20-5-24
	- Implemented pushWalls by basically mimicking what I achieved with the door code. So the effect will be achieved by using offsets, and leveraging the fact that the user will not be able to see any other side of the wall when it's being pushed in. At the end at 100%, the game replaces the offset texture with a real wall.
	- Still having trouble rendering the offset effect, will be fixed tomorrow.
21-5-24
	- After dabbling with threads and not seeing much improvement, I ran the program through GCC and noticed how many threads were being created and destroyed each frame. So I have reimplemented the threading a bit, making a pool which are constantly accessed - render times immediately shot up and I gained an extra 5fps from that alone. Soon I will tackle the sprite threaded render as well using the same principles, which worked before but some z fighting was apparent. 
	- Adding some mechanics to end a level, show an intermission screen and load a next level.
	- Push walls are almost 100% functional now and look and feel just like Wolf3D's.
	- Ummed and arred about getting some sort of file format to store level data from, ultimately deciding it was outside the scope of the raycaster.
	(DONE)TODO - attenutation isn't wokring properly for audio
	TODO - make some more constness in code - not nearly enough
	(NOT GOING TO FIX)TODO - decals go through sprites (This might actually be ok)
	(DONE) TODO - Saves and Loads
	(DONE)TODO - Load config file settings.cfg

22-5-24
	- Added ability to load settings from a settings.cfg file.
	- In addition, allowed for changed resolution and detail modes (resolution scaling) - this required a major refactor. Made RENDER_WIDTH and RENDER_HEIGHT macros to call a the renderResolution function, makes it easier for refactor.
	- The importance of loading in settings from .cfg early mustn't be overstated. On future projects, I will design a settings loader early, avoiding #defines and allowing for bug detection etc with easy changing of the file and not having to recompile.
	(DONE) TODO - 1 pixel underdraw for secret walls
	(DONE)TODO - enemies see player through door
	TODO - shoot sound and rate attached to framerate?

23-5-24
	- I considered last night only making the 3D render be affected by the "resolution scale", and in a proper game this would be what you would do, but for my purposes affecting the whole renderering pipeline is sufficient.
	- Adding startup command line eye candy, reminiscent of Doom and Duke3D's start up.
	- I've been really impressed by the quality of life improvements that the settings.cfg gave my programming workflow, so I've decided I'd like to implement some more:
		- auto kill of process upon remake inside gdb (DONE)
		- Pre-compile c files into .o, so that building a new version is very quick if I've only changed one module.
		- set up an auto backup of raycasting-c to a zip which saves to some folder at a specified time each day.
	- After toying around with the idea of making a dedicated mapBuffer which then gets drawn to the main colour buffer, I've decided just to keep it simple and scale sprites and textures naturally onto the screen.
	- Bit tired this afternoon, will do some busywork to remove compilation warnings
	(DONE?)TODO - shooting secret wall segments (during?) or after secret press gives a SEGFAULT. 
	TODO - see about how we could record and playback demo's, using the functionality inside timer.h

24-5-24
	- After tossing it around in my head overnight, I came to the conclusion that only the 3D buffer should be scaled in quality, not the hud or menus. So after a bit of refactoring, quality now can shift and leave the hud, map and menu systems alone.
	- 3D buffer now changes directly based on changes in detail, and the other buffers are left untouched.
	- I am LOVING this new system where I have precompiled .o files for all my other c files. So fast, so easy.
	- Fixed vertical push walls not working correctly. (Was a logic problem in the case in secret.c)
	- might just leave the 1 pixel off thing for secret pushwalls as a "feature".

26-5-24
	- Made sprite.c multithreaded and refactored the code in that module, splitting it in 2 and making it slightly easier to traverse. Performance gains were marginal at best however, but I'd like to test with a high number on sprites on-screen at once.
	- ALmost fixed the pixel fill wipes.
	- Disabled changing detail when GAMESTATE != GAMESTATE_RUNNING

27-5-24
	- In the doldrums of a bug today that finally got fixed -- The pixel wipe function now works for all detail settings. Render pipeline is a bit messy.
	- Rest of the day is implementing and expanding the level editor. Adding Doors, pushWalls, Exits

28-5-24
	- Further implementation and refinement of the Level editor, in preparation for adding final features to the game's C source.
		Final map features will be:
			(DONE)1) Routing for enemies to walk
			(DONE)2) Key Doors
			(DONE)3) Exit switch
		Then we can go through the level ending screen stuff and handling next level
		Menu stuff	
	- Editor now capable of displaying all the placeable features of our engine.

29-5-24
	- Bug related to a segfault in texture reading wall.c had something to do with the format of the png - it was not compatible with the program.
	- upng images must be saved in RGBA it seems - Updated texture.c to make the check and issue warnings.
	- Waypoints are now loading into the game, refactored file.c a bit.
	- Brightness can now be changed via settings.cfg
	- Exit switches now function - texture changes, sound effect activated
	(DONE)TODO - certain door sounds aren't opening properly.

30-5-24
	- Brightness is now controlled via SDL and setting the colorBufferTexture just before the final stage of the rendering pipeline.
	- New RenderState allows for us to dim and brighten the entire game when needed.
	- Refactored graphics.c in order to separate out the drawing functions.
	- Refactoed main.c and split everything up in a nicer way.
	- end lecel code is coming along. Animated sprite drawn, static characters drawn, midi changed, states changed.
	- I tried to refactor the rendering pipeline to use blending SDL textures - i'd been initting a bunch of textures that I  never used. It was a 
	pretty horrendous failure.
	- Midis are now loaded by the level loading mechanism.
	- Added level Name, number, par time to mapData.txt files. Program now reads it in.
	- End Level screen now has increasing counter on time in level, to be interrupted and skip to the end just like doom.
	(DONE) TODO: Refactor input.c - dedicated functions for different things.

31-5-24
	- Refactored input.c and made it more modular.
	- Finished preliminary work on the first end level screen, we have working tickers for all metrics (time, kills, secrets, treasure)
	- Spent a lot of time with AI generating some sprites and textures. Sprites was pretty good, some half decent results, textures are going to be harder.
	- Tried out some more render resolutions - interesting but buggy results.
	- Thought of a cool game mechanic that wouldn't take long to implement - each level will have a secret "Romanov Dynasty relic" that the player has to hunt down and collect.

1-6-24
	- Generated a bunch of much better looking images, and after some touch up are perfectly appropriate for the game.
	- Almost finished level end code
	(DONE)TODO - Fix up maps so that they're centered.

2-6-24
	- Full refactor of the map drawing code - much cleaner, streamlined, all map changes can be in one place. Map works like it should now.
	(DONE)TODO - Add Map Names to bottom right, map num, time elapsed, kills, treasure, secrets (decided to just have map name an timer)
	- Refactored text drawing function so it can be scaled.
	- Redid some more sprites.

3-6-24
	- Decided to add Windows into the game, as static billboard sprites - this will be the last in-engine feature I will make. When implemented, advance game version to 0.4.
	- After 2 hours of messing around with static sprites, I've decided that it's above my paygrade. So i'm trying a new approach using the raycaster to achieve the effect.
	- After a further 2-3 hours of messing about, I ahve almost got it working. I have transparency on a wall, but it's not aligning properly.

