David Haywood's Homepage
MAME work and other stuff
December 10, 2011 Haze Categories: General News. 52 Comments on Positive MAME memories of 2011?

So, we’re a third of the way through December now, which means the close of another year is drawing near and I was wondering what positive memories people have of MAME progress in 2011.

For me the year started out with a bang, there was a flurry of activity surrounding the Seibu COP emulation, and of course the emulation of the sound on many Toaplan 2 games (Fix Eight, Dogyuun, V-Five etc.) but after that the early promise really faded away. SSV, STV and Megatech improvements were all piled high with the January stuff, as was the initial adding of the Pinball sets setting the ball rolling for later mechanical additions.

In terms of overall MAME core changes it’s hard to see any real overall benefits, there were lots of driver code cleanups, but in the end many of them simply lead to more broken titles, and as a result we’ll likely be leaving the year with a number of significant regressions on drivers which worked fine when we entered it. On the flip side some long standing regressions have been fixed, but when an equal amount seems to be getting broken it’s hard to really feel progress. The majority of real core improvements have come in areas more heavily used by MESS, such as greatly improved usability of the softlists.

Kale has shown a few glimpses of brilliance in fixing long standing problems like the Riding Fight sound he recently demonstrated, improvements to the Nichibutsu text layer protection simulation, and support for missing effects which had gone unnoticed there, but by and large the majority of his noteworthy progress has also been in MESS (Saturn + various Japanese computers)

In terms of added games we’ve seen a bunch of Korean titles such as Candy Puzzle, Penfan Girls, Mr. Kicker and Baryon, the last of those being the Korean shooter from Semicom which I added, the first two being Eolith games which are sadly still silent due to the unemulated QDSP sound system they use.

I guess the thing that really got everybody talking and showing an interest in MAME again was the Cave SH3 board emulation, but unfortunately that ended up being removed again. I’d still put it down as a positive memory tho because it came at a point where MAME was really starting to fade into obscurity, and the only comments about it were rather aggressive ones from people who didn’t like the idea of the emulation of Mechanical machines and recent additions related to them (despite the first step of this process being added in January with the Pinball sets)

A handful of rare things have shown up, Taito’s Galactica, Mama Top’s Birdiy, and some really interesting prototype revisions of already supported NeoGeo games, but again compared to other years nothing that knocks your socks off. I was meant to have one such thing to display here, a unique, groundbreaking title from the 80s, but due to circumstances beyond my control that’s going to have to wait until next year at the earliest now. I guess Galaxy Game *could* be considered one such title, the first arcade game with a ROM / ‘CPU’ based program and all which from a purely historical and technical point of view is very significant but it’s cultural value is almost nothing.

Namco System 22 got some attention, bringing a number of titles closer to a state that could *really* be considered working, but the main issue people still call the driver up on is still present, the pesky Time Crisis helicopter.

There was a lot of work on the HLSL rendering, which for me is no more than a novelty, some people like it however, so it’s worthy of a mention.

From a coding point of view, the main codebase has been more stable in 2011, which makes it easier to pick up and code for without having to relearn everything. For newcomers it gives a better chance of learning how things work. That’s a positive, but the lack of newcomers even with a more stable codebase is still a negative.

I think the negatives probably still outweigh the positives for this year, and I’m starting to get the impression that even within the team it’s more than just me who thinks current policies are bizarre. I see a recent CPS1 submission enabled the CPS changer sets, which for each set is just a single line addition, no real cost and a nice benefit for users / developers alike, they were promptly disabled again because they’re not true arcades.

Maybe 2011 has one or two final surprises left yet, and maybe 2012 will start all guns blazing just like this year did, there’s certainly no shortage of things left to do, or things that could be done easily I can’t really say I’m optimistic at this point tho, there just seems to be a complete lack of talking points, and a lack of progress of the magnitude needed to inspire people. I definitely have one last little surprise myself, *if* things work out, but I can’t promise that will be done for the end of this year..

