Change log - July 05, 2017 - 09:11

Changes:


* forgot this (nw)

* retofinv.cpp: Added original Taito 68705 MCU dump, and moved bootleg 68705 MCU to its own new set. Renamed sets to make it more obvious which are bootlegs. Verified XTAL, dividers, and SN76489A type from pcb pictures. Added missing original Taito CLUT PROMs and fixed video/retofinv.cpp to be able to use those or the single bootleg CLUT prom. [Corrado Tomaselli, Lord Nightmare, brizzo]

* sm510: small update to mw56 svg (nw)

* fix clang build (nw)

* Merge pull request #2439 from inflatador/patch-1

* Updated Fedora instructions

DNF is now preferred package manager. alsa-lib-devel is also required (not mentioned in the current docs)

* remove unnecessary verbosity (nw)

* output: Fleshed out network output provider so it has the same capability as legacy Win32.

Clients:

- Connect to TCP port 8000

- Messages will be of the form "verb = value\1"; if your client is busy when MAME is spamming, you may get multiple messages glued together by \1 separators. (\1 was chosen because it's neutral on Win32/Mac/Linux).

- You will get a "hello = 1" message upon connection to MAME, and a "mamerun = 0" message when MAME shuts down.

- You may send "send_id = n" to MAME to get IDs where n=0 means ROM set name of current game, 1-? = output node names

- MAME will reply "req_id = string\1"; this is currently the only case where a string will be returned instead of an integer value.

- A working example POSIX client will be released soon.

* segas16b.cpp: Replace goldnaxe2 317-0122 i8751 bad dump with working redump from undamaged chip. Remove MCU HLE code. [anonymous, digshadow]

* stlforce.cpp: switched to oki map instead of ROM_COPY (nw)

* removed some redundant flags: MACHINE_IS_SKELETON = MACHINE_NOT_WORKING | MACHINE_NO_SOUND (nw)

* Merge pull request #2438 from DavidHaywood/040717

* actually after all that the CPU is a duplicate of another one, just with a different number (nw)

* Merge pull request #2437 from DavidHaywood/040717

decrypted SEGA_315_5133 used by 'seganinja' set, promoted it to working [David Haywood]

* case (nw)

* decrypted SEGA_315_5133 used by 'seganinja' set, promoted it to working [David Haywood]

it's the exact same code revision as the other sets, just for a different encrypted CPU, so not actually that interesting.

* Merge pull request #2435 from DavidHaywood/200617

* m62 proposed cleanups.

I've split the massive PROM blocks up into something more modern looking (the single PROM region with all PROMs crammed into it is an old MAME thing from when only one PROM region was allowed) I've also made use of multipe palettes and gfxdecode structures (required using rgb32 to draw to)

As the boards are quite modular this seems to make sense to me, the palette for the sprites and the sprite decode + sprite palette decode is the same across all games while the rest of the hardware changes a bit. This should allow us to step closer to converting the various boards into devices and piecing them together more logically.