4-6-24
	- Advanced program version to 0.4 - Windows have now been implemented in the game engine, and with that, the last rendering feature is complete. Looks so cool :)
	- This is achieved by casting a ray, and making note of all windows that it traverses. Then, renderer renders the furthest hit wall, then goes back from the last recorded window hit and draws.
	- (DONE)TODO - workout rendering kinks - 
		(DONE)clipping textures, 
		(DONE)sprite Z axis,
		(DONE)doors, 
		(DONE)pushwalls, 
		(DONE)LoS, 
		shooting mechanics, 
		(DONE)Windows in windows and recursiveness
	- TODO - Get map editor fully functioning and start build Level 1 - HOLDING CELLS
	- Really we should be finalising texture positions, item positions in textures etc. It's going to get more messy
	- OR think of a better way of texture planning that doesn't involve shifting numbers continuously
		-In the doldrums with this. Deciding to create a new texture struct that keeps pertinent info about the texture together.
	- We have 255 to work with - Which is textures, which doors, which system textures, etc.)
	- Start planning our maps

5-6-24
	- Almost pulld my hair out over a seg fault bug. After HOURS of debugging and giving up in frustration at one point, I traced it back to me assigning a upng height as its width. I had to step through each value in the buffer position calculation
	to discover it (including comparing the buffer size to the file size, and determining where the runoff occurred).
	- Feels good to get my program running again. This time I have simplified the process considerably -
		- Made a texture_t struct containing width, height, a pointer to the buffer and the texture name.
		- modified upng.c to include the filename inside the upng struct.
	This has allowed me to ditch a bunch of array matching which were bloating the code significantly and opening the door to mroe bugs down the line.
	Now, we have a codified texture system with certain limitations:
		0-4 are system textures.
		5-69 are Wall Textures.
		70-85 are Door and doortrack textures.
		86-99 are spritesheets.
		100+ are reserved for misc non-level textures (menu items, help screens etc.)
	- Exits can now have secret exits (3rd number will refer to the next level)

6-6-24
	- Work continues bringing the editor to a working state. Really excited as now I can start designing some real levels.
	(DONE)TODO - All map Tiles can be correctly placed
	(DONE)TODO - Waypoints can be correctly placed
	- Working hard to make the editor completely usable. I am not a fan of Javascript.
	
7-6-24
	- Not much to say - Working very hardon the editor, but since JavaScript is not my forte, not really interesting.
	- Map editor now functional! Already starting to find some engine bugs to fix. Development on the proper game should speed up immensely now.
	- Fixed small bug in map editing code rendering different subquadrant triangle textures.
	- Map now correctly renders windows, and fixed a few bugs related to the scaling factor. This now allows us to hide the gun in full screen map, and the map can draw behind the HUD.
	- Sprites now draw behind windows correctly. - Separated the drawing threads into two distinct parts, and rendered the sprites in the middle.
	- Windows can now see through into new windows. (Fixed rendering order by comparing distances of wall hit to Window hit)
	- (DONE)TODO - Sprites in front of windows are rendered correctly - I think this can be done by using the amount of windows and creating "zones"

8-6-24
	-Code is very buggy right now. Here is current TODO to debug:
		(FIXED?)TODO: raycast.c can include multiple of the same window ID - shouldn't happen
		(DONE)TODO: renderSprite is crashing due to sprite.distance - why?
	- This is an absolutely brutal problem. I've spent hours banging my head against the wall, and trying to debug with little success.
	- In times like these I seriously feel out of my depth - when I feel this way, I take it back to grug basics. I start commenting every line, and abstracting out everything into functions.
	- I often will make a new file and make getters and setters for everything.
	- Time will tell as to whether I can fix this. UPDATE-  fixed. Grug basics and massive commenting/abstraction to functions worked a charm!
	- (DONE)TODO: Doors can be seen through windows.
	- (DONE)TODO: Pushwalls can be seen through Windows.
	- As far as I can tell, all rendering problems with windows are now sorted.
	
9-6-24
	- Texture creation and E1M1 creation!
	- TODO - EDITOR: ticking blocking does nothing (This only happens if the sprite is not actively selected with the RED square around it.)
	- TODO - GAME: Midi doesn't loop
	- (DONE)TODO - implement key doors
	- (DONE)TODO - implement game messages at top right of screen.
	- (DONE)TODO - EDITOR: Enemies cannot be placed in editor.

11-6-24
	- (DONE)TODO - Reset is not resetting everything, getting some corruption - crashing
	- (DONE)TODO - fix startup bug with a line that's massively out of bounds drawing - this causes an infinite loop for the frame in which it's being drawn, causing delays. - was caused by a bool being defined as a float.
	- Found out you can inspect different threads inside the debugger and create void pointers and cast to different types - can Access data of many different things now!
	- Fixed switching resolutions with new render mode for windows - was an old parameter name that was misleading
	- (DONE)TODO - Decals not working.

12-6-24
	- Messaging system now works in-game.
	- Sprite decals are now (for the most part) fixed and working again - it was a rendering order.
	- Found out a way to print out a Buffer's contents into the console of gdb!
	- (DONE)TODO - Optimise the render pipeline a little bit - especially when no windows are in frame, we can go to the simpler render pass with traditional methods
	- (DONE)TODO - Enemies can see player through doors.
	- (DONE)TODO - patrol routes
	- (DONE)TODO - improve SDL input feel 
	- (DONE)TODO - Loading textures/sounds dots in console to indicate position, same with level.
	- TODO - Finish E1M1
	- TODO - Work out any further editor kinks, NaN for example
	- TODO - V1 of the chickenometer
	- (DONE) Window hit registration - maybe we make it a 50% chance for bullets to pass through windows, 50% chance for it to richochet?
	- Or, if it hits within the middle 40% of the window wall, it's guaranteed to pass it?
	- Thinking about patrol routes, we can probably simplify considerably.
		- Each route waypoint is a 64x64 tile.
		- Each route waypoint modifies the walk direction of whatever entity goes in it, as long as they're not in chase mode.
		- This simplifies the logic check to a simple isMapAt on each tick, and modifies the walk based on our predetermiend walk directions. Also, it can simply be stored in the map array of the .txt

13-6-24
	- Loading bars for telling you status of texture and sfx loading are now in the engine.
	- Sprite decal bug fixed (root cause not too sure, but made a conditional check to not render invalid buffer positions). used my kickass debugging functions to figure it out.
	- Windows now can be shot through (My solution was a tied-knot, but it worked first time to my amazement - treating decals like sprites.)
	- (DONE) TODO - something is causing the mapLinesToDraw array to be filled (or at least report as such).
	- TODO - Fix up FPS Metrics again
	- (DONE) TODO - Check decal arrays better - encountering overflows.

14-6-24
	- (FIXED) Decal array getting filled up - why?
	- TODO - on reset level,set gun sprite to its initial position.
	- (DONE)TODO - keys don't reset when player dies.
	- (DONE)TODO -Enemies flinching when shot should be highest priority to all other states

15-6-24
	- Spending the day on Linux today - will get cross-platform compatibility up to date, and will use Valgrind for perf and mem leaks
	- First problem - .txt files made in windows have different return symbols etc. that Linux doesn't use. Need to include specific use cases for reading in files in Linux.
	- FFS - why is software development this fucking obnoxious? I cannot get past the texture reading because the realloc makes the first of the array some unreadable memory address.
	- Did a bit of error fixing.

16-6-24
	- Not sure why, but address sanitizer allows the program to continue.
	- Accounting for situations when sound doesn't load.
	- Fixed an out of bounds memory access write bug in drawTecturedRect and drawTexturedQuad/Triangle
	- Today I wil profile on Valgrind to see where I can speed things up some more.
	- Created a function that will print expected repeated values on the same line without clogging the console.
	- Valgrind fruits:
		- Some uninitialised values that were causing problems before (sprite->seen, sprite-> distance) now should have proper intilisations.
		- found memory that wasn't freed (textureFileNames in textures.c)
		- rays in rays.c is never freed in main program!
		- decalBuffer in sprite_decal was never being freed. Freeing when decal is generally freed.

17-6-24
	- Patrol enemies now patrol properly, even when partially clipping into other sprites.
		It will be mapmaker's responsibility not to make this look jank, there will be no clipping for sprites while in these behaviours
	- (DONE?)TODO: Enemy painstates should be consistent. Right now only triggers behind windows... (?)
	- Want to start to implement pathfinding for enemies today. They are going to look and behave much smarter and the game will take on a new dimension.
	- Want to go with A* pathfinding.
	- Got a basic implementation in, made some pen and paper simulations to solidify understanding.
	- May as well use this situation to:
		1) Let enemies open doors.
		2) Use A* for sound propagation emulation.
	- Enemies now fire through windows and can see player through windows.
	- Fixed a bug I had introduced with walkspeed

18-6-24
	- Okay a* pathfinding file built first pass, impressive.
	- A* is working!* Having some small issues, but enemy reliably moves toward player.
	- Going to implement debug labels for enemies in fullscreen map.
	- Added a way to give debug coordinates in the fullscreen map for enemies.
	- TODO - If the map's open and a player dies, the map draws over the death effect. -- FIX: Close map upon death?
	- So A* is working as it should, but it's obvious my spriet movement as it stands needs to be more sophisticated. Enemies are clipping on walls when they shouldn't be.
	- (DONE?) TODO - Improve enemy LoS in hard to determine cases (edges of walls etc.) - fixed by forcing enemies into the center of the tile.