Does anybody feel there is significant progress in 2011 that I’ve missed? Any personal highlights from this year that you’d like to talk about? Maybe you’re overjoyed by progress in something I haven’t mentioned? (Return of Lady Frog sound?, Support for various Gambling / Medal video based games etc.) I’d be interested in hearing thoughts on this year from other people. We’re talking about everything after the MAME 0.141 release, so maybe Xidy making available several more of the early Exidy ROMs (again in January) is something you consider a highlight (even if it isn’t related to MAME progress directly), I certainly wish more companies would come forward and do that!


You can follow any responses to this entry through the RSS 2.0 feed.

The PSX drivers became usable, now that the CDROM has been supported! I think there were also updates to the PSX music drivers that were ported from pSX.

I personally thought the HLSL stuff was the most interesting. It could be used to ‘show’ what the games originally looked like. If anyone were to get into the detail of what monitors were usually associated with particular pieces of hardware. Right now it seems to be up to the end user how it looks. There should be at least a good default for each game. Unfortunatly at this point so many boards have been ripped out of their cabs it may be a hopeless task…

Well again the PSX driver is more MESS specific than anything else, at least the CD Support.. that said most games still don’t actually work, either crashing or having no inputs, also the driver regressed pretty badly with regards to some quads being drawn in a fairly recent update..

For HLSL I doubt there is a huge difference in the actual monitors used back in the day, I’d hazard a guess in most cases it was simply whatever they could source at any given time. In some sense it’s neat, but the actual effect still seems a bit artificial to me, it’s a simulation of perfect imperfection when the real hardware had many unaccountable variances often simply based on what was being displayed

I think your being a bit harsh as regards HLSL…

Compared to the previous filters in MAME, it’s a revelation, for the first time in 6yrs, since i ditched my old 19′ Mitsubishi Diamond Pro CRT, i like the way the games ‘look’ on screen…because they look like they used to on my old CRT, which was a close as possible to the way they looked in the arcade.

For all their advantages, Modern LCD’s suck donkey balls for Retrogaming the 8 & 16bit era.

I just wish some clever person would port it to work on BSNES & the Saturn Emulator SSF.

2011 to disastrous shell mameui
Completely killed by the speed and added a bunch of other problems.

Waiting for that finished the mega-TECH but never saw the success = (

Well, as I’ve said.. HLSL is very much ‘each to their own’

For some systems it really helps, for quite a few things in MESS it’s near essential to have it working properly because they *rely* on the artefacting for additional colours etc. In general tho, I have a big old TV in the other room, and no matter how HLSL is tuned it doesn’t ever really feel like that display to me. (Maybe because it’s PAL and all of this work is based around NTSC standards, but really it just seems to be something else, the way the screen ‘warps’ based on the brightness of what gets displayed doesn’t happen at all, nor the brighter than bright type effects you get, which are more typically associated with vector monitors but still seem to happen somewhat with old TVs and high contrast images) Some people like it, personally I keep it turned off unless a system *requires* it. I don’t think I’m being harsh about it, I mentioned it in the list of positives at least, even if it’s not for me ;-)

Megatech can be finished when somebody gives a perfect recording of exact behavior when the timers run out etc. Right now I don’t know if to reset the systems, if there should be flashing and/or additional sounds to alert you that a new coin is needed soon, I know nothing, so I can’t progress.

Yes MAMEUI has fallen apart in 2011, but I was looking for positives, that’s without doubt one of the huge negatives of the year, but I already have plenty of those ;-)

I wish close this your Mame and Mess year report with sweet chinese song :http://www.youtube.com/watch?v=DuHMCFYIC9E&feature=related, afetr all i think who love video games, love oriental world. Thanks for efforts Haze.

Well, as long as development don’t completely stop, I’m quite happy. Almost all games that I used to play in the arcades are already there, so I see anything new as a very welcome bonus (obviously I’ll be watching every new “U” release, ’cause I became kind of addicted).

