Zetheros
Developer
Hey Exanimates,
Last week we talked about implementing the new pathfinding, and how we had all these difficult extra requirements, but we thought we had them all figured out. Well, perhaps we focused too much on these extras and overlooked just how "complex" the basic problem is.
As we got deeper into the implementation details we realised that building a multiple layer 3D map from "polygon soup" is indeed pretty complicated to begin with. Some fairly simple approaches might work 99% of the time, but there are problem cases with multiple overlapping disconnected paths, inaccessible regions etc. that massively increase the complexity of the problem. We explored a bunch of approaches in detail and kept running into some cases where they would fail. We've been seriously worried that we just wouldn't be able to find a solution that is fast, run-time and doesn't burden level designers with extra work and technical challenges. It's a lot, and why games usually rely on a combination of hand crafted and precomputed solutions.
We did persist though and finally arrived at a solution that we think still meets all our requirements too. It's not simple, it's going to take some work, but assuming we didn't forget anything we should be able to get through it fairly quickly.
Meanwhile we're making good progress on the new area, adding more new and unique props, fleshing out the details. The early levels are also getting some updates with new elements, and we're finally shortening the maze section of the first level. There's a bunch of new items and many existing items are getting a major overhaul and being converted to the procedural system. Hopefully we can get a good amount of the updates to existing content in some earlier test builds to make them a bit more enticing.
Have a great week!
-the BM team
Last week we talked about implementing the new pathfinding, and how we had all these difficult extra requirements, but we thought we had them all figured out. Well, perhaps we focused too much on these extras and overlooked just how "complex" the basic problem is.
As we got deeper into the implementation details we realised that building a multiple layer 3D map from "polygon soup" is indeed pretty complicated to begin with. Some fairly simple approaches might work 99% of the time, but there are problem cases with multiple overlapping disconnected paths, inaccessible regions etc. that massively increase the complexity of the problem. We explored a bunch of approaches in detail and kept running into some cases where they would fail. We've been seriously worried that we just wouldn't be able to find a solution that is fast, run-time and doesn't burden level designers with extra work and technical challenges. It's a lot, and why games usually rely on a combination of hand crafted and precomputed solutions.
We did persist though and finally arrived at a solution that we think still meets all our requirements too. It's not simple, it's going to take some work, but assuming we didn't forget anything we should be able to get through it fairly quickly.
Meanwhile we're making good progress on the new area, adding more new and unique props, fleshing out the details. The early levels are also getting some updates with new elements, and we're finally shortening the maze section of the first level. There's a bunch of new items and many existing items are getting a major overhaul and being converted to the procedural system. Hopefully we can get a good amount of the updates to existing content in some earlier test builds to make them a bit more enticing.
Have a great week!
-the BM team