19-6-24
	- Enemy does not stop firing when LoS check succeeds despite code that should vary it up.
	- Okay i've decided to Grug Debug the enemy code. I've refactored everything into new files for los, rotation etc.
	- Let's map out the frame 

20-6-24
	- Refactoring the enemy logic.

21-6-24
	- (DONE) TODO - Sprite change for new rotationDraw is not being reflected quick enough when the value spriteDraw does change.

22-6-24
	- (DONE)TODO - mapLinesToDraw goes bananas, probably with an uninitialised value. -- This when msf was 0 - lines wouldn't have anywhere to draw and so went bananas

23-6-24
	- After hours and days of endless debugging, I am almost getting satisfactory results with the pathfinding. There was a bug where I was looping through sprites to find out of a blocking sprite
	  was occupying a cell, and I was repeatedly getting answers that a cell with a barrel was not blocked. I finally realised that there were 2 sprites overlapping in that cell, and that the function
	  looking at the first sprite, seeing that it wasn't blocking, and thus returning early that there was no block at that cell.
	- I've also gotten the map to display the line and the path that the enemy will walk, as well as instructed the enemy to always walk to the center of the node. This is almost perfect, but not quite
	just yet. (I will brute force this with a jank movement towards center point).
	- Guards now open doors that aren't locked.
	- (DONE) TODO - Full screen map after die doesn't flash the black when new level loaded.
	- (DONE) Secrets uncovered should regenerate the node map, as the map structure has changed If not, players can go into secret areas and enemies won't follow them in.

24-6-24
	- (DONE) TODO - Returning invalid direction at certain parts of a*. This is a mapping issue - make sure no sprite is blocking single squeeze points.
	- (DONE)TODO - Think about how to get enemy centered in tile - Brute force jank will be what I do here. Will code on the train home.
	- Some general fixing of pedantic warning and constness.
	- Upgraded program version to v0.5!

25-6-24
	- Fixed map rendering of multiWalls.
	- Changed event handling from one event per frame to multiple events using a while loop - didn't know this existed!
	- Movement now is a lot sharper and quick.
	- Noticing sprites are disappearing with some sort of mouse input?
	- (DONE)TODO - Sprites close up, even when transparent, tank the framerate and looks ungainly.
	- Fix up FPS counters so I know clearly where bottlenecks are.
	- Implemented a nice refactor based off the LBA source code I saw - using function pointer that changes depending on what detail level has been set, so we save many lines of code and millions of function calls getting the detail mode each pixel draw.
	- If player goes out of range of an NPC's pathfinding, the pathfinding should remember the last best path to the player, get there, then initiate the bouncing off wall routine
	- Idea - WHat if we precomputed what section of a sprite is transparent or not, and skip a whole chunk of the check routine for transparency by the pre-computation at startup?
	- Think - Where can we preprocess data before runtime? Where am I doing the same thing over and over?

26-6-24
	- Successful implementation! Improved framerates by whopping 9-10fps in some cases (20ms per frame). We divided all sprites in a spritesheet into 4 zones and preprocessed them at startup.
	 If there were any pixels in each zone that weren't transparent, then that zone was marked to be drawn.
	- TODO - I still see a bit of phantom slowdown in the program that isn't the usual suspects.
	- (DONE)TODO - player goes out of bounds or into an unreachable place by a*, enemy should still get a fix on last known point.
	- Bit jittery, but pathfinding looks pretty solid for now.

27-6-24
	- TODO - Menu systems should be able to start a new game and quit.
	- Refactored a bunch of code.
	- Created some new textures.

28-6-24
	- Back to the Win 10 development environment today. Created a bunch of new textures and Sprites.
	- tested out what I will call "fences" which are exactly what they sound like - wall textures marked as windows, but act like a fence. I will make some more code to make hit detection for these go straight through them for 100% of the texture.
	- Playing fantastically!
	- TODO - Dogs follow pathfinding.
	- TODO - A* variation?
	- (DONE)TODO - Door keys are not being saved in editor for door types.

30-6-24
	- AI thinkers have to be changed so we don't have all enemies thinkinga t once, or at least drawi	ng their lines??
		- So implemented a simple LoS check, one line only from player to enemy. If closer, more advanced LoS cone scan will activate.
		-
	- (DONE) TODO: Decal arrays need debugging.
	- (NOT DONE)TODO: Fence placement has to be really careful due to the "hack" I have used to remove the inner side wall (see transparent texture 3)
		-I think I have to take a pragmatic compromise here and design my levels to hide this flaw in the engine that would require significant engineering to fix. It does go beyond the scope of the original idea of windows, anyway.
	- (DONE) TODO: Avtomat weapon refire rate needs to be more refined.
	- (DONE) TODO: MultiWall dropzones should be cleared upon clear.
	- (DONE) TODO: Deleting Enemies/Sprites? -- Now possible by pressing the delete key when entity is selected in the editor.

1-7-24
	- Door saving code is not working in the editor.
	- (DONE) TODO - Door locked message displays twice.
	- (DONE) TODO - MIDI is crashing game on death.
	- (DONE) TODO - Fences take 100% of the bullets going through them.

2-7-24
	- (DONE) Debugging the decals array today.
		Okay so some decals are not being removed, despite being created.This seems to be hte cause of the bug.
		What was happening is if the sprite was not drawn fro whatever reason, the timer was not increased.
		So I changed it so the timer always increased every game loop regardless of being drawn or not.
	- (DONE) I think I have added the ability to accurately show hit detection for bullets through windows.
			-Actually this is not perfect yet... See TEXTURE_WOOD_WINDOW_1
	- Many enemies in LoS all calling the costly function check LOS - Line drawing?
	- Dogs open doors?
	- Patrol isn't working as it should.
	- Map can draw faster with the sprite optimisation?

4-7-24
	- (DONE) LoS checks need to be performed once a second to inform enemy behaviour, not once every frame.
	- (DONE) Enemies should play alert sound upon spotting player.
	- (DONE) Typing in player's angle will not allow more than one keypress (same with x, y).

5-7-24
	- Generated many new sound effects.
6-7-24

	- Today I want to process the majority of the issues with enemy pathfinding and behaviour.
	- (DONE)Load up new, smaller sound effects to facilitate easier loading into program.
	- (DONE) Fix up the input parameters for the Textboxes in the editor. - fixed - renderMap was being called on each keystroke.
	- (DONE)CurrentFrame not reflecting rotateDraw/RotateMove?
	- Enemies following same a*

7-7-24
	- Okay problem with the delay in the currentFrame is to do with "timeInCurrentFrame", this is causing a delay in the change.
		-Fixed: Added a switch bool and if that is activated, it skips the frameDuration check.

8-7-24
	- (DONE) Enemies are not aggressive enough.
	- (FIXED) Enemies move while in pain state.
	- TODO - If enemies cannot move for whatever reason, instead of spazzing out, he should take potshots at player if player is in simple LoS check.
	- (FIXED) TODO - ENEMY_STAND does not do checks every second for LoS - error was in enemy_los.c - player intersections not triggering
	- TODO - If more than a certain distance from the player, stepChecks get larger to save processing power. (no need for high precision when player is on the other end of the map)
	- (DONE)TODO - Window hit registration needs a revisit - Added a loop condition so if a window is found deep into the loop, it won't register a hit at the surface of the window.

9-7-24
	= (DONE) TODO - Enemy Aggression must be investigated today.
	- a* is broken for multiple enemies - see 3rd enemy in alcove. - Ok it's to do with check blocking nodes command

11-7-24
	-This is what was causing it:
		    bool isBlockableSpriteHere = isBlockingSpriteInCell(i, j, true);   // false if no blocking sprite found
	Having this set to true meant that certain nodes were compeltely blocked off, and that if the enemy was static they were stuck in their tile.
	There really wasn't a good reason to have this on, but we do need to implement a situation where if an enemy is in a tile, then the pathfinder skips that tile.

