Mame 0 153 Rom Set Items
Contents. Release Date MAME 0.153 was released on 7 April 2014. Contributors The known contributors for this version are, in alphabetical order:.
Specific Contributions The known contributions for this version are, in the order specified in the whatsnew:., and completed a massive architectural change: There is no longer a single, global palette. Instead, palettes are now devices, and drivers and devices can have as many independent palettes as needed.
Apr 09, 2014 Mame and MESS 0.153 is out. And I just the got 0.152 rom set and chd's. Assuming you got.152 from the same place you get.153. MAME 2014 Reference Set (Complete MAME 0.159) Console ROM Reference Sets: Atari 2600 VCS 2015 Reference Set (Complete ROM Hunter v11).
This update required extensive changes to many parts of the core, UI, OSD layers, drivers and devices. and updated the core so there is no longer a single, global array of decoded graphics. Decoded graphics are now owned by devices using a new device interface, called, unsurprisingly, devicegfxinterface. A 'gfxdecode' device (which does nothing but inherit this interface) provides the old, driver-level decoding functionality for drivers with video emulation not based on devices and for devices that haven't been updated to the new interface (which is currently most of them). Also, graphics layouts can now reference shared RAM regions by tag the same way as ROM regions, eliminating the need to set up RAM-based graphics manually (note that drivers/devices are still responsible for marking gfx elements dirty when the RAM they are decoded from is written to)., and did a huge effort to modernized core and drivers, almost every file has been touched. fixed a bug preventing memory display of INT8/UINT8 in debugger.
added missing interfaces to netlist. Mame -debug pongf now does not crash any longer and displays some state. Still more work needed. Further improved the netlist state saving.
updated tx0.c: Modernized cpu core. removed legacy code in netlist and separated out 74153 and 74107(A).
updated superfx.c: Modernized cpu core. updated the Truco-Tron driver: Reversed and documented almost all PCB's ICs. added a '-u' (for unix) flag to srcclean to produce unix-style endings. updated spc700.c: Modernized cpu core. added a dry-run option (-d) to srcclean. Know what you are doing.
updated mc68901.c: use aer register to select which transition to generate an interrupt on and added preliminary transmit and receive support. updated se3208.c: Modernized cpu core. provided an alignment of netlist with MAME standards. Inputs to netlist are now registered using by MCFG. as mame subdevices and exhibit the usual writeXX/readXX methods including a PORTCHANGED method. This allows direct linking to inputs. updated sc61860.c: Modernized cpu core.
updated ssp1601.c: Modernized cpu core. updated diserial.c: don't transmit anything when driven by an external clock and nothing has been queued. Call setdataframe before calling setrcvrate as otherwise it acccesses msynchronous before it is set leading to random behaviour.
Changed parity & stop bits to an enum (you can now pass in 1.5). Updated the uarts that were testing for 1.5 stop bits to pass that in, but there are probably others & 1.5 stop bits is converted to 2 by diserial. The 68681 requires stop bits to be specified in clocks, so this will change in the future.
Replaced synchronous flag with start bit count, as some uarts can use a start bit in synchronous mode & that whether there is a start bit is all the flag is currently controlling. Updated rs232 terminal to allow startbits, stop bits 1.5 to be specified (although that is currently not supported by diserial) and individual transmit and receive baud rates. updated scmp.c: Modernized cpu core.
updated atarist.c: push all gpio values to allow interrupts to work. updated unsp.c: Modernized cpu core. provided AVR8 improvements:. Added ELPM instructions.
Added fuse bits macros. Added reset logic to decide initial program counter based on fuse bits configuration. Added initial support for ATMEGA1280 and ATMEGA2560. Use register names in the disassembly of IN and OUT instructions. updated tms32010.c: Modernized cpu core. and updated segaorun.c: Added motor and lamp outputs. updated via6522.c: Converted to use devcb2.
