PDA

View Full Version : MAME v142



Demon Cleaner
4th April 2011, 13:43
New MAME release:


MAMETesters Bugs Fixed

* 04286: [Sound] deadang and clones: Missing ADPCM Audio (Aaron Giles)
* 04288: [Crash/Freeze] goodejan, goodejana, totmejan, sengokmi: Crash after OK (Aaron Giles)
* 04290: [Crash/Freeze] honeydol: Crash after OK (Aaron Giles)
* 04291: [Crash/Freeze] Many sets in 8080bw.c: [debug] Assert before OK (Aaron Giles)
* 04289: [Crash/Freeze] twinadv, twinadvk: Crash after OK (Aaron Giles)
* 04287: [Graphics] wintbob: Video missing (Aaron Giles)
* 04295: [Documentation] stdragon: The game description is "Saint Dragon". (Fujix)
* 04285: [Misc.] tenkaid: Missing NO_DUMP label for rom tydg002.u8 (hap)
* 04294: [Graphics] jrpacman, jrpacmbl: pac-man and the ghosts are not visable. (hap)



Source Changes

Converted Super Shanghai Dragonīs Eye to use the common deco video files functions, fixing a rowscroll bug in it [David Haywood]

Fix font loading in sdlmame. BDF files are just loaded as TTF by TTF_OpenFont. Load is now protected by a magic check. Also fixed a bug in SDL1.3 keyboard mapping and changed SDL1.3 code to be compatible with recent 1.3 API changes. [Couriersud]

sdl-config flags are now separated into INCFLAGS and COMFLAGS in sdl.mak. This fixes "make depend". [Couriersud]

softlist.c: split the field (added some time ago to handle compatibility requirements) into and . the former stays tied to the whole software entry, the latter gets inherited by each part. [Fabio Priuli]

Added device_t::memory() to fetch a reference to the memory interface, or assert if not present. [Aaron Giles]

Split address_space::install_[legacy_]handler into install_[legacy_]read_handler, install_[legacy_]write_handler, and install_[legacy_]readwrite_handler. [Aaron Giles]

Added variants of address_space handler installers which donīt take mirror or mask parameters, since this is by far the most common case. [Aaron Giles]

Deprecated API cleanup. [Aaron Giles]

cpu_suspend ==> device_suspend
cpu_resume ==> device_resume
cpu_yield ==> device_yield
cpu_spin ==> device_spin
cpu_spinuntil_trigger ==> device_spin_until_trigger
cpu_spinuntil_time ==> device_spin_until_time
cpu_spinuntil_int ==> device_spin_until_interrupt
cpu_eat_cycles ==> device_eat_cycles
cpu_adjust_icount ==> device_adjust_icount
cpu_triggerint ==> device_triggerint
cpu_set_input_line ==> device_set_input_line
cpu_set_input_line_vector ==> device_set_input_line_vector
cpu_set_input_line_and_vector ==> device_set_input_line_and_vector
cpu_set_irq_callback ==> device_set_irq_callback

device_memory(device) ==> device->memory()
device_get_space(device, spacenum) ==> device->memory().space(spacenum)
cpu_get_address_space(cpu, spacenum) ==> cpu->memory().space(spacenum)
cputag_get_address_space(mach, tag, spacenum) ==> mach->device("tag")->memory().space(spacenum)
cputag_get_clock(mach, tag) ==> mach->device("tag")->unscaled_clock()
cputag_set_clock(mach, tag, hz) ==> mach->device("tag")->set_unscaled_clock(hz)


Created new enum type address_spacenum for specifying an address space by index. Update functions and methods that accepted an address space index to take an address_spacenum instead. Note that this means you canīt use a raw integer in ADDRESS_SPACE macros, so instead of 0 use the enumerated AS_0. [Aaron Giles]

Standardized the project on the shortened constants AS_* over the older ADDRESS_SPACE_*. Removed the latter to prevent confusion. Also centralized the location of these definitions to memory.h. [Aaron Giles]

Converted DECO BAC06 tilemap (used by dec0.c) to be a device, reducing code duplication [David Haywood]

Deprecate the old memory_install_* macros. Dynamic installation is now handled directly by calling methods on the address_space, which have been expanded with aliases to cover all previous situations. In addition, variants with no mirror or mask value are provided to cover the common cases: [Aaron Giles]

memory_install_read*_handler(space, begin, end, mirror, mask, handler)
==> space->install_legacy_read_handler(begin, end [, mirror, mask], FUNC(handler))

memory_install_write*_handler(space, begin, end, mirror, mask, handler)
==> space->install_legacy_write_handler(begin, end [, mirror, mask], FUNC(handler))

memory_install_readwrite*_handler(space, begin, end, mirror, mask, rhandler, whandler)
==> space->install_legacy_readwrite_handler(begin, end [, mirror, mask], FUNC(rhandler), FUNC(whandler))

memory_install_read*_device_handler(space, device, begin, end, mirror, mask, handler)
==> space->install_legacy_read_handler(*device, begin, end [, mirror, mask], FUNC(handler))

memory_install_write*_device_handler(space, device, begin, end, mirror, mask, handler)
==> space->install_legacy_write_handler(*device, begin, end [, mirror, mask], FUNC(handler))

memory_install_readwrite*_device_handler(space, device, begin, end, mirror, mask, rhandler, whandler)
==> space->install_legacy_readwrite_handler(*device, begin, end [, mirror, mask], FUNC(rhandler), FUNC(whandler))

memory_install_read_port(space, begin, end, mirror, mask, port)
==> space->install_read_port(begin, end [, mirror, mask], port)