I am very grateful to all of you Mame developers. No matter how much misunderstanding there is between yourselves (developers), I’ll always respect and admire you all. I specially admire the passion that you (Haze) are showing towards Mame and Mess future. But as an outsider, I don’t have enough information to position my self about it or pick sides. I can just wait and hope for the best.

Sh3 emulation was an incredible surprise, mainly because I have never seen these games before (and I love shmups). I compiled a tinyMame just for it (so I don’t have to add it back to every new Mame version) ’cause I really liked it.

The pinball and fruit machine roms addition gives me a good perspective of what can become the artwork system (I really hope someone grab this for further development).

The main thing I really long to see in Mame, are the laserdisc games (Dragon’s Lair, Space Ace, Mad Dog, etc). Hell, how much more Digital Leisure can milk these games? Everyone that like these games already have their DVDs and Blurays. What else can they do now? A new remaster of Dragon’s Lair in a next gen media with even greater resolution? Shit, I would probably buy it (but I think they could have done way better with the Bluray). These are the only games that I played in the arcades that aren’t already “mamed”. Yes, I know Daphne and Singe, but these aren’t as reliable or precise as Mame.

Other things that I would appreciate, are the proper or perfected emulation for Cps3, Igs, M2, M3, x86 systems, etc. But you know what… you guys already gave us more than I I thought would be possible.


I think the Digital Leisure excuse for MAME not supporting DL is just that.

MESS already quite happily supports several console versions (or did, the SegaCD version seems to have broken at some point) and it’s essentially the same game, just with worse video quality.

The LD video quality is also going to be substantially worse than the material they’re selling today too, and in MAME CHD format, at a far large file size.

It’s not like they’re still selling the actual arcade version, whatever they’re selling today has been reworked to simply run on a DVD player (the game logic really is that simple..)

LD emulation has basically just stalled, the rips that MAME has are of questionable quality (see the MT bug reports) and very few people actually built the setup to rip them. I think any other excuse you’re given is, as I’ve said at the start, simply that, an excuse. Notice how the preliminary Dragon’s Lair support is still there, and always has been there. If there was an official request to not emulate it (like with the Cave games) that would have been removed. Namco are still selling rehashes of Pacman every year and that doesn’t get stripped out because the arcade, in it’s original form is a different product.

I agree, HLSL was the high point. Or at least the high point that hasn’t been since removed. I spent some years repairing and calibrating monitors and had a hella lot of fun playing with it. I think it’s way more authentic than anything we’ve seen yet.

But like you said, in the grand scheme it doesn’t seem that important.

Haze, you said “the way the screen ‘warps’ based on the brightness of what gets displayed doesn’t happen at all”. That’s called high voltage regulation and should be easy enough to emulate.

Showing white puts a real load on the CRT power supply/FBT/HOT which reduces the voltage needed to scan…I think it mostly shrinks the horizontal but the entire picture can be affected, depending on how weak the regulation is.

I suppose you could emulate the entire CRT monitor, flyback transformer and all, wow!

And what about my idea to use IR impulses for the sound? Someone will do it one day.

Ahh, I guess you know more of the technical side of it than I do then ;-)

but yeah.. that.. right now it just feels more like I’m looking at the image through a broken pair of glasses, rather than having something real and dynamic, much like every image enhancement filter.

It probably is one of the true highs of the year tho, but I think that speaks more about the worrying lack of actual tangible emulation improvements.

Sorry to infiltrate a different post, but I only just thought more about this recently, earlier this year you said in regards to Fantasy Zone II “By ‘Won’t be supported in MAME’ I mean it won’t be supported in MAME using the PS2 .pak files. The extracted data has been used on other standard System 16b boards (with appropriate RAM mod applied) and sold before, so that might end up in there one day in that form.”
Can you point me in the direction of who has done that (or any web post about it)? I would like to do it myself :) Cheers!

PCBs have popped up on ebay and the like now and again, I don’t have any more specific details tho I’m afraid

Other great work in 2011: hap’s fixes for various drivers, your work on the Data East video hardware, Phil Bennet’s fixes to mk4. And, as you already wrote, Kale’s incredibile improvements.
Sadly I’ll remember 2011 mostly for the dwindling involvement of Aaron Giles.

