|
MAME 0.192
#371471 - 11/29/17 04:30 AM
|
|
|
MAME 0.192
Even if you’re still feeling burned from Singles’ Day, Thanksgiving, and/or Black Friday, MAME 0.192 is here, and there’s plenty packed into this update. At long last, the MCU for Ping Pong King is simulated, making the game playable, and Flower now runs better than it ever did. Taito’s abstract maze game Marine Date is also emulated better than ever, although there are still some issues with collision detection. We’ve got newly dumped prototypes, including a prototype of Battlecry, and a more complete version of Grudge Match that appears to come from a location test in Italy. Many of the Aristocrat MK5 gambling machines have been redumped so the ROM checksums pass and they boot. Missing graphics in the Merit Megatouch games have been fixed.
But we know that rumours about the next part of this announcement have made lots of arcade fans excited: IGS PGM2 software is showing signs of life. I’m sure at least some of you are wondering how this was achieved. Usual suspects Morten Shearman Kirkegaard and Peter Wilhelmsen (recently featured here in connection to their success in dumping Gaelco protection programs) have built another FPGA-based rig that allowed them to dump the internal programs from the ARM CPUs for Oriental Legend 2 and Knights of Valour 2 New Legend. On top of this, there are some nice performance improvements to MAME’s MPEG audio decoding that should benefit other systems as well.
In computer system emulation, we’ve got a number of graphical fixes for FM Towns, heaps of improvements for the Tatung Einstein, No-Slot Clock support for the Apple //e family, and support for some Brazilian CoCo clones from Prológica and Codimex. There are some big updates to the PC software lists, too. The ACI Destiny Prodigy, Mephisto RISC 1MB and Mephisto RISC II chess computers are now working. Finally, the Interpro drivers have numerous improvements, including preliminary keyboard/video support.
That’s just scratching the surface – there are far more bug fixes, newly dumped system and performance improvements. You can get the source or Windows binaries from the download page.
MAMETesters Bugs Fixed
- 00525: [Speed] (metro.cpp) bangball: When you finish a level, sometimes the game runs like slow-motion. (Angelo Salese)
- 02084: [DIP/Input] (saturn.cpp) sokyugrt: Emulation starts with 1 or more credits inserted. (Angelo Salese)
- 03719: [Graphics] (meritm.cpp) magat5, megat6, megat3, megat4 and clones: “Run21” missing graphics for black-suited cards. (AJR)
- 03804: [Color/Palette] (ssozumo.cpp) ssozumo: Wrong colors in the title screen. (Angelo Salese)
- 03805: [Color/Palette] (lasso.cpp) wwjgtin, photof: Wrong colors. (Angelo Salese)
- 04366: [DIP/Input] (saturn.cpp) shanhigw: The game starts with 2 credits inserted. (Angelo Salese)
- 04651: [Misc.] (gei.cpp) geimulti: Thirteen game ROMs are loaded, only nine are listed in-game and playable. (Osso)
- 05761: [Crash/Freeze] (dwarfd.cpp) All sets in dwarfd.c: [debug] Assert shortly after start. (AJR)
- 06193: [Speed] (shanghai.cpp) shanghai, shangha2 and clones: When playing, the gameplay and sound are very fast. (Angelo Salese)
- 06518: [Graphics] (metro.cpp) blzntrnd: Sprite layer disappears entirely after completing single-player game. (Angelo Salese)
- 06527: [Sound] (taitosj.cpp) elevator: Elevator Action sound balance is incorrect. (hap)
- 06694: [Sound] (midtunit.cpp) mk: Missing music. (hap)
- 06735: [Original Reference] (einstein.cpp) einstei2: Cursor not visible in 80-column B/W screen! (Dirk Best)
- 06736: [Original Reference] (einstein.cpp) einstei2: Character set selection (M004) in 80-column B/W screen! (Dirk Best)
- 06742: [Graphics] (pacland.cpp) pacland and clones: Sprite priority problem in invincible state. (hap)
- 06745: [Crash/Freeze] (marinedt.cpp) marinedt: MAME crashes when starting game. (Angelo Salese)
- 06747: [Media Support] (einstein.cpp) einstein: Formatting on double-density drive fails! (Dirk Best)
- 06749: [Graphics] (toaplan1.cpp) outzonecv: Screen is misaligned 16 pixels to the left. (hap)
- 06754: [Color/Palette] (zx.cpp) ts1000: Timex Sinclair 1000 screen displays black on cyan (should be black on white). (star2root)
- 06755: [Core] (zx.cpp) ts1000: The Timex Sinclair 1000 emulation only has the option of having 2k of RAM. (star2root)
- 06759: [DIP/Input] (seta.cpp) calibr50: Speed difference when rotating player. (AJR)
- 06766: [Sound] (spacefb.cpp) spacefb and clones: DAC sounds not working 100%. (Tafoid)
- 06771: [Compiling] (coco12.cpp) coco3, dragon, etc.: Genie fails to detect nested dependency of coco_t4426 on mc14411. (hap)
- 06772: [Gameplay] (dragon.cpp) Dragon machines run too fast! (hap)
New working machines
- Akai Katana (2010/ 8/13 MASTER VER.) [Bicycle Repairman]
- Pac-Eight [rod_wod, The Dumping Union]
- Player's Edge Plus (PS0239) Jackpot Jewels Slots [Brian Troha]
- The Big Joke (Version 0.00) [Coolmod, The Dumping Union]
New working clones
- '99: The Last War (bootleg) [ShouTime]
- Action Fighter (System 16B, unprotected, analog controls) [David Haywood]
- Action Fighter (unprotected) [David Haywood]
- Action Fighter (unprotected, analog controls) [David Haywood]
