MAME philosophy - was: Build pc question

[Haze]

saint's note - forked from this thread: http://forum.arcadecontrols.com/index.php/topic,164155.0.html

It's funny how triggered some people are by the merge.

I'll repeat, it was a *response* to the project starting to die, because the arcade userbase was aging, and not really supportive of the project.

RA gained most of its ground in the years where we stubbornly *refused* to expand the scope, while they were offering things that appealed more to the current generations alongside the MAME cores.

We left it about 5 years too late in that sense, the writing was already on the wall, for every single thing we did, almost the only feedback we got from what was left of the arcade community was "why didn't you do that instead" or "that sucks, it no longer runs on my 10 year old system"  In those years we could have been building a better reputation with new communities (and in many sense, we're almost there now for some of them)

Even to this day, most of the RA arcade userbase that apparently we should care that we 'lost' only chase after objectively awful older versions when it comes to the arcade emulation, because they can save money on hardware and don't actually care about things getting better.  They were never really supportive of the project or where it needed to go.  Somebody insisting on a 5-10 year old version is not helpful, those were just the users who moaned whenever we tried to make anything better before, it's impossible to consider them to be 'lost' users as we have to be able to move forward.

What we've seen on other older arcade focused MAME forums is yes, they've died out, they were dying out for years, the thing that saved MAME from the same fate was changing trajectory, or it too would have gone the same way.

Re: cataloguing stuff, as I've said before, nothing stopping the communities from doing that outside of MAME, MAME can show custom filtered lists if people create them.

[Vocalitus]

Besides there are still unemulated hardware systems that have not been saved or preserved, but the software has to some degree.  Laseractive is a good case in point.  With the plethora of Sega addons that was created for the system, are already emulated.  I am not sure if the LDV1000 is emulated in MAME, but the laserdisc software has been archived constantly now for a couple of years.  The driver has been dumped and we can only wait for someone qualified and capable who can connect the dots, the same goes for the MSX Palcom laserdisc system (emulated already) and all the non adult laserdisc released games need to be saved too.  MESS will give the correct direction for the project, but it will not require a huge shift in processing power. Storage maybe.  10TB HDD?     

[Howard_Casto]

It's funny how triggered some people are by the merge.

I'll repeat, it was a *response* to the project starting to die, because the arcade userbase was aging, and not really supportive of the project.

RA gained most of its ground in the years where we stubbornly *refused* to expand the scope, while they were offering things that appealed more to the current generations alongside the MAME cores.

We left it about 5 years too late in that sense, the writing was already on the wall, for every single thing we did, almost the only feedback we got from what was left of the arcade community was "why didn't you do that instead" or "that sucks, it no longer runs on my 10 year old system"  In those years we could have been building a better reputation with new communities (and in many sense, we're almost there now for some of them)

Even to this day, most of the RA arcade userbase that apparently we should care that we 'lost' only chase after objectively awful older versions when it comes to the arcade emulation, because they can save money on hardware and don't actually care about things getting better.  They were never really supportive of the project or where it needed to go.  Somebody insisting on a 5-10 year old version is not helpful, those were just the users who moaned whenever we tried to make anything better before, it's impossible to consider them to be 'lost' users as we have to be able to move forward.

What we've seen on other older arcade focused MAME forums is yes, they've died out, they were dying out for years, the thing that saved MAME from the same fate was changing trajectory, or it too would have gone the same way.

Re: cataloguing stuff, as I've said before, nothing stopping the communities from doing that outside of MAME, MAME can show custom filtered lists if people create them.

You don't purchase a reference book , turn to the table of contents and it says...... "There's no table of contents but communities outside of this publishing company are welcome to create one"  I've never been "triggered" by the merge... I was actually one of the few that defended it initially, but as it became apparent that mame was going to add even more stuff, the majority of which will never run due to having some sort of electro-mechanical component, and the necessary quality of life improvements weren't coming with it, I realized that it was the end of a strong, dedicated community trying to support the project.   If you save the patient by chopping off their arms and legs, blinding and deafening them did you really save the patient?

[Haze]

It's funny how triggered some people are by the merge.

I'll repeat, it was a *response* to the project starting to die, because the arcade userbase was aging, and not really supportive of the project.

RA gained most of its ground in the years where we stubbornly *refused* to expand the scope, while they were offering things that appealed more to the current generations alongside the MAME cores.