12-7-24
	- Noticed now for the 2nd time that pathfinding issues are occuring because a sprite I placed int he level is 1-2 pixels overlapping into another tile. So, I will implement a "snapping to tile" for enemies and
	sprites, which can of course be modified but will be the default to avoid these easy to miss pathfinding issues.
	- Now we have added some rounding for the initial placement onto tiles, which can of course be modified by the mapmaker down the line.
	- (DONE)TODO - Federov is not synced
	- TODO - RotationDraw for centretile alignment not always correct (Improving)
	- (DONE) Light intensity does not get saved in the map editor, neither does map number

14-7-24
	-(DONE) Map name and midi should properly be saved now.
	- Fixed Enemy Aggression - It was due to a bug in LoSCheckconditions - This only gets called once a second on an enemy, but the conditions to shoot should happen after 3 seconds. But deltaTime accumulation is only able to happen once per second.
	- Now ChaseTimer will accumulate each loop of updateEnemies().
	- Fixed ENEMYPAIN - it needed a timeInCurrentFrame reset when the enemy goes into their pain state.

15-7-24
	- Dog chomp range fixed. Dog attacking is a bit jank which will be addressed today.	
	- Dog won't miss his bites now due to accuracy, but perhaps if the player cycles back they can avoid the bite
	- (Fixed) Dog will start moving after chomp even if chomp frame is still occuring.
	
16-7-24
	- E1M2 almost finished.
	- Plays really nicely. As soon as I stop spazzing enemies when pathing is interrupted, I'll advance to v0.6.

17-7-24
	- More sprite replacements.
	- Fiddling with frankensteining a few sprites from other games into my own Bolshevik guards.
	- (DONE) End of level counter for (time, ?) should increase more quickly, and wraps around for some reason.
	- Last known good node seems to be broken...
		Seems to be getting set as the enemy, not the player.
		Seem to have fixed this, it's much better now. Enemies now only getting stuck on other enemies.
	- Completed my first enemy spritesheet for the guard.	

18-7-24
	- Fixed up the height of the guard. He was too tall before.
	- With this spritesheet, which took about all said and told 5 hours to make, I will advance the game to 0.6.
	
23-7-24
	- Spent the last 5 days on and off making the complete v1 spritesheet for our dog. It's now done.
	- (DONE)TODO - Time timer increments correctly in the end screen.
	- Made the bonus tickers work. Level end works nicely right now.

24-7-24
	- More sprite generation. Want to push getting the intermission mechanism sorted out.
	- Added in map screen. Decided that conquered levels will replace the Bolshevik red flag with the russian tricolor.
	- New levels now load from completion of a level! 
	- Mouse now doesn't shoot in intermission screen. (In future - )
	- Using 02 optimisation.
	- TODO: Investigate phantom slowdown, not being detected in my usual counters.

25-7-24
	- TODO - New level load doesn't change music
	- (DONE)TODO - New level load breaks pathfinding - I think this is because sprites aren't getting dumped?
	- (DONE?) TODO - New LEvel Load doesn't transition with a frameDuration
	- (DONE) TODO - New level load doesn't load in background colours
	- Implement waving flags (DONE), red blinking squares to denote where we're going next map (DONE), next map pauses to show player which map they're entering(DONE).

26-7-24
	- Level load transition is now really smooth.
	- TODO - Fix up all midi file playing problems.
		- Now, Midi won't crackle and snap as it is playing alongside level loading. Loads and plays cleanly.
		- Midis now switch between level loads.
	- TODO - Fix up the end screen (Bonus spaces, repeated menu select sounds played)
	- Start E1M3 - Train Station
	- (DONE) Design and implement the flags as spritesheets.
	- End level screen shows some nice animation with some flags denoting conquered levels, that change to the tricolor once the player has finished that level.
	  We can also do some small map animations down the line as well - Rivers trickling, shorelines breaking, clouds moving
	- TODO - enemies no longer give points when killed -why? They used to?
	- TODO - 2nd end level acts funky - likely a full reset has not been performed.

29-7-24
	- Crash when speedrunning past enemies at the exit,inside the A* pathfinding. I think this is to do with leaving enemies alive chasing you, then having a map design flaw where they can't
	 get to you, and then trying to exit to level.
			Thread 1 received signal SIGSEGV, Segmentation fault.
			reconstructPath (node=0xfeeefeeefeeefeee) at src/enemy_move_pathfinding.c:253
			253             x1 = (node->col * TILE_SIZE) + (TILE_SIZE / 2);
	- The use of static variables in my end level screen meant it became cumbersome to reset the whole module for a new end level - so I have redesisgned it using statics at the top of the .c source.
		Fixed the issues above with the end level counter and also made it a lot cleaner to work with.
	- Will start today to implement the normal program operation. if Argc == 1 it will go through a start up mode, displaying my software company's logo and the main title card.
	  After that, the main menu should load up, and in the background a demo should run.\

30-7-24
	- Added company intro and intro splash screen routines for raw program load without any specific map number specified.
	- Created cutscene.c/h and some boilerplate for this module to handle static screen loading like this.
	- TODO - Add Demo playback capability to play.

31-7-24
	- Basic demo recording and playback capability now added! Needs quite a bit of fine-tuning but i can get input from a file to manipulate my game engine.
	- So there's something sitting at a deeper level desyncing the inputs.

1-8-24
	- Redesigned the demo recording abilities, now events will be genereated for all keys each time.
	- PRoblem - Some keys have analogs of each other, and if one is on (say the W key), the UP key being off isntantly switches it off.

2-8-24
	- Okay so we've made some engine changes. No longer are logic calculations also tied to deltaTime, they'll now be on fixed increments called FIXED_TIMESTEP.

3-8-24
	- So I've learnt these entirely new concepts called the accumulator and the fixed timestep, and I've had a bit of a nightmare of a time trying to shoe-horn these concepts into my engine successfully.
	- Right now, the engine is all over the place, it runs but goes into death spirals, is patchy etc.
	- Spent about an hour realising that this code:
	        while (getAccumulator() >= FIXED_TIMESTEP) {
            update();
            deductFromAccumulator(FIXED_TIMESTEP);
        }

	has a fatal flaw: If the accumulator is lower before the first update(), the update function is never called first, and the render function has n othing to update and so hangs.
	My fix:
	    do {
        update();
        deductFromAccumulator(FIXED_TIMESTEP);
    } while (getAccumulator() >= FIXED_TIMESTEP);
	My first ever use case for do-while loops!

4-8-24
	-This demo recording system might actually beat me. I have been at this for days and am tying myself in knots.

5-8-24 
	- Okay we are at the grindstone - 
	= Problem is tha tinputs are polling extra frames and spilling over, making all calculations off
	- I've been defeated. I simply couldn't make this work conventionally. Since this project is not a commercial project, 
	  I am simply going to take the easy way and simply record exact inputs per frame along with keyStates.

6-8-24
	- Hallejulah,praise the lord! we have a working demo file!
	- It's beeeauuutttiful, although a bit cumbersome. We could probably get the filesize down a bit.
	- we can finally start to move forward to the complete game state.
	- (DONE)TODO - demo end on p parameter should drop to command line with real tics and game tics printed at the end.
	- TODO - Demo record - should end on death + on levelexit should drop to cmd line with stats.
	- (DONE)TODO - Some keys should be usable during a demo playback.
		- Map, detail, escape, f10
	- Expanded trigger functions to include an option to use levelFrames to trigger future functions - I believe this may be key in accurate shot playback. (It was)

7-8-24
	- TODO - keys don't carry over between levels
	- (FIXED?) Segfault in a* between levels.
	- TODO - music should loop
	- Adjusted bullet damage profile.
	- Created a new UI look using a sharp, dark font.
	- Also created a Logo for the game.

8-8-24
	- (FIXED)TODO - minus and plus keys don't change resolution in demo playback.
			 Menu keys don't operate either.
	- (DONE)TODO - Upon normal game load,we want the splash screen to show, then after x seconds, demo and the menu should show up.
		   - Demos should load from folder /demos, and they should play in a random order. After each demo finishes, a secondary splash screen
		   	 with credits should display, then another demo, then splash screen 2? After this, the company log and splash screen should play again
			 play again. This time, new demos are chosen and the whole loop continues again.
	- Refactoring some states in game to be logically coupled in state.h - startupParameter, mapOpen etc.
	- Refactor the starting parameters for the game.
	- Actually, level demos will run too fast or too slow(a roughly 42 sec demo can end in 24 seconds with full speed).

