HalfNES Version Notes:

0.014 (11/1)
Completely rewrote the sound core, it´s an order of magnitude better sounding than it was.
Slower,too, but that was necessary. Now updates sound every time something changes.
Envelopes are correct too. Can make it faster by moving the counters forward in larger
increments since it doesn´t generate samples at 2 Mhz anyway. Also added in frameworks
for options dialog, about box, and command line args support.
Missing still:
Sweep unit, noise channel loop flag (MM2 Quickman stage), all of the DMC stuff, vertical scrolling,
just about everything about the PPU,really. And half of MMC1 roms still don´t boot,and if I want to
make the options window et cetera work right I´m going to need to re-abstract the interface away
from the actual emulation code. Why did I change that in the first place? Plus all the other older stuff.
0.013 (10/27)
Haven´t worked on this for a while. Audio code has length counters,envelopes, and noise channel,
but it really needs to be rewritten to not use wavetables anymore and update maybe once a line or something.
Even better would be a catchup type design.
Anyway, other things that still don´t work:
-Hardware pitch sweep
-the LINEAR counter for the triangle
-Envelope decay sounds terrible, especially when it loops (Super Arabian)
-Noise frequency does nothing
-Other specific sound bugs: Commando titlescreen, SMB jumping sound
-MMC1 mirroring switching
-Vertical scrolling. Had it almost working but tore it out again when I just could NOT read the attribute tables properly.
-Games: Kid Icarus, Antarctic Adventure, Milon´s Secret Castle, anything MMC3
-Video rendering should be at least scanline granularity, ideally pixel at a time catchup whenever a register is written; same for audio engine.
-Sprite hits should be real
-CPU doesn´t add a cycle penalty when crossing page boundaries. Mostly because i´m not sure what counts as this. This auses most of blargg´s tests to fail. Also the CPU core is very slow, see if you can find a better way to write the switch statement or at least see if you can make it compile into a jump table.
-I need a settings window, at least to map keys and change sound volume
-Also need SRAM saving for Zelda
-MMC3

Once (if) all this is in I can call it Version 0.1

0.012 (10/4/good buddy)
MEGA MAN 2 works! Mapper code was fine the whole time but CHR RAM write + MMC1 = disaster.
So time to start fixing the sound code.
Also, let MMC1 games change mirroring.
Plus... vscrolling still doesnt work.
Also there seems to be something seriously wrong with PRG RAM.
Maybe it´s just that Zelda etc expect certain blank fill for a new game
or maybe writing just doesn´t work at all.
Figure it out, but not tomorrow since gotta study!
0.011 (10/3/10)
Side Scrolling Works! But Vertical Scrolling is
Broken and Galaga looks Terrible. Check For Other Regressions.
I Really Need To Start Fixing Those.
also : get the sync better since audio is a full second behind.
and still sounds bad. anyway.
0.01 (10/2/10)
Deserves the version bump, because MMC1 CHR switching works now and I can play Tetris!
(In other news, progress has pretty much stopped. I wonder why.)
Still, bugfixes, the audio was an octave too low and I never noticed, etc.
Still left with the Big 3 of a proper APU, scrolling, and MMC1 PRG.
Also you can push Left and Right at the same time and I need to make that not possible.
And I really need a volume control as well. Someday.
0.009
Finally found the nasty gfx corruption bug - or let´s just call it an unimplemented feature. Pass more PPU tests,
sound is horrid so I´ll start working on that but wanted to save a copy first because i haven´t been and now I can´t find versions where games
used to work. Well, win some lose some. SMB actually works now! Just needs background sprites, scrolling, and some palette fixes to be playable.
Too bad Antarctic Adventure and Milon´s don´t get as far as they used to.
0.008? (9/26/10)
Palettes! Also more speed.
broken: mappers,still some cpu bugs left,scrolling,8x16 sprites,speed when not running a nrom game. Sound is still horrid.
0.007 (9/24/10)
well, i´ve done a lot of stuff to this. Plenty of things (sort of) work.
Sound, sprites, controller reading, screen refreshes...
But now I REALLY need to do some refactoring, so I can pass
the Graphics context of my window to the PPU to be rendered upon properly
and get REAL sprites and scanline effects.
But wanted a separate version first.
0.006 (9/22/2010)
Well, i did this instead of... you know... studying, but it works!!
I finally get an image and it doesn´t crash!
Well all right, pallettes aren´t in, sprites aren´t in, scrolling
isn´t in, controller support isn´t in, Donkey Kong Jr crashes,
and to refresh the displayI have to minimize the window,
but what the hell! It works!
I think I´m justified in skipping a version number.
0.004 (9/21/2010)
CPU Core is faster and actually works properly on something that´s not
NESTest! (all the interrupts were egregiously broken.)
Cycle count still doesn´t quite match up but whatever.
GUI is in (sort of), PPU palette is in, CPU actually works now, so the
next thing is going to be doing the nametables. and I have GOT to find
a faster way than just directly writing each pixel to the image.
0.003 (9/7/2010)
CPU Core sucks less! Passes nestest up to line 5000 of the log or so
(and at that point it starts running multi - byte NOPs so screw that)
PPU is barely started, need to read up on Swing, but for now at least
Balloon fight loads up far enough to do some writing to the PPU.
So i´ll start work on that when I have time. (prob. not until next week)
0.002 (9/4/2010)
I guess i´m going to call this 0.002 since i´ve got the CPU (sort of) working now
All (standard) opcodes but ROL and ROR are implemented (still buggy).
In particular:
-RTS goes to an utterly wrong place
-Some branches still keep triggering even when they shouldn´t
-Speed is probably gonna suck
PPU is NOT implemented yet beyond the status and control registers and even then only vblank and interrupts.
I haven´t even started thinking about mappers or controller yet.