We left it about 5 years too late in that sense, the writing was already on the wall, for every single thing we did, almost the only feedback we got from what was left of the arcade community was "why didn't you do that instead" or "that sucks, it no longer runs on my 10 year old system"  In those years we could have been building a better reputation with new communities (and in many sense, we're almost there now for some of them)

Even to this day, most of the RA arcade userbase that apparently we should care that we 'lost' only chase after objectively awful older versions when it comes to the arcade emulation, because they can save money on hardware and don't actually care about things getting better.  They were never really supportive of the project or where it needed to go.  Somebody insisting on a 5-10 year old version is not helpful, those were just the users who moaned whenever we tried to make anything better before, it's impossible to consider them to be 'lost' users as we have to be able to move forward.

What we've seen on other older arcade focused MAME forums is yes, they've died out, they were dying out for years, the thing that saved MAME from the same fate was changing trajectory, or it too would have gone the same way.

Re: cataloguing stuff, as I've said before, nothing stopping the communities from doing that outside of MAME, MAME can show custom filtered lists if people create them.

You don't purchase a reference book , turn to the table of contents and it says...... "There's no table of contents but communities outside of this publishing company are welcome to create one"  I've never been "triggered" by the merge... I was actually one of the few that defended it initially, but as it became apparent that mame was going to add even more stuff, the majority of which will never run due to having some sort of electro-mechanical component, and the necessary quality of life improvements weren't coming with it, I realized that it was the end of a strong, dedicated community trying to support the project.   If you save the patient by chopping off their arms and legs, blinding and deafening them did you really save the patient?

MAME is an A-Z of everything, the level of cataloging we do is provide a brief description.  It's closer to an encyclopedia.  You have your table of contents, you just don't like it because it doesn't go into the detail you want.

We have no interest in that, we provide all the technical metadata anybody could need to write their own set of filters, without resorting to more subjective labels.  If anything the current line between 'Game' and 'System' will go away too, as even that is entirely subjective and a relic of the past (a Game and Watch currently gets called a System, while an arcade update firmware gets called a Game) so they'll likely all become 'Machines'

I really don't think it's too much of an ask to say the communities should maintain such things, and if anything the expectation that we should also do such is one of the things I find to be toxic; we're doing the part we're interested in (the emulation), demanding we do more comes across as rude.

If the community that cares about such things so much puts in the effort, great, if they don't, they don't, and I guess there can't really be that much demand.  99% of what we do is self-motivated, so what I'm saying to you is no different than if there's a system I'm interested in that I want to see emulated.

[Haze]

no reason, at least in my opinion anyway, to tell someone to go bleeding edge on a cab with an emulator either.

and we're not.

My 4ghz i7 from about 6 years ago still does a good job for the most part, it handles the latest netlist sound stuff etc. with a little room to spare.  That's very much not 'bleeding edge' it's downright old.  That's what we're recommending, tech that is over half a decade old.

Bleeding edge is the Apple M1, that is absolutely demolishing it in benchmarks (for cases where recompilers aren't needed)

We've actually become much more moderate with this over time.  Back in the day even a 2 year old machine would have been considered grossly unsuitable because the real world speed of what you could buy was doubling every year or two, yet people complain more about this now than they did back then simply due to those underpowered Pis etc.

These days even the current AMD offerings are competitive against Intel, but even when they weren't we were being attacked from all sides for refusing to recommend AMD when we were just being honest; you got about half the performance at the same clock speed, so they weren't even good value.  Now that current AMD is competitive if you do want a cheaper modern option, people seem to want to extend that recommendation to the older ones that weren't.

[jennifer]

It is no secret Jenn is no fan of Mame, However they did do what they set out to do, save the arcade games, and with Rom dumps those lost games are avaliable even today.The Devs are not there to kiss butt, and the better of them should be commended for the commitment, on pretty much a labor of love after all it is freeware...As for building a computer, not rocket science even Walmart sells them.

[Osirus23]

These days even the current AMD offerings are competitive against Intel, but even when they weren't we were being attacked from all sides for refusing to recommend AMD when we were just being honest; you got about half the performance at the same clock speed, so they weren't even good value.

Were you doing the inverse during the Netburst era?

[Howard_Casto]

It's funny how triggered some people are by the merge.

I'll repeat, it was a *response* to the project starting to die, because the arcade userbase was aging, and not really supportive of the project.

RA gained most of its ground in the years where we stubbornly *refused* to expand the scope, while they were offering things that appealed more to the current generations alongside the MAME cores.