Removed alot of read handlers where the values could more easily (and often were being) pushed. CA1 is input only, so he has removed the output callback. The lisa via hookup has been changed as it had functions named as CA2/CB2 hooked up to CA1/CB1 and CA1 output no longer exists, it appears to have been wrong since before the code was merged with MAME. Added support for shift out free running at t2 rate, cb1 output now lasts for the specified number of clocks and the signal goes low at the end. updated tms32025.c: Modernized cpu core. enhanced the netlist parser and cleaned pong.c. Also added a folder nlexamples which contains standalone netlist examples.
updated tms32051.c: Modernized cpu core. updated mquake.c: Added notes and NODUMP placeholder for I/O MCU. updated the SDL2.0 port: Fix relative mouse motion. This makes Missile Attack playable again with the Logitech marble trackball. fixed PNP model and added test netlist (bjt.c). updated tms7000.c: Modernized cpu core.
made Pong have complete emulation of discrete logic for runQ and SRST signals. Rewrote switch2 device to be based on resistances as well. provided AVR8 cpu changes.
Timers wip. Add TIMSK. reads - Refactor timertick calls. Remove all redundant register init statements.
We simply set all 0x200 registers to zero. Debugging timers. Better syntax for disassembly of RJMP destination address. Use register names in disasm of CBI, SBI, SBIC, SBIS instructions. Use register bit names in disasm of instructions SBI, SBIS, SBIC, CBI.
Mame4all Rom Set
WIP on avr8 EEPROM read/write support. updated the Netlist code:. fixed 7474 save state. remove a printf. simplified netlistlistt. increased reschedule count to 15 - diodes are dog-slow. made reschedule count a solver parameter.
simplified solver code. solver time now also resets.
updated the fdfdc device: Fix the FM sync detection. updated the n68681 device: modernized and killed config struct in favor of devcb2. provided a fix for pixel gaps between multi-screen games in HLSL mode. merged pet/vic20/c64/plus4/c128 user ports, the pinout is slightly different on each one but there are cards that are compatible with each of the different machines. Removed c64 version of vic1011 and moved protovision 4 player adapter from bus/c64 to bus/vic20. hooked up the 6551 to the plus 4 user port. added user port pinout from CBMII manual.
fixed commodore quick load by passing in the address space instead of using machine.firstcpu. added PET user port joystick adapter. updated segas32.c: Added alien3 and darkedge readmes. updated the Netlist code:. added missing parameter to analog input device. parser now parses NETMODEL.
fixed a bug in switching bjt implementation. and fixed noise period on AY8910. prevented clock bit to be deleted after CRC in FM. and used proper AY8910 mixing model for MSX.
Volumes set to match MSX hardware. provided BBC Micro keyboard updates/fixes:. added numeric keypad on Master series. fixed CAPS LOCK and SHIFT LOCK.
removed dipswitches from Master series. improved dipswitch descriptions.
provided a Naomi/Triforce documentation update. converted PIA6821 to DEVCB2, removed portar/portbr/ca1r/ca2r/cb1r/cb2r functions as they only returned what you'd written to the chip, only a couple of those were ever used (by audio/exidy.c). Removed some ca./cb. read callbacks that just returned what had already been pushed or always returned 0.
• Control attributes such as Friction, Bounciness, Damping, Mass, Velocity, Spin, and Torque. • Use the multi-threaded NVIDIA® PhysX® engine* to create static, dynamic, and kinematic rigid-body simulations directly in the Maya viewport, and gain the ability to match a runtime solution. 
As there are no longer any cb2 read callbacks that functionality has been removed. Reduced the number of callers to setainput by calling portaw & made the WRITE8HANDLER call that as well, instead of the other way round and having to lookup an address space that is never used. updated the Netlist code:. Added Ebers-Moll BJT model. Currently only for NPN. Fixed convergence issue in the solver.
Accuracy for Ebers-Moll must be better than 1e-5 to not generate nano-second spikes. Typical sound applications should be able to run with less since the spikes are not audible. updated tlcs90.c: Modernized cpu core. removed calls to mpu4state.state = machine.driverdata; by moving code into mpu4state. Removed runtime lookup of devices that were already looked up in mpu4state.
updated docastle.c: hook up slave CPU IRQ according to schematics. added a '-listdevices' option to nltool to list all available devices.
