the state of mame

[danny_galaga]

Hands up anyone who thinks of Mahjong games when they think classics?

[Haze]

Hands up anyone who thinks of Mahjong games when they think classics?

Based on the posts I've had on my blog, and the requests we've had for some of them I think if you asked this question on a Japanese forum you'd be flooded with replies.

The early Dynax and Nichibutsu ones especially are considered 'classics' in Japan, and were just as popular, if not more popular than the likes of Street Fighter 2 was over here.  There is a reason they made so many of them and near enough every major manufacturer has a portfolio of them.  There have been just as many donations made *specifically* to buy Mahjong boards as anything else over the years.

On similar lines, if you asked on a Japanese forum 'Who considers Marble Madness a classic' you'd probably get hardly any replies, because it's an American developed game and they really didn't care for those as much.  (A trend which continues to this day, just look how difficult Microsoft found it to break into Japan)

A fairly good example as to what Japan considers 'classic' is when they pay homage to the titles in future games, and how often they copy / imitate them.  If you fire up Jaleco's "Game Tengoku - The Game Paradise" (which is a shooter set in an arcade) and set it to Japan mode you'll see that the initial level includes a whole bunch of Mahjong games running in the arcade cabinets in the background.  You'll also find hundreds of other games using Mahjong tiles in some form, and even some characters made out of them.

It amazes me that people still try to use this argument to bash MAME, the project, and what gets emulated.  It's a global project, representing a global industry, not something tailored to your exact tastes.

[Vigo]

Hands up anyone who thinks of Mahjong games when they think classics?

Based on the posts I've had on my blog, and the requests we've had for some of them I think if you asked this question on a Japanese forum you'd be flooded with replies.

The early Dynax and Nichibutsu ones especially are considered 'classics' in Japan, and were just as popular, if not more popular than the likes of Street Fighter 2 was over here.  There is a reason they made so many of them and near enough every major manufacturer has a portfolio of them.  There have been just as many donations made *specifically* to buy Mahjong boards as anything else over the years.

On similar lines, if you asked on a Japanese forum 'Who considers Marble Madness a classic' you'd probably get hardly any replies, because it's an American developed game and they really didn't care for those as much.  (A trend which continues to this day, just look how difficult Microsoft found it to break into Japan)

It amazes me that people still try to use this argument to bash MAME, the project, and what gets emulated.  It's a global project, representing a global industry, not something tailored to your exact tastes.

Don't take this in a rude way, just a question. I honestly don't know how you guys are structured, but if it is the primarily Asia fanbase that considers these endless Mahjong titles of great value, is the development of these titles, the man hours, worked on by a designated Asia MAMEdev team? To me, I would think that a donation to buy Mahjong boards does not relegate that N. America and European support must be doing any further development to the endless sea of Mahjong titles out there.

I also wouldn't think that comparing your everyday Mahjong title with Marble Madness is accurate. I think it would compare much better to video slot machines. There are so many video slot machine games in N. America, and tons of people play them, but there really are not any classics that had the crowds watching on. Mahjong games were more or less your typical gambling games without the payouts.

[Haze]

Hands up anyone who thinks of Mahjong games when they think classics?

Based on the posts I've had on my blog, and the requests we've had for some of them I think if you asked this question on a Japanese forum you'd be flooded with replies.

The early Dynax and Nichibutsu ones especially are considered 'classics' in Japan, and were just as popular, if not more popular than the likes of Street Fighter 2 was over here.  There is a reason they made so many of them and near enough every major manufacturer has a portfolio of them.  There have been just as many donations made *specifically* to buy Mahjong boards as anything else over the years.

On similar lines, if you asked on a Japanese forum 'Who considers Marble Madness a classic' you'd probably get hardly any replies, because it's an American developed game and they really didn't care for those as much.  (A trend which continues to this day, just look how difficult Microsoft found it to break into Japan)

It amazes me that people still try to use this argument to bash MAME, the project, and what gets emulated.  It's a global project, representing a global industry, not something tailored to your exact tastes.

Don't take this in a rude way, just a question. I honestly don't know how you guys are structured, but if it is the primarily Asia fanbase that considers these endless Mahjong titles of great value, is the development of these titles, the man hours, worked on by a designated Asia MAMEdev team? To me, I would think that a donation to buy Mahjong boards does not relegate that N. America and European support must be doing any further development to the endless sea of Mahjong titles out there.

I also wouldn't think that comparing your everyday Mahjong title with Marble Madness is accurate. I think it would compare much better to video slot machines. There are so many video slot machine games in N. America, and tons of people play them, but there really are not any classics that had the crowds watching on. Mahjong games were more or less your typical gambling games without the payouts.

Asia / Japan don't really seem to believe in the open source culture as much unfortunately, so contributions in terms of actual code from that part of the world are more limited.  It's clearly a cultural thing, and when you do get Asian / Japanese emulators their approach to emulation is often very different eg. if you look at the NeoGeo CD emulators they've *rewritten* most of the bios code, to avoid emulating it.. In worse cases they will rewrite parts of the game for extra speed, rather than emulating it.. They will integrate the roms in the actual emulator, and encrypt the whole package to stop people reverse engineering it then package the whole thing with a ton of ad-ware so that they can make money of of people using their emulators.  Don't get me wrong, there have been contributions from Asia / Japan which have been valid but the cultural differences, and coding standards / practices make them fewer.  The *primary* development of MAME has always been European, which is ironic when you consider that only a tiny tiny handful of the supported games originate in Europe.  If European devs only worked on European produced games there wouldn't be much MAME at all ;-)

From a 'writing the emulation code' perspective it makes very little difference if you're emulating a Mahjong game or not.  The challenges are still the same, the hardware can still be interesting, and they can help fix bugs / improve things elsewhere.  The Mahjong games on Psikyo's SH2 based hardware were of great help in fixing the SH2 timers / interrupts which in turn also improved the music in several other games relying on that logic.  Improving hardware emulation is always beneficial to the project, and anything that helps that, be it emulation of a Mahjong game, or a console system in MESS is a step forward.  The best devs really aren't fussy over the nature of the games they emulate.  As I've said before, the role of Mamedev is to emulate, not discriminate, and there is a level of responsibility that comes with that to not simply ignore something because you don't like it.

Mahjong seems to have a bigger role in mainstream Japanese arcades than slot machines do over here.  Here slot machines more or less replaced arcade games in many places, in Japan they exist together.  I'm not really sure I'd say they were equivalent.

That said, even for slot machines, there are ones people consider to be 'classic'.  The original Cherry Master for example is really akin to the 'Pacman' of that genre.  It's simple, but to many still iconic, which is why it's copied / rehashed so often.  There are still hacks of this very same game being sold on newly manufactured PCBs today, and fresh 'classic' versions are pushed out just as Namco re-release Pacman every couple of years.  Overall there is a lot of hatred towards slot machines and the like tho, because many see them as the 'killer of arcades'  

For Australia, the early Aristocrat games are considered 'classics' too and have a substantial fan-base.

For a more central European base, Magic Card is another one that regularly gets requested because it doesn't yet work properly in MAME.

Just like any genre, you have a period of games, which are really stripped down to the basics, nothing too fancy, but advanced enough to be functional and memorable.