- Battlecry (Prototype) [ShouTime, progetto-SNAPS, Patrick Wheeler, Sean Sutton, Surgeville, Paul Vining, Marisol Nuñez Serrano, Rod_Wod, Jeffrey Gray, John Wilke, gamerfan, Smitdogg, The Dumping Union]
- Battlecry (Version C) [ShouTime, progetto-SNAPS, Patrick Wheeler, Sean Sutton, Surgeville, Paul Vining, Marisol Nuñez Serrano, Rod_Wod, Jeffrey Gray, John Wilke, gamerfan, Smitdogg, The Dumping Union]
- Bucky O'Hare (ver AA) [Bill D., The Dumping Union]
- Codimex CD-6809 [Felipe Sanches]
- Grudge Match (V00.90, Italy, location test?) [Vernimark, f205v]
- Gunbird 2 (set 2) [Hyid Choi]
- Head On N [ShouTime, Andrew Welburn, Rob Carr, Paul Vining, InsertMoreCoins, Smitdogg, The Dumping Union]
- Kaypro 10 - 1983 [rfka01]
- Otrona Attaché 8:16 [Al Kossow]
- Prológica CP400 Color II [Felipe Sanches]
- Prológica CP-500 (PVIII REV.3) [Felipe Sanches]
- The 26th Z [ShouTime, Paul Vining, The Dumping Union]
- Thunder Dragon 3 (bootleg of Thunder Dragon 2) [ShouTime]
- Troopy (bootleg of Mr. Kougar) [MikeMcBike]
Machines promoted to working
- ACI Destiny Prodigy [Edström]
- Cromenco MCB-216 [Robbbert]
- Flower (US) [Angelo Salese]
- Mephisto RISC 1MB [CB-Emu, RolandLangfeld, Sandro Ronco]
- Ping Pong King [Angelo Salese]
- Time Attacker [Angelo Salese]
Clones promoted to working
- Cromenco CB-308 [Robbbert]
- Flower (Japan) [Angelo Salese]
- Mephisto RISC II [CB-Emu, RolandLangfeld, Sandro Ronco]
New machines marked as NOT_WORKING
- Add Arbor Ambassador [Robbbert]
- ADDS Viewpoint 122 [Robbbert]
- ADDS Viewpoint 60 [Robbbert]
- Altos II Terminal [Robbbert]
- Ampex Dialogue 80 [Robbbert]
- AT&T 630 MTG [Robbbert]
- C.Itoh CIT-220+ Video Terminal [Robbbert]
- Convergent Mightyframe [Robbbert]
- Data General Dasher D461 [Robbbert]
- DEC VT-240 [Robbbert]
- DISER Lilith [Robbbert]
- Golgo 13: Juusei no Chinkonka (Japan, GLT1 VER.A) [Guru]
- Hazeltine Esprit [Robbbert]
- HP 2100 [Robbbert]
- HP-2622A [Robbbert]
- HP-700/92 [Robbbert]
- InterPro 2000 [Patrick Mackinlay]
- InterPro 2400 [Patrick Mackinlay]
- InterPro 2500 [Patrick Mackinlay]
- InterPro 2700 [Patrick Mackinlay]
- Kaypro 1 [rfka01]
- Kaypro 2/84 [rfka01]
- Kaypro 4x [rfka01]
- LSI M3 [Robbbert]
- Micro-Term 420 [Robbbert]
- Micro-Term 5510 [Robbbert]
- Milwaukee Computers MC-1200 [Robbbert]
- Motorola AMPS Car Phone [Robbbert]
- Motorola Powerstack II [Robbbert]
- Mushiking The King Of Beetles - Mushiking IV / V / VI (World) [Darksoft, f205v, MetalliC, rtw]
- Onyx C5000 [Robbbert]
- Qume QVT-102 [Robbbert]
- Qume QVT-103 [Robbbert]
- Qume QVT-190 [Robbbert]
- Qume QVT-201 [Robbbert]
- Qume QVT-70 [Robbbert]
- Relisys TR175II [Robbbert]
- Solbourne Series 5E Computer Workstation [Robbbert]
- Sony PlayStation 2 [Robbbert]
- Televideo TS-3000 [Robbbert]
- Televideo TVI-912C [Robbbert]
- Televideo TVI-955 [Robbbert]
- Televideo TVI-965 [Robbbert]
- Terco 4490 Mill CNC Control [Edstrom]
- Three Rivers PERQ 1A [Robbbert]
- TRS-80 DT-1 [Robbbert]
- Vectrix VX384 Graphics Processor Terminal [Robbbert]
- Visual 100 [Robbbert]
- Visual 102 [Robbbert]
- Wyse WY-50 [Robbbert]
- Xerox Alto I [Robbbert]
- Zentec Zephyr [Robbbert]
- Ziatech ZT-8802 SBC [Robbbert]
- unknown ACD computer [Robbbert]
- unknown TTI SCSI host adapter [Robbbert]
New clones marked as NOT_WORKING
- Action Fighter (FD1089A 317-unknown, analog controls) [David Haywood]
- Action Fighter (FD1089A 317-unknown) [David Haywood]
- Action Fighter (System 16B, FD1089A 317-unknown, analog controls) [David Haywood]
- Action Fighter (System 16B, FD1089A 317-unknown, set 1) [David Haywood]
- Action Fighter (System 16B, FD1089A 317-unknown, set 2) [David Haywood]
- Commodore PC 40-III [rfka01]
- Compaq Portable III [rfka01]
- Gran Trak 10/Trak 10/Formula K (older) [TTL] [Ed Fries, Tim Giddens, Andy Welburn]
- Hazeltine Esprit III [Robbbert]
- Kaypro 4/84 plus88 [rfka01]
- Kaypro New 2 [rfka01]
- Landing High Japan (Ver 2.02 O) [ShouTime]
- Martial Masters (ver. 102, 101, 101TW) [skate323k137, The Dumping Union]
- Sanyo MBC-28 [rfka01]
- Siemens PCD-2 [rfka01]
- Wyse WY-150 [Robbbert]
- Wyse WY-160 [Robbbert]
- Wyse WY-30+ [Robbbert]
- Wyse WY-55 [Robbbert]
- Wyse WY-60 [Robbbert]
- Wyse WY-85 [Robbbert]
New working software list additions
- archimedes: ALPS - Adventure Language Programming System, APEC Utils Disc v1.3, Archimedes Hard Disc Utilities Disc, Euclid v2.06 [Nigel Barnes]
- bbca_cass: 100 Programs for the BBC Microcomputer [Nigel Barnes]
- bbcb_cass: Fairy Tale, Instant Recall, Starter Word Splits, WestQuest 1847 [Nigel Barnes]
- bbcb_flop:
Acornsoft Demo Disk v1.0, Revs Demo, Superior Software - Demonstration Disc, Tree of Knowledge [Nigel Barnes] BeebSID 18 - Club Hits [PitfallJones]
- electron_cart: Starspell, Starspell & Starword, Starspell & View, Stop Press [Nigel Barnes]
- fmtowns_cd: AnotherTOWNS - Anata~ Free Software Collection, Dragon Shock, NHK Jissen Eikaiwa, Okumanchouja II, Oshare Cooking II, Tom Snyder's Puppy Love 2, Toushin Toshi II [r09]
- hp85_rom: Mass storage ROM [F.Ulivi]
- ibm5150:
B.A.T. (CGA/EGA/Tandy versions) (Quest & Glory Compilation), Teenage Mutant Hero Turtles: The Coin-Op! [ArcadeShadow] La Collection Amstrad PC - Arkanoid / Super Tennis, Oliver & Company [breiztiger] 4D Sports Boxing (Euro, Alt), A Line in the Sand, A Nightmare on Elm Street, Advanced Dungeons & Dragons - Curse of the Azure Bonds (V1.1), Advanced Dungeons & Dragons - Curse of the Azure Bonds (V1.2), Advanced Dungeons & Dragons - Pool of Radiance, Advanced Dungeons & Dragons - Pools of Darkness, Advanced Dungeons & Dragons - Secret of the Silver Blades (1.10), Alpine Tram Ride, QNX Demonstration (Apr 17th, 1989), Word 1.15, Word 2.00, Word 5.0 (Swedish), Word 5.5 (3.5", German), Word 5.5 (3.5", Italian), Word 5.5 (5.25"), Word X1.06 (September 7, 1983 beta) [Justin Kerk]
- ibm5170:
Ecstatica, Fables & Fiends - The Hand of Fate (German), Heart of China, LucasArts Classic Adventures, Rise of the Dragon [ArcadeShadow] WarCraft - Orcs & Humans [breiztiger] hal91 0.2.0, HAL91 0.4.5, PC Tools 7.1 (5.25" HD), PC-MOS Version 5.01, PC-MOS/386 Version 1.02, PC-MOS/386 Version 3.00, Photon microGUI Development version 1.10 for QNX 4.2x, Photon microGUI Runtime version 1.10 for QNX 4.2x, PicoBSD 0.41 (Dialup), PicoBSD 0.41 (Dialup) (Polish), PicoBSD 0.41 (ISP-alpha), PicoBSD 0.41 (Network), PicoBSD 0.41 (Network) (Polish), PicoBSD 0.41 (Router), PicoBSD 0.41 (Router) (Polish), PicoBSD 0.43 (Dialup), QNX 4.23, QNX 4.24, QNX 4.25, QNX Demo (Modem v4.00), QNX Demo (Network v400), QNX Demo Floppy (Modem Version V4), QNX Demo Floppy (Network Version V4), QNX Internet Appliance Toolkit Demo (v1.0), QNX Internet Appliance Toolkit Demo (v1.0a), QNX Internet Appliance Toolkit Demo (v1.0a Beta), QNX Internet Appliance Toolkit Demo (v1.1 - Russian), QNX Internet Appliance Toolkit Demo (Modem v3.03), QNX Internet Appliance Toolkit Demo (Network v2.0), QNX Internet Appliance Toolkit Demo (Network v3.03), QNX TCP/IP Runtime version 4.23, Serial Terminal Linux (0.1), Serial Terminal Linux (0.2), Serial Terminal Linux (0.3) [Justin Kerk]
- mtx_rom: MTX NewWord, Hisoft PASCAL [Nigel Barnes]
- px4_cart: Dialtext 4 [Robbbert]
Software list items promoted to working
- bbc_flop_arm: ARM Evaluation System Discs [Nigel Barnes]
- ibm5150: Ikari Warriors (alt) [breiztiger]
New NOT_WORKING software list additions
- dragon_cart: Dragon 32 Soak Test [Nigel Barnes]
- electron_cart: Advanced Plus 3, Advanced Plus 4, Click, Cumana Floppy Disc System, Pegasus 400, Sound Expansion v1, Sound Expansion v3 [Nigel Barnes]
- fmtowns_cd: Suikoden - Tenmei no Chikai [r09]
- ibm5170: 1830 - Railroads & Robber Barons, 5th Fleet, Alfalinux 0.3.1 -Laetitia-, Brutalware Linux 1.1, DLX Linux Demodisk V1.0, Phindows, tomsrtbt (1.7.185), tomsrtbt (2.0.103), Win32s (1.15), Win32s (1.15a), Win32s (1.25), Win32s (1.25a OLE), Win32s (1.30) [Justin Kerk]
- lynx: Alpine Games [anonymous]
- mtx_cart: MTX Node Ring [Nigel Barnes]
- mtx_cass: Demonstration Tape, MTX Draughts [Nigel Barnes]
- px8_cart: Clock, DAKComm, dBase II, Epson BASIC, Geneva Connection, Intext 1.0, Intext 1.1, Othello, Oval IRM, Portable Calc, Portable WordStar, SuperCalc, The Traveling Appointment Manager, The Traveling Expense Manager, The Traveling Time Manager, Turbo Pascal (unofficial ROM), CP/M Utilities [Nigel Barnes]
- vsmile_cart: Disney Cenerentola - I desideri magici di Cenerentola (Ita), Disney Il re leone - La grande avventura di Simba (Ita), Disney La sirenetta - Ariel e la crociera magica (Ita), Disney/Pixar Ratatouille - Le nuove ricette di Remy (Ita), Disney/Pixar Up (UK), Kung Fu Panda - Path of the Panda (UK), Scooby-Doo! - Avventura a Funland (Ita), Zayzoo: An Earth Adventure (UK) [tag2015]
Translations added or modified
- Dutch [Jos van Mourik]
- Greek [BraiNKilleRGR]
- Portuguese [Pedro Simoes]
- Spanish [A. Viloria]
Source Changes
ygv608.cpp fixes and improvements: [Agiri] Fixed rotation and scrolling. Added scrolling-by-row (used in NCV1 Xevious Arrangement). Fixed scaling effect (used in NCV2 Pacman Arrangement staff roll).
mice updates – boots into monitor now: [AJR] Generate Baud rate through 8155 (clock kludge required). Factory settings for DSW7 to configure baud rate, data bits and parity. Hooked up Rx and Tx interrupts. Split into separate sets for each CPU-specific monitor program.
i8155: Substantially rewrote timer emulation. [AJR]
ringking: Fixed misconfigured NMI gate – works again now. [AJR]
isbc8030: Remove hacky and superfluous UI-configurable Baud clock. [AJR]
n8080.cpp: Directly set 8035 IRQ line rather than generating fake pulses (restores some sounds in helifire). [AJR]
Changed 9615 baud configurations to the nominal 9600 baud now that Z80SIO is more tolerant. [AJR]
Made SDL input less eager to generate double-click events when mouse doesn’t move between clicks. [AJR]
m68000: Reinstated registration of SR for debug expressions. [AJR]
mc68901: Improved USART behavior for polled operation. [AJR]
v9938, v9958: Converted to use RGB32 bitmaps, fixed V9938-onV9938 transparent overlay in meritm.cpp. [AJR]
Improved wraparound handling for relative inputs. [AJR]
Attempted to fix PPC DRC scheduling bug (MT06615). [AJR]
mc68681: Distinguish SCN2681 and MC68681 DUART types. [AJR]
vt220: Hooked up SCN2681 based on schematics. [AJR]
stvvdp1.cpp: Fixed line/polyline drawing (Die Hard Arcade wireframe map). [Angelo Salese]
dec0.cpp: Sly Spy updates: [Angelo Salese] Emulated sound protection state machine for Sly Spy and removed ROM patch. First pass at emulating Sly Spy RNG device at $31c00d – used extensively for randomisation, especially with bosses.
