Change log - December 31, 2016 - 18:06 PM

Changes:


* xbox: graphic improvements: (nw)

add support for packed float vertex color r11g11b10f

corrected drawing of indexed vertices primitives

add "xbox nv2a_wclipping" debug command

* Merge pull request #1898 from mamehaze/291216

* new WORKING game - Hangzo (Japan, prototype) [ShouTime (+others)]

* Fix compile for I8039 test code in mario sound. (nw)

* Hook up luigi walking sound in netlist mario sound implementation.

Mario driver now uses netlist audio implementation instead of discrete implementation. The previous discrete sound emulation has not been removed yet because it still contains a lot of documentation. [Couriersud]

* -stuntcyc: Fixed some NOR gates that should have been NANDs, fixes the playfield. [Ryan Holtz]

* gkigt: hook up nvram and add initialized defaults for ms3 & ms72c. (nw)

* Fix OSX build - no idea why it wasn't breaking before (nw)

* Merge pull request #1895 from ajrhacker/osbexec_fix

* osbexec: Fix IRQ-related regression; simplify handling using input_me…

* osbexec: Fix IRQ-related regression; simplify handling using input_merger_device (nw)

* cstr() ==> c_str() (nw)

* Align noexcept usage. Rename register_con to add_terminal for clearity.

Fix bug introduced with last commit. (nw)

* Analog to digital proxy rework. (nw)

Properly handle connected inputs when creating a-d proxy. Aligned a-d proxy to d-a proxy class structure.

* Merge pull request #1815 from zx70/master

Support for earlier COS versions on the RM380Z

* ScrollUP fix for the Otrona Attachè

The implementation of the scroll register is partially delegated to the specific machine video driver.

This same approach was used on the ITT3030 emulation and is probably ok for this peculiar controller.

* Scrollup fix for the TMS9927 family

This fix will avoid to reconfigure all the CRTC parameters on any cursor shape change.

This invasive "reset" procedure was resetting the scroll counter !

* Update aussiebyte.h

* Update aussiebyte.cpp

* Update rm380z.h

* Update rm380z.cpp

* Merge pull request #3 from zx70/zx70-patch-3

* Making the RTC chip react on the Aussie Byte

* Merge pull request #2 from zx70/zx70-patch-2

* Making the RTC chip react on the Aussie Byte

* Merge pull request #1866 from sparrowred/master

* correct clone relationship for zoopj (nw)

* reverted last change (nw)

* Merge remote-tracking branch 'upstream/master'

* correct clone relationship for zoopj (nw)

* Merge pull request #1872 from npwoods/imgtool_coco_rsdos_cleanups

* [Imgtool CoCo/RS-DOS] Miscellaneous cleanups and C++ modernization

* Merge pull request #1881 from system11b/S11darkw

* cvs.cpp: popmessage in some games only appears with VERBOSE=1 now

* Merge pull request #1883 from mamehaze/xxx

added decapped dump of the d8751 on quizard4 sets [Team Europe]

* also decapped / dumped an MCU for the original quizard (1) [TeamEurope]

* note that the dumped Quizard 4 MCU is German region.

* fix flags on these, the quizard3 and 4 game versions has never worked past coin-up to my knowledge. quizard4 does boot now tho, it was using the wrong sim before.

* replace with better dump (nw)

* added decapped dump of the d8751 on quzard4 sets [Team Europe]

(also added NO_DUMP listings to the other - MCU might control region tho, so possibly needs further thought)

have NOT hooked this up properly so not 100% sure it's good - timer interrupt code ends up jumping to unused area

000B: ljmp $016A

016A: clr tr0

016C: mov th0,#$FD

016F: mov tl0,#$12

0172: setb tr0

0174: push acc

0176: push psw

0178: mov a,r0

0179: push acc

017B: mov a,r1

017C: push acc

017E: anl c,/it1

0180: djnz r1,$013A

013A: rr a

013B: add a,r1

013C: mov r0,#$83

013E: mov p0.7,c

0140: acall $0012

0012: mov r7,a

0013: reti

0142: ljmp $E430

E430: nop

E431: nop

E432: nop

E433: nop

* Merge pull request #1892 from ajrhacker/soundlatch_pending

* New callback for generic_latch_8/16_device (nw)

- Add MCFG_GENERIC_LATCH_DATA_PENDING_CB to raise and lower a line automatically as the latch is written and read. This should make sound IRQs easier to deliver and reduce the incidence of HOLD_LINE in drivers.

- Update a few drivers to use this new callback.