It's also interesting to see how slot machine mechanics integrate into normal games too.  You've only got to look at something like the Casino Night levels in Sonic 2 to see that elements of the whole 'slot machine' culture are iconic enough to end up as a mechanic in a more modern game, even Japanese developed ones.  Kaneko's Wani Wani World (a personal favourite) incorporates such elements too.

Pinball games are another example, some people see them all as utterly worthless, others will no doubt be able to name what they consider to be 'classics' in that genre too.

Asking if anybody considers Mahjong games classic on a US centric forum isn't going to give you many answers tho because it's a game heavily tied to Japanese culture, which doesn't translate well outside of Japan.

[Warborg]

Ok...  This is just a question for my clarification and is probably a bit OT from the actual development of MAME, but it's something I've always wondered (so please, don't take this as any form of argument or criticism)...

I saw it mentioned again earlier, the idea that MAME is to document old arcade games so that they are not lost to obscurity, but no emphasis is made on their playability...  I guess I don't understand that, documenting games that their whole purpose IS to be played and nothing else.  It would seem to me like making sure we document classic movies, but not be concerned with actually being able to watch them, just have documentation to keep their memory alive?  So like I said, not being argumentative in any way, I am just curious as to why?

[Vigo]

Hands up anyone who thinks of Mahjong games when they think classics?

Based on the posts I've had on my blog, and the requests we've had for some of them I think if you asked this question on a Japanese forum you'd be flooded with replies.

The early Dynax and Nichibutsu ones especially are considered 'classics' in Japan, and were just as popular, if not more popular than the likes of Street Fighter 2 was over here.  There is a reason they made so many of them and near enough every major manufacturer has a portfolio of them.  There have been just as many donations made *specifically* to buy Mahjong boards as anything else over the years.

On similar lines, if you asked on a Japanese forum 'Who considers Marble Madness a classic' you'd probably get hardly any replies, because it's an American developed game and they really didn't care for those as much.  (A trend which continues to this day, just look how difficult Microsoft found it to break into Japan)

It amazes me that people still try to use this argument to bash MAME, the project, and what gets emulated.  It's a global project, representing a global industry, not something tailored to your exact tastes.

Don't take this in a rude way, just a question. I honestly don't know how you guys are structured, but if it is the primarily Asia fanbase that considers these endless Mahjong titles of great value, is the development of these titles, the man hours, worked on by a designated Asia MAMEdev team? To me, I would think that a donation to buy Mahjong boards does not relegate that N. America and European support must be doing any further development to the endless sea of Mahjong titles out there.

I also wouldn't think that comparing your everyday Mahjong title with Marble Madness is accurate. I think it would compare much better to video slot machines. There are so many video slot machine games in N. America, and tons of people play them, but there really are not any classics that had the crowds watching on. Mahjong games were more or less your typical gambling games without the payouts.

Asia / Japan don't really seem to believe in the open source culture as much unfortunately, so contributions in terms of actual code from that part of the world are more limited.  It's clearly a cultural thing, and when you do get Asian / Japanese emulators their approach to emulation is often very different eg. if you look at the NeoGeo CD emulators they've *rewritten* most of the bios code, to avoid emulating it.. In worse cases they will rewrite parts of the game for extra speed, rather than emulating it.. They will integrate the roms in the actual emulator, and encrypt the whole package to stop people reverse engineering it then package the whole thing with a ton of ad-ware so that they can make money of of people using their emulators.  Don't get me wrong, there have been contributions from Asia / Japan which have been valid but the cultural differences, and coding standards / practices make them fewer.  The *primary* development of MAME has always been European, which is ironic when you consider that only a tiny tiny handful of the supported games originate in Europe.  If European devs only worked on European produced games there wouldn't be much MAME at all ;-)

From a 'writing the emulation code' perspective it makes very little difference if you're emulating a Mahjong game or not.  The challenges are still the same, the hardware can still be interesting, and they can help fix bugs / improve things elsewhere.  The Mahjong games on Psikyo's SH2 based hardware were of great help in fixing the SH2 timers / interrupts which in turn also improved the music in several other games relying on that logic.  Improving hardware emulation is always beneficial to the project, and anything that helps that, be it emulation of a Mahjong game, or a console system in MESS is a step forward.  The best devs really aren't fussy over the nature of the games they emulate.  As I've said before, the role of Mamedev is to emulate, not discriminate, and there is a level of responsibility that comes with that to not simply ignore something because you don't like it.

Mahjong seems to have a bigger role in mainstream Japanese arcades than slot machines do over here.  Here slot machines more or less replaced arcade games in many places, in Japan they exist together.  I'm not really sure I'd say they were equivalent.

That said, even for slot machines, there are ones people consider to be 'classic'.  The original Cherry Master for example is really akin to the 'Pacman' of that genre.  It's simple, but to many still iconic, which is why it's copied / rehashed so often.  There are still hacks of this very same game being sold on newly manufactured PCBs today, and fresh 'classic' versions are pushed out just as Namco re-release Pacman every couple of years.  Overall there is a lot of hatred towards slot machines and the like tho, because many see them as the 'killer of arcades' 

For Australia, the early Aristocrat games are considered 'classics' too and have a substantial fan-base.

For a more central European base, Magic Card is another one that regularly gets requested because it doesn't yet work properly in MAME.

Just like any genre, you have a period of games, which are really stripped down to the basics, nothing too fancy, but advanced enough to be functional and memorable.

It's also interesting to see how slot machine mechanics integrate into normal games too.  You've only got to look at something like the Casino Night levels in Sonic 2 to see that elements of the whole 'slot machine' culture are iconic enough to end up as a mechanic in a more modern game, even Japanese developed ones.  Kaneko's Wani Wani World (a personal favourite) incorporates such elements too.

Pinball games are another example, some people see them all as utterly worthless, others will no doubt be able to name what they consider to be 'classics' in that genre too.

Asking if anybody considers Mahjong games classic on a US centric forum isn't going to give you many answers tho because it's a game heavily tied to Japanese culture, which doesn't translate well outside of Japan.

Thanks, Haze. That answered my question. Very Informative!

[Haze]

Ok...  This is just a question for my clarification and is probably a bit OT from the actual development of MAME, but it's something I've always wondered (so please, don't take this as any form of argument or criticism)...

I saw it mentioned again earlier, the idea that MAME is to document old arcade games so that they are not lost to obscurity, but no emphasis is made on their playability...  I guess I don't understand that, documenting games that their whole purpose IS to be played and nothing else.  It would seem to me like making sure we document classic movies, but not be concerned with actually being able to watch them, just have documentation to keep their memory alive?  So like I said, not being argumentative in any way, I am just curious as to why?

In terms of what? Performance?  You just have to hope that machines catch up, to do things properly sometimes they just have to be slow, unless you start rewriting / not emulating significant parts of the game to avoid emulating some of the hardware.  At that point it stops being emulation at all, and becomes more like a port.

In terms of controls?  Half of X2s arguments here have been that MAME cares *too* much about this kind of thing, and not enough about the original controls.  MAME is designed so that the developers can test it / use it / play it (which is necessary to make sure the emulation works!) on a regular PC.  It's made as accessible as possible, and it's only the special use cases that suffer.  For most cases in MAME you can fire up the emulator, and play the games with a keyboard / mouse / x360 pad.  Most things in MAME are figured out by looking at the behavior of the game when it's running, it's therefore vital that developers actually be able to play through the games, using ordinary hardware in order to find good use-cases that test the emulation code, and improve the overall emulation.  You can't tell that you've emulated something properly just from staring at the title screen!