marinedt.cpp: Rewrote driver from scratch, fixing multiple issues that persisted in previous implementation. [Angelo Salese] Added sea bitmap and flashing when shark appears. Improved sprite/tilemap colors – now more accurate to the references. Fixed priority – both sprite layers are actually above text tilemap. Added player 2 trackball inputs and global coin lockout. Streamlined and improved collision detection (still not perfect, so still marked NOT_WORKING).
flower.cpp: Rewrote driver from scratch, fixing multiple issues that persisted in previous implementation [Angelo Salese] Fixed visible area to 288x224 – fixes leftmost columns (inside the big ship). Corrected background layer priority (outside and inside the big ship). Fixed zoom points for 32x32 sprites.
gladiatr.cpp: Added preliminary sound emulation and MCU simulation for inputs in Pink Pong King. [Angelo Salese]
cyclemb.cpp: Added preliminary audio for Cycle Maabou/Sky Destroyer. [Angelo Salese]
Added/fixed HMC20 and screen raw parameters for multiple drivers. [Angelo Salese] bogeyman.cpp, chanbara.cpp, dec0.cpp, dec8.cpp, firetrap.cpp, mystston.cpp, ssozumo.cpp
mainevt.cpp: Fixed sprite shadows for Devastators. [Angelo Salese]
chqflag.cpp updates: [Angelo Salese] Improved shadow/highlight code – fixes black rain and lights at dawn. K051937 sets up a register when background should be dimmed – fixes start/end pouring event colors. Fixed dimmed screen condition bug after soft reset.
amusco.cpp: Improved colors. [Angelo Salese]
metro.cpp: Added sprite X/Y center point registers , and removed screen check hack. [Angelo Salese]
Rewrote Imagetek i4100/i4220/i4300 video display processor family, and hooked it up to metro.cpp. [Angelo Salese]
zx.cpp: Fixed ts1000 display and added extra RAM options. [Barry Nelson]
attache: Added Attaché 8:16 ROMs, and revision H BIOS; added preliminary communications between Z80 and 8086. [Barry Rodewald]
fmtowns: Separated palette into a 256-colour palette and two 16-colour palettes for each layer – fixes colours in Abel. [Barry Rodewald]
r9751: Added more SMIOC commands – disktool is now able to print startup text. [Brandon Munger]
mrisc: Fixed opening book. [CB-Emu]
cosmac: Added WAIT/CLEAR input lines. [Curt Coder]
cdp1852: Added DO read function. [Curt Coder]
tmc600 updates: [Curt Coder] Added Euro bus skeleton, fixed printer handshake, and fixed RAM/ROM region sizes. Implemented video according to actual hardware and fixed cursor blink rate. Implemented TMC-700 real time clock.
pgm2 improvements: [David Haywood] Fleshed out driver, creating devices for the Atmel ARM AIC and a derived IGS036 CPU. Emulated IGS037 video hardware – covers sprites (with zooming), tilemaps (including rowscroll), and the priority system. Still marked NOT_WORKING due to missing memory card emulation (requires simulation of undumped MCU).
mpeg_audio: Use pre-calculated cosine table for DCT decoding – yields significant performance improvement. [David Haywood, Aaron Giles]
arm7: Fixed BLX (LO) thumb opcode in ARM7/9 – return address was offset by one halfword. [David Haywood]
dsk_dsk: Don’t segfault on empty tracks – fixes einstein 4game. [Dirk Best]
deco146: Fully handle sound latch internal to the device. [Dirk Best]
einstein updates: [Dirk Best] Created “Tatung Pipe” bus and converted 80 column device (use -pipe tk02 to use). Added cursor rendering, support for alternate character, and save state support set to the 80 column device. Removed redundant einstei2 driver. Added user port bus interface with support for speech cartridge and mouse. Corrected default floppy drive types, and added support for more types. Added ADC0844 device, emulated analogue joystick, and fixed minor issues with memory map. Marked Einstein 256 ROM dump as bad. Added support for the Silicon Disc RAM drive as a pipe device (appears as drive 3). Fixed Centronics port – printing now works. Hooked up RS232 port. Improved interrupts and add WIP Speculator support.
px4: Fixed keyboard input from cold start. [Dirk Best]
floppy: Added TEAC FD-30A 3 inch floppy drive. [Dirk Best]
Added a generic Z80 daisy chain device, for use in drivers with non-Z80 peripherals. [Dirk Best]
midvunit: Use ADC0844 device. [Dirk Best]
micro3d updates: [Dirk Best] Use ADC0844 device and fixed throttle regression. Attached RS232 port to monitor port.
adc0844: Added support ADC0848 variant and made some adjustments to differential mode. [Dirk Best]
74123: Use line handlers for single lines. [Dirk Best]
wave: Look up speaker count and cassette device relative to owner – fixes cassette speakers inside devices. [Dirk Best]
headonn: Fixed colors and added DIP switch info. [Dirk Best]
headon2: Fixed colors and cleaned up. [Dirk Best]
spacetrk: Marked as imperfect graphics due to missing star field and background gradient. [Dirk Best]
via6522: Added some interrupt logging. [Edström]
prodigy: Added clickable layout, making the machine playable; updated keypad logic to use ioport_array. [Edström]
t4426: Replaced a bad dump caused by corrosion with a new verified dump. [Edström]
epc: Fixed load order for newer ROM (now gives same error as older ROM), and set EGA as default as it works. [Edström]
z80scc: Ported improvements from z80sio and cleaned up code. [Edström]
hp9895: Fixed NMI handling and secondary addressing in PHI chip. [F.Ulivi]
hp85: Added support for I/O slots and emulation of HP82937 module; added support for internal printer. [F.Ulivi]
Fixed metadata fro Brazilian CoCo clones, and renamed LZ Color64 ROMs to document labels and PCB locations. [Felipe Sanches]
Added Codimex CD-6809 FDC (1986) device. [Felipe Sanches]
trs80dt1 updates: [Frank Palazzolo, Robbbert] Fixed keyboard key sense to fix key repeat. Switched refresh to 60Hz to match with 60Hz jumper setting. Added temporary hack to get the last three columns to display.
i8275: Added check to avoid potential buffer overflow. [Frank Palazzolo]
m68000: Added support for 68010/68020/68030 bus error stack frames formats – fixes common CPU type detection routine. [Jean-François DEL NERO]
Fixed DS128X/DS1288X main frequency divider – output frequency must be fixed to 1Hz. [Jean-François DEL NERO]
smc91c9x: Implemented SMC91C9X Ethernet controller connected mode with WinPcap or TUN/TAP tunnel. [Jean-François DEL NERO]
cv1k.cpp: Demoted all sets to imperfect timing, set audio output to mono, and documented DIP switches. [MetalliC]
ymz77x improvements: [MetalliC] Added pan, main volume level, and clip limit control. Emulated YMZ774 chip at feature level used by PGM2-platform games.
