You should really learn a lesson in regards to the nag screens. For over a decade users said they were dumb and don't serve any purpose. Finally you remove them and act like all those cries of foul somehow didn't count but now that the mame team has come to that conclusion it somehow matters more. Long story short the masses are usually right, just listen to them and save yourself years of complaints.
not really, they served a purpose, as I said they were removed because we now feel the project is on better legal footing; with the non-arcade stuff there are thousands more legal use cases we can easily point at. This is not 'admitting we were wrong' or that the 'masses were right' it's moving with the times as the project has evolved. The very same reasons are the reasons I'm now happy to relicense any code I've contributed as plain old BSD without the non-commercial restriction, there are more legitimate reasons why somebody would want to sell the software rather than it looking like we were in bed with those selling illegal MAME cabinets.
also, on the same subject, despite giving users some rope and taking the copyright nag away idiots here are STILL removing the warning screens telling you about imperfect emulation / non-working drivers. Those screens are absolutely essential to the user experience, they're NOT nags, they exist so that people know when they're running a broken driver, those screens help us (avoid false bug reports), and help improve the user experience by letting them know to expect problems, so please excuse us if we don't especially care for your 'needs' either after years of complete disregard for ours.
yes, we're passing on the work, it makes more sense that way, people can categorize things however they want then, let somebody create an .ini file, import it into your frontend.
What we've basically seen over the years is that what most 'game players' want is simply a 'good game worth playing!' flag, so no, we're not going to listen to them, because that's the opposite of what the project is about; if it wasn't then we would have called it quits after emulating Pacman, NeoGeo, CPS, and Mortal Kombat.
I'm telling you what is happening, we don't categorize games into genres, we don't really need to categorize machine types either, some stuff is more difficult to use than others, that was always the case (some of the gamblers have really contrived setup sequences) As I said, you're likely to see different ways of operating things like CPS3 in the future, where you do select the cartridges and CDs from the software lists etc. maybe the existing shortcuts to run the games will remain, maybe they won't, if they don't, how does it differ from a non-arcade system? we're making MAME into a more powerful piece of software, as a result there will be more of a learning curve, we're not going to be restrained by the third party software or make it easy for such software to ignore the progress.
With things like the lua scripting capabilities and other such enhancements like the web interface we've actually given more power than ever to people driving MAME from external programs with a view to maybe making the modern MAME experience a bit easier by allowing them to provide things like 'auto boot' scripts to the program which can be used with some of the non-arcade systems etc.
FWIW I don't think this is 'hurting' MAME, allowing ourselves to go too far down the 'games machine' path (just select a game and play) is exactly what projects like RetroArch seem to be trying to do, tie everything down to simple but incredibly limited concepts that can be operated using just a joypad on a console. The strength of MAME is moving in the opposite direction to that, which is why we're never going to consider integrating a RA target officially, it just constrains what we can do; if somebody is doing a frontend for MAME they have to understand that MAME is an ambitious project and any frontend will require maintenance. Simply looking for ways to try and filter out that need by only showing simple stuff isn't really the correct solution.
Since making these changes we've seen more contributors getting involved than ever, it hasn't hurt the project, it's revitalized it.
I get that you're not too keen on where things are going, but it won't be long before there really are as many different 'types' of things emulated as there are genres of game, if it has a CPU it is a valid target, and even beyond that some things without CPUs are valid targets too if a full schematic can be drawn up and a netlist created. Trying to categorize things is not a job for the dev team, providing the emulator for them is. Trying to categorize things in the past actually caused immense damage to the project and preservation efforts when we were actively excluding things we felt didn't fit in a certain category and people started scrapping them rather than dumping them.