Everything in MAME stands as a decent reference for if somebody else DOES want to do it their own way.  Things might not always be 'playable' at 60fps in MAME, or have the exact interface some people want, but MAME is a good reference, and people can do things their own way in their own projects based on the information we're giving them.  The DEmul guys are making rapid progress in Model 3 in their own emulator and I'm sure there are many ways in which MAME has helped them make that progress even if the games aren't playable in MAME yet (and are unlikely to run at 60fps even if they did boot properly)  Even in the cases where MAME does try to make things easier for PC users, there is sufficient information in the code (IMHO) for somebody to modify something to work with whatever input hardware they need.

If MAME was to take nasty shortcuts, people would just copy the nasty shortcuts, and things would never really be emulated.  Going back to the NeoCD stuff, every single emulator I've seen is more of less lifting the same 'avoid emulating the CD controller, bypass most of the bios' hacks that the original Japanese NeoCD emulators did.  As a result they're full of tiny hacks for specific games.  There is little excuse for this!  When the system actually ends up being emulated in MESS (which is really just MAME for everything that doesn't fit in MAME) it will be done properly, and thus provide a far better reference to how things work.  If people would then rather copy that it's up to them, but having an emulator that does things properly IS important.

To sum things up, I'm not really sure where you're coming from.

The emphasis is on documenting things, and figuring out how they work, but that by nature implies allowing a level of playability, otherwise the former is also impossible!

The emphasis on documentation just means that MameDev don't consider something to be 'finished' just because it's 'playable'  If there is room for improvement (emulating things more precisely/correctly, even if there is no visible difference) then that remains a target.  It also constrains MAME to trying to emulate things accurately, rather than the project allowing developers to fix bugs present in the original games, or otherwise 'enhance' things.  I'm sure you prefer your Pacman with the kill screen, than if MameDev patched it out ;-)

[Warborg]

Ok...  This is just a question for my clarification and is probably a bit OT from the actual development of MAME, but it's something I've always wondered (so please, don't take this as any form of argument or criticism)...

I saw it mentioned again earlier, the idea that MAME is to document old arcade games so that they are not lost to obscurity, but no emphasis is made on their playability...  I guess I don't understand that, documenting games that their whole purpose IS to be played and nothing else.  It would seem to me like making sure we document classic movies, but not be concerned with actually being able to watch them, just have documentation to keep their memory alive?  So like I said, not being argumentative in any way, I am just curious as to why?

In terms of what? Performance?  You just have to hope that machines catch up, to do things properly sometimes they just have to be slow, unless you start rewriting / not emulating significant parts of the game to avoid emulating some of the hardware.  At that point it stops being emulation at all, and becomes more like a port.

In terms of controls?  Half of X2s arguments here have been that MAME cares *too* much about this kind of thing, and not enough about the original controls.  MAME is designed so that the developers can test it / use it / play it (which is necessary to make sure the emulation works!) on a regular PC.  It's made as accessible as possible, and it's only the special use cases that suffer.  For most cases in MAME you can fire up the emulator, and play the games with a keyboard / mouse / x360 pad.  Most things in MAME are figured out by looking at the behavior of the game when it's running, it's therefore vital that developers actually be able to play through the games, using ordinary hardware in order to find good use-cases that test the emulation code, and improve the overall emulation.  You can't tell that you've emulated something properly just from staring at the title screen!

Everything in MAME stands as a decent reference for if somebody else DOES want to do it their own way.  Things might not always be 'playable' at 60fps in MAME, or have the exact interface some people want, but MAME is a good reference, and people can do things their own way in their own projects based on the information we're giving them.  The DEmul guys are making rapid progress in Model 3 in their own emulator and I'm sure there are many ways in which MAME has helped them make that progress even if the games aren't playable in MAME yet (and are unlikely to run at 60fps even if they did boot properly)  Even in the cases where MAME does try to make things easier for PC users, there is sufficient information in the code (IMHO) for somebody to modify something to work with whatever input hardware they need.

If MAME was to take nasty shortcuts, people would just copy the nasty shortcuts, and things would never really be emulated.  Going back to the NeoCD stuff, every single emulator I've seen is more of less lifting the same 'avoid emulating the CD controller, bypass most of the bios' hacks that the original Japanese NeoCD emulators did.  As a result they're full of tiny hacks for specific games.  There is little excuse for this!  When the system actually ends up being emulated in MESS (which is really just MAME for everything that doesn't fit in MAME) it will be done properly, and thus provide a far better reference to how things work.  If people would then rather copy that it's up to them, but having an emulator that does things properly IS important.

To sum things up, I'm not really sure where you're coming from.

The emphasis is on documenting things, and figuring out how they work, but that by nature implies allowing a level of playability, otherwise the former is also impossible!

The emphasis on documentation just means that MameDev don't consider something to be 'finished' just because it's 'playable'  If there is room for improvement (emulating things more precisely/correctly, even if there is no visible difference) then that remains a target.  It also constrains MAME to trying to emulate things accurately, rather than the project allowing developers to fix bugs present in the original games, or otherwise 'enhance' things.  I'm sure you prefer your Pacman with the kill screen, than if MameDev patched it out ;-)

Thank you for your detailed response Haze, I just wanted to have a more complete grasp on what the perspective was regarding playability, and that answered a lot of it.  One question I do have is why isn't it possible to use current 3D acceleration with MAME wile still emulation, maybe just translating the original calls to current 3D hardware?  Mind you, I'm not a programmer at all, the last time I tried programming was BASIC on a C64...  More of a hardware guy myself... 

Well...  Computer hardware and cars, love me some fast cars... 

[Haze]

Thank you for your detailed response Haze, I just wanted to have a more complete grasp on what the perspective was regarding playability, and that answered a lot of it.  One question I do have is why isn't it possible to use current 3D acceleration with MAME wile still emulation, maybe just translating the original calls to current 3D hardware?  Mind you, I'm not a programmer at all, the last time I tried programming was BASIC on a C64...  More of a hardware guy myself...  

Well...  Computer hardware and cars, love me some fast cars...  

Getting modern 3d hardware to accurately represent graphics in a deterministic way (which won't differ from system to system) is near enough impossible.  Video drivers are nasty hacked up messy things and hardware implementations + capabilities differ from card to card, it' simply not that reliable.

There has been talk of using the pixel shaders on the next generation of cards to handle rendering, because they should be sufficiently powerful to do so, and give accurate output, however, with nobody to code such a system it's still quite unlikely to happen, and MAME would always have to have a software fallback for systems which don't have the hardware.

For the majority of cases where people think 3D acceleration would help however, it wouldn't.  A lot of the system requirements still come from the CPU emulation, which for various reasons is a lot slower in MAME than other emulators.

[LeedsFan]

 I'm sure you prefer your Pacman with the kill screen, than if MameDev patched it out ;-)

Haha!   I remember when Donkey Kong finally got most of it's sounds emulated correctly and didn't need samples.... and people complained that the running sound of Mario sounded wrong. That was because they'd got so used to hearing the sampled sound in Mame for so long that when it was emulated correctly it didn't sound right to them. But I remember playing the game way back in 1982 and I always thought the sampled sound of Mario was too high pitched. In fact you can still hear the old sampled sound on some of the video clips used for frontends... then when the game starts you hear the difference (assuming your Mame is up to date ofc)

[Xiaou2]

and it's only the special use cases that suffer.

