I'd love some help from anyone who's interested in working on improving the project with me, so I'm submitting my GitHub project page link here for anyone interested.
Back in 2004, computers were much slower and so performance was a much larger concern with emulators. I was worried that many fan translations, ROM hacks, and homebrew games were depending on bugs in the other emulators to function, and feared this could lead to a loss of those games in the future, should emulators one day improve. And so I started bsnes with the goal to writing an emulator that was as accurate as possible. Needless to say, I underestimated the complexity by a few orders of magnitude.
In 2010, I led a fundraiser to decap coprocessors inside of game cartridges to extract their firmware. This allowed bsnes to be the first emulator to reach 100% game compatibility [1]
In arguably the most surreal moment of my life, it turned out that one of these coprocessors (NEC uPD7720) was used in professor Stephen Hawking's voice machine. Having written the only emulator for this processor, his assistants reached out to me and I was able to donate a portion of code to their project, which was ultimately successful in replacing the original failing voice machine with a much more portable version. [2]
In 2012, to improve emulation of cartridge PCBs, I started a goal of purchasing (and reselling after to break even) every SNES game cartridge. I started with the USA collection (725 games), finished it, moved onto the Japanese collection (1,450 games), finished that, and now I've been working on the European collection (540 games.) I've spent around $30,000 and 1200+ hours working on this goal.
Something I don't like to talk about, but it happened: once a person graciously lended me 100 PAL games worth around $5,000 - $10,000. It was lost in the mail. After a long, stressful five-month wait, I reached out to the media and within a few weeks, the package was recovered, and the games were analyzed and returned to their owner intact. [3]
A few years back, bsnes grew to emulate dozens of systems, and so I made the decision to move the multi-system emulation platform to a new project called higan. [4] This project is likely far too ambitious, but it keeps me occupied.
With bsnes once again a standalone SNES emulator, over the past year I've set about trying to create an easier to use, friendly emulator that is more performance oriented. One of the key insights was developing a method of parallelizing the video rendering in such a way that does not add any input latency.
This parallel renderer provided for another first: the ability to analyze the entire screen before rendering the first pixel. A programmer by the name of DerKoun came along and realized that by interpolating at higher precision the affine transformation coordinates of the screen, that 256x240-resolution SNES games could be upscaled to basically any resolution, even 3840x2160. The results were something no one else believed possible, and it received a good deal of press as a result.
DerKoun has continued with this project, adding new features to SNES emulation such as widescreen support, true color gradients (to remove banding from the system's 15-bit color palette), etc.
Recently, in an effort to try and preserve more history, I've gone back to emulate the older emulators themselves, so that the oldest fan-made SNES software can still be played.
It's never been bsnes' goal to be #1, and over the years I've contributed code, notes, and explanations to basically every known SNES emulator. Both free software and commercial, both open source and closed. My underlying goal has always been to improve emulation quality for everyone.
Thanks for reading!
[1] https://www.tested.com/tech/gaming/44376-16_bit-time-capsule...
[2] https://www.sfchronicle.com/bayarea/article/The-Silicon-Vall...
[3] http://www.nintendolife.com/news/2017/02/usps_recovers_lost_...
[4] https://github.com/byuu/higan
[5] https://arstechnica.com/gaming/2019/04/hd-emulation-mod-make...