I’d like to see Aaron more involved if he actually finished things he’d started properly.

Unfortunately you can trace a fair number of major regressions (both performance, and general bugs) back to things he started, but never finished.

That would be fine if he fixed them in a timely manner, but every time he popped up it seemed to be with a new set of changes, to something else, and in most cases I honestly don’t feel there was even a pressing need for the change.

In that sense Kale have brought about a far greater stability, and at least put effort into finishing off some of those changes (or at least updating drivers to more current standards)

As a driver developer you can get away with jumping from one task to another because the changes you make only affect the driver in question. As a core developer it tends not to work so well. If you kill the performance of a MAME function, or introduce undesirable behaviour across the entire core it can affect everything.

+ the whole problems with MAMEUI

This is also true of cases where you make mass changes and fail to take into account the code in forked projects, maintaining a fork (or something like MESS) is much easier when there is stability.

It would definitely be nice to see more contributions from Aaron again, but, especially if working in a limited capacity I think he needs to carefully consider the scope and impact of any changes he makes, especially if he’s not going to be able to revisit them for a significant amount of time. He’s probably best suited to actual driver work, maybe working out how to re-read those LD games without problems, or support for more of them. As a point of reference, when Nicola comes back to make contributions they’re always really good because they’re usually just small snippets of code to solve a long standing problem, with the direct result of real emulation improvements.

As for my contributions, I didn’t really want to mention the Data East stuff too much, it was good in terms of code cleanup and it fixed some bugs, but it also introduced some bugs*, plus it’s still not quite finished yet, several drivers couldn’t be converted over due to the convoluted way the sprite handling is stuffed into the video code at the moment. The Seta cleanup was better, but offsets and such still need to be verified, plus the alternate handling hacks still bug me a bit (maybe there is a pin to switch modes?)

* at least I’m guessing they’re bugs, not ‘non-bugs’

About Aaron’s work: I agree with you that in the last years his contributions have gotten a bit less “finished”: he updates the core and just seems interested to show how to update the drivers accordingly, letting others the chore of doing it. Sadly this has led to deprecat.h lasting for years, lots of CPU cores and device still being old style, ecc.. And obviouly the bug you mention, even though I really wouldn’t be sad if frame-skipping would be removed from the core. I consider MameUI rotting a positive development. It’s really time to let separated front-end gain users.
The thing is, under Nicola and you the core hadn’t really gotten updated and it had gotten old and patched. So Aaron had the enormous task of bringing it to modern standards. Every major update to the core has been done by him, apart perhaps the discrete sound system (by couriersud and Derrick) e some memory and ui changes (by OG). Prior to those, who was the last dev to touch the core not named Aaron? Phil Stroffolino with pdrawgfx? It more dev would help with the core, Aaron’s task wouldn’t be so impossible. But, as you said, there are few devs active nowadays. My final question: is it better to let the core rot or is it better to update what you can and deal with bugs and problems introduced later? I feel there’s no clear-cut answer, but I tend to think the latter option is slightly better.

Well, the most important thing is that the core serves as a platform for the rest of the emulation.

There have been many updates over the years, and Aaron has been involved in many of them, but there definitely reached a point where the core was capable of handling the needs of nearly everything, and the changes being made were entirely of the refactoring variety, rather than addressing the cases where it wasn’t capable. Rewriting basic things like the rom loading, which have served the project well for years aren’t going to actually help emulate anything.

If you contrast this to earlier changes, such as the blit time rotation, and simplification of the overall video functions (no need to manually mark colours as used, areas of the screen as dirty etc.) then the benefits in writing drivers are closer to non-existent.

C++, and the device system were properly beneficial, however having multiple types of devices, and declaring a brand new system legacy within months of it being introduced (and a significant number of drivers being converted over) also did nothing but fragment standards over the codebase, and introduce confusion for people wanting to code.