 Special cases, as in losing 100% Arcade ACCURACY.

 You can write your +3500 words of BS EXCUSES all you want, but it does not add up.

 Its a complete contradiction to the purpose of mame, which is to ACCURATELY PRESERVE these machines for future generations.

 Yet, without ACCURATE CONTROLS, the game is NEVER Accurate. PERIOD!

 Its not accurate in how the game is balanced.
 Its not accurate in how the player experiences these games.

 Theres no good reason why these emulated machines should not have ability to hook into CORRECT Arcade ACCURATE CONTROLS.  The main reason is of course nothing to do with developer ease.  Its to due with Care.  Devs Dont care about the machines they are preserving... which is why they dont go the full distance and get things 100% CORRECT.

 Which again, is why there is a need for Money to pay developers to care, and get things 100% Preserved.  Not half A**ed like it is currently.

 

[nitz]

  I'm sure you prefer your Pacman with the kill screen, than if MameDev patched it out ;-)

Haha!   I remember when Donkey Kong finally got most of it's sounds emulated correctly and didn't need samples.... and people complained that the running sound of Mario sounded wrong. That was because they'd got so used to hearing the sampled sound in Mame for so long that when it was emulated correctly it didn't sound right to them. But I remember playing the game way back in 1982 and I always thought the sampled sound of Mario was too high pitched. In fact you can still hear the old sampled sound on some of the video clips used for frontends... then when the game starts you hear the difference (assuming your Mame is up to date ofc)

I have no doubt that the emulated sound in current mame DK is closer to the original than the samples were - but frankly, it sounds like ---steaming pile of meadow muffin---. And it sounds like ---steaming pile of meadow muffin--- whether you're used to the samples or not. It may very well be how a real game sounds, I wouldn't know, but for now I will be using an old mame for this one.

Similar thing with Berzerk. There was a change awhile back that put this whiny sound in the game that I don't like, so I play that on an older mame as well.

If I think about it, there is probably some other stuff in mame that bugs me also, but I've made my point.

HOWEVER in spite of that, I like and respect the devs approach to emulation, and I am all for mame being as accurate as possible! If you start making little exceptions for this and little exceptions for that, where does it end? There is always old mame and other emulators if you need them.

[Haze]

Theres no good reason why these emulated machines should not have ability to hook into CORRECT Arcade ACCURATE CONTROLS.  The main reason is of course nothing to do with developer ease.  Its to due with Care.  Devs Dont care about the machines they are preserving... which is why they dont go the full distance and get things 100% CORRECT.

You seem to think you know everything about everything, and why things are done as they are, despite not being involved at all.

You continue to post this same BS all the time, despite being told by developers of the project why things are done as they are.

You continue to only want to believe your own viewpoint, no matter how far from the truth it may be.

As I've said before MAME is open, the code is there, if you have special use cases for the code, it's there for you to use.  I don't have a hundred odd novelty controllers to hook up to the PC to do development with, nor does any other developer.  It *is* about developer, and general purpose user ease.  People have to be able to use MAME to test it, so that bug reports can be given, and actual emulation of the hardware improved.

This is the truth, regardless of if you want to believe it or not.  Compromises are made for the benefit of the development and accessibility of the project only.  This has been the case for over 10 years.

[Haze]

 I'm sure you prefer your Pacman with the kill screen, than if MameDev patched it out ;-)

Haha!   I remember when Donkey Kong finally got most of it's sounds emulated correctly and didn't need samples.... and people complained that the running sound of Mario sounded wrong. That was because they'd got so used to hearing the sampled sound in Mame for so long that when it was emulated correctly it didn't sound right to them. But I remember playing the game way back in 1982 and I always thought the sampled sound of Mario was too high pitched. In fact you can still hear the old sampled sound on some of the video clips used for frontends... then when the game starts you hear the difference (assuming your Mame is up to date ofc)

I have no doubt that the emulated sound in current mame DK is closer to the original than the samples were - but frankly, it sounds like ---steaming pile of meadow muffin---. And it sounds like ---steaming pile of meadow muffin--- whether you're used to the samples or not. It may very well be how a real game sounds, I wouldn't know, but for now I will be using an old mame for this one.

Similar thing with Berzerk. There was a change awhile back that put this whiny sound in the game that I don't like, so I play that on an older mame as well.

If I think about it, there is probably some other stuff in mame that bugs me also, but I've made my point.

HOWEVER in spite of that, I like and respect the devs approach to emulation, and I am all for mame being as accurate as possible! If you start making little exceptions for this and little exceptions for that, where does it end? There is always old mame and other emulators if you need them.

For DK It probably doesn't sound perfect yet, I think even Derrick recognizes that, in some ways it does offer a better, more realistic sound but there are still problems with it sounding too muffled.  Most samples have been post-processed and enhanced tho, so it's never going to sound exactly like them, and due to the nature of these circuits you're probably never even going to find 2 boards that sound quite the same.  It's a complex challenge to get it right, but MAME is always kept up to date with the latest findings, it will get there eventually.  Again it's an area only one dev really understands, one on which it's very easy to get burnout from working on it, and one where there are countless games which need attention.

Berzerk on the other hand sounds close to perfect to me.

To give you some perspective, this is why emulating the older discrete games is an insane challenge; the sound systems on these things are just a couple of circuits designed to play sounds.  Even a simple game like pong is 10x more complex.  Something like Monaco GP is absolute madness.

[EwJ]

I have no doubt that the emulated sound in current mame DK is closer to the original than the samples were - but frankly, it sounds like ---steaming pile of meadow muffin---. And it sounds like ---steaming pile of meadow muffin--- whether you're used to the samples or not. It may very well be how a real game sounds, I wouldn't know, but for now I will be using an old mame for this one.

IIRC DK sounded like ---steaming pile of meadow muffin--- in every arcade I saw it in many moons ago.  If you are finding, in MAME, that DK sounds all garbled, tinny, slightly delayed, and as if you're barely hearing it through static and what seems like the death croaks of a dozen frogs in a blender, it is likely far closer to original than you might suspect.  Enjoy it! Revel in it! This is arcade history and you are experiencing it just as if you dropped a quarter into it in an 80's arcade.

If I think about it, there is probably some other stuff in mame that bugs me also, but I've made my point.

What's your point?

[Vigo]

Theres no good reason why these emulated machines should not have ability to hook into CORRECT Arcade ACCURATE CONTROLS.  The main reason is of course nothing to do with developer ease.  Its to due with Care.  Devs Dont care about the machines they are preserving... which is why they dont go the full distance and get things 100% CORRECT.

You seem to think you know everything about everything, and why things are done as they are, despite not being involved at all.

You continue to post this same BS all the time, despite being told by developers of the project why things are done as they are.

You continue to only want to believe your own viewpoint, no matter how far from the truth it may be.

Pretty good 3 point summary of X2, Haze! 

Methinks all this post rage about inaccuracies could be directed to make some good developments to MAME....oh wait...has that been suggested before? 