We left it about 5 years too late in that sense, the writing was already on the wall, for every single thing we did, almost the only feedback we got from what was left of the arcade community was "why didn't you do that instead" or "that sucks, it no longer runs on my 10 year old system"  In those years we could have been building a better reputation with new communities (and in many sense, we're almost there now for some of them)

Even to this day, most of the RA arcade userbase that apparently we should care that we 'lost' only chase after objectively awful older versions when it comes to the arcade emulation, because they can save money on hardware and don't actually care about things getting better.  They were never really supportive of the project or where it needed to go.  Somebody insisting on a 5-10 year old version is not helpful, those were just the users who moaned whenever we tried to make anything better before, it's impossible to consider them to be 'lost' users as we have to be able to move forward.

What we've seen on other older arcade focused MAME forums is yes, they've died out, they were dying out for years, the thing that saved MAME from the same fate was changing trajectory, or it too would have gone the same way.

Re: cataloguing stuff, as I've said before, nothing stopping the communities from doing that outside of MAME, MAME can show custom filtered lists if people create them.

You don't purchase a reference book , turn to the table of contents and it says...... "There's no table of contents but communities outside of this publishing company are welcome to create one"  I've never been "triggered" by the merge... I was actually one of the few that defended it initially, but as it became apparent that mame was going to add even more stuff, the majority of which will never run due to having some sort of electro-mechanical component, and the necessary quality of life improvements weren't coming with it, I realized that it was the end of a strong, dedicated community trying to support the project.   If you save the patient by chopping off their arms and legs, blinding and deafening them did you really save the patient?

MAME is an A-Z of everything, the level of cataloging we do is provide a brief description.  It's closer to an encyclopedia.  You have your table of contents, you just don't like it because it doesn't go into the detail you want.

We have no interest in that, we provide all the technical metadata anybody could need to write their own set of filters, without resorting to more subjective labels.  If anything the current line between 'Game' and 'System' will go away too, as even that is entirely subjective and a relic of the past (a Game and Watch currently gets called a System, while an arcade update firmware gets called a Game) so they'll likely all become 'Machines'

I really don't think it's too much of an ask to say the communities should maintain such things, and if anything the expectation that we should also do such is one of the things I find to be toxic; we're doing the part we're interested in (the emulation), demanding we do more comes across as rude.

If the community that cares about such things so much puts in the effort, great, if they don't, they don't, and I guess there can't really be that much demand.  99% of what we do is self-motivated, so what I'm saying to you is no different than if there's a system I'm interested in that I want to see emulated.

Yeah but see you can't have it both ways.  You'll spout mame's mission statement to defend one thing and then say "we just do what we want" in defense of another.  An arcade game, or a console is more than just the code... it's everything from the promotional materials to the controls and everything in-between.   Do I expect the current devs to add all of that?  No.  Do I expect them to have a system in place where volunteers can go in and officially archive everything possible?  Hell yes.  There are projects out there already, reach out and make it official.  On the documentation side though, yes it should all be in mame or available as an official, constantly updated document from the mame website.  Let's just say that some average user with not a lot of knowledge of emulators runs across "The X-Files" in the game list.  Unless they really understand the xml file and can gleem the clues from it they don't know if that's a console game or an arcade or what.  We know it's a DMD Pinball, so wouldn't adding a single category line that says "Pinball / DMD" be extremely helpful to end users and not take all that much effort or space to add in?  If you don't want to do it that's fine... you guys show me an example of how you want it done and I'll get together with the catver people or find some volunteers and do it.  I never said, "hey you do this" but rather "hey you should allow this"  There is a huge difference. 

[formula409]

Categories are stupid because they're subjective. What is Strider? Is it a "platformer?" You jump around on platforms. But it's not really the same as what people think of when they think of a platformer. What's Earth Worm Jim? It's sort of a platformer, but it's also sort of a run n' gun. Speaking of which, what is a run n' gun exactly? Ask three people and get three different answers. MAME is and should be focusing on the stuff that's empirically verifiable.

I fully support the MAME/MESS merger. Look, let's get real here. It's over for arcade games. Every arcade game made after 2000 pretty much blows. Metal Slug 3 was basically the swan song. After that you have a bunch of boring Cave shooters that are all basically the same, and lame cheap Chinese/Korean cash ins that are worse than 20 year old games they're ripping off.

The only real place for MAME to go at this point is improving accuracy. There is very little new ground to claim that's of any interest to end users. The only real glaring omissions in MAME at this point are good Model 2 and Model 3 emulation. Other than that, the real meat is somehow figuring out how to get more accurate emulation like doing CPS wait states and so on if that's even possible.