Simplified code. fixed Z80DART parity selection. updated pulsar.c: hooked up Z80DART correctly.
provided preliminary patches for Donggul Donggul Haerong (still doesn't work, but at least boots now). corrected his previous patch for Donggul Donggul Haerong, game still doesn't work. updated the i8251 core: as per the manual, disabling Tx mid-byte should wait for transmission to finish.
updated v810.c: Modernized cpu core. updated s2650.c: Replaced FLAG fake IO port with a DEVCB2 callback. Added writesense method and use it for pipbug serial port, Hunchback (Galaxian Hardware) and replace calls to setinputline(1 in lazercmd.c & meadows.c. updated exp85.c: Added an RS232 port instead of hard coding the serial terminal.
added preliminary sound to thunderlbl. Unsurprisingly seems to share the same hook-up as wiggie and superbon, the other thunderl hacks in the driver. Marked as IMPERFECTSOUND cause the music keeps restarting after a few seconds after inserting a coin. Seems fine during gameplay, though. fixed compiling with Visual Studio 2012. added the correct, half-swapped dump of cpupg5.11 to spyhunt/spyhuntp.
implemented near-to-proper reset for all netlist devices. Reset now works for improved matrix solving. added 82s153's from Ropeman (Roc 'n Rope bootleg) and also corrected the rom names and locations. In addition, corrected some of the names of the one United States Donkey Kong Jr. updated isbc.c: converted to use RS232 port instead of hard coded to using serial terminal.
removed some dead code and ambiguous functions in the netlist code. introduced the concept of 'logic families'.
Currently, standard devices default to TTL. Going forward, this enables the family type (e.g. LS, S, HCT.) to be a device parameter. Currently, no changes are introduced. This is architecture work. created a mame sound device for netlist. Fixed an issue causing wild oscillations.
Currently, pong has sample code for using the sound device. Uncomment 'TESTSOUND' in pong.c to hear a constant sound from a 7400 multivibrator. Sound is still WIP, but the proof of concept has been done. added music and save states to bmcbowl. added music and save states to koftball. updated the tms9927 core: Added VSYNC callback.
updated the SDL port: Temporarily restore compatibility with Qt versions below 4.7., and added proper internal rom to 'svg' parent set, it resets on coin now, need to investigate. updated the tms5501 core: Rewrote the emulation to use devcb2 and diserial.
updated flopimg: Added preliminary support for the 8-N-1 serial encoding scheme used on Compucolor II floppies. updated diserial: Added framing error detection. updated the Netlist code:. Working Ebers Moll model. That's a significant step ahead. Simple 2x2 RHS SPICE stamps now supported by two terminal devices. This was implicitly contained in the design, but setmat now shows how a two-terminal device fits into a SPICE approach.
Introduced direct solvers for net groups with 1 or 2 nets. Introduced specialized solvers for N=1,2,3,4,5 by using templates. Nice performance increase for BJTs. fixed bad graphics in donghaer. updated the n68681 core: Implement OP3 toggle on timer expire and input delta interrupts. updated h2680.c: moved contents of.inc into.c to removed #define pollution when including h6280.h. updated z80.c: converted c.
updated ata/ide: Added support for 8 bit data transfers using the set features command, instead of using memmask as ATA doesn't work that way. replaced some calls to machine.firstcpu with mmaincpu. replaced some calls to machine.firstcpu-pcbase with machine.describecontext. updated bankdev.c: don't change the current bank on reset, because another device might have already changed the bank when it was reset. updated clcd.c: Implemented MMU making it usable. Hooked up rtc, rs232, centronics and sound. LCD colors are matched up from this photograph.
The character rom is not dumped:. So a fake font is used which takes the unused upper case font in the kernel rom as a starting point and adds lower case characters, trying to make them match the available screen shots:.
