The X and Y flip flags can be applied through the stack consisting of:
- Object attributes
- Animation frame attributes
- OAM Data
Each of these negate eachother.
Confused yet? The same stack is traversed to obtain the final tile ID,
with an added layer on top for the base GFX offset and the offset for
the dynamically loaded GFX requested by the object!
wBattleAnimDelay is populated with the values passed to `anim_wait`.
Having the pic animation macros be the same as the oam animation macros
isn't really turning out so well... I wonder if we should split them up
eventually.
These are used where the head or the feet of the player/enemy have to be
moved in an animation, and shouldn't overlap. These aren't actual GFX
and should be loaded with the proper commands, and they're always loaded
at the end of the VRAM area.
Furthermore, I've defined BATTLEANIM_BASE_TILE, which is the tile from
which battle animation graphics may start to load. This value was picked
to make sure at least an entire pokemon pic fits in the area before it,
even though it doesn't seem very used...
Using an underscore before macro variables avoids cluttering up the
global namespace. While this isn't much of a problem right now, it's
good practice that'll keep us from problems in the future.
Tried simplifying things a bit, although not nearly as much as I wanted.
Ideally, we'd either have one of two situations:
- A single set of calculations based on values depending on the
direction of the connection
- A bunch of "generic" calculations done before applying simple
modifiers to them in the final `if` block
Right now it's an icky mix of both and I'm not really sure what to make
of it.
It now only takes one numerical parameter, which is the offset of the
target map relative to the source map, much like in AdvancedMap. This
makes it easier to make connections and avoids having to calculate these
values by hand, and/or "mess with the values 'till it works", as many
have been doing thus far. It's just one, easy-to-understand value.
To convert from the old macro to the new macro, just take the fourth and
the fifth parameter, and calculate `<4th_param> - <5th_param>`. The
result is the value required for the new macro.
These are by far not always checked each week, and as such shouldn't be
called that.
Since they're almost always used through the `bit` instruction, it's
very inconvenient to just make wDailyFlags a `dw` instead.
Some character codes were erroneously named after their text command
counterparts. This has caused a lot of confusion with naming their
functions and with other things.
I've also removed the `dict2` macro and expanded the `dict` macro. This
really isn't something we should be doing for macros but I can't deny it
looks a lot neater than repeated code.
Out of the current two most prevalent official conventions, this one was
chosen, because they're used in a direct counterpart to this game, and
thus feel more accurate.
`hittarget` was misleading, as it doesn't actually "hit" the target, it
only plays the animation. As such, I've renamed it to `moveanim`.
`checkfaint` has nothing to do with checking if the user or target has
fainted, instead, it applies wCurDamage onto the target's HP, hitting
the substitute if applicable. I've renamed the command to `applydamage`
and all of its subfunctions accordingly.
I have no idea why this was a thing (do people store this repo on FAT32
flash drives or something?), but quite a bit of files had a permission
of 755. This isn't really a problem, but it's inconsistent and weird.
Since we're aiming for consistency across the codebase, I believe it
includes a uniform way to refer to the creatures this game consists of
in the labels of the code.
The only exceptions to this rule are labels referring to things named
through the use of the <PK><MN> or <PKMN> characters, in which case PKMN
is used.
Most of this was already consistent enough™, I just picked the
convention with the most occurences and fixed the outliers.