SH3/4 DRC: Mark DRC cache dirty on reset. [MetalliC]
arm7: Handle DP-commands ROR <Rs> case where Rs=32/64. [MetalliC]
arm7: LDR/LDRM should switch to Thumb mode on ARMv5T. [David Haywood, MetalliC]
ti99_8: Added Hexbus write timing logic. [Michael Zapf]
bbcb: Added Computer Village 1797 FDC (LVL Dos) – not working. [Nigel Barnes]
electron: Added Acorn Plus 1 expansion device, moved cart slots to plus1, and replaced OS ROM with good dump. [Nigel Barnes]
bbc: Added ARM Evaluation System co-processor device. [Nigel Barnes]
cfa3000: Added input devices on userport, analogue, and 1MHz bus. [Nigel Barnes]
bbcm: Allow switching between internal and external tube ports. [Nigel Barnes]
mtx updates: [Nigel Barnes] Fully implemented both ROM- and RAM-based memory maps. Added Finnish and Danish keyboard ROMs. Implemented ROM extension board and cartridge slot. Added support for ROM 2 subpages on ROM extension board. Implemented cassette motor control.
accomm: Added CMOS RAM, serial and econet ports, implemented RAM/ROM switching according to manual, and documented full memory map. [Nigel Barnes]
interpro updates: [Patrick Mackinlay] Added basic support for the system bus variously referred to as SR, SR bus, SRX and CBUS. Added initial GT graphics card implementation sufficient to boot the diagnostic monitor in graphics mode. Added high-level emulation of the InterPro keyboard with US English layout. Added support for InterPro 20x0, 24x0 and 27x0 systems. Added support for Ethernet, mouse, and serial DMA. Improved timer, interrupt and DMA handling. improved comments/documentation
bt459 improvements: [Patrick Mackinlay] Added support for blinking (pixel planes and cursors). Fixed dual-cursor and X Window mode. Corrected screen output alignment.
28fxxx: Added device emulating 28F010 Flash memory and family (not compatible with JEDEC Flash protocol). [Patrick Mackinlay] Converted Interpro to use this device.
clipper: Floating point improvements: [Patrick Mackinlay] Switched to softfloat implementation. Implemented floating point macro instructions. Implemented floating point exceptions and rounding.
apple2: Added preliminary support for “SSB Apple” speech board. [R. Belmont, Apple II Documentation Project]
apple2: Added preliminary support for AE TransWarp accelerator card. [R. Belmont, Apple II Documentation Project]
apple2: Added No-Slot Clock support to //e, //c, and friends. [R. Belmont]
apple2cp: Improved “MIG” custom chip handling for apple2cp. [mgcaret, R. Belmont]
apple2cp: Corrected MIG RAM size to 2K (was 0.5K). [mgcaret, R. Belmont]
apple2c: Fixed Slinky register return when no expansion RAM is present. [R. Belmont, mgcaret]
ds1315: Added support for DS121x parts where the RTC sits in a ROM’s address space. [R. Belmont]
arm7: ARM v5TE support improvements: [R. Belmont] Added support for high vector option. Fixed v5 BLX to save the return address in R14. Implemented BLX Rn form. Fleshed out ARM946ES model, added movable DTCM and ITCM support.
nds: De-skeletonized driver: [R. Belmont] Added IPCSYNC. Added WRAM banking and more mirroring. Added timers, interrupt management, partial DMA, and ARM7 halt-until-IRQ.
ibmpcjr: Added quiksilver fast-boot BIOS. [Robbbert]
mmagic: Added sound samples. [Robbbert]
Aristocrat MK5 updates: [Roberto Fresca] Redumped the following sets so they have correct checksums and boot: Adonis (BHG1508, US), Boot Scootin' (GHG1012-02, US), Bumble Bugs (CHG0479-03, US), Cash Chameleon (DHG4078-99, US), Enchanted Forest (JHG0415-03, US), Golden Pyramids (AHG1206-99, US), Magic Garden (AHG1211-99, US), Margarita Magic (EHG1559, US, set 1), Party Gras (BHG1284, US), Sweethearts II (PHG0742-02, US), Sweet Liberty Deluxe (AHG1575, US), The Gambler (EHG0916-02, US), Tropical Delight (PHG0625-02, US). Added BIOS / USA Set Chip v4.04.08. Reformatted the game drivers to be more readable.
fmtowns: Fixed kanji ROM offset calculation and scrolling in 16-color mode. [r09] Kanji in fixed ROM font now displays correctly, making boot messages intelligible. Scrolling imperfect but better than before – fixes scrolling/positioning issues in multiple games.
hcd62121 updates: [Ricardo Barreira] Made improvements based on experiments with hardware using a CFX-9850G. Fixed mistakes in CL flag calculations, and added more unknown instructions. CFX-9850G is now emulated just accurately enough for most of the RUN mode to work well.
mrisc: Added PROM dumps and removed the ARM boot hack. [RolandLangfeld, Sandro Ronco]
arm7: Added rudimentary instruction prefetch buffer. Fixes GBA NES Classics games. [Ryan Holtz]
gba: Made GPIO ports pass ROM through on read by default – fixes Doom 2 and Duke Nukem Advance. [Ryan Holtz]
e132xs: Reworked all opcodes to reduce runtime branches and unnecessary work – now considerably faster. [Ryan Holtz]
sm7238: Implemented reverse video, 80/132 column switching, and alternate font. [shattered]
vt240: Fixed palette and screen size fix. [shattered]
monty.cpp: Wake Monty Plays Scrabble and Master Monty from HALT instruction using reset. [smf]
z80: Implemented HALT output and clear HALT condition during reset. [smf]
c65 updates: [smf] Reordered keys and renamed I/O ports to match system specification. Added caps key and two extra keyboard columns, although only one of them may have made it into hardware. Added support for switching between 40/80 columns with F1 and blink/underline/highlight/reverse attribute. Allow switching between upper case/lower case character ROM with shift+Commodore key.
