* Add A1000 keyboards, with "new" MCU program as bad dump until we get the real 6500-1 dump - gets the correct number of keys at least
* Fix Amiga 600 keyboard matrix errors
* Don't allow 555 to retrigger
* Add some more notes about keyboards
* Add more notes and pinout for 6500/1
* Implement Mitsumi Amiga 500, 600, and 2000/3000/4000/CDTV keyboards
* Add unlabeled keys to UK layout
* Restrict available keyboards depending on system type
* Note that C-A-A reset is now broken on "big box" Amigas as MAME doesn't implement it properly, and the hack providing a fake dedicated reset line has been removed
6502 MCU: fix execute loop
6500/1: implement as device with onboard peripherals
Fix some bogus comments
All PLDs were unprotected.
On my set, rom #10 (dduxb10.bin, sprites) has addr 00004D8A with 0F, and on MAME's is 8F. It's probably a bitrot, but really not sure which one is good.
The new PROMs maybe related to priority (bad on this set).
- Replace callback configuration macros with devcb3 (slightly shortening function names in the process)
- Change configured clock to be twice the dot clock (as on hardware), rather than equal to it
- Allow the VDP to configure the screen in device_config_complete (taking into account 60 Hz/50 Hz variations), making MCFG_TMS9928A_SCREEN_ADD_NTSC/_PAL macros obsolete
- Add notes on clock outputs
msx.cpp: Use a template to simplify configuration of all the MSX1 VDP variants (nw)
bbcbc: Increase clock speed to a more plausible rate
WARNING: Withheld for now due to Vas's quibbles with the similar approach used in 577f198564.
- Allow device finder to be used as an argument for set_screen (nw)
screen: Calculate physical aspect ratio whenever required, not in device_config_complete, since the renderer caches the result anyway (nw)
cdp1861, cdp1864: Eliminate the "magic reference" constructors, doing their work in device_config_complete instead (nw)
zsg2.cpp: Remove misleading comment [nw]
zsg2.cpp: saturate the emphasis filter output. Right shifting will
cause unwanted noise [nw]
zsg2.cpp: Discharge the output filter slowly when the cutoff is set
to 0. Fixes popping noises in raycris song 12 [nw]
Use a simplistic multiply to simulate contrast control. InterPro systems write to a "virtual contrast DAC" register/latch of some kind, which presumably adjusts the reference voltage produced by the Bt438 before it enters the RAMDAC; this approximation is very crude, however it's simple and produces a satisfactory visual result.