[ivwshane]

Thank you for your detailed response Haze, I just wanted to have a more complete grasp on what the perspective was regarding playability, and that answered a lot of it.  One question I do have is why isn't it possible to use current 3D acceleration with MAME wile still emulation, maybe just translating the original calls to current 3D hardware?  Mind you, I'm not a programmer at all, the last time I tried programming was BASIC on a C64...  More of a hardware guy myself...  

Well...  Computer hardware and cars, love me some fast cars...  

Getting modern 3d hardware to accurately represent graphics in a deterministic way (which won't differ from system to system) is near enough impossible.  Video drivers are nasty hacked up messy things and hardware implementations + capabilities differ from card to card, it' simply not that reliable.

There has been talk of using the pixel shaders on the next generation of cards to handle rendering, because they should be sufficiently powerful to do so, and give accurate output, however, with nobody to code such a system it's still quite unlikely to happen, and MAME would always have to have a software fallback for systems which don't have the hardware.

For the majority of cases where people think 3D acceleration would help however, it wouldn't.  A lot of the system requirements still come from the CPU emulation, which for various reasons is a lot slower in MAME than other emulators.

How hard is it to implement something like GPU hardware acceleration for increased processing power?

[Rick]

Which again, is why there is a need for Money to pay developers to care, and get things 100% Preserved.  Not half A**ed like it is currently.

This brings to mind a couple of thoughts.

Firstly, "put up or shut up" comes to mind.  If you're going to go 'out of pocket' to support the developers, then do so.  If you're going to step up and do some programming, then do so.  Otherwise, this type of opinion is not appreciated, and is not the consensus of the community at whole.

Secondly, who are you to say that the solutions being provided are 'half assed?'  (Oh, and thank you for not trying to fool the spam filter.  Do we need to quote the rules again?)  Who do you think you are, calling out those who are working - out of personal expense and LOVE for the community and MAME as a whole - to give YOU something more than you had yesterday?  You'd be better to say 'thank you', and just step out of their way.  Really.

 

[Haze]

Thank you for your detailed response Haze, I just wanted to have a more complete grasp on what the perspective was regarding playability, and that answered a lot of it.  One question I do have is why isn't it possible to use current 3D acceleration with MAME wile still emulation, maybe just translating the original calls to current 3D hardware?  Mind you, I'm not a programmer at all, the last time I tried programming was BASIC on a C64...  More of a hardware guy myself...  

Well...  Computer hardware and cars, love me some fast cars...  

Getting modern 3d hardware to accurately represent graphics in a deterministic way (which won't differ from system to system) is near enough impossible.  Video drivers are nasty hacked up messy things and hardware implementations + capabilities differ from card to card, it' simply not that reliable.

There has been talk of using the pixel shaders on the next generation of cards to handle rendering, because they should be sufficiently powerful to do so, and give accurate output, however, with nobody to code such a system it's still quite unlikely to happen, and MAME would always have to have a software fallback for systems which don't have the hardware.

For the majority of cases where people think 3D acceleration would help however, it wouldn't.  A lot of the system requirements still come from the CPU emulation, which for various reasons is a lot slower in MAME than other emulators.

How hard is it to implement something like GPU hardware acceleration for increased processing power?

In a way which is general purpose enough to benefit everything, flexible enough to be a basis of a cross-platform solution, while still being able to give guaranteed pixel perfect (or at least identical across cards / drivers) output for any given driver, and retaining proper 2d/3d mixing abilities, framebuffer readback, and correct timing etc.?  Hard, very hard.

That's why the general talk for it is more akin to doing a 'software renderer' on the GPUs using the shaders etc.

 

[Vigo]

Thank you for your detailed response Haze, I just wanted to have a more complete grasp on what the perspective was regarding playability, and that answered a lot of it.  One question I do have is why isn't it possible to use current 3D acceleration with MAME wile still emulation, maybe just translating the original calls to current 3D hardware?  Mind you, I'm not a programmer at all, the last time I tried programming was BASIC on a C64...  More of a hardware guy myself... 

Well...  Computer hardware and cars, love me some fast cars... 

Getting modern 3d hardware to accurately represent graphics in a deterministic way (which won't differ from system to system) is near enough impossible.  Video drivers are nasty hacked up messy things and hardware implementations + capabilities differ from card to card, it' simply not that reliable.

There has been talk of using the pixel shaders on the next generation of cards to handle rendering, because they should be sufficiently powerful to do so, and give accurate output, however, with nobody to code such a system it's still quite unlikely to happen, and MAME would always have to have a software fallback for systems which don't have the hardware.

For the majority of cases where people think 3D acceleration would help however, it wouldn't.  A lot of the system requirements still come from the CPU emulation, which for various reasons is a lot slower in MAME than other emulators.

How hard is it to implement something like GPU hardware acceleration for increased processing power?

In a way which is general purpose enough to benefit everything, flexible enough to be a basis of a cross-platform solution, while still being able to give guaranteed pixel perfect (or at least identical across cards / drivers) output for any given driver, and retaining proper 2d/3d mixing abilities, framebuffer readback, and correct timing etc.?  Hard, very hard.

That's why the general talk for it is more akin to doing a 'software renderer' on the GPUs using the shaders etc.

 

Keep in mind i'm no programmer here, but is it a possibility to only focus on building support piece by piece for only one series of graphics card for the time being? I think there would be a ton of people willing to get an exact graphics card to gain the benefits of GPU acceleration.

I remember on some old PC games, the developers only worked on building support for 1 card. It made me think of this as a possible solution. And if hardware acceleration got buggy on emulation, it could possibly be turned on or off on a game by game basis. I would think that would be no greater compromise on emulation than the frameskip option.

[saint]

MAME is cross-platform though. You'd have to do a platform specific build for a specific video card I think. Don't see that fitting in with MAME's origins/direction.

[MagicManPA]

I'd just like to say that while MAME is not perfect, nothing else in life is either.

I am glad MAME is available for me to use & I enjoy it every time I use it. I don't know how to do programming so kudos to the folks that made it for us to use. 

[Vigo]

MAME is cross-platform though. You'd have to do a platform specific build for a specific video card I think. Don't see that fitting in with MAME's origins/direction.

I'm not going to argue that.   I don't know all that goes into getting hardware acceleration for even a single card working, but I am guessing that if it were as simple as I put it, it would have been attempted already. Something like MAME, I can understand how it would be exponentially complex. I don't exactly see how it is a method that destroys the integrity of the emulation, but I live in a simple world that thinks that using a graphics card for this kind of application would only be utilizing the card's processing power.

[Haze]

Keep in mind i'm no programmer here, but is it a possibility to only focus on building support piece by piece for only one series of graphics card for the time being? I think there would be a ton of people willing to get an exact graphics card to gain the benefits of GPU acceleration.

I remember on some old PC games, the developers only worked on building support for 1 card. It made me think of this as a possible solution. And if hardware acceleration got buggy on emulation, it could possibly be turned on or off on a game by game basis. I would think that would be no greater compromise on emulation than the frameskip option.

Honestly, I'm not sure how it would work.  As you might have noticed, graphic cards in general are becoming more general purpose processors which is what I believe MAME is hoping to take advantage of.

It might end up being something like the recompiler system, whereby MAME defines a intermediate 'language' to say how the video hardware works, and then has backends (including a fall-back one in C) to generate code for the target platforms.  That would be slower than a specifically crafted solution for a specific card, but would prevent MAME from being tied to a specific piece of hardware entirely.