There appears to be several bugs in the roms that affect the rtc. and added unknown encrypted Royal Card. This game is running on Dino 4 encrypted hardware, with a CPU+PLCC daughterboard. Decrypted program address + data, but code still jumps into $48xx range where there's no valid code. Decoded and partially decrypted the graphics set. updated the-n68681 device: make loopback operate at the current baud rate instead of instantaneously. Fixes dectalk DUART self-test.
updated z180.c: Modernized cpu core. updated tourvis.c: Correct 1 cart name and add newly dumped U4-43 BIOS. smf converted MSM58321 to DEVCB2, uses level logic rather than edge detection for strobes (as per datasheet), added nvram interface as there are some configuration bits to save, added support for 12 and 24 hour. It defaults to 12 hour, but you can configure it in the driver to default to 24H.
You can configure what year 0 is in the driver, which is used when setting the date at startup. updated sharc.c: Modernized cpu core. added SCREENRAWPARAMS to Express Raider, fixed screen size to 256x240. updated z8000.c: Modernized cpu core. updated the m6502 core: expose SYNC pin through devcb2 for more flexibility. updated tugboat.c: correct AY chip type and clock, some cleanups. and updated AICA: EFREG is mapped every other word, not every word.

Fixes memory trashing in all Naomi/DC/AW games. changed name from “VIC20 user port” to 'PET user port' as that is where it originated and reformatted the source file. updated qix.c: Let's not regenerate the entire palette every time we do a partial update. updated the mos6551 device: support Rockwell 6551s, which show different register contents from MOS/CSG parts on reset.: updated nemesis.c:. Hooked up RC filters to AY8910s. Fixed konamigt layout aspect ratio. Some corrections to address maps and inputs.
updated popeye.c: added RC filter (MT 3726), cleanups. updated the flipscreen routines: tilemap flipping is now calculated relative to the center of the visible area rather than to the total screen size, and the generic flip screen functions no longer modify the visible area.
These changes ensure that in most cases flipping should Just Work for drivers that use MCFGRAWPARAMs, games that have a programmable CRTC, and games that have raster effects, and should fix many longstanding flip/cocktail-related regressions. added Bingo (1983). PCB seems bootleg, but the game looks legit. Worked from the scratch a whole set of inputs and button-lamps support for this game.
Changed the poker41 description to Four in One Poker (as seen in the official brochure). Added game and technical notes. provided improvements to Unknown Royal Card on Dino4 hardware:. Fully decrypted the graphics set.
Added technical notes. added Bingo PCB layout and technical notes. fixed rallybik title screen color. updated toaplan1.c: remove sound reset hack, unify gfx decoding, various cleanups.
and redumped 317-0024 (FD1089B) used by Time Scanner. updated twincobr.c: fix flipscreen after core changes, cleanups. updated FD1089 documentation., and added skeleton drivers for some discrete games using ROMs, and added some documentation. provided centronics WIP using write lines and DEVCB2. Added an output latch and input buffer device to allow configuration in MCFG. Renamed centronics device 'printer' to 'image'.
FM-7 parallel port joystick works again. Added support for using PC LPT control lines as inputs (not tested). C64 geocable now just passes individual lines through, changed the strobe line to what is documented online (not tested). Converted MSX & exidy sorcerer parallel port DAC to use covox device instead of using a configuration switch (not tested). Sorcerer parallel port is more like the commodore user port, so should be converted to it's own slot where one option is a centronics card. updated Magic Card II: Override the wrong visible area and centered the game through a dedicated CRTC config interface.
converted SN76496 ready to DEVCB2. updated dkong.c, docastle.c, galaxian.c, mario.c, zaxxon.c: fix flipscreen after core changes. updated bagman.c, bublbobl.c, chance32.c, galaga.c, mappy.c, pacman.c: fix flipscreen after core changes. Also, try (again) to improve sprite-bullet alignment in bosco. added generic barcode reader emulation: the code accepts sequences of digits in UPC-A, EAN-13 and EAN-8 format, stores them in both byte format and in pixel format; it is up to the driver to implement the correct transfer protocol.
added UI menu to enter barcodes via keyboard. It only gets activated if the running machine has a barcode device.
and updated Truco-Tron: Identified the last unknown IC (U15). Added Technical notes. updated diimage: Added 'readonly' feature to the software list to denote write protected images.
created src emu bus rs232 & src emu bus midi directories and separated rs232 and midi devices, changed h89 to use an rs232 port to communicate with the serial terminal to instead of connecting it directly. added ability to add optional or required device/ioport/memoryregion/. And similar device finders to any device, and also created ability to set tag inside static configuration of device so you can reference it in driver machine configuration. converted i8251 to DEVCB2. provided Funworld driver updates:. Added Luna Park (set 1, dual program).