My only real disappointment with MAME is that the Jaguar and Saturn drivers never went anywhere. It really irritates me that there isn't a single way to play Tempest 2000 today via emulator that doesn't suck. It's such a glaring omission. It's one of THE games. And this isn't just about MAME. There's literally nothing. Every Jaguar emulator sucks. The best one is probably that single game t2k emulator that Dio made like a decade ago that's closed source and therefore worthless.

[Howard_Casto]

No categories are kind of universal if they are done with care and that's not what I'm talking about anyway.  I'm talking about Hardware categories, like Arcade, E.M., Pinball, console, portable, ect....   It STILL doesn't matter even if you argue that they are subjective.   A categorization that some individual doesn't agree with is still better than none at all.  It's like these idiots that argue universal health care is bad because it may be slightly worse than private health care.  Spoken like a person that's never been without health care.  I assure you, sub-par is always better than none. 

[dmckean]

and we're not.

My 4ghz i7 from about 6 years ago still does a good job for the most part, it handles the latest netlist sound stuff etc. with a little room to spare.  That's very much not 'bleeding edge' it's downright old.  That's what we're recommending, tech that is over half a decade old.

Bleeding edge is the Apple M1, that is absolutely demolishing it in benchmarks (for cases where recompilers aren't needed)

We've actually become much more moderate with this over time.  Back in the day even a 2 year old machine would have been considered grossly unsuitable because the real world speed of what you could buy was doubling every year or two, yet people complain more about this now than they did back then simply due to those underpowered Pis etc.

These days even the current AMD offerings are competitive against Intel, but even when they weren't we were being attacked from all sides for refusing to recommend AMD when we were just being honest; you got about half the performance at the same clock speed, so they weren't even good value.  Now that current AMD is competitive if you do want a cheaper modern option, people seem to want to extend that recommendation to the older ones that weren't.

The mainstream crowd that likes to tinker pretty much exclusively uses PIs, Odroids and other similar low cost, low-powered boards and you have to go back 14 years to find versions of MAME that run well.

On this forum, we often want to use older PC hardware because it's easier to get 15Khz out of it to run our CRTs.

AMD mini PCs that use the 4400G or 4650G APUs would run modern MAME really well, but LCDs sort of suck for classic arcade and console games and medium sized OLED panels are practically non-existent for under $4k. I do hope the 3:2 aspect ratio panels catch on as it would be better for us.

I know that none of this is really MAME's fault but it should be easy to see why it's dwindling in popularity.

[bobbyb13]

Maybe MAME isn't dwindling in popularity so much as it has a lot of competition these days-
AtGames, A1Up, MiSTer...
And would any of these other entities even exist if it hadn't been for MAME to begin with?
I doubt it.
Those who cared would all be fighting over the few extant bits of real hardware.

The "growth" in the retrogaming arena is in people who never played any of these games in the wild.
Nostalgia for me, something new and quaint for many (most?) others.

Can't expect much from the masses who don't understand what any of this stuff really is to begin with- and for various reasons just want to buy it rather than build it.

For most "good enough" is the only mantra- fidelity be damned.  Especially since the real thing never existed in their lives.

Every discipline has its purists.  Those who learn and care will seek the best and so I think MAME will continue as long as the crew who actually understand coding this stuff continue to care.
I only understand a fraction of how it all works, but what has already been accomplished is pretty awe inspiring from my seat.

I have only gratitude for the crew of people who (decades ago!) took it upon themselves to preserve this stuff, and also for those brainy enough to add things to make it even better.

Let them do as they wish.

And if they command that I need a certain piece of hardware to pull it off using this stuff, then so be it.

I know that it is somewhat of a moving target, and that there are minimum requirements listed is the mamedev docs already, but couldn't there be a spec sheet for appropriate hardware 'per build' of MAME version?

[ivwshane]

I didn't read the original thread but I feel like giving my two cents as an end user who has been into emulation since 2000 with my first experience with emulators being with neorage, may add some perspective. From neorage I moved to snes emulators and eventually to mame (specifically mameUI) as my emulator of choice for about ten years (I took a break from mame before the mess merger). I explored emulation on non window machines, specifically the raspberry pi's and was amazed at the many builds that were easy to use and that included a nice UI.
I briefly went back to mame but was overwhelmed at the difficulty of setup. I thought the inclusion of mess was going to be a godsend as having to interact with multiple separate emulators was a hassle. Unfortunately I haven't touched mame in a while because of the difficulty of setting it up.