Of course, this might be why one of the reason the MAME recompiler doesn't give quite such good results anyway, and only a handful of cores have even been converted over to it (SH2 and MIPS) and only 3 backends (aside the C one) exist, those being x86 / x64 and PPC.

It's no easy task, and to get it right will require a lot of planning and a lot of expertise in these areas.  If the scope is too limited drivers will quickly outgrow it, and be back to where we started.

[Xiaou2]

More Excuses.

 720 is a prime example.

 It used to be able to be controlled with an arcade CORRECT controller.  However, nobody else could play it who didnt have a 720 controller, or didnt build their own 720 controller (which IS very possible - "IF" the game supports it in mame).

 Just about Any other game in mame can have MULTIPLE controllers to make them work:  IE: Marble Madness can be played with:

1) Keyboard
2) Mouse
3) REAL TRACKBALL

 But 720?   Nope.  You have to settle for the most INACCURATE, butchered, and mechanically unsound method of all.

 Now, give us a good reason why MAME would KILL 100% Accurate WORKING
controls of a machine, and NOT allow it as an OPTION to the users?

 Its called a "SWITCH".  Just like any other thing in mame which can be disabled or Enabled.  Mame is just too lazy to implement it.. because mame does not care about 100% PRESERVED ACCURACY.

[Xiaou2]

Firstly, "put up or shut up" comes to mind.  If you're going to go 'out of pocket' to support the developers, then do so.  If you're going to step up and do some programming, then do so.  Otherwise, this type of opinion is not appreciated, and is not the consensus of the community at whole.

 Firstly, I HAVE donated to various money collecting ventures to buy Rare PCBs, several times, despite my poor financial standings.  You?

 I have Also managed to borrow a collectors Discs of Tron artwork from his Pristine working Environmental machine, which I scanned in 4 parts, and hand edited, which ended up in Mame.  And THEN, had to scan in like 15 pages of Schematics, and send them via snail mail to a Dev to get the lights to work correctly.

 Now, what have YOU done for Preservation sake?

 As for the Consensus, plenty of people who would like to play Atari's 720, would easily disagree.  And though Im not even a big fan of the game, even I realize the IMPORTANCE of Preserving the ability to Correctly Interface a real controller for 100% Arcade Accuracy.

 Plenty of people have complained about various controller options lacking, and over time, many of them have been solved.  However, there are still many cases which have been butchered (720), ignored (Shifters), etc.

 The Devs dont need a 720 controller to make a 720 controller work.  That is Hazes BS line to try to fool the ignorant.  720 used to work fine with a real controller. 
All they need to do, is bypass their OCD for 10 seconds, and code in a few switches.

 Various Examples:

 Arcadecontrols = yes
 Arcadecontrols = no
(easy fix for a game like 720, which has only one special controller that isnt available)

 ArcadeController1 = Yes / No
 ArcadeController2 = Yes / No
(Could be used for multiple controls in a game, without need for individual names)

 720controller = Y / N
 4wayShifter  = Y / N
 2wayShifter  = Y / N
(Define each control option available at the start, and actually enforce them)

 More robust solutions could be implemented, but a simple solution could easily fix the problem until a time when some Dev cares enough to do things better.

[Rick]

Firstly, I HAVE donated to various money collecting ventures to buy Rare PCBs, several times, despite my poor financial standings.  You?

Nothing, actually.  I wasn't actually aware this had occurred, as I'm still pretty new here.

Now, what have YOU done for Preservation sake?

Hmm.  Again, nothing.

But, here's where we differ:   I respect those who have contributed to MAME development, and I don't post like a pompous, elitist, self-serving ass who thinks he's entitled to have his opinions on the direction of the development actually heard and followed.  You post like your donations or ideas mean you deserve something in return.

It just simply doesn't work that way.  The Development Team doesn't exist to feed your fancy, and they'll do it their own way, regardless of how 'smart' or 'right' you think your ideas are.  There's no sense getting all butthurt about it, either.  Even if you have the be-all, end-all solution to all of MAME's problems for today and the future, it doesn't mean they're going to build it as you say, or make it work as you say.  As it's been said a million times before in this thread, if you want it changed, pick up a programming manual and do it yourself.  Nobody's stopping you.  Just don't expect the developers to do anything, just because you've kicked in some coin, or scanned in some images.

[saint]

Gentle reminder to keep this thread civil. (This post goes to no one specifically) Thanks!

--- saint

[Haze]

I also think we need to be honest here and realize how many old games were running nearly identical hardware.  I imagine once they figured out how to emulate a Z80 and dump a few ROMs they cracked 90% of the games from the 1980s.  You don't get that luxury with anything newer.

Actually of newer stuff is running on well documented modern hardware, so it's easier.  That's why it's so shambolic that PSX and N64 based emulation in MAME is so bad.

If you think simply emulating a CPU core grants you perfectly working games then you have a lot to learn.  There are still games running on a Z80 that have been an incredible amount of work to emulate; see the recent progress on Heavy Unit for example.  The processor is merely one component, the rest is 99% undocumented and understanding the rest is pretty much the entirety of the work I've done for MAME over the last 10 years, giving hours to it every day.

Then of course you've got the fact that even something as 'simple' as the Z80 has undocumented behavior, even as recently as last year some things were being fixed on it, chances are nothing relies on those features (and they seem to differ from manufacturer to manufacturer) but again it goes to show that there is always room for improvement.  There are apparently some gambling games that DO check manufacturer specific behavior of some of the 6xxx series CPUs as a form of copy protection!

[Haze]

blah blah blah blah blah..

1) To properly implement something, yes, developers need the controls

2) Aaron would like to keep switches down to a minimum, he's stated this in the past.

3) Again, if you have a special use case, there is nothing preventing you from either changing it yourself, or paying for somebody to change it yourself.  Just don't expect Mamedev to do it for you.  It's a special case, nothing more, nothing less.

Your expectations of the dev team are _ridiculous_.

Would you expect a deaf person to fix the sound emulation in something?  Given what you're saying, I'm guessing the answer is yes, because they could look at the waveforms of course!

Introducing unmaintainable, untested code into the project is suicide.

[nitz]

Berzerk on the other hand sounds close to perfect to me.

Thanks for your response Haze!

Just to be clear, the whining in Berzerk I'm talking about is this (lifted from MAWS):

- 0.125u9: Update S14001A core [Lord Nightmare]: Improved filtering to be more accurate to the chip, changed internal audio renderer to produce 4 bit sound as the real chip does. Moved filter outside of the main rendering loop into the glue code. Note: With this update, you may hear a whining noise that was not previously present. This is accurate to the chip! The real chip would almost always have an analog filter placed on its output to suppress this whine.

Maybe your soundcard/speakers suppress the whine more, or maybe your definition of "sounds close to perfect" is "accurate to the chip". I'd rather sacrifice accuracy though for sound that I find less annoying. Others may prefer accuracy even if it is annoying, or they may not find it as annoying as I do. It's all personal preference.

If I think about it, there is probably some other stuff in mame that bugs me also, but I've made my point.

What's your point?

I guess I was just saying that I can kinda sympathize with people who feel mame is too accurate because this often introduces stuff we don't like (games needing crazy PC specs being the main gripe). But as I stated, I still think "accuracy" ultimately equals "better" in the long run, even if we don't like it sometimes, because if little hacks are allowed to creep in just to make things run faster or suit personal tastes, it'll end up a being a big mess.