One of the reasons I’ve pushed for tighter MESS integration is because a lot of the systems there highlight areas where the core could do with being improved, areas which actually test functionality, and expose weaknesses in it to address due to a natural requirement for far greater accuracy, and good performance code. You *can’t* survive on hacks in the drivers, which is what parts of the MAME codebase survive on now. Some of the recent changes are nice in principal, but come at such a performance cost using them simply isn’t practical, especially now we’re only seeing marginal improvements in CPU speeds and such.

Actually the recent disabling of co-threads (introduced to improve the ability to do certain things the core doesn’t otherwise allow) could be seen as a further regression in functionality, still can’t really understand that decision myself, even if it wasn’t being used yet.

Simply rewriting something doesn’t stop rot, you have to have specific test cases in mind, to test for both functionality and performance of new core code while being mindful of how it scales from older systems right through to newer ones.

With the CaveSH3 stuff for example I was finding about 20% of the CPU time was being spent in the memory system, that was closer to 50-60% when the blitter reads were being channelled through it too. For things like Saturn you still see the perfect scheduling of the CPUs can drag framerates down to nothing in places, it almost calls for a core system to be in place which can cause MAME to treat 2 paired CPUs (as you have in that case) as an atomic unit as far as the overall scheduling is concerned, a way of treating the 2 cpus as if they were 1 to the rest of the system. You need interesting new designs, ways of handling things etc.

The actual core UI, that could do with work too, why is it all still on the same thread as the emulation? The interface becomes unresponsive when the emulation does. Why do we still have a single windowed interface with no way to offer a consistent display, using (by default) oversized kiddie like fonts when what you need to do is present technical information, and fit a good amount of data on screen for usability purposes? (again especially important to MESS) The artwork system makes my eyes bleed, I was looking for good ways to do the fruit machine stuff with it, but some of the existing code to handle it there is so badly thought out (from the point of view of say a high performance game engine) it makes me wonder what people were smoking when it was coded ;-)

The debugger? Why isn’t it in some way more modular? Why can’t drivers specify their own debugger windows with custom views for things like VDPs? Stepping backwards?

Some kind of video system recompiler? different targets, C / HLSL etc.?

Better core level support for systems which change resolution per-scanline? (AFAIK they only ever do horizontal) but the concept of a core bitmap having an x and y dimension is out of line with how real video output works (no fixed x resolution) right now drivers have to manually upscale everything to fit the most common divisor. Of course there are systems which mix different layers at different resolutions (PC-FX) so again it’s by no means an easy problem to find a ‘one size fits all’ solution to, but there probably should be some core level helper support to build on at least.

Real core changes are *HARD*, ones that actually bring real improvements, but if you’re truly talking about stopping rot and making real progress you need people with an actual vision and ability to make changes which are actually useful, simply rewriting bits of code to offer the same functionality as before isn’t that.

Now don’t get me wrong, Aaron HAS made such changes in the past, as have others, and if we still had the core of 10 years ago we’d be in REAL trouble right now, but an awful lot of what’s been going on for the past 2-3 years has amounted to little but code churn in terms of bringing about needed improvements and often spurred on change of questionable value (I’m really not a fan of the new banking for example) At the same time, lack of a tighter MESS integration has IMHO made some issues more ignorable than they should be.

Also I think for some of these decpract.h removals Kale has he’s lacking a real Vision too, duplicating the same bits of code in drivers, when really he could have replaced the interrupt hack with a simple set of lines in the machine structure which takes a scanline and interrupt level, just as the old ones did but with 2-3 lines instead of 1. For the basic use cases he’s replacing he’s actually making the code more ugly, it’s not like these are complex systems with dynamic pixel clocks which would require manual code. Of course, that’s just a difference of opinion, but I would definitely have added a core macro for those cases.