13-8-24
	- Been busy for a few days, but in the meantime here's what's happened:
	- A lot of refactoring for states.
	- Demo now plays automatically from intro.
	- (DONE)TODO - Regulate the speed of demo playback - on my PC when plugged in it runs far too fast, which it shouldn't do.
	
14-8-24
	- Ok, I  _Think_ I have a proper intro loop established, needs a bit of polish but should work.
	  Demo loop is established. Now for polishing...
	  	- Some good polish now. A few transitions to fix up. Sometimes background doesn't draw correct cols.
	-  (FIXED)I noticed that loading from demo, the ceiling colours don't load?
		- So they're read in properly, but I don't think the buffer flashes in the demo mode?
	- (FIXED)All demos should reset player stats in between levels.
	- (FIXED)HUD is showing after first demo finishes.
	- Still getting mapLinesToDraw array filling up - Not sure exactly hwat is causing it, something isn't getting flushed?? (FIXED?)
	- Time is coming close to doing the main menu, and along with that loads and saves.
	- With saves and loads we need to save all playerStats, most of the player info, and all active enemies and things in level.

15-08-24
	- FIXED - Timestep should now be accurate regardless of CPU speed. This was due to my use of a do-while loop, which would mean that the game would
	  be forced to update once every game loop REGARDLESS of whether it was ready.
	- New input mode = -1, no input at all allowed. This will be applied at the Pelican Blue Software splash.
	- FIXED - Ceilings/floors draw correctly. Music fixed. Automap closes automatically at end of demos, messages cleared after demo finish.

17-08-24
	- New font to replace the old Wolf3d one.
	- New Difficulty icons.
	- New Splash screen/

18-08-24
	- Menu now has headings.
	- Difficulty icons now implemented in game, along with difficulties to choose from.

19-08-24
	- Now players can start a new game using the in-game difficulty selection. BOLSHEVIKTOR mode starts the user at half health and with half the bullets, as well as some other gameplay tweaks.
	- Draw menu now returns a new enum MENU_ACTIONS, which the main.c handles as it has complete control over the scope of the game, whereas menu.c is very limited in what it can access without some forward decl shenanigans
	- (TODO) Starting a new game while a demo is running is not working properly, will fix.
	- TODO: Why do midis still crackle when a new one is queued up? Seems unprofessional and sounds amateurish.
	- (DONE)TODO: Tonight working on the main menu, allowing modals for sliders.
		- God menu.c is a rat's nest.	
	- After quite a bit of fiddling and refactoring, I can get a bar drawn on screen.
	- (DONE)TODO: Menu Titles should always be centered.

20-08-24
	- Nice code refactor this morning shortened codebase by several hundred lines, made several constants computed once to save processing power and made code more const correct.

21-08-24
	- SFX volume now processes correctly.
	- Brightness slider now works and modifies brightness correctly!
	- Resolution and Volume sliders now work and modify correctly as expected (Albiet a bit code-heavy in input_fns.c)
	- (DONE)TODO - Menu escape key should go up one level but is getting out of the menu.
	- TODO - Occasional fuckup of the renderer when changing render resolutions.
	- (DONE) TODO - settings.cfg should save once the out of the options menu.
	- (DONE)TODO - Implement Quit and some cheeky options goading player back ala Wolf3D.

22-08-24
	- (DONE)TODO - Extract a general pixel drawing function to further reduce code bloat when drawing things to screen.

23-08-24
	- Menus are much more work than they seem. Really nothing is left automatic, everything must be programmed manually.
	- Quit now displays a modal with a random taunt, and a YES/NO dialog to quit game. Pressing Y or N will also work.
	- Menus now work really well, and with the exception of Loading and Saving game, everything works as expected.
	- Gonna do a checkpoint here, and advance the version to 0.7. Now I'm going to embark upon tightening the code up so i'm not constantly repeating boilerplate code.
	- Successful refactor has reduced code by 15% in some files, now using drawFromTexture().

24-08-24
	- (FIXED) TODO - Music volume modified from main volume.
	- (DONE) Today We'll save the settings.cfg file.

26-08-24
	- Started mapping out the save file and the data we'll need to save. Creating new arrays to hold the original level data that can be compared with current level data, and if any changes
	are noted that is put into the save file.
	- Some very nice optimisations and further refactoring.

27-08-24
	- Work has started on solidifying a save game file format.
	- Populating a save game variable with appropriate differences from original and current sprite and enemy arrays.

28-08-24
	- Work continues - Serialising save data today.
	- My god saving and loading is a whole other pandoras box I'm opening. This will take days to implement.
	- Okay - Function to serialise game data is finally complete, at least in conception. 

29-08-24
	- Okay it compiles alright!
	- (DONE, but only save does something atm)TODO - Assign F5 to quicksave, F9 to quickload.
	- TODO - write a quick program that can read the contents of the loaded save file to the console, separate from the main game. THis will allow quicker debugging of the code.
	- Actually just gonna test inside program with breakpoints - I can successfully read back in a save file (Files are like 200kb! (mostly color buffer)), and when all checks are satisfied, I'll connect it up to a level load.
	- Well, you win some dev offshoots and you lose some - I think I am just going to store the enemies and sprites arrays in the binary of the save file - it's not worth that added complexity otherwise. I had some fancy code
	would try to differentiate between the two but it ended up being a dead end, as the game is constantly recalculating inside this array, reshuffling it etc.
	- Updated the file pool code so now I have two pools - one for demos and one for save files.
	- Let's make it so I can see the files when I go to load 
	- Let's also get a successful level load of a save from a file.

30-08-24
	- Screenshot is now redrawn on screen! Now we are simply sampling from the main buffer and saving that to a save file, halving the normal resolution. Looks fine in menu.

1-9-24
	- Some further refactoring and constness-ing. Strings can now be different colours. 
	- Fixed level time issue.
	- Load game screen almost done, needs a bit of tweaking.
	- TODO - in msf 0.2, lines outside of boundaries still draw onto the buffer.
	- TODO - Map - overdraw on window tiles?
	- TODO - Map - Some items visible before being seen?
	- Some menu UI visual improvements.
	- Today - Load in a save game.

2-9-24
	- Doors must be saved as well - forgot that.
	- Everything is basically hooked up, now to see a successful level load.
	- WOW! First go, everything loads - new level loads immediately. Will proceed to do some tests.
	- Getting save screen in menu good.

3-9-24
	- Continuing save menu code.
	- Want to implement arrows for moving up and down the modal for loads and saves, and others.
		- Arrows now draw, but now they need to light up when there's more to scroll, and not when there's no more.
		- Now they roughly light up, but the yonly need to light up when there's more to draw. Additionally, new save slots need to load in place of others if there's a bigger list.
		- Ok game saves successfully through the menu, creating a new file.
		- TODO(DONE) - Game overrides old save.
		- Menu behaves appropriately, loading in new saves as it goes.


==============================================================
Starting today, I'm going to implement something similar to Carmack's .plan files:
Prefix 	    Meaning
No prefix 	mentioned but not fixed or implemented on that day
* 			completed on that day
+ 			completed on a later day
- 			decided against on a later day
==============================================================

4-9-24

	* Map lines drawing out of bounds in msf 0.2
	* Saves and loads fill up a max 2 spots at a time in menu and intelligent scroll
	* Saves can overwrite old files.
	* Certain map tiles are bigger than others for no reason. (Check if that +3 in drawMap_rectangles)
	+ Player movement needs to slide down the walls

5-9-24

	* Mouse integration with menu UI elements
		*- Main Menus
		-- Sliders
		+- Binary dialog.
	+ Mouse rotation removes sprites from 3D Buffer.
	+ Mouse movement should be smoothed and linked to X axis sensitivity from settings.
	+ Mouse readings in demo files. (Really only needing left click to shoot)
	- Slider type has a click box
	+ Control menu option to use mouse to control (changing A and D to auto strafe.) (maybe making this default?)

6-9-24

	* Unselected drawn menu items rendered grey.
	* X axis sensitivity slider support in menu.
	* Fullscreen boolean slider support in menu.
	* Render mode slider support in menu.
	* Always run support in controls (and in menu)?

