So instead of fixing the bug they actually replicated it into a second function. :(
It's too bad that we cannot enable proper writing of NaNs and Infs because it makes most JSON formatters emit an error on this undefined extension of the format.
* This was causing the first input packet to only be sent to the playsim after the first tic occurred.
* Previous: player 1 of 1 (1 nodes); E1L1: Hollywood Holocaust; PlayClock: 4; PlayClock: 4; PlayClock: 12;, etc.
* Current: player 1 of 1 (1 nodes); E1L1: Hollywood Holocaust; PlayClock: 0; PlayClock: 4; PlayClock: 8;, etc.
* Fixes#900.
* The setting of `Player::bIsFiring` needs to be done before setting a new weapon.
* Tidy up some weird reverse engineering `while (1)` loop which GDX doesn't do.
* Since game maintains its own clip capacities, we need to use them instead of the generic handler for the games that do not.
* Clean up the game-side code in `CheckClip()` to use `min()`.
* Repair issue with pistol clip calculation that was using modulo when it shouldn't. A full weapon with 300 will divide into 6 with no remainder, therefore nothing in the clip.
* Fixes#906.
* Rename a few variables for better clarity as to what's going on.
* Add new inline to the backend `getWrappedIndex()` that properly wraps around negative values based on the supplied maximum using modulo.
* Directly set the weapon when doing the next/prev tests to avoid unnecessary tests to see whether the player has the weapon or not.
* Ensure's a smooth return to 0 so the player's actual pitch can be targeted towards Ramses.
* In situations where the player had their view pitch changed in the ticker, this would have never reset.
* When I read this against the generic weapon pickup code, it looked similar enough but the grenade's logic is reversed. That is, after picking up the weapon, equip the player. The generic weapon handler equips the player, then gives them ammo.
* Fixes#904.
* It needs to come after the function it's been moved into, but its confusing why its calculating a new pos delta when there's one on the stack.
* By directly nesting it in, it's a lot clearer.
* We can just pass the needed value into `doPlayerPitch()`.
* For Ramses, no need to set `nDestVertPan` to the map value, just use the map value.
* Avoids an excess `atan2()` call by way of `maphoriz()`.
* Use `maphoriz(-11)` instead of some voodoo degree amount so it's clear when comparing sources.
* Don't stop interpolating while doing Ramses.
* Apply Ramses pitch amount to `nVertDestPan`, just like the original.
* Don't reset all angles when initialising Ramses, just set the yaw angle.
* Don't return player's view to center when doing Ramses, use the delta between `nDestVertPan` and the player's pitch, using same formula as original.
* Comment in how normal/non-ramses `nDestVertPan` setting should be 8x the sprite Z move delta, not 2x. I know it's not correct, but its been this way for a long time and we probably should CVAR it.
* When doing the player's death pitch, it's 40.156 degrees (horizon of 108), not 38 degrees (horizon of 100). This was a misread of mine way back when.
* Was dropped in 572869f502.
* Game's crouch testing necessitates that this value be pre-cached.
* Interestingly, this value is 62.5 whereas the game's view height is 55. Change in late development to make the world feel larger?
* Remove some extern'd CVARs that didn't need externing.
* Change `player_struct::centeringView()` bool into proper method to handle the operation.
* Operation now properly unsets `SB_CENTERVIEW` if conditions aren't met instead of relying on implied behaviour from the rest of the playsim.