It’s not just MAME of course, lots of recent software has suffered in similar ways over recent years. For all the praise Windows 7 gets, it’s a TERRIBLE operating system when it comes to simply things like file management, simple operations which should be fast can take forever compared to something like Total Commander, and even XP handled them just fine.. Aren’t these meant to be the very things an OS does? the basic fundamentals? I think the only reason it gets praise is because it’s not as slow as Vista (even if it inherits many of the problems) and most people don’t even use an OS as an OS anymore just a way to access the internet. Windows 8 looks even worse, but we’ll see.

Just my opinions of course.

Why don’t you, Haze, help with the core? Some things you said seems to be really simple.

A thing that looks to take forever to be performed is the drawing. Following the code, it seems to pass through 6 differents systems to finally OSD draw.

I don’t think i’d make a good core dev, my main interests are in figuring things out and understanding the way the software/hardware interact, I’d lose interest very quickly doing core work.

Even during the time I was in charge core work was delegated out, much as it is now under Kale, who is also somebody who prefers to work on drivers.

The advantage of this is obviously the project co-ordinator has an idea of what the driver authors actively working on emulating things actually need, the disadvantage is that there isn’t always somebody to actually do the core work.

As for drawing, it’s never something I’ve found to be a major bottleneck, at least not unless you’re talking really low end systems. Given the lack of real advances in technology in recent years there is very little difference (in terms of raw CPU power) between a system you can pick up for nothing, and a top end, so I’m not overly concerned about that, MAME was never really meant to be an application for mobile platforms, and I don’t think you can buy a desktop where the draw / blit code is a bottleneck.

Tilemap management code and such for cases with RAM based tiles is sub-optimal, yes, maybe for those cases it would help to have tightly optimized custom drawing routines for drawing the likes of ROZ layers without the tilemap cache, but again it’s an area I think you’d need some kind of recompiler to generate the specific code you need for each driver in order to do it efficiently, otherwise you’re going to be making a lot of per-pixel function calls to look up tiles etc. so doing it in a generic way would add a lot of overhead when in reality you’d want the cleanest, most minimal code paths for each use case, which could differ wildly between systems. Obviously the core can’t have every potential use case hardcoded in, but if it could generate the most efficient possible code for any specified use case that would be handy.
Of course generated code doesn’t agree with things like the Google Native Platform stuff, that port already had to rip out the CPU recompilers, and worrying I do think come a couple of Windows versions we may find ourselves forced to work in such environments.

Actually I believe there are some problems with clipping calculations and the current zooming functions, I noticed when working on Data East that manually checking to see if a sprite was on a line gave me a big speedup, which would indicate that the clipping code in the core drawing functions might not be doing their job.

I guess the rest of the draw code does become a problem for systems with complex artwork layouts, I already mentioned that, although my main concern there wasn’t the depth of the calls, but the way they seemed to be (at first glance at least) allocating and freeing huge amounts of memory on every single draw call rather than caching things up front.

Yeah, drawing is not really a problem, though i remember the changes log file notifying the drawing system been changing to make the things look “nicer” (sorry I can’t really precise the things “said” in the file).

Anyway, people saying HLSL was the better thing for this year, but, in my opinion, HLSL seems to go against the point of the project. Well, it is good at some point, if “shading” with hardware, why not processing and drawing vertex with it?

BTW, why some drivers losing speed every new MAME release? (Talking most precisely about Neogeo)
Is the driver itself suffering changes? Or these are things related to core?

in 2011 for arcade emulation, don’t forget the model 3 driver emulation and Demul has fully emulated the Naomi and atomiswave system

There’s no reason the NeoGeo should have seen any significant slowdown since the last time it was rewritten to be fully scanline based, which was quite a long time ago now, I’d be rather surprised to hear otherwise.

and as for model 3 / Demul … I was asking specifically about MAME..

What others do is irrelevant to that, aside from showing that MAME is no longer the emulator offering the pioneering new progress everybody talks about. MAME hasn’t even managed Model 2 at a level comparable to the standalone yet.

In other words, I can’t really count those as positives of MAME development ;-)

Re: UltimateMame I might put out a u3 build, although I still don’t think there have been any worthwhile improvements in MAME/MESS the last few updates, just a lot of things broken and some makeshift patches to fix other things that broke in the previous update..