7-9-24

	+ Demo record mode shouldn't display startup logo
	+ Demo record doesn't accept inputs
	+ apply buffer zone to horz/vert movement sliding
	* always run setting isn't being loaded
	+ exit taunt string should be centered
	+ always Run option menu not working/applying

8-9-24

	== COMMENT ==
	Making smooth sliding along walls was far trickier than I initially expected, but it boiled down to simply prohibiting movement along one axis,
	to not also the buffer zone to be penetrated.

9-9-24

	* Caps lock and always run mode should work in concert.
	* Hide mouse cursor in-game when mouselook is enabled.
	* new Movement code clipping through locked doors.
	* no further mouse movement possible when at edges of screen.
	+ blocking sprite gliding is not perfect - see 1st hallway barrel -- This is important as players can enter areas they shouldn't with it
	- shift key should downshift when always run is on.
	* Intro loads as normal for both demo record and playback.
	+ Demo loads sequences should be random.
	+ Loaded games in the same level should not restart the music.
	+ Quickload does not load the game.
	+ Sound starts again when level loads from next level queue.
	+ Menu modal should close after new game selection
	+ New game should start with a fade out and fade into new game.
	+ End level screen needs a facelift re. new font, code organisation.
	+ Player shouldn't keep keys from level to level.

	== COMMENT ==
	The weird "sprite disappearing bug" is definitely nestled within findVisibleSprites().
	Okay bug fixed, I simply hadn't been normalising mouse angles (but turning angles with keyboard were).
	Refactoring the main function to make an easy-to-understand launchpad.

10-9-24

	+ Royal treasure is not getting saved in saved games.
	* Save games should handle the royal treasure as well.
	* Implement royal treasure in end level screen.
	+ Flags are not updating when a level has been conquered.
	- Games cannot be loaded during intermission screen.
	* Buffer overflow when going into the loading screen too quickly and trying to load a save
	+ Level read fails should exit the program entirely. 
	+ (quick)loads don't redraw the BG_STATIC buffer (if they're loading from a different level).
	+ Improvements needed to the music module.
	+ New texture needed, (64*9) * 64, for the higher res versions of the royal treasures.

	== COMMENT ==
	Refactoring level_end_screen.c for conciseness and clarity, and in line with rest of program.

11-9-24
	+ Picking up the gold key deactivates the silver key in the UI.
	* If ammo goes above 99, UI doesn't display it.
	* End level screen is not resetting.
	+ Train station red flashing square should be better centered/with better proportions
	+ Weapon firerate, and their sprites, should be on FIXED_TIMESTEP
	+ 3rd Enemy type should be designed now, holding a Avtomat.
	- Left click for save and load games loads and saves a game appropriately.
	+ Pathfinding from level to level doesn't reset 100% correctly?

	== COMMENT ==
	Working on sprite sheet for 3rd enemy type. It's time consuming, but needs to be done.

12-9-24

	== COMMENT == 
	Sprite sheet about 15%, 33% done, 60% done. 80% done.

13-09-24

	* Add in 3rd Enemy Type.
	+ Third enemy type, Elite guard, fires burst of three bullets.
	* They also run somewaht faster than the regular guards, and have 50% more hp.
	* These guards will drop an Avtomat upon being killed.
	* If player already owns Avtomat, they can stll pick up weapon as long as their Ammo < 99.
	* Do not inform player of acquiring the gun in messages (instead inform of clip), nor do grin face.
	+ ENEMY_ELITE should be shorter slightly.
	+ Change colour of weapon in UI.
	+ Ranged enemies don't come within 1 map unit of player ever and just keep taking pot shots at player. (stops them looking like giants).
	* Spritesheet initial done.

14-09-24

	+ Pistol spritesheet replacement.
	* Secret walls shouldn't intent into walls.
	* Saved games don't save sprites and wall pieces seen on the map already.
	* Saved games don't save Window seen status.
	+ Since dogs can't open doors, they are perpetually stuck trying to access player.
	* New sprites for UI keys, fix their positiions in UI

	== COMMENT ==
	Indentation bug that's been present for months is in pushWallLogic, line 426.
	Bug fixed. Ended up taking me 6 hours for something fairly simple - horzHit or vertHit snapping walls to grid.
	That had been an engine bug for at least 4 months.
	
15-09-24

	+ Loading a game after finishing a level doesn't reset the level end screen.
	* Disable clicking to save game when level end screen is in effect.
	+ Menu should pause level end screen's tickers.
	+ Space bar should advance tickers, and spamming it should finish tickers early.
	+ Handle game loads during intermission (Disable menu action entirely)
	* After the main tickers, we should have a sound effect and a highlight if a piece of royal treasure has been found.

16-09-24
	* Cheats are going to be added into the game.
		* pbdqd - invincibility
		* pbclip - clipping disabled
		* pbinvis - enemies don't react to player.
		* pbkfa - all keys, 99 ammo and all guns
		+ pbitem# - adds an item to a player.
		* pbdt - all map discovered.
		+ pbclev# - change level
	- Bolsheviktion 3D blue drawing in console at program start sometimes stops drawing after the text is drawn.
		(This seems like a bug inside MSYS, not worth fixing)
	+ pbclip causes segfault from a* pathfinding.
	+ Enemies don't obey patrol waypoints.
	+ Sound propagation for aggroing enemies

	== COMMENT == 
	What a painless implementation - cheats should've been added MUCH earlier in development cycle for testing.

17-09-24
	* Patrol from South walking north doesn't work properly.
	+ Use BFS to implement sound propagation in sndpropagate.c
		* - Basic implementation compiles.
		* - Sort out segfaults
		+ - Handle doors.

	== COMMENT == 
	Patrols are in much better shape now, but took hours of debugging and headscratching.
	Sound propgation is almost in engine, shouldn't be too taxing on the system.

18-09-24
	+ SndPropagation:
		* - Handle windows.
		* - Aggro enemies.
		* - Handle depth.
		+ - Reload propagation map on map edits (i.e. secret pushwalls).
	* Sprites are not drawing on first pixel at top or the left of screen on lower quality settings.
	+ ENEMY_ELITE should dance using the 2 frames reserved for them while being hit.
	+ ENEMY_ELITE should drop as soon as their hp reaches 0. (fixed with 24-9 sndpropagate fix)
	- drawBox/Draw line doesn't draw a complete box. (Good enough)
	+ Consolidate level loading code into one function.
	+ Load level loading on a function trigger
	+ Load 2 -> load 1 -> end level (level 2 still loaded into end level screen)


	== COMMENT ==
	As level loading code refactor will be quite a major restructure of the code base, I'll do a code checkpoint here.
	Hopefully after the refactor we can standardise the ideas of clearing memory, fading in and out, and music loading.
	When testing, we must check for:
		Fade to and from black.
		Music loading.
		Memory leaks

19-09-24
	+ Loading games from regular game intro loop continues demo play.
	+ Dogs, if they have no LoS, Should bounce around like nutters off the walls.
	+ Functions with only internal linkage should only be static

20-09-24
	* In lower resolution mode, messages at top of screen are spread out disproportionately.
	* Unlocking a locked door retriggers pathfinding grid rebuild
	* All weapon frame control is under a constant.
	+ Inlining and static:
		+ attack
		* enemy_move_pathfinding (Massive init speedup)
		* sndpropagate
		* textures (3-5% ish speedup)
		* wall_window
		* hud
		* utils
		* enemy_move
	+ Lower resolution modes perform worse (??? very strange)
	+ Hud in lower resolutions have a small gap between the bottom and the HUD (rounding error)
	* Dogs no longer get stuck at closed doors.
	* Enemies should no longer get stuck and glitch out in position.
	* Time in dying frames sped up. (0.15->0.13)
	* Enter spamming should get players straight into a new game with medium difficulty mode.
	* Labels in load game (and save game) are not aligned if not using default resolution.
	* Sound Effect load in the console are missing some #.
	+ Save game, when left open and demo ends, will start to show menu saving options.
	+ Fix save and load game screenshot ratios.

	== COMMENT ==
	Obstruction flag?? - Modifications need to be made in enemyChase for dog behaviour.
	Ok we can set a flag to not run a*, but rather bounce off walls as default, for x levelFrames.
	After the counter reaches 0, we try again using a*. This is effective and is good enough for this game.