chdman: Use DeviceIoControl to get disk length – allows chdman to access physical drives on Windows 10. [Ted Green]
Added 32- and 64-bit population count utilities. [Vas Crabb]
Moved object finder resolution before device_start (github #2759). [Vas Crabb]
Added an SDLC consumer device that logs SNA frame headers and data. [Vas Crabb]
Replaced hacky Kaypro keyboard that bypassed the SIO with emulated Kaypro 10 keyboard. [Vas Crabb, rfka01, TeamEurope]
kaypro: Switched from z80dart to z80sio, hooked up BRG to SIOs, hooked up direct connections from SIOs to RS232 ports. [Vas Crabb]
z80sio updates: [Vas Crabb] Decoupled from device_serial_interface – it can’t support synchronous modes, on-the-fly register updates, and other features. Made asynchronous receive behave more like real device: check that start bit persists for half a bit interval, sample data bits mid-interval, handle invalid stop bit as described in Zilog manual, check parity and latch overrun and parity errors. Re-implemented break detection. Implemented SDLC transmission including bit stuffing, transmit CRC, abort, and underrun/end-of-message behaviour. Generalised synchronous transmission to other modes. Completely overhauled interrupt logic – vectors should be correct now. Improved handling of transmit and receive buffers and control lines. Implemented different auto-reset receive errors in MPSC vs SIO. Implemented enough of synchronous reception to pass uts20 loopback tests (not accurate, but helps understanding). Disabled automatic CRC transmission in 1-byte sync mode (necessary to pass uts20 loopback tests).
e132xs: Templated most Hyperstone register-register and immediate opcode handlers to reduce code duplication. [Vas Crabb] Also fixed behaviour in some corner cases including privileged register traps and SR as source in subtraction.
uts20 updates – all self-tests now pass without ROM patches: [Vas Crabb] Hacked in permanent loopback on SIO channel A TxD to RxD, and SIO channel B TxD to RxD and WAIT/READY to DCD+RTS. Implement parity poison and NMI on parity error feature. Corrected NVRAM width to four bits.
namcops2.cpp: Added dump of Time Crisis 3 V291 I/O board. [Darksoft, Mitsurugi]
naomi.cpp: Dumped WaveRunnerGP motor board. [Darksoft, Mitsurugi]
skyskipr: Added DIP switch locations. [einstein95]
vicdual: Added sound samples for tranqgun and bonus sound sample for depthch. [Jim Hernandez]
Fixed some typos. [Martin Lindhe]
pgm2: Dumped IGS036 internal ROMs for Oriental Legend 2 and Knights of Valour 2 New Legend. [Morten Shearman Kirkegaard, Peter Wilhelmsen]
pntnpuzl: Added 8798 MCU dump. [Reznor007]
Added Commodore PC 30-III BIOS v2.00 and alternative Ericsson PC BIOS. [rfka01]
Dumped obj roms for Seibu CATS system. [ShouTime, Bill D., The Dumping Union]
Redumped Taikyoku Base Ball (previously called Champion Base Ball Part-2 (set 2) and only partially dumped). [ShouTime, progetto-SNAPS, Patrick Wheeler, Sean Sutton, Surgeville, Paul Vining, Marisol Nunez Serrano, Rod_Wod, Jeffrey Gray, John Wilke, gamerfan, Smitdogg, The Dumping Union]
Identified Lives and Demo Sounds DIP switches for Sky Destroyer. [sjy96525]
Corrected years for Spikeout: Final Edition, Ehrgeiz, Golgo 13 Kiseki no Dandou and Daytona USA. [sjy96525]
Fixed failure to boot for Tekken Tag Tournament (World, TEG2/VER.C1, set 2) and Tekken Tag Tournament (US, TEG3/VER.B). [sjy96525]
Corrected inputs for CPS3 multi-game bootlegs – Street Fighter III games and Red Earth / Warzard require six buttons. [sjy96525]
namcos10.cpp: Updated documentation. [Guru]
|
|
|
|
Re: MAME 0.192
[Re: Vas Crabb]
#371491 - 11/29/17 10:45 PM
|
|
|
An "huge" update
Thanks for the information
PS: perfect timming!
|
|
|
|
Re: MAME 0.192
[Re: Vas Crabb]
#371500 - 11/30/17 03:00 AM
|
|
|
Thanks to everyone who helped create the Hyperstone DRC! The games are running 3-5% faster now.
|
|
|
|
|
> Thanks to everyone who helped create the Hyperstone DRC! The games are running 3-5% > faster now.
There is no DRC yet..
and 3-5% seems rather low? I thought we were looking at about twice as fast here? MG was talking about 240% -> 400% type boosts.
|
|
|
|
Re: MAME 0.192
[Re: Haze]
#371504 - 11/30/17 04:43 AM
|
|
|
Last I checked it was in the 30-40% range of improvement for me. An impressive speed up, given that as you say there's no DRC yet.
|
.
|
|
|
Re: MAME 0.192
[Re: Haze]
#371505 - 11/30/17 07:16 AM
|
|
|
> and 3-5% seems rather low? I thought we were looking at about twice as fast here? MG > was talking about 240% -> 400% type boosts.
It does seem low, but there was some performance loss when I had to back off some of the more aggressive optimizations. So the final figure ended up closer to 240% -> 360% for elfin on my machine. The gains vary depending on the game, and whether it's already idle-skipped or not.
In Diet's case, he's most likely listing the total gain in the "average speed" output, and in a previous post he mentioned some games running as slow as <50%. I wonder, too, if perhaps the games are bound more by the memory system or video emulation on his machine.
|
|
|
|
Re: MAME 0.192
[Re: MooglyGuy]
#371507 - 11/30/17 07:55 AM
|
|
|
> In Diet's case, he's most likely listing the total gain in the "average speed" > output, and in a previous post he mentioned some games running as slow as <50%. I > wonder, too, if perhaps the games are bound more by the memory system or video > emulation on his machine.
Well, elfin is dgPix hardware with the VRender0 video chip - that seems to eat a significant chunk of performance when testing on my i7 notebook.
|
|
|
|
Re: MAME 0.192
[Re: Vas Crabb]
#371514 - 11/30/17 03:13 PM
|
|
|
> > In Diet's case, he's most likely listing the total gain in the "average speed" > > output, and in a previous post he mentioned some games running as slow as <50%. > I > > wonder, too, if perhaps the games are bound more by the memory system or video > > emulation on his machine. > > Well, elfin is dgPix hardware with the VRender0 video chip - that seems to eat a > significant chunk of performance when testing on my i7 notebook.