Complete Mame Rom Chd Set
Running in modified CMC hardware, with video RAM 4000-4FFF mirrored in 6000-6FFF and color RAM 5000-5FFF mirrored in 7000-7FFF. This game has the highest address line of the program tied to DIP switch #1, so you can select between 2 different programs. Both programs write to videoram either to each video RAM ranges. Added proper program ROM banking and connected to DIP switch #1. Added Luna Park (set 2). This one writes to normal CMC video RAM. Cleaned-up the machine drivers.
Derived clocks via #define. Added technical notes. added default NVRAM to Crystal Colours (CMC hardware), allowing to boot with clean meters/settings. added preliminary support for National Semiconductor MM58167 real-time clock/calendar. provided Jubilee Double-Up Poker (TMS9980) updates:.
Corrected the crystal value and derivate clocks via #DEFINE. Improved memory map. Hooked the CRT controller, but the init sequence seems incomplete. Created the accurate graphics banks. Found and mapped the video RAM.
Hooked the ATTR RAM. Assigned the correct graphics banks to the proper drawn tiles.
Find and mapped an input port. Started a preliminary workaround to demux the input port. Added technical notes. and provided Naomi updates:.
added all known IC locations from sets missing them. corrected some bad Sega chip numbers and IC locations. used real chip dump for mvsc2.
fixed ROM loading for Animal Basket. updated gottlieb.c, kyugo.c, madalien.c, rallyx.c, taitob.c, zodiack.c: fix flipscreen after core changes. provided Jubilee Double-Up Poker (TMS9980) updates:. Demuxed the input system. Hooked an cleaned all inputs, except the coin in (missing). Added NVRAM support. Added technical notes.
updated Jubilee Double-Up Poker: Fixed the NVRAM size. Video and working RAM are both halves of the same device (TC5517AP-2, tied to a battery). Also added technical notes. converted palettet and paletteclient to classes. General palette.c cleanup. provided more Jubilee Poker updates. Hooked CRTC properly.
Adjusted the screen size and visible area according to CRTC values. Adjusted the screen pos 8 pixels, to get a bit centered.
Added technical notes. updated galivan.c: add buffered spriteram to fix MT 05465; cleanups. provided Jubilee Double-Up Poker updates. Added a default clean NVRAM. Found and implemented the credits input. The game is now working!
Still no sound. fixed display of device tags in internal profiler. abstracted netlist sources. This is a first step to support libraries (e.g. updated v60.c: Modernized cpu core. rewrote 6850 based on datasheet using external CLOCK & removed MIDI clock hacks.
Standardised COM8116 DEVCB2 callbacks. GUS gf1device device derives from acia6850device to gain access to the tx & rx irq state, rather than exposing it with a public method. updated upd7810.c: Modernized cpu core. added IGS036 decryption support.
made timers once again execute at the end of each timeslice instead of at the start, as they did prior to 0.141u2. Fixes MT 4308. fixed rendering issue in psyaria and raycris attract. provided Gold Fruit improvements: Created a dedicated machine driver and decode properly the first graphics bank, since graphics are encoded/packed in a different way. This game has the main rate fixed to 40%, and coins A+B fixed with 100 credits by pulse (maybe used as remote credits inputs). updated bking.c, toypop.c: fix flipscreen after core changes.
updated the-cat driver: Removed hacks required by the old 68681 implementation; keyboard works again. updated Mundial/Mondial (Italian/French): Implemented the program banking properly. Now you can choose the program through a DIP switch. removed local lists of midi devices as they are currently all the same. converted PIT8253 to DEVCB2. updated the-n68681 device: fix a few uninitalized variables, beefed up optional logging.