21-09-24
	* Ambush flagged enemies should not be awakened by sound propagation.
	+ Difficulties should be added in game/editor.
		* - Added into editor.
		+ - Enemies faster and have more hp.
		* - Specific difficulty flags for objects/ enemies to appear/not appear (default appear all)
		+ - Starting hp and bullets less.
		+ - Points gained increased or decreased.
		+ - Enemies do more (or less) damage.
		+ - Save games track difficulty (and are displayed in the save/load modal)
	* Add elite to editor.
	+ Diminishing damage based on distance.
	+ Perfect the HUD face and draw a few more reduced health passes.

22-09-24
	+ Editor not flushing old windows
	* Editor not loading in previous skill settings.
	* Save game is available during GAMESTATE_INTRO & GAMESTATE_RUNDEMO_1, GAMESTATE_RUNDEMO_2

23-09-24
	+ What happens when lives reaches 0? (HI SCORE table?)
	* Work on face array (80% done)
	* Selecting Bolsheviktor disables all input
	- Debug text (using SDL_TTF) should be scaled according to resolution.
	+ Enemies are no longer flinching (or flinch time is too long)
	* Halve ENEMY_ELITE damage output
	* Starting bullets should be just what is in a standard NEGANT clip. (7?)
	* Tweak player damage

24-09-24

	* Reset game correct ammo
	+ if a settings.cfg is not found - generate one, corrupted - regenerate
	- Elite guards zig-zag in pathfinding instead of finding the shortest possible route.
	* Editor: Choosing a new enemy type does not update the sprite.
	* Get rid of the "animated" parameter in enemy_t - it's deprecated.
	* Level 1 has a lot of features exclusive to hard and bolsheviktor mode.
	* Buffer overrun when reading in very long level strings.
	* Editor: Adding new doors saves as undefined if not chosen from dropdown.
	+ Knife can be spammed too quickly.
	* Doors shouldn't close on corpses.

	== COMMENT == 
	Very strange - attacking with a knife triggers appropriate pain state, but shooting them does not.
		Possibility 1 - Something is disabling ENEMYSTATE_PAIN just after being set in loop.
	Ok what was happening was that sound propagate was picking up the bullet fire, and "awakening" the enemy again.
	Placed some conditions to forbid "reaggroing" enemies already aggroed.

	It's really playing like a closer and closer polished game. It'a actually quite fun.

	Musing on different timers in game:
		getTicks() is for performance calculations and cosmetic randomness.
		getDeltaTime() is for timed operations in game (i.e. menu interaction)
		getLevelFrames() is for deterministic level engine activity.

25-09-24
	Upgrade a* - 
		* priority queues
		movement patterns
		* memory allocation efficiency.
	- Enemies don't draw at the first y pixel, similar to sprites before at very low resolution (Fixed/not important)
	+ sndpropagate freezes the game while processing (was a bug in the weapon drawing code)
	* Pain state slightly reduced in duration
	* Finally update timers to analyse performance on modules.
	* Rip out SDL_TTF and use internal text drawing functions to BUFFER_UI & perfmeasure.h

	== COMMENT ==
	Now all text in game has no dependencies, rendering straight to buffer. So that means I can remove the SDL_TTF library
	as it is no longer needed.

26-09-24
	+ Disable cheats for Bolsheviktor mode.
	* Fix blood decals.
	* Precompute common calculations for defs.
	* Implement extra life pickup

	== COMMENT ==
	Blood decals are not being considered as visible. Turns out there was a bad value flipping the x and y to negative.
	Continued work for E1M3 - Introduce the Elite guard.

27-09-24
	* 40% speedup in renderColorBuffers() by precomputing bools that won't change in loop
	* 20% speedup in renderWallProjectionThread().
	* Big speedup in drawRect due to direct buffer manipulation
	* Various speedups in graphics_draw.c and all pixel drawer functions
	* Scale map in small resolutions
	* Zoom map in and out
	+ Demo recordings, when you die, should end the recording.
	+ Demos not taking into account weapons switches using keys.
	- Weird artifact of an arrow on demo record (Demo record not going to be in final release)
	* When at 0 ammo and ammo is picked up, player switch to pistol.
	+ Starting a new game when there's already an active game should prompt a yes/no dialog.
	* Map lines drawn are drawn in old code.

	== COMMENT == 
	After hours of debugging, we finally have an enhanced map that scales with resolutions. I also took advantage of the refactor
	to add map zooming in via <> keys.

28-09-24

	+ Implement invisibilty and invincibility into the game.
		+ Invisibilty should add a masking effect over the weapon sprite	
		+ invincibility should modify BUFFER_3D to be fully white.

	- Segfault in sprites when recording demo and changing detail:
		Thread 1 received signal SIGSEGV, Segmentation fault.
		0x00007ff753fb2b92 in drawSprite (renderSprite=renderSprite@entry=0x5ff9f0) at src/sprite.c:416
		416                         texelColor = renderSprite->textureBuffer[texWidthByTexOffsetY +
		(Never could reproduce this)
	* Double shot fired with mouse click.
	+ AVTOMAT sometimes skips over fire frame.
	+ Shadow on ammo box sometimes clips and glitches on lower details.
	* mouse zooms in via mousewheel the map.
	+ Level end tickers play snd effect too slowly.
	+ Starting new game from main screen demo does not change the level number.

	== COMMENT ==
	Sndpropagate glitch from 25/9 is very strange. It seems to cause backlog in A* processing. Disabling map helps.
	Demo has some very minor desync issues, and I suspect it's how demos are processed that are causing this.

29-09-24
	* Implemented invincibility
	- Some doors are staying open even when they shouldn't (see map03) (Sprite placement near these doors)

30-09-24
	* Implemented invisibility - adapted from Carmacks' Doom invis idea (fuzztable)

1-10-24
	* Fixed a few out of bounds access bugs
	* Invincibility doesn't display in full map mode.
	* Add padding to system messages.
	* map level info x pos needs to be correctly positioned.
	* Map enemies should correctly face their cardinal positions (not relative pos to player)
	* F10 during gameplay should open a confirm window.
	* Rough Complete E1M3 with some new textures

2-10-24
	* Program version now 0.8.
	* Code tightened up.
	* E1M3 v1 finished.

3-10-24
	* Decal duration shortened, runs tighter.
	* Reset level resets music twice
	* Double midi playback
	* Dying in demo record stops record
	+ new cutscene with message writing - like Doom intermission
	+ L/R attenuation for sound effects
	+ Demo desync with amount of generated events
	* Demo structure simplified
	+ Demo mode on death fades to red, only THEN back to menu or quit game
	* Sound effects play through fn pointer

	== COMMENT ==
	Two extra frames are being written into the demo at the end of recording. Don't know why.
	Going to redo the demo file structure, "events" and "frames" are too confusing.
	Simplified syntax quite a lot for demos.
	Playback of demo with 'p' parameter can linger for x second after death before quitting back to screen
	going back to main game intro loop.

4-10-24
	* Error sound effect.
	* Simplified and added spatialisation to sound effects.
	* Generated more items and textures. Prepare to draw E1M4.
	* Decal buffers draw bizarre colours at low resolutions.

5-10-24
	* On death with regular mode doesn't kill music.
	* On death goes back to main when it shouldn't.
	+ Hi score on 0 lives left.
	+ E1M1->E1M2 die (no music stop, no death sound effect)
	+ EndLevelScreen not being given new information.

	== COMMENT ==
	Refactored into death.c and death.h to modularise this buggy feature and make it easier to debug.

8-10-24
	+ Implement Hi Score screen and hook up 0 lives to it.
	+ Level start from command line still does a fade-in like it was rendering the company logo.

9-10-24

	* Texture loading line print doesn't print last # for certain texture nums.
	+ GAME OVER cutscene
	+ Story so far Cutscene
	+ Level string selection from command line shouldn't segfault - kill program if not found.
	+ Kill pathway from reading functions all the way up to main game loop.

10-10-24
	+ When enemies die in map, they don't display their death frames.