It's VRender0- (Minus) which in hardware terms is basically just a framebuffer with some extra write modes for handling transparency. I assume in this case VRender0- is actually just a software library compiled into the game code.
It shouldn't be eating up a significant amount of performance in MAME's video code, mostly memory writes and the Hyperstone.
Actual VRender0 is Crystal System, where the video is actually hardware accelerated. (Same features but done in hardware, not a software lib?)
Then there's the Voodoo / Medix GX PC based board on which Funky Ball runs which they called 'VRender 2Beta Rev4'
|
|
|
|
Re: MAME 0.192
[Re: MooglyGuy]
#371520 - 11/30/17 06:31 PM
|
|
|
> > and 3-5% seems rather low? I thought we were looking at about twice as fast here? > MG > > was talking about 240% -> 400% type boosts. > > It does seem low, but there was some performance loss when I had to back off some of > the more aggressive optimizations. So the final figure ended up closer to 240% -> > 360% for elfin on my machine. The gains vary depending on the game, and whether it's > already idle-skipped or not. > > In Diet's case, he's most likely listing the total gain in the "average speed" > output, and in a previous post he mentioned some games running as slow as <50%. I > wonder, too, if perhaps the games are bound more by the memory system or video > emulation on his machine.
Most of the Hyperstone games including Elfin, only got 3-5% faster. Jump Jump has improved the most out of all the games I tested, going from 77-79% to 92-94%. I was using D3D in the MAME I am comparing to, and am now using GDI because D3D has been discontinued. That might explain why most games only got a slight speed increase.
|
|
|
Tafoid![Administrator Administrator](//www.mameworld.info/ubbthreads/images/adm.gif) |
I keep on testing.. testing.. testing... into the future!
|
|
|
Reged: 04/19/06
|
Posts: 3138
|
Loc: USA
|
|
Send PM
|
|
|
|
> > > and 3-5% seems rather low? I thought we were looking at about twice as fast here? > > MG > > > was talking about 240% -> 400% type boosts. > > > > It does seem low, but there was some performance loss when I had to back off some > of > > the more aggressive optimizations. So the final figure ended up closer to 240% -> > > 360% for elfin on my machine. The gains vary depending on the game, and whether > it's > > already idle-skipped or not. > > > > In Diet's case, he's most likely listing the total gain in the "average speed" > > output, and in a previous post he mentioned some games running as slow as <50%. > I > > wonder, too, if perhaps the games are bound more by the memory system or video > > emulation on his machine. > > Most of the Hyperstone games including Elfin, only got 3-5% faster. Jump Jump has > improved the most out of all the games I tested, going from 77-79% to 92-94%. I was > using D3D in the MAME I am comparing to, and am now using GDI because D3D has been > discontinued. That might explain why most games only got a slight speed increase.
D3D has not been discontinued - DDRAW was Not sure why you don't have D3D, but you need Direct X9.0c installed to run it. Even -video bgfx would be an improvement over GDI. That even makes pacman slow.
|
|
|
|
Re: MAME 0.192
[Re: Tafoid]
#371527 - 12/01/17 01:04 AM
|
|
|
> D3D has not been discontinued - DDRAW was > Not sure why you don't have D3D, but you need Direct X9.0c installed to run it. Even > -video bgfx would be an improvement over GDI. That even makes pacman slow.
It won't work on GPUs that don't support NPOT any more, including the infamous Silicon Image/Intel GMA950 and GMA945. Users with these GPUs should try OpenGL.
|
|
|
|
Re: MAME 0.192
[Re: Vas Crabb]
#371539 - 12/01/17 05:24 AM
|
|
|
> > D3D has not been discontinued - DDRAW was > > Not sure why you don't have D3D, but you need Direct X9.0c installed to run it. > Even > > -video bgfx would be an improvement over GDI. That even makes pacman slow. > > It won't work on GPUs that don't support NPOT any more, including the infamous > Silicon Image/Intel GMA950 and GMA945. Users with these GPUs should try OpenGL.
OpenGL does not work on my computer either. The menu starts with the text replaced with white squares, and when a game is loaded the whole screen turns white. BGFX works, but is much slower; Elfin runs around 28% with it, and around 65% with GDI.
|
|
|
|
|
> OpenGL does not work on my computer either. The menu starts with the text replaced > with white squares, and when a game is loaded the whole screen turns white. BGFX > works, but is much slower; Elfin runs around 28% with it, and around 65% with GDI.
No offense, but your computer must be a real old piece of junk. It'll hopefully still benefit when there's a Hyperstone DRC, but just, man. If your computer is so old that none of the video output options except GDI work, how do you even do anything on the modern Internet?
|
|
|
|
|
|
|
Re: MAME 0.192
[Re: MooglyGuy]
#371543 - 12/01/17 01:53 PM
|
|
|
> > OpenGL does not work on my computer either. The menu starts with the text replaced > > with white squares, and when a game is loaded the whole screen turns white. BGFX > > works, but is much slower; Elfin runs around 28% with it, and around 65% with GDI. > > No offense, but your computer must be a real old piece of junk. It'll hopefully still > benefit when there's a Hyperstone DRC, but just, man. If your computer is so old that > none of the video output options except GDI work, how do you even do anything on the > modern Internet?
those crap old chipsets are actually still very, very common, I still know a lot of people with 'junk' like that for their main PC, because for internet etc. they just use a smartphone.
they typically end up finding that RetroArch works better on such hardware and ditching proper MAME for good unfortunately.
|
|
|
|
|
> OpenGL does not work on my computer either. The menu starts with the text replaced > with white squares, and when a game is loaded the whole screen turns white. BGFX > works, but is much slower; Elfin runs around 28% with it, and around 65% with GDI.
https://tenor.com/uqSQ.gif
|
|
|
Dullaron |
Mitchell or Dunard Garaus
|
|
|
Reged: 07/22/05
|
Posts: 6134
|
Loc: Fort Worth, Tx
|
|
Send PM
|
|
|
Re: MAME 0.192
[Re: Haze]
#371547 - 12/01/17 07:15 PM
|
|
|
I pointing out that the Official MAME is way better than MAME for the RetroArch. So many issues on the MAME RetroArch. It lags on some of the scanlines effect. Tempest and some others doesn't run very well on the speed. It just the way RetroArch is written. RetroArch isn't clean for the MAME port.
Official MAME vs MAME for the RetroArch. Official MAME is a winner.
Like I said before I only uses Official MAME on my Windows.