Actual progress on getting an official ‘complete’ target seems to have died again, predictably.

It would be nice to end / start the year with a bang, and some kind of official endorsement of MESS with combined, as well as the single MAME/MESS binaries from the official MAME site would be something along those lines, but it doesn’t look likely, which is disappointing. Instead they’d rather just create their own hacks of PCE roms and shove them in with a timer…..

2 questions:

– How come Wyvern F-0 still hasn’t been added?

– Any chance of more Cybertank progress?

Thanks Haze for your continued support.

Wyvern F-0 is being hoarded by Guru due to some payment dispute with the Dumping Union guys. It’s bullshit, Guru has offered various developers the chances to work on it as long as they hoard it too. All the reputable ones have turned him down. It will also need decapping FWIW, it’s a Taito MCU protected game, and their protections are usually complex with subtle behaviors.

It will probably end up showing up anyway for next to nothing in Italy like half the rare Taito games and they’ll all just look stupid anyway. The main value of the game is the unique cabinet anyway, the game has a pseudo 3d effect with glass+mirrors to project the sprites higher than the ground. The board is worthless without the cab, the cab is worthless without the board.

Cybertank, I was rather curious as to why the inputs don’t work properly ingame, without them working it’s quite hard to develop. Could do with some game code analysis.

I was afraid some dispute was going on, it has been so long since the game was obtained. Like you said it will probably show up somewhere else at this rate.


I don’t know if raiden 2 can be done in 2012…

Raiden 2/DX emulation will happen sometime within the next decade or two, or the day after the world ends in late 2012 etc. :)

I Hope see in 2012 all the Pinball and Fruit Games Emulated.


Never happen……

Even Gals Panic 2 and for now…………
I don´t see nothing from 2010-2011.

Only 2009 have some suprises, like Gals Panic 3, Oedo Fight and others.

you can play raiden 2 and dx with psx version at moment.Its work fine on one between ssspsx or psx1.13 emulators.I don’t remember exactly that now.

Raiden DX is also one of the few PSX titles that works pretty well on the MESS driver (or the UltimateMAME build) you can even play with a vertical screen mode just like the arcade if you tell MAME to rotate the image and set the game to TATE mode. The only thing you’re missing is the ugly CG intro ;)

Itazura Tenshi was a pleasant end of 2011 surprise.

Yeah, and it even looks like an interesting game, unique concept… although from the dates in the driver and everything was actually done in August and held back until now ;-) Guess that’s another reason the Christmas / Jan period is busier :-p

hey, the UltimateMAME build was a positive step in the right direction 2011 IMHO

Possibly, it depends how much traction it gains really.

Kale has hinted that after 0.145 the actual working trees might be merged (so MAME private and MESS SVN just become actual mirrors of each other)

Whether that means the source trees will be distributed complete (hopefully) I don’t know, but if so it would be a step in the right direction, and make it possible to have an official ‘complete’ target in the source like the one demonstrated in my last UltimateMAME update.

I notice Olivier has been doing most of his work exclusively in the MESS tree (where all files are present) already.

Mr. Haze, can you say something about PGM emulation? i was waiting for more progress on this but i guess is a little “on hold” for now, i mean, i still interested better Martial Master emulation, and in Demon Front, Spectral vs. Generation or DoDonPachi 2, even if everybody said DDP2 suck ass because it wasn’t developed by Cave but by IGS itself.

And yeah i guess the most interesting thing of MAME in 2011 was the latest SH-based Cave games emulated.

I for one am a bit disappointed by the lack of progression with MAME this year. Part of that I suppose comes down to me being frustrated at the lack of direct talks with whoever’s doing the developing, which is why I appreciate it when Haze makes informative posts and works on something that’s notable.
The lack of progress with the Raiden II driver is also a shame, though I’ve seen comments and some stuff get added to it in the GIT every so often it’s like everyone’s looking and then walking around it. If it was made by Cave I bet it’d be getting the attention it deserved.