[Haze]

Maybe your soundcard/speakers suppress the whine more, or maybe your definition of "sounds close to perfect" is "accurate to the chip". I'd rather sacrifice accuracy though for sound that I find less annoying. Others may prefer accuracy even if it is annoying, or they may not find it as annoying as I do. It's all personal preference.

Possibly, I just have a pair of cheap headphones on my dev PC at the moment, which might not be capable of reproducing the sound.

This was as much the case with the cabinet speakers / acoustics.  In a noisy arcade you probably wouldn't have a clue where a high pitched whine was coming from, even if one was present, the actual CRT monitors can generate just as much of a whine!  (I'm glad my hearing is worse now than when I was younger, because I found some TVs impossible to watch, or even be within 20 meters of for this reason)

There isn't really much MAME can do about this if it is a correct output tho, especially if there are other games that rely on the frequencies output for correct sound reproduction without the whine.

Of course, the benefits of emulating the actual speech chip are quite clear.  Previously MAME had a set of samples for Berzerk, they might have sounded nice and had imperfections, but they were for Berzerk only.

Recently both a German set of Berzerk, and the rare prototype Moon War showed up.  With only samples for Berzerk these would be have been left without proper speech, and in the case of moonwar it would have been nearly impossible to even record it; only 2 parts of it have shown up, the speech board in the US, many years ago, and a main board (lacking the speech board) in Germany of all places.  It's possible they were once part of the only remaining boardset for the game!  (the German Berzerk set is wonderful btw, the robots sound so much more intimidating speaking German)

Due to the fact the speech chip is emulated, rather than using samples both these automatically ended up with correct speech support in MAME.

The likes of Gorf suffer from a similar problem right now, the only speech support is through samples, so the foreign language versions simply don't play the correct samples.  The German sets of that and Wizard of Wor should also talk in German!

[danny_galaga]

Hands up anyone who thinks of Mahjong games when they think classics?

Based on the posts I've had on my blog, and the requests we've had for some of them I think if you asked this question on a Japanese forum you'd be flooded with replies.

The early Dynax and Nichibutsu ones especially are considered 'classics' in Japan, and were just as popular, if not more popular than the likes of Street Fighter 2 was over here.  There is a reason they made so many of them and near enough every major manufacturer has a portfolio of them.  There have been just as many donations made *specifically* to buy Mahjong boards as anything else over the years.

On similar lines, if you asked on a Japanese forum 'Who considers Marble Madness a classic' you'd probably get hardly any replies, because it's an American developed game and they really didn't care for those as much.  (A trend which continues to this day, just look how difficult Microsoft found it to break into Japan)

It amazes me that people still try to use this argument to bash MAME, the project, and what gets emulated.  It's a global project, representing a global industry, not something tailored to your exact tastes.

Don't take this in a rude way, just a question. I honestly don't know how you guys are structured, but if it is the primarily Asia fanbase that considers these endless Mahjong titles of great value, is the development of these titles, the man hours, worked on by a designated Asia MAMEdev team? To me, I would think that a donation to buy Mahjong boards does not relegate that N. America and European support must be doing any further development to the endless sea of Mahjong titles out there.

I also wouldn't think that comparing your everyday Mahjong title with Marble Madness is accurate. I think it would compare much better to video slot machines. There are so many video slot machine games in N. America, and tons of people play them, but there really are not any classics that had the crowds watching on. Mahjong games were more or less your typical gambling games without the payouts.

Asia / Japan don't really seem to believe in the open source culture as much unfortunately, so contributions in terms of actual code from that part of the world are more limited.  It's clearly a cultural thing, and when you do get Asian / Japanese emulators their approach to emulation is often very different eg. if you look at the NeoGeo CD emulators they've *rewritten* most of the bios code, to avoid emulating it.. In worse cases they will rewrite parts of the game for extra speed, rather than emulating it.. They will integrate the roms in the actual emulator, and encrypt the whole package to stop people reverse engineering it then package the whole thing with a ton of ad-ware so that they can make money of of people using their emulators.  Don't get me wrong, there have been contributions from Asia / Japan which have been valid but the cultural differences, and coding standards / practices make them fewer.  The *primary* development of MAME has always been European, which is ironic when you consider that only a tiny tiny handful of the supported games originate in Europe.  If European devs only worked on European produced games there wouldn't be much MAME at all ;-)

From a 'writing the emulation code' perspective it makes very little difference if you're emulating a Mahjong game or not.  The challenges are still the same, the hardware can still be interesting, and they can help fix bugs / improve things elsewhere.  The Mahjong games on Psikyo's SH2 based hardware were of great help in fixing the SH2 timers / interrupts which in turn also improved the music in several other games relying on that logic.  Improving hardware emulation is always beneficial to the project, and anything that helps that, be it emulation of a Mahjong game, or a console system in MESS is a step forward.  The best devs really aren't fussy over the nature of the games they emulate.  As I've said before, the role of Mamedev is to emulate, not discriminate, and there is a level of responsibility that comes with that to not simply ignore something because you don't like it.

Mahjong seems to have a bigger role in mainstream Japanese arcades than slot machines do over here.  Here slot machines more or less replaced arcade games in many places, in Japan they exist together.  I'm not really sure I'd say they were equivalent.

That said, even for slot machines, there are ones people consider to be 'classic'.  The original Cherry Master for example is really akin to the 'Pacman' of that genre.  It's simple, but to many still iconic, which is why it's copied / rehashed so often.  There are still hacks of this very same game being sold on newly manufactured PCBs today, and fresh 'classic' versions are pushed out just as Namco re-release Pacman every couple of years.  Overall there is a lot of hatred towards slot machines and the like tho, because many see them as the 'killer of arcades'  

For Australia, the early Aristocrat games are considered 'classics' too and have a substantial fan-base.

For a more central European base, Magic Card is another one that regularly gets requested because it doesn't yet work properly in MAME.

Just like any genre, you have a period of games, which are really stripped down to the basics, nothing too fancy, but advanced enough to be functional and memorable.

It's also interesting to see how slot machine mechanics integrate into normal games too.  You've only got to look at something like the Casino Night levels in Sonic 2 to see that elements of the whole 'slot machine' culture are iconic enough to end up as a mechanic in a more modern game, even Japanese developed ones.  Kaneko's Wani Wani World (a personal favourite) incorporates such elements too.

Pinball games are another example, some people see them all as utterly worthless, others will no doubt be able to name what they consider to be 'classics' in that genre too.

Asking if anybody considers Mahjong games classic on a US centric forum isn't going to give you many answers tho because it's a game heavily tied to Japanese culture, which doesn't translate well outside of Japan.

Good point, but if the japanese fan base of Mahjong games really don't care about accurate emulation, why bother doing it in MAME? Seems to me neither East or West would appreciate all that work...

[Warborg]

Firstly, I HAVE donated to various money collecting ventures to buy Rare PCBs, several times, despite my poor financial standings.  You?

Nothing, actually.  I wasn't actually aware this had occurred, as I'm still pretty new here.

Now, what have YOU done for Preservation sake?

Hmm.  Again, nothing.