Uses that MAME for the RetroArch on my Android devices only. (Until Official MAME come out for the Android. I will switch to this after it release.)
|
W11 Home 64-bit + Nobara OS / AMD Radeon RX 5700 XT / AMD Ryzen 7 3700X 8-Core 3.59 GHz / RAM 64 GB
|
|
|
Re: MAME 0.192
[Re: Haze]
#371555 - 12/01/17 11:49 PM
|
|
|
> those crap old chipsets are actually still very, very common, I still know a lot of > people with 'junk' like that for their main PC, because for internet etc. they just > use a smartphone. > > they typically end up finding that RetroArch works better on such hardware and > ditching proper MAME for good unfortunately.
Prove to me, with photos and screenshots, that RetroArch will run on PCs that have video hardware which doesn't even support non-power-of-two textures, and I will revise (but not recant) all of the following. Having said that:
You consistently rail against the universe for people turning to RetroArch, yet you're now also railing against even the slightest change that involves bringing MAME's rendering pipeline into this century, which - if implemented simultaneously with a UI overhaul, and the planned support for things like 3D art assets in layouts - could actually stand half a chance at clawing back some of those users. Can you please clarify your standpoint such that, to me at least, you don't sound like a hypocrite who, typically, abounds with criticism for what everyone else is doing while being convinced that his own feces is without scent?
I mean, really, Haze. You're choosing this thread - this thread, of all of them - to moan about how MAME devs are driving people towards RetroArch? Can you point out which emulator it is that could make Hyperstone games run full speed on Diet Go Go Fan's machine? Because from where I'm sitting, there aren't any. And if he (or she, what do I know) happens to switch to RetroArch, then he will effectively be using an older version of MAME, which is slower, because this is a thread in which we're discussing my optimizations that made a driver faster in this version of MAME, you implacable knob.
|
|
|
|
MAME vs. RetroArch [was Re: MAME 0.192]
[Re: Dullaron]
#371565 - 12/02/17 07:31 AM
|
|
|
Is anyone actually moving from MAME to RetroArch?
Lots of people are moving from standalones to RetroArch, but considering how shitty and old the MAME driver is for RetroArch I'm really skeptical to consider them a direct threat.
It's more like the majority of emu people aren't interested in what MAME has to offer to begin with. Arcade and Apple II gamers are getting older and older. The console emulation on MAME is poor (that should be fixed) and the UI is kind of intimidating (I don't care if this is 'fixed'). RetroArch seems to attract people interested in consoles primarily, and MAME being there is just a bonus.
|
|
|
Dullaron |
Mitchell or Dunard Garaus
|
|
|
Reged: 07/22/05
|
Posts: 6134
|
Loc: Fort Worth, Tx
|
|
Send PM
|
|
|
Re: MAME vs. RetroArch [was Re: MAME 0.192]
[Re: anikom15]
#371566 - 12/02/17 09:53 AM
|
|
|
> Is anyone actually moving from MAME to RetroArch? > > Lots of people are moving from standalones to RetroArch, but considering how shitty > and old the MAME driver is for RetroArch I'm really skeptical to consider them a > direct threat. > > It's more like the majority of emu people aren't interested in what MAME has to offer > to begin with. Arcade and Apple II gamers are getting older and older. The console > emulation on MAME is poor (that should be fixed) and the UI is kind of intimidating > (I don't care if this is 'fixed'). RetroArch seems to attract people interested in > consoles primarily, and MAME being there is just a bonus.
MAME for RetroArch is straight up poor. Slash and hacked to death. Bloody mess. ![](//www.mameworld.info/ubbthreads/images/graemlins/oops.gif)
I rather use the official MAME Android port than this horrible Frankenstein version. MAME Dev doesn't have the time to finish the project on this port. Still in alpha stage.
|
W11 Home 64-bit + Nobara OS / AMD Radeon RX 5700 XT / AMD Ryzen 7 3700X 8-Core 3.59 GHz / RAM 64 GB
|
|
|
Re: MAME vs. RetroArch [was Re: MAME 0.192]
[Re: anikom15]
#371568 - 12/02/17 01:20 PM
|
|
|
> Is anyone actually moving from MAME to RetroArch?
Moreover, is anyone important - i.e., potential contributors of value - moving from MAME to RetroArch?
From my perspective, the people moving from MAME to RetroArch are generally the annoying, irritating users against whom we've railed for literally years: The ones who are only interested in playing free games.
Despite Haze consistently proclaiming gloom and doom for the project, anyone can look at the list of contributors and commits on Github, and see as plain as day that the number of contributors to the project has absolutely exploded over the past two to three years.
The "free games" crowd are flocking to RetroArch, which if anything has left MAME free to establish itself as the "everything else" emulator, and I'm rather happy with that. I don't know if Haze has forgotten or what, but for years MAME devs have been consistently frustrated by users only interested in using it for free games, and now the optics have changed to make it more obvious to potential contributors what MAME is all about. That seems to be paying off in a big way, just not by the metric Haze uses, which is the number of users who will throw rose petals everywhere he walks.
Yea, though I walk through the valley of the shadow of death, I will fear no evil.
|
|
|
|
Re: MAME 0.192
[Re: MooglyGuy]
#371569 - 12/02/17 02:06 PM
|
|
|
> > those crap old chipsets are actually still very, very common, I still know a lot of > > people with 'junk' like that for their main PC, because for internet etc. they just > > use a smartphone. > > > > they typically end up finding that RetroArch works better on such hardware and > > ditching proper MAME for good unfortunately. > > Prove to me, with photos and screenshots, that RetroArch will run on PCs that have > video hardware which doesn't even support non-power-of-two textures, and I will > revise (but not recant) all of the following. Having said that: > > You consistently rail against the universe for people turning to RetroArch, yet > you're now also railing against even the slightest change that involves bringing > MAME's rendering pipeline into this century, which - if implemented simultaneously > with a UI overhaul, and the planned support for things like 3D art assets in layouts > - could actually stand half a chance at clawing back some of those users. Can you > please clarify your standpoint such that, to me at least, you don't sound like a > hypocrite who, typically, abounds with criticism for what everyone else is doing > while being convinced that his own feces is without scent? > > I mean, really, Haze. You're choosing this thread - this thread, of all of them - to > moan about how MAME devs are driving people towards RetroArch? Can you point out > which emulator it is that could make Hyperstone games run full speed on Diet Go Go > Fan's machine? Because from where I'm sitting, there aren't any. And if he (or she, > what do I know) happens to switch to RetroArch, then he will effectively be using an > older version of MAME, which is slower, because this is a thread in which we're > discussing my optimizations that made a driver faster in this version of MAME, you > implacable knob.
I'm reluctantly suggesting he try RA, because I've heard success stories for cases where MAME no longer works, you're reading far too much into what I said.
|
|
|
|