This function impl does not change any GL drawing state, and neither
should Vulkan, so the constraint that this function not be called while
a graphics context is active is not necessary.
This prevented hashing of non-const qualified types. The contract for
std::hash already implies that it is a callable for an lvalue reference
of the template type.
The reference, compare mask and write mask for each face direction in
the stencil test is now dynamic pipeline state and are implicitly set to
default values when a pipeline is bound. This is implementable using
Vulkan dynamic pipeline state bits and so there is no reason not to
provide it.
In the OpenGL implementation of RHI, this requires tracking some stencil
state internally in the graphics context because the stencil state
functions require multiple inputs that do not cleanly map to the Vulkan
equivalents.
- Use easing functions for the rubberbanding values themselves, for more fine-control over the end-points and the curves.
- Fixed K_UpdateRubberband being broken from using shift instead of divide.
- Rename `debugbotpredict` to `debugbots`, it now displays some of the bots' botvars values as well. If the player isn't a bot, it will show them from the first place bot (for inspecting the rubberband value).
- Bot turning buff was increased, from x1.25 to x2.0. Noticed that Tails bots were failing turns on Popcorn Workshop. I personally think that Tails bots shouldn't be failing very many turns unless if they're finding objects, and constant wall-bumping should be reserved for Eggman and Metal :P
This removes a call to G_DestroyParty from CL_ClearPlayer.
The problem with calling it from there is that the
consoleplayer of a splitscreen is removed first, the local
party is cleared. Then G_LeaveParty assert fails when the
next splitscreen player is removed because the console's
party was already cleared.
- Lv.1 has a significanty lower minimum top speed (from 75% -> 50%)
- Lv.MAX has a significantly higher minimum top speed (from 112.5% -> 125%)
- Every bot level has a slightly lower maximum top speed (Lv.1: 100% -> 80%, Lv.MAX: 220% -> 200%)
- Uses easing functions now to make the distance rubberbanding applies to, and the rubberbanding curves, much more customizable.
- If the player's MT_KART_LEFTOVER is being followed instead of their flung corpse
- Flicker with the object.
- Else, don't draw anything
Fixes the teleporting corpse minimap tracking when you NO CONTEST.
- Center generalised text without the "got through round" backing
- For spectators observing a Round Queue, show "ROUND (n)" at the top instead of the more generalised fallback strings
- Uphill check is ignored if you have slope resistance, or if you're moving fast enough to overpower it.
- Offroad check is ignored if it wouldn't slow you down.
- Made slow acceleration check have a higher range.
Before, it did level up based on how much the status quo was messed up (if a loser bot got pushed in front, or a rival bot got pushed ahead).
Now, bots will additionally level up based on the best of the 4 players' overall performance, and the performance of the human player that beat them. This means that back of the pack bots will level up more often and the rival won't level up if you place poorly.
The new formula ensures ALL bots are guaranteed at least 1 level up if you come in first place, whereas the old formula had a decent chance of not leveling up some bots. However, loser Chao getting boosted from 12th into 1st, or your rival getting sent to 7th, will still wildly piss them off :B
- k_hud.c: Add K_DrawSticker function (was previously named M_DrawSticker)
- Seperated the confirm handling into the menu ticker (fixes the confirm delay being framerate dependent)
- Updated to use the button graphics