Baryon: Future Assault was actually quite interesting for me. I found that the original PC game was also a Korean-made game going by the staff credits in the demo’s readme but since the arcade game just loops over and over and there’s no credits one can’t really find out whether the arcade port was made by the same people or not.

I should also mention there is/was a user on the Shmups Forum who got MAME working something close to natively under OSX as a native Xcode 4 project. The download link to it got taken off and the source is ‘available on request’ last I heard, but this would be interesting to see if development on this could continue and we got a fully native arcade emulator that made the most of the Apple software and hardware.


Better Martial Masters emulation? It should be perfect. It’s running the original ARM code and everything. The only thing that might not be 100% is the sound, but to be honest it doesn’t sound that bad either.

Anything by IGS newer than DDP2 can not be emulated properly. That includes Demon Front and Spectral Vs. Generation. The internal ROMs are completely protected (in hardware) against reading out, you’d have to rewrite your own ARM protection code, at which point it isn’t *emulated* at all. DDP2 should be possible, but I’ll know more on that soon. (actually I wanted that one done for Xmas, but simply haven’t had the time)


Raiden 2 is Raiden 2, it’s not as if Seibu aren’t as famous as Cave, they’re probably moreso, and it’s probably the single most requested title of the last 10 years. The simple fact is it’s very well protected. If the Cave games had been as well protected they wouldn’t be emulated (they had ample opportunity, especially with the single board PGM hardware, but failed to take advantage) People look, and then due to the sheer number of factors and parts involved in the protection give up rather quickly.

I don’t know much about native Mac development, but I thought most people used SDLMame there now, and found it to be better than the old ‘native’ MacMame ports anyway?

Haze what about Gals Panic 2?
Have the same proteccion like IGS games or something like that?
And the picture girls in game is very difficult to get them too?
I can´t wait to get that game playable at onces. It’s only this game won´t work i think :/

Gals Panic 2 protection probably isn’t that hard.. just nobody really cares about it, including me. It’s just a Kaneko MCU that does data move/copy operations and possibly some decompression.

Ok, but people shouldn´t care…. No!
There are a lot of people asking about this.
You don´t see because no one ask.
It’s a game man common.
I wish to play this game.
I was 6 when i saw on my country,
But i never see the game at all only Arcade Machine shut down.
And besides i can´t play in that age + 16 only :/


Thanks for your info, speaking of Martial Masters, no, the game isn’t perfect, i heard and there are videos confirming that some combos cannot be performed on MAME, some timing issues or something like that, a problem with the gameplay itself.

I’m doubtful of such issues, afaik the protection isn’t even involved in that, it mainly deals with the CPU AI.

Maybe just different game revisions.

(or maybe one of the different region settings.. maybe the game attempts to simplify the combos in some regions?)

How do we motivate LM to work further on the Votrax SC-01 speech synthesizer?

I don’t know if motivation is the problem.. or if he just has his hands in too many baskets to actually finish anything ;-)

Double Wings is another good shooting game,no one has fixed yet…

I still swear you must be somebody trying to sell MAME. People have written entire reviews of Double Wings without even noticing the rare protection fault(s) that still exist as you have to really play the game in a specific way to still break it. (Let bosses time out instead of actually trying to kill them) You just seem to report ‘bugs’ or ‘things to fix’ all the time, which you appear to have no personal connection or real knowledge of, as if they’re coming from somebody else.

I remember being really excited about Heavy Unit but I’m not sure if that was at the end of 2010 or beginning of 11.
Also QSound support being worked at at the start of the year but I don’t think anything came out of that?
Except those I was excited for the new Model 3 Supermodel emulator but I guess that’s not really mame.

Heavy Unit progress was from the very end of 2010, can’t remember if the first public build with it was the start of 2011 tho..

By continuing to use the site, you agree to the use of cookies. more information

The cookie settings on this website are set to "allow cookies" to give you the best browsing experience possible. If you continue to use this website without changing your cookie settings or you click "Accept" below then you are consenting to this.