 |
|
Downloads
|
 |
|
|
|
Forum
|
|
| Mitglieder Online | | Keine Mitglieder online. |
You are an anonymous user. You can register for free by clicking here |
| |
|
|
|
Review of the moment
|
|
|
|
|
Von M.E.S.S. - dem Multi Emulator Super System - ist gestern spät Abends die Beta v0.61.1 erschienen. Obwohl die Versionsnummer eher auf kleine Bugfixes schließen lassen würde gibt es diesmal eine ganze Elle von Änderungen und Neuerungen, siehe auch den erweiterten Text.
New System Drivers Supported (in no particular order):
-----------------------------------------------------------------
- Amstrad family (fixed and reenabled) [Raphael Nabet]
System Driver Changes:
-----------------------------
- [COCO1] Update CRC for bas10.rom. Previous dump was bad. I would like to send
a pecial "Thank you" to Sellam Ismail (http://www.vintage.org) for finding a
Color Computer that contained Color BASIC 1.0. [tim lindner]
- [COCO1/2/3] Moved disk handling code over to new FormatDriver system;
allowing transparent disk image handling for JVC, VDK, and DMK file formats.
[Nate Woods]
- [COCO] Fixed a recent stupid bug in m6847 timing that had an effect on
certain timing situations (bug #249) [Nate Woods]
- [COCO3] Fixed a palette timing bug that affected mid-frame video rendering.
(bug #253) [Nate Woods]
- [COCO3] Fixed a bug where hi-res text characters with the high bit set were
rendered as garbage. [Nate Woods]
- [COCO3] Fixed a bug where $FF9D and associated registers were being updated
before vblank, when they should have been updated after. (bug #252) [Nate
Woods]
- [COCO1/2/3] Made sure that SAM (and GIME on the CoCo 3) registers are
properly reset after a crash. [Nate Woods]
- [DRAGON32/64] Made artifacting disabled by default, as this is how PAL systems
would normally be. [Nate Woods]
- [DRAGON64] Fixed a bug that prevented the system to run (the dragon64 driver
was using the coco memory map) [Nate Woods]
- [DRAGON64] Adjusted the memory mapping scheme, enabling the DragonDOS
cartridge and 64k mode to coexist. Thanks to Simon Hardy for tracking down a
better algorithm. [Nate Woods]
- [A5200] Fixed a regression in 0.61 that prevented the keypad from working.
(bug #246) [Nate Woods]
- [A7800] Suffixed the name of the a7800 driver with NTSC and marked the PAL
version as a clone of the NTSC version. [Nate Woods]
- [GBCOLOR] Implemented background priority. This fixes objects appearing over
the background when they shouldn't be. [Anthony Kruize]
- [TI99/4A] Added support for BwG clock and floppy controller [Raphael Nabet]
- [SNES] Memory map is more accurate, adding support for HiRom carts. Fixed
DMA transfers. Fixed missing tiles. Added preliminary mode 7 support. Added
subscreens, including colour addition/subtraction. Fixed vertical flip for
large objects. Improved vertical/horizontal latches. Improved ROM loading
with better header and ROM type detection. Improved the SPCSkipper so a lot
more games actually run now. [Anthony Kruize]
Source Changes:
---------------------
- The MESS core will not automatically open software images in a clone system
drivers directory anymore, under the belief that this functionality belongs
in the OSD code. [Nate Woods]
- All MESS drivers need a SYSTEM_CONFIG (formerly COMPUTER_CONFIG) parameter
specified. Also changed the implementation of the SYSTEM_CONFIG_* macros to
provide a more flexible system for representing the data.
Added the ability for devices to be declared within the SYSTEM_CONFIG block.
Consider the older way of declaring devices deprecated.
Added new functions for iterating through devices, in order to better
abstract the internal representation of devices:
device_first()
device_next()
device_find()
Also renamed GameDriver dev field to dev_ in order to break any source
dependencies on the dev field. [Nate Woods]
- Added new memory pool code that implements memory pools that can be freed in
one step. [Nate Woods]
- Overhaul of code that maintained the list of loaded images; moved to
mess/image.[c|h], and changed the prefix of them from device_ to image_ (as
opposed to the device_* calls that are used to access the IODevice
structure). Also, implemented these new calls for drivers to use:
image_malloc() (similar to auto_malloc(), but for the lifetime of devices)
image_strdup() (uses image_malloc() to allocate strings)
image_exists() (tests for the existance of a loaded image)
image_filetype() (accessor for file extension)
For the most part, there is no longer any reason for MESS drivers to use
malloc() and free(), so calls to these should be eliminated over time. [Nate
Woods]
- Removed IO_RESET_ALL, as it was broken and an ugly hack to begin with (bug
#239) [Nate Woods]
- In the opening display software info box, if the filename and the goodname
(as determined by CRC) match up, the goodname will not be displayed in
order to reduce screen clutter. [Nate Woods]
- Removed all calls to device_close(IO_CASSETTE) in system drivers, instead
provided a cassette_exit() function that drivers can specify, removing this
responsibility from system drivers. [Nate Woods]
- Implemented preliminary paste functionality; it attempts to decode the
driver's input tables and attempts to create a mapping between ASCII and
input signals. This process is far from perfect at this time. In the
future, it will be possible for drivers to customize the way that paste works
for a more perfect translation. [Nate Woods]
- Added new PORT_KEYx macros that should be used in place of PORT_BITX when
mapping keyboards. These macros also specify unicode values for the input
codes for natural keyboard support. [Nate Woods]
- Added a new osd_keyboard_disabled() call. If this returns non-zero, then
the core will supress input of type IPT_KEYBOARD. This is useful if a front
end wants to implement a natural keyboard. [Nate Woods]
- Added a new osd_parallelize() call, for basic SMP acceleration. [Nate Woods]
- Renamed osd_dir.h to osd_mess.h, to further reflect that it is more than just
directory calls. [Nate Woods]
- [Win32] New -threads parameter, to specify the maximum amount of threads used
for osd_parallelize(); defaults to the number of processors in the system.
[Nate Woods]
- [Win32 GUI] Printers are now treated separately from the other devices; you
do not specify printer ouput like you do disk images. Printer output now
gets specified in the configuration tab of the system properties. The added
benefit is that the same output file can be used repeatedly. [Nate Woods]
- [Imgtool] Imgtool floppy drivers can now be "templated"; a driver for a
particular file system format can point at a structure that lists allowable
disk image formats. This allows for the addition of new disk image format
drivers without touching the imgtool drivers. [Nate Woods]
- [Imgtool] Fixed a bug in the FormatDriver handling that caused some file puts
to not be properly written. [Nate Woods]
- [Imgtool] Added Color Computer DMK disk image support. [tim lindner]
|
| |
|
| NEWS | Benutzeranmeldung | 1 Kommentar |
|
| | Wir sind nicht verantwortlich für Kommentare unserer Benutzer |
|