memory_install_read_bank(space, begin, end, mirror, mask, bank)
==> space->install_read_bank(begin, end [, mirror, mask], bank)

memory_install_rom(space, begin, end, mirror, mask, ptr)
==> space->install_rom(begin, end [, mirror, mask], ptr)

memory_install_ram(space, begin, end, mirror, mask, ptr)
==> space->install_ram(begin, end [, mirror, mask], ptr)

memory_unmap_read(space, begin, end, mirror, mask)
==> space->unmap_read(begin, end [, mirror, mask])

memory_nop_read(space, begin, end, mirror, mask)
==> space->nop_read(begin, end [, mirror, mask])


uiimage.c: Improved File Manager usage (in the internal UI) by catching char inputs and searching for matching names among files and directories [Fabio Priuli]

some work on the 3 games in dec8.c which use the bac06 tilemaps[David Haywood]

Converted Act Fancer, Trio the Punch and Mad Motor to use the common deco video functions [David Haywood]

Converted Stadium Hero driver to use the common Deco video functions [David Haywood]

Cleanup of machine.h. Shuffled some fields around, and moved several to private member variables with accessors: [Aaron Giles]

machine->m_respool ==> machine->respool()
machine->config ==> machine->config()
machine->gamedrv ==> machine->system()
machine->m_regionlist ==> machine->first_region()
machine->sample_rate ==> machine->sample_rate()


Converted Pocket Gal to use the common deco video routines [David Haywood]

Started moving common deco MXC06 sprite controller code to itīs own device files, reducing code duplication [David Haywood]

Converted Act Fancer to use common deco sprite code [David Haywood]

Converted Vapor Trail to use the common deco sprite functions [David Haywood]

Converted Stadium Hero, Cobra-Command, Psycho-Nics Oscar to use the common deco sprite functions [David Haywood]

Removed redundant machine items from address_space and device_t. Neither machine nor m_machine are directly accessible anymore. Instead a new getter machine() is available which returns a machine reference: [Aaron Giles]

space->machine->xxx ==> space->machine().xxx
device->machine->yyy ==> device->machine().yyy


Globally changed all running_machine pointers to running_machine references. Any function/method that takes a running_machine takes it as a required parameter (1 or 2 exceptions). Being consistent here gets rid of a lot of odd &machine or *machine, but it does mean a very large bulk change across the project. [Aaron Giles]

Remove redundant item cpu from address_space, in favor of space->device(). [Aaron Giles]

Two hash_collections can only match if they have at least one matching hash. [Aaron Giles]

If we get an error parsing options, see if there was no command and what looks like an invalid system name. In that case, provide suggestions for the system name instead of reporting the options error. [Aaron Giles]

Move overridable device->rom_region(), device->machine_config_additions(), and device->input_ports() to protected methods. Added non-virtual wrappers around them. [Aaron Giles]

Implemented karnov style sprites in its own file, shared with karnov and dec8 drivers [David Haywood]

Fixed multiwidth sprites in decmxc06.c, removed code which appeared to serve no purpose other than to break the birdtry gfxdecode in machine/dec0.c [David Haywood]

Added new macro MACHINE_CONFIG_DERIVED_CLASS() which works just like MACHINE_CONFIG_DERIVED() except you can specify an alternate driver_device class. Used this in the 8080bw.c games which require an _8080bw_state, but derive from mw8080bw_root which has the base class mw8080bw_state. [Aaron Giles]

lordgun.c: Corrected Dipswitch names and added dipswitch locations to Lord of Gun & Alien Challenge [Brian Troha]

tumbleb.c: Corrected Dipswitch names and locations for Choky! Choky! [Brian Troha]


New clones added

Fashion Gambler (Set 2) [Team Europe]
AAcrobatic Dog-Fight (USA) [Tirino73]


New games marked as GAME_NOT_WORKING

Jyunai - Manatsu no First Kiss [The Dumping Union, alien_mame]
BiKiNikko - Okinawa de Ippai Shityaimashita [The Dumping Union, alien_mame]

Harrison
4th April 2011, 23:47
This seems to have been a fairly big update as the rom torrent joined at 82.1%, and the overall size is 1GB larger at 27.6GB, compared to the previous release. Normally they join around the 96% completion mark.

And the CHDs have grown 12GB from 125GB to 137GB.

So the complete MAME set, with the Roms, Extras, LD CHDs and CHDs it is now totaling 252GB. I wouldn't like to be anyone trying to get hold of this for the first time. But if anyone is the offer is always open for you to send me a blank HDD and I will do a copy for you.

Demon Cleaner
7th April 2011, 18:07
MAMEUI 142 seems to have problems:


"Warning: MameUI .142 is being released but it is still taking 1.2GB in memory when loaded, this is compared to 215MB of .141. The x86 version may crash out if used w/out sufficient resources. The DEVs continue to investigate."

Harrison
8th April 2011, 00:21
That sounds like a huge memory leak in the new version!

I've updated my set to the current release but hadn't yet copied it over to my emulation PC or updated MAMEUI yet. I think I will wait until they have fixed this bug for now.

Menace
26th May 2011, 12:03
They seem to get worse as they get bigger-CHD size and rom size is high and you need a powerfull PC for some of the CHD games,i am glad i just play retro arcade games upto around the 1990`s lol.

Harrison
27th May 2011, 13:06
There are still some from the late 80's and early 90's that need a fairly powerful PC to run, especially the 3D Sega games like Virtua Racing, or Namco Ridge Racer, which I was pleased to discover runs pretty well now when I tried it last time.

Demon Cleaner
28th May 2011, 10:22
Naomi ROMs are tough to run.