removed diserial connect. Converted thomson, mc1502 & wangpc to use DEVCB2 instead. updated coco.c: uses rs232 port instead of bitbanger. fixed a netlist crash on clang due to ATTRNORETURN. Also fixed some debug statements.
updated baraduke.c, pacland.c: fix flipscreen after core changes, add SCREENRAWPARAMs. updated namcos1.c, namcos86.c: fix flipscreen after core changes, add SCREENRAWPARAMs. updated pdp1.c: Modernized cpu core. and updated the Witch driver: Changed both CPU clocks to 3 MHz, Both YM2203 clocks to 3 MHz, and confirmed the ES8712 clock (coming from pin 14 of M5205) in 8 MHz.
All measured and traced from the PCB. Implemented all clocks through #define. added free version of uni bios v3.1. and converted several drivers that have multiple different video chips bankswitched into the same address space, or have RAM bankswitched with ROM, to use the address map bank device (aka bankdev). changed popeye to use palette computed from schematics.
Also implemented interlaced field support. This is actually read by the game. made Popeye have netlist filtering and amplification. implemented CD4066 device - Quad Bilateral Switch. fixed compiling on unusual OS X configurations. provided Cherry Wheel improvements:. Created new memory map due to hardware differences.
Added default NVRAM. Partial MCU simulation.
Without it, only cherries appear and the player always wins. Added proper button-lamps support. Added technical notes. updated Super Nove (Playmark): Created proper machine driver and improve the second graphics bank decryption/decode. updated i386.c: Modernised cpu core. added inputs from the scratch for Super Star '97. Also improved inputs from other games of the driver.
Added technical notes. redumped Salary Man Kintaro. reatly improved M4 cart handling (added needed flash commands, fixed encryption on/off logic). fixed f355 EPR was overdumped (mirrored 2 times). completed dump of vtenis2c, ROM test passes now. fixed M1/Actel and Namco carts IC layouts, ROM test passes now. changed Namco 2f-2b IC names to.2d,2c etc, without fl0-fl4, because its the second IC label on the PCB (like 'IC17S ROM1' on Sega carts).
re-parented World Kicks. changed IC labels according to cart PCBs.
did a minor docs update. updated Super Star '97: Added more translations and technical info. Created a new graphics bank with different palette for the only girl graphics. fixed pong (rom based).
Minor changes no netlist code. updated upd7810.c: Added ADC support. fixed the colors of cb3d rearranging the split values stored in 4-bit bipolar PROMs and adding the missing one.
Promoted to working state. added deviceoutputinterface and macros MCFGOUTPUTINDEX/MCFGOUTPUTNAME which can be used to specify which output in the layout a device should use. updated CPS1.c: Willow, Strider and Ghouls and Ghosts have some changed factory setting defaults as stated from manual. Also noted differences in bonus live coverage where applicable.
updated the omti8621 device: Converted to proper ISA card. updated nbmj8688.c: Refactored to use HD61830 device. hooked up sound in bnstars.c by borrowing code from ms32.c and adding the second YMF271.
and updated Cherry Bonus III (set 4, encrypted bootleg). Fixed both graphics banks decryption/decode.
changed Battle Cross joystick type to be 8-way. corrected spyhunt labels / positions. updated drawgfx.c: Eliminate the need for a dummy 1:1 lookup table when drawing to ind16 bitmaps. provided Win Cherry improvements:. Decrypted/decoded all graphics banks. Preliminary memory map.
updated the UI graphics viewer so that it can now display multiple palettes and multiple gfx decoders in games that have. and updatedcinemat.c: redumped spaceftr. added Royal Card (French). This set is original, but running in a bootleg board.
