I think this article should be the first to start the thread with
Some people are probably wondering why the engine is reusing the id Tech 4 (BFG) codebase (which uses polar opposite architecture compared to Gamebryo (Gamebryo is oriented on standalone games while id Tech engines are mod-based)) rather than reusing only OpenMW codebase alone or use some of the modern engines (like Unity, Unreal or CryEngine/Lumberyard)
Well, initially the engine was intended to use only OpenMW codebase and there was some work made in that direction (2017). Unfortunately, it wasn't successful and after some time it was dropped. After that, there was some time to think about a way to keep the development of the project fun and reduce the probability of burnout (while working alone) (2017-2018)
The decision was to resurrect the Doom 3 BFG Edition codebase and use it as a foundation for the engine. There are a couple of key reasons for that:
- Despite the features it has, id Tech 4 is a "dark horse" in id Tech engines family. It's not that popular compared to previous generations so at least there will be a new life for the code as part of the SugarBombEngine
- BFG Edition has some Id Tech 5 improvements (for me, mostly, the new job-based system which allows multiple threads to be load balanced) which are very useful in modern time
- Id Tech 4 (BFG) is a really solid foundation for a modern game engine (if you don't have that much time to write an engine from scratch, the codebase is a really good starting point because it consists of working battle- and time-tested code which has a lot of core features other more complex systems are based on)
- Doom 3 game code has some similarities with Fallout 3 mechanics, so it feels natural to use it as a foundation for F3 recreation
- Also, Doom 3 game code already is a working solution for an FPS/TPS, so we only need to combine it with RPG mechanics from OpenMW (of course, it will require to put more effort to properly recreate all F3 mechanics, but still, nonetheless, with Doom 3 + OpenMW game code we'll get major part of F3 mechanics reimplemented)
- OpenMW and D3BFG licenses are compatible which allows to freely combine them together (while Unity/Unreal/CryEngine aren't GPLed and has EULAs which forbids to do some things or not fully open-source despite them having more features and maintainers)
After that, before the E3 2018 event (May 2018), an unconfirmed Bethesda's plan was leaked into the Web. The Fallout 3 Remaster was mentioned in the list, so, in case that will be truth info, there will be not much sense in manually reimplementing the game. In addition, the fan-made recreation on F4 engine (Capital Wasteland) was also frozen a couple of months earlier that year. Bethesda presented a new Fallout instead and remaster was made for Fallout Shelter (which was just 3 years old). Time passed by and F76 came out. It was definitely not what people expected and there was no hope for an official F3 remaster/fan-made reimplementation (Nov 2018)
So this is where the current attempt/generation comes in. OpenMW took 10 years and a lot of contributors to achieve what they got. And F3 is even more complex game, so potentially it could take the same amount of time (even if using OpenMW code) or more and probably even more amount of contributors. Of course, it will be hard to recreate such a project (especially alone), but, at least some steps could be made in that direction. Even if the project won't be a success and end up abandoned, some other people will be able to use the work done to go further
