These were retrieved from the Mac port (v0.0.7), which kept these
functions.
Unfortunately, this port is missing local variable names, so we can't
use it to make InitBack accurate.
The CMake file allows you to compile the accurate branch with
whatever version of Visual Studio you have lying around, without
having to clumbsily convert the VS2003 project.
I've tested this with VS2019, VS2003, and VS6. VS6 is goofy - it's
missing a few types and constants, and it's not smart enough to
realise that ints and longs are the same in ILP32 data models. I've
added a few small hacks to address this. Might undo them. Who knows.
For now, I want to support VS6 because Mint compiled CSE2 with it
before, and because VS6 uses `msvcrt.dll` as its C runtime, which
apparently comes pre-installed in Windows, as opposed to all those
other annoying runtime versions that require they be installed
separately (which is why MinGW targets it specifically).
Also, VS6 *should* give us Win95-compatible builds. The internet says
MSVC2003 is Win95-compatible too, but Mint claims the vanilla EXE
doesn't run on there. I imagine it has something to do with its
static runtime library (VS2003 links the static one by default for
some reason).
Visual Studio 2017 doesn't like them being 0 and 1: the compiled EXE
uses the small icon as the taskbar icon. Changing them to 101 and
102, like CSE2 did before the accurate-portable-split, fixes this.
This used to be a feature before the accurate-portable split, I'm
just restoring it.
Previously, while the EXE itself had a unique icon, the window and
taskbar both used the generic 'small' icon, which the original EXE
only used for the window.
SDL2 gives us a way to assign separate icons to each, but it's a
little clunky: it's Windows-only, requires the icons be in .ico
format, and needs them to be embedded in the EXE as resource files.
Also, for some reason, SDL2 doesn't let us refer to them by name - we
have to use their numerical ID.
The original code didn't account for it failing, which leads to
annoying bugs happening whenever it does fail.
Now, the game just closes, like it does with any other init error.
There's no evidence Tags.h contained these variables (if anything, it
appears that Pixel manually declared them in every file that used
them).
gg Pixel
This matches the original EXE. I didn't catch it while doing the
ASM-accuracy checks because absolute addresses don't match yet.
Thanks to Gabe for noticing this.
Now the SDLSurface backend survives window resizes (also triggered by
alt-tabbing while in fullscreen), and the SDLTexture backend properly
regenerates its textures after a fullscreen alt-tab in DirectX mode.
Storytime: Cucky's original SDL2 port work involved using SDL2's
threading API to emulate the original WinAPI threading.
I can't be assed with that stuff, so I used the same trick Cucky did
for the Wii port, and hooked Organya up to the SDL2 audio callback.
This actually opens up the possibility for perfectly-synchronised
Organya playback. By that I mean, instead of needing a super
low-latency audio callback, I can have the callback synchronise its
audio mixing with Organya itself. I haven't done it yet, I plan to
soon.
