> Phil is also pretty well-known for doing hacky proofs-of-concept that are in no way > ready to be accepted into MAME. Remember how long it took for the Fairlight CMI and > Konami 3DO M2 drivers to be picked up by someone who wanted to bring them up to > scratch, and we still haven’t seen the Casio Loopy stuff, either.
The Casio Loopy stuff is generally down to Phil wanting to get the driver 100% correct before submitting it, and having a tonne of things to juggle in real life. I personally disagree with the idea of waiting, but hey, it's his choice to make.
As the person who updated the CMI IIx and M2 drivers to modern APIs, I'd rather you didn't use them as reasons for throwing someone under the proverbial bus, because your analysis is just incorrect. The current CMI IIx and M2 drivers are pretty much identical to how they were at when Phil put them to the side, it's just that after so many years of them not existing upstream, they needed roughly a decade's worth of updating to the modern MAME core APIs. What we have now isn't based on some "hacky proof-of-concept", they're literally the drivers that Phil himself wrote.