Added a default NVRAM to get the game working. Also added technical notes. fixed the duart ip pins so they don't accidentally clear each other's state on devicewriteline; this fixes the Canon Cat printing in MESS. added basic implementation of the NSC810 RAM-I/O-Timer device, and plugged it into the Husky Hunter 2 driver.
removed unselectable slot options (SLOTINTERFACEINTERNAL) from the listxml output. added an osdtruncate call to truncate files. moved core template container classes up from emutempl.h to coretmpl.h:. these classes now no longer take a resourcepool; everything is managed globally - this means that objects added to lists must be allocated with globalalloc. added new autopointer template which wraps a pointer and auto-frees it upon destruction; it also defaults to NULL so it doesn't need to be explicitly initialized. moved taggedlist template to tagmap.h.
redid the low-level memory tracking system:. moved low-level tracking out of emu emualloc into lib util corealloc so it can be shared among all components and used by core libraries. globalalloc and friends no longer use a resource pool to track allocations; turns out this was a wholly redundant system that wasted a lot of memory. removed globalresourcepool entirely. added globalfreearray to delete arrays allocated with globalallocarray.
added tracking of object versus array allocation; we will now error if you use globalfree on an array, or globalfreearray on an object. added new utility helper conststringpool which can be used to efficiently accumulate strings that are not intended to be modified. Used by updated makelist and software list code.
updated png2bdc and makelist tools to not leak memory and use more modern techniques (no more MAXDRIVERS in makelist, for example). deprecated autostrdup and removed all uses by way of caller-managed astrings and the software list rewrite. rewrote software list management:.
removed the notion of a softwarelist that is separate from a softwarelistdevice; they are one and the same now. moved several functions into deviceimageinterface since they really didn't belong in the core software list class.
lots of simplification as a result of the above changes. added button-lamps support to all Cherry Bonus III sets.
Also created almost realistic control panel internal layout. Added technical notes about Cherry Bonus III memory map. provided Double-Up Poker (Jubilee) improvements: Added partial lamps support (some are missing), and a workaround for discrete sounds trigger offsets. Also a zillion of technical notes about findings.
updated pgm.c: simplify tile gfx ROM loading/decoding. updated the i8089 device: added channel priorities. update the m6502 core: fixed some unofficial opcodes. fixed MT05065. Double Crown PLDs converted to binary. added lamps support to Golden Star (IGS) sets. Also fixed the inputs accordingly.
Mame Roms That Work
removed pseudo-palettes from video/generic.c. All drivers that were using them now have an actual palette. updated shanghai.c: Add PCB layout and ROM locations for Shanghai sets. provided a new driver for American Music Poker (Amusco) based on a 8086 CPU + CRTC 6845 + PSG SN76489.
added palette to k053250 device, fixing xexex crash. Also modernized its coding conventions (dynamicarrays, devicevideointerface to attach screen, etc.). provided HD63484 updates:. converted to use a 16-bits address map. implemented RPR, MOD, SCLR, SCPY, LINE, PLG, PLL, CRCL, ARC, FRCT, PAINT, DOT, PTN, GCPY commands.
added pattern RAM emulation. updated wdfdc: Refactored to use devcb2 callbacks. rewrote 6551 emulation.
Changed clock to be the o2 clock instead of the baud rate xtal, supports external clocking, generates transmit interrupts when idle, added echo mode, dcd/dsr/cts input handling, parity & stop bit checking. Tested and fixed hookup to commodore lcd & plus 4. updated offtwall.c: Corrected the rom names and locations of the upright version of Atari's Off the Wall and also added a dump of one of the GAL's (a GAL16V8A-25LP) on the boardset which was the only one without the security fuse blown. updated the-upd765 device: Refactored to use devcb2 callbacks. made gfxelement::decode private; fixed drivers that were calling it directly. added proper IC locations to tmnt2 sets. added Korean?
Tile rom to newly dumped Shanghai3 set (was present on the board, but unused in our sets? Probably for use with undumped Korean program roms and leftover from conversion). provided a h63484.c compile fix.
added compile time check for 64-bit intel compiler to delegate.h (to prevent assert(sizeof(mfp).
