![]() Raytracing solutions have silly hardware requirements, are still cumbersome to use and not as detailed as you might think. AO just darkens corners that shouldn't have been lit in the in the first place, and real time solutions are glitchy screen space things that don't work well with a forward renderer. Outdoors we had to use another solution entirely, with its own limitations.Īny kind of baking is a content pipeline nightmare and of course can't handle dynamic lights or objects anyway. It was also limiting the size and layout of our environments, turning off beyond a certain distance or depth. Without direct light everything looked very flat and disconnected, and light would still leak even through doors. We had something crude in place so that we could handle light and darkness without light leaking through walls, but that was it. The other big topic was global illumination. Here's an example of better materials and shading, as a direct comparison to the previous version: Our new material system is pretty much fool proof, and can handle complex materials much more efficiently. We need artists to make correct materials easily and intuitively, and without consuming a bunch of resources. Our material system was cumbersome and complex to use, overhauling it is something we've considered many times. But there quite a few things we wish we had done better, wanted to improve or raised questions for the future. The main goals here were to reduce driver CPU bottlenecks, also a growing problem as GPU performance races ahead of CPUs, handle hundreds of detailed unique procedural objects efficiently, and get away from visual artefacts that were ruining our shiny new art. And you know us, if we're going to do something, we're going to do it properly. To solve all these issues, we had to build a new rendering engine. Much as we wanted to avoid it, we needed a fundamental change. We also keep pushing what we do with content and graphics, and these issues were preventing us from moving forward. Our next update was supposed to just be some improvements and additions to arena gameplay systems and procedural weapons, making a new rendering engine is not something we saw ourselves doing at this stage in development, or any really, but we've been running into compatibility issues, driver updates have been steadily degrading our unconventional renderer's performance and causing ugly artefacts. ![]() They can be found in %APPDATA%\Exanima (put this in the run or file explorer address box). You may also want to back your saves just in case something goes wrong. To see the game the way it's supposed to you should start a new game. Story saves technically still work, but a lot has changed or been replaced and existing saves will have incorrect lighting, old props, items may look wrong and various other weird things. If anybody has a reason why this wouldn't be the best option for both the playerbase and Bare Mettle, I'd like to hear it.We have tried to maintain compatibility with old saves, especially for arena which will receive an actual wipe soon, however the hub has changed and before you update if you have any items lying around in your hub you should put them in your arsenal or they will be lost. If that were to be the next update instead of a new dungeon level, I would actually feel like waiting a year would be worth it. Unless Madoc is literally hex editing the whole thing, the dev tools already exist and just need a bit of proofing before publishing. ![]() If funds aren't an issue, then at least they would be an extra few thousand hours towards sui generis in the same time that they would instead be a couple hundred. After one year of community modding, they could easily skim what's available, compile the best of it, and have an entirely new product to sell if funds are any issue for them. Imagine the countless hours of unpaid development time that could be offered up by a dedicated community, that bare mettle is currently missing out on by updating with one dungeon level per year. ![]() Unless multiplayer or some other fundamental-changing implementation is going to happen, mod support should be the highest priority.
0 Comments
Leave a Reply. |