I always thought mames goal of accurate emulation was a worthwhile one and I didn't mind needing better and better hardware to keep up. However what eventually turned me off was also what I originally like about mame, the additional information it provided about the games. Now I know what I'm talking about is actually outside of what mame provides but if we are discussing why mames popularity is weakening its important to understand what made it popular in the first place.

For someone first getting into emulation the goal is to play your old favorite games but as you grow out of that phase you start to appreciate things other than just being able to play games for free. For me, clicking on a random game and being able to read up on the history of the game and seeing screenshots or even video snaps was really informative and introduced me to games I wouldn't have bothered to play and therefore appreciate. Having game info is nice but having categories is even better, the more categories the better. Why? Because it allows for purposeful exploration and at the end of the day, emulation is pointless if games/systems aren't played. It would be like having the library of congress with no card catalogue or even being alphabetically organized. Sure books would be preserved but for what purpose?

Anyway, the point is that the extra information may not seem relevant to emulation but its very much relevant to any type of repository and without that extra information you just have a data dump of which new users won't want to be bothered to use because there are easier alternatives out there that will allow them to play their favorite games quicker and it will alienate veteran users as it doesn't provide anything extra to feed their curiosity or thirst for knowledge.

To use another analogy, think of a museum where the artifacts have no labels. Sure some of the stuff is cool to look at but without context most people will be in and out quickly with no one really staying around to learn more.

If I was a game developer or someone who worked on a game on a system that is no longer active or doesn't exist, having my game emulated is great but I would get even more satisfaction out of people actually playing it and people won't/can't play a game they've never heard of.

That's my two cents. Take it or leave it but either way I'll still be grateful for anyone willing to put work into the scene, especially knowing that other than personal satisfaction they probably get nothing out of it.

[Howard_Casto]

I couldn't have said it better myself.  Mame shouldn't just be about the inner-workings of a board, it should be about the experience.  We preserve things so they can be enjoyed by future generations, which is why museums have exhibits instead of being nothing but a massive warehouse closed off to the public.  I'm not being pushy, I'm not trying to turn mame into something that it isn't..... I just want butts in the seats..... these games need to be played. 

[bobbyb13]

I couldn't have said it better myself.  Mame shouldn't just be about the inner-workings of a board, it should be about the experience.  We preserve things so they can be enjoyed by future generations, which is why museums have exhibits instead of being nothing but a massive warehouse closed off to the public.  I'm not being pushy, I'm not trying to turn mame into something that it isn't..... I just want butts in the seats..... these games need to be played.

And therein you both bring up points that hadn't occurred to me yet.

When I stumbled onto emulation it took me months to figure out how to even put together a (somewhat) functional system even with crappy controls- and that was just to prove to myself that I was capable of pulling it off.
I can't code for squat, but I know my way around computer hardware pretty darn well, and it still took a lot of trial and error.

And the don't ask don't tell of roms aside, I figured the agony of sorting all this stuff out was simply dues to be paid to join the club- and that this was the intent of those who had built that which made it all work.
Put in some sweat, or find another hobby.

Not that there was any malice in it of course, and the community here is overwhelmingly helpful in contrast to some of what I first encountered.

To Howard's point of getting butts in seats, I can scroll though Badmouth's "All Killer..." compilation and pick a pile of great games to play- because I know what the hell they are already from pumping half my paper route money into the original machines.

For those under the age of 35 who probably never had the chance to do that, how can anyone expect them to know what they are looking at (especially without categories, gameplay explanations, cabinet pics, snaps, etc.)

I had a lot of fancy looking, flashy, noisy machines to lure me into playing these games-
How do you recreate that when (sadly) most of those sweet cabinets have become worm food?

[Haze]

Mame shouldn't just be about the inner-workings of a board, it should be about the experience.

Except that is what MAME is.  That's what interests us, that's what we work on.

The rest of the experience is very much defined by the community, what gets built around it.

We do the parts that fit our skills / interests / needs.  You do the parts that fit your skills / interests / needs.

Together, they form the overall experience.

Maybe one day, if somebody shows up, committed to maintaining such filters on a long term basis, they'll be brought onboard, asked to update them before each release.  If not, it remains an external task.  If they start, and drop out again, it becomes an external task.

Personally I'd like to see the cheats etc. imported into the main repository so we can add to them as we develop, as needs arise, but in that case it seems the community does want to maintain them externally rather than submit them.

The MAME experience and even development is much more decentralized these days, even if commits go to public central repository.