3D games. This meant that a virtual machine was not a valid option since its max settings were not enough to emulate the conditions our players would be in. I spent a long time testing and researching virtual machines, and figuring out how to build for Linux to run it on the virtual machine, but in the end, nothing was able to work. Our only option then going forward is to run the game manually on a low-end computer to test the performance.

That was everything for this week. Hopefully, next week will be more promising for the performance testing. Either way, I will be posting an update once again with either success, failure, or more likely, a combination of both.