But, here's where we differ:   I respect those who have contributed to MAME development, and I don't post like a pompous, elitist, self-serving ass who thinks he's entitled to have his opinions on the direction of the development actually heard and followed.  You post like your donations or ideas mean you deserve something in return.

It just simply doesn't work that way.  The Development Team doesn't exist to feed your fancy, and they'll do it their own way, regardless of how 'smart' or 'right' you think your ideas are.  There's no sense getting all butthurt about it, either.  Even if you have the be-all, end-all solution to all of MAME's problems for today and the future, it doesn't mean they're going to build it as you say, or make it work as you say.  As it's been said a million times before in this thread, if you want it changed, pick up a programming manual and do it yourself.  Nobody's stopping you.  Just don't expect the developers to do anything, just because you've kicked in some coin, or scanned in some images.

Sorry, just have to say it...   PWNED.

[Hituro]

Wow, such a heated discussion! 

All I have to say though, is that regardless of how things are run, I'm at least glad mame exists.  if it wasn't for mame, I would not have had the drive to build my own arcade cabinet that houses a multitude of these games.  Games which I've also went out and bought just to support the saving of these old classics.  I know there are some games that I would love to play that don't work properly, but it doesn't really bother me too much, some of them I'm able to find as PC versions, and although they might not be 100% to the arcade originals, I can still enjoy them and remember the good old days.  On top of that, if it wasn't for mame, I wouldn't have stumbled upon a ton of games that I've never even heard of that were great classics that I myself have grown to enjoy.  Just remember, if it wasn't for mame we probably wouldn't have half the people we have today, working on ways to perfectly emulate arcade game play.  Like Andy from Ultimarc, or many of the awesome cabs that I've browsed through on the web.  And more importantly, this forum may not have ever existed.  Just something to think about

[Haze]

Good point, but if the japanese fan base of Mahjong games really don't care about accurate emulation, why bother doing it in MAME? Seems to me neither East or West would appreciate all that work...

You can't really bundle up everybody into the same group.

The Japanese / Asian developers have a different philosophy but the actual work is appreciated greatly by the community as a whole (including PCB owners, who are otherwise hindered by the lack of openness shown)

The benefits beyond that are the ones I've already mentioned?  Like having proper sound in the likes of Cyvern?  You've got Mahjong games to thank for that....

When we get around to adding sound in the fighting game 'Rabbit' the two Mahjong games using the same custom chip will no doubt help there too.

The same can be extended to adult games and others.  If it wasn't for a bootleg Pocket Gal Deluxe showing up (if for example MAME ignored adult games, so buying it was never considered) it's possible that Boogie Wings and Diet Go Go would not yet be working because that seemingly pointless bootleg, and comparisons between the bootleg and the original sets allowed us to figure out the encryption on everything using the DECO102 custom.

The Korean language quiz game running on Raiden 2 hardware has so far been proving to be an essential reference for improving the emulation of Raiden 2, because like many 3rd party games they use the protection in more obvious ways than Seibu, thus allowing it to be figured out more easily, piece by piece.  This might still take years to do (or it could end up being done next year, who knows) but once that's all properly figured out a whole bunch of Seibu games from the early 90s, including the 'classics' Seibu Cup Soccer and Raiden 2 / DX will work.

Even outside of emulation, Toaplan, who produced quite a few popular shooters before everybody went their own ways to form the likes of Cave, started off by bootlegging Mahjong games because they were popular enough that there was a high enough demand even for games which were just bootlegs of existing ones.  They admit this in an interview somewhere.

Don't underestimate the value of 'worthless' games ;-)

[Mikezilla]

Its so funny, the only game I have ever even heard of that you mentioned was Raiden 2. And the Mahjong games of course.

[Mikezilla]

Well said PBJ. Couldnt agree more. 

[Haze]

Its so funny, the only game I have ever even heard of that you mentioned was Raiden 2. And the Mahjong games of course.

I guess that's just representational of the era of games you experiences.

Cyvern is, like Raiden 2, a shooter, and considered a very good one at that
http://www.shmups.com/reviews/cyvern/index.html
(oh yeah, and I fixed the laying issues he mentions in the review with the help of the Mahjong games on the HW too)

Rabbit was one of EAs few forays into the arcade market, it was also released on the Saturn, and isn't especially bad.

Boogie Wings is one people remember fondly for the 'pick up elephants and swing them around' type gameplay.


Toaplan did Flying Shark (which everybody should know, although it displays Taito because Taito distributed it), and eventually Batsugun which is considered the game from which pretty much all modern 'bullet hell' shooters are derived.

There are all things people repeatedly asked after, and many of them are either only emulated properly due to some mahjong / adult games helping, or, only exist because some Mahjong games supported the company.

Nichibutsu were no different either, they gave you late 80s classics like Terra Cresta and Terra Force, but they made most of their money making Mahjong games!

Your argument that you've never heard of any of the games doesn't really make the point any less valid, I can tell you that emulating and producing Mahjong games has underpinned both the emulation and production of many games which have been enjoyed greatly by audiences outside of Japan.

Anyway, my other points are still valid.  There is a strong shift in the attitudes of up and coming programmers.  None of them seem to do it for enjoyment anymore.  To the very best devs MAMEDev has had it didn't matter to them if they were emulating Galaga or Doki Monkey Panic, the interesting part was figuring things out and getting them running.  The challenge of it all.  The developers of old would be relishing the challenge of getting Cool Riders ( Sega Cool Riders Attract ) running simply because it's such a weird piece of hardware.  (if it looks familiar basically that's because it's a 'sequel' to OutRunners)  The younger generation couldn't care less about the technical challenges of it, or the enjoyment of the coding, and figuring things out properly.  They see programming as a job, nothing more, nothing less.

I missed most of the 80s games, I was too young at the time to play the majority of them, that didn't matter.  I have no particular affection for 99% of the things I've emulated, and the ones I did care about (due to playing home ports mostly) were already emulated, yet I still enjoyed working on the project.  The same could be said for most of the good developers (if you only care about a single game / system contributions tend to be limited, and experience gained more limited)  People who have 'favourite games' and only care about their 'favourite games' were probably never going to make good developers in the first place, you can't be a good MAME developer with such narrow vision.

[Bootay]

I would also agree that the vast majority of the games that I grew up with and would want in my cab are already emulated quite well in Mame. (With the exception of Dragons Lair and Space Ace...I have a love for these games even though some would argue their controls are sucky) I could really live without most of the 3D games as I can play the majority of them on the PS1 or PS2. Sure some could argue that I could pick up Namco Museum on PS1 and play Pac-Man too..but my point? I am old, when I think arcade I think Donkey Kong and Pac-Man and heck, even Mortal Kombat and Killer Instinct. I don't think Tekken and Soul Calibur. These games I played on the consoles and not so much at the arcades. This is just me though. I understand we have younger people here who would love to get Tekken running arcade perfect with 3D acceleration. But me? I am happy with what has been achieved so far. I have been using Mame since there were only 20 games emulated (middle 1990s I think). I never would have guessed that the project would still be going on today. I am glad that it is though. And I am still waiting for Dragons Lair and Space Ace. 

(I know Daphne emulates these games but I prefer 1 emulator instead of multiple ones, plus I believe Daphne only does re-encoded mpegs from the DVDs and I would rather see a straight arcade LD rip like MAME did with Cliff Hanger)