Top Blade V ElSemi Carket Ball ANY, David Haywood Mega Phoenix Enricnes (from AUMAP), David Haywood Aaargh (Arcadia) Unigame, David Haywood Happy 6-in-1 (ver. Strykr wrote: For the gamer (which are the only people I imagine use mame) Their really should be a seperate build for gamblers and fruits etc, but in reality what games are people waiting for? There arent really i'm waiting for tbh, all the good uns are already working. I'd love them to get something like GTI club running on a modest PC setup, and a couple of the other meaty drivers, and perhaps the model 2 and 3 games running well, apart from that I havent upgraded since.149 Like I've said elsewhere, it's not really up to us to judge the quality of the games. Many of the gamblers are rather rare these days, and finding good quality pictures / artwork of some of the non-video ones is close to impossible. You could argue they're a waste of space, other people might even argue some of the recent prototype dumps are a waste of space (unfinished versions of games - why would a gamer care about a broken unfinished version of a game?) Personally I think they're all worth supporting and glad we do (and very thankful for anybody who spends their time and money dumping + emulating them) You'd be surprised by how many people have different interests!
As has been mentioned already people will upgrade if we improve support for something that interests them, I suspect a lot are still waiting on Raiden 2. I don't think anybody interested in Raiden 2 will refuse to upgrade come the day we fix it just because 100 gambling games were added in the meantime;-) As for 0.153, you have the Cave stuff as observed, which is in a better state than when it was removed a couple of years back (less audio glitches) although still has imperfect audio emulation. There's the Yukiwo prototype that was dumped by Unigame etc. That will be of interest to anybody who likes Master of Weapon due to the vast number of differences. Aaargh, another Unigame dump and a proper 2 player version of the game (apparently the Amiga version was single player only?) Mega Phoenix is a VERY rare Spanish arcade game, a remake of Phoenix on a more modern platform, it popular on many of the 8-bit and 16-bit machines we had over here but the arcade version was practically unknown, sound isn't quite right yet but gameplay is fine. Top Blade V is a rather unique 'Spinning Top' type battle game, doesn't really appeal to me beyond a few plays but it's something different.
X2222 is a shmup, although lacks sound, the roms for it were provided by the programmer, unfortunately apart from lacking sound the 'complete' build displays debug numbers at all times, and the earlier prototype is missing several levels and doesn't have the boss graphics finished so they display garbage. It's an interesting insight into a game that was probably never released tho. Carket Ball is a cute Korean Pool / Cards mashup game, like Pocket Gal etc. But different. The cartridge version of S.V.G.
Spectral vs Generation is supported, it's a more complete version of the game than the PCB version supported in 0.152 Galaxian Growing Galaxip is one of the more interesting Galaxian bootlegs / clones, with a strange 3 stage ship growth as you progress through the levels. Disco Boy (Promat license?) plays more like a traditional shooter than the existing set, maybe less interesting but easier to understand. Those are some of the personal highlights in terms of 'new games' at least. Thanks for the detailed update Haze. I, for one, hugely appreciate the time and effort that you and the rest of the MAME crew put into the behemoth of a project that it has become, and I'm sure millions of other people out there share my sentiment too. I always look forward to a new release of MAME and am never disappointed when one comes along, no matter how many or few new games it supports.
After all, who can moan when such an amazing project is totally free of charge to the end user. Please keep up the great work that you guys and girls do. 'although still has imperfect audio emulation' re: the cave games, I meant slowdown emulation, it was late when I was writing that and there's no edit button here. The main issue with Namco System 10 is the encryption, beyond that it's not really any different to any other Playstation 1 based platform.
I'm a little surprised there hasn't been more progress because there are some clear patterns in the ROM data in terms of byte counts etc. But that can be deceptive, and given that some of our best eyes have looked at it (Olivier, and I believe Andreas) and are yet to come up work a working solution to the encryption issue I imagine it's trickier than it appears (this is often the case) There was a small discussion about it a few days ago regarding attempts to collect data on the real PCBs for studying, however it was concluded that something on the hardware is performing checksums of the roms (probably the protection device) making modifying them more difficult because it refuses to run modified code.
It is one of the more significant challenging.platforms. left to emulated, and no, it hasn't been abandoned, it's just unexpectedly tricky compared to the other Playstation based ones, surprising really because platforms with lots of puzzle games are usually the budget platforms!