11-10-24
	- Create a default map that loads when it cannot find other maps. (E1M0.txt)
	+ Saves should be sorted by their most recent first
	* 7-8% fps gain for drawPixel refactor

13-10-24
	
	On game over - check if total score eligible for top 10 scores.
		+ if so, add it in.
		+ if not, show YOU as a temp 11th slot on the high score table.

14-10-24

	+ Make sure all game modes, when 0 lives (either from command line or regular start) exit gracefully and load main demos.
	+ Test writing of the high score and reading to file.
	* Darken the game over image
	+ Finalise the static HUD colours
	* Loading game while demo running carries over input mode
	* Loading game while demo running processes ghost frames
	* Loading game while demo running can carry over keyStateFire

15-10-24

	+ After high score, main intro loop not rendering
	+ E1M4 blueprint
	- EDITOR: Clearing Patrol Waypoints bugged.
	Horizontal sprite check not passing

16-10-24
	* Rendering bug in BUFFER_3D in level select from cmd line
	Enemy movement overhaul:
		* - Enemies now dice roll attacking again only when not in player range.
		* - Enemies in closest range to player will continually attack.

17-10-24

	+ Music stutters when repeating
	* Disable render quality changes during death
	+ Main volume doesn't immediately attenuate music
	* Fix pixel overflow in low details for sprite drawing
	* When switching from keyboard to mouse+keyboard, sensitivity should be auto adjusted
	* Story so far modal access, Keyboard skipping
	* New game from game over and pre-loaded level crashes as it can't find a levelNum
	+ Occasional segfault in loading game menu:
		Count: -17891602
		NewSave: -17891602
		Thread 1 received signal SIGSEGV, Segmentation fault.
		drawFromBuffer (srcBuffer=0x0, destBuffer=BUFFER_UI, originalWidth=160, originalHeight=90, newWidth=160,
			newHeight=90, startX=139, startY=272) at src/graphics_draw.c:525
		525                 const color_t texel = srcBuffer[srcYByOriginalWidth + srcX];
	* Level name should be always run through uppercase maker.
	+ Doors and windows seen on map before officially seen.
	+ Intermission screen design. (To be played between M4 and M5)

	== COMMENT == 
	It seems like calculations are incorrectly accessing an earlier part of the object texture, specifically the medkit texture.
	Bad access to (2687, 60) on the long texture.
	Bug fixed.

18-10-24

	* Add custom soundfont loading support and choice via settings.cfg.
	* All map elements now are discovered properly at the right time.
	* Cap Map zoom
	+ Segfault at invisibility pickup

	== COMMENT ==
	I suspect that we're not returning early for some raycasting conditions properly, thus we're exntending checks further than we need to.
	Once fixed, Windows and doors should stay hidden, but also, there should be a performance boost we we're not casting over what we can see.
	Okay the issue is this: Both rays are cast - vert and horz rays. One is always discarded. BUT, the one that is discarded still triggers the window and
	door to be seen, even if it's not chosen as the final ray.
	solution: Move the window and door seen checks to after we have chosen our ray.
	Each wallHit_t keeps a record of the windows it hits. Make modifications to keep record of doors hit as well
	Bug fixed.

19-10-24

	* Code speedups
	* Added NULL check to wall.c texture retrieval.
	* Increase window retrieval time by storing array of pointers to windows.
	+ Add early exit conditions for broken maps etc. with exit() function.

20-10-24

	+ Some window render corruption issues in BUFFER_3D and  * BUFFER_MAP
	* Allocate more audio channels for playback
	* getSprite() huge optimisation.
	+ Player can interact with doors while inside them, breaking them.
	* Cheeky toilet easter egg.
	* New sound Effects for enemy guard and enemy elite.
	+ Player can glide through closed doors (See E1M1 secret treasure room)
	+ Shooting non-aggroed enemies skips their aggro sound.
	+ Window FPS drops can't be seen in the perf meters for some reason. (See E1M3 secret train)

21-10-24

	* Rendering of invuln in menu flickers.
	+ Train station end switch are back to front.
	* If an invuln is picked up in one level, then the next level will still render that effect.
	* Sprites are not picking up well. (improved)
	* in end Level screen, all level names must be rendered all caps.
	* Large number of new sound effects for guards.
	* Implement fast door retrieval from grid matrix, similar to window.
	+ Rework level end knifeswitch texture
	* AVTOMAT v1 hud weapon
	* Saved sprite array is not saving/loading properly.
	+ Pathfinding isn't working too well. (acceptable, not perfect now.)
	+ Set gun to original position upon death.

22-10-24

	* new typedef - uint - unsigned int
	* 40% speedup in ray_window.c

	== COMMENT == 
	Midi stutter on replay is due to the fact that fluidsynth is loading the file again from memory. Since Fluidsynth likely is multithreaded,
	the game advances 2 or 3 frames during that load. And thus, for 2 or 3 frames, the last part of the buffer is played,
	hence stutter.

	This wasn't the case. I think this signals time I learned about the midi file structure.
	Midi file structure and manipulation is a waste of precious time right now.
	Bug fixed. I had to create a custom callback function that seeks to 0 when one tick away from finishing the song.
	New sound effects and looping music finally give the game some AURA

23-10-24

	- Increase MAX_WINDOW_HITS (Short term solution) (Fixed underlying issue)
	* After loading game, Menu's selector cursor is off (not reset from the load menu).
	+ Code refactoring in raycast.c to fix window errors.
	* Slowdown when not firing for a long time and then firing.
	* Sound effects beefed up for weapons.
	* Decals should go into render order.
	+ This can probably be pre-computed in ray.c:
		const float fishEyeFix = cosf(ray->rayAngle - player.rotationAngle);
	+ Segfault in decal on lower resolutions (320x200)
	+ Weapon sprite off center in lower resolutions.
	* Music does not play on reset.
	+ start from level -> die to end of game -> demo loop -> after first demo, game quits

	== COMMENT == 
	Decided to find out source of phantom slowdown __on fire of weapon__. A knife does not slow down.
	Problem seems to be in animateWeapon()
	Okay - I think it is not that function per se, but the calling of that function is going wacko - drawing many times more
	than needed.
	Yes that's the problem - 333 calls of this function when the weapon is fired. SO there is a logic error calling this function/.
	Bug is in animateWeaponSprite();
	Bug fixed.

24-10-24

	* All ray variables are now precomputed at the beginning of the pipeline.
	* Scrubbing code clean using very strict compilation settings.
	* Weapon draw can skip at least 25% of y segments of the weaponsprite as nothing will ever be in it.

25-10-24

	+ Slime trail on certain window tiles
	+ Segfault when compiling in O2.
	+ Rework weapon animation to ensure full animation rather than skips when 2 levelFrames are drawn per cycle.

	== COMMENT ==
	I think another window's id and thus texture is being loaded wrongly as a window hit on an edge-case.
	Bug in window projection function.
	So O0 obscures the bug and doesn't trigger the slime trails.
	So under O0, ray isn't considered to have a window (as intended). Under O1, ray is considered to have a window.
	The slime trails have the exact same distance as the last window read.
	They're testing the exact same point.
	No assignment of our slime trail is occurring in raycast.c. Must be an addition down the pipeline.
	This is another very difficult bug to fix.
	Full window pipeline has to be elaborated to fix this bug. lignering window hit or something in wall_window

26-10-24

	* castAllRays moved to the render function (was being called each update!!) (BIG IMPROVEMENT)
	* streamlined redundant data resets for castRay
	+ clev cheats
	+ output to .log setting
	+ After invuln wears off, bg needs to be redrawn (stays white (50% of the time?))
	+ no decals for door shooting

	== COMMENT ==
	Wondering why castAllRays is  happening multiple times per frame? (It should, as a renderer, logically only happen once per frame?)
	Bug fixed. Out of bounds access in processWindowHits().

28-10-24
	* pbitem cheats
	* Debug enemy map info text should scale to zoom
	- uncap FPS? (True FPS now shown as 2nd number in perf counter)
	+ EDITOR: Improve responsiveness on large levels.
	* Player can shoot enemy behind door.
	* Slight off position for weapon draw.

29-10-24
	* pbclev forces pistol starts.
	* Disable pushwall activation from the direction that they'd naturally move toward.
	+Logging - memory allocation
	+Logging - printing to .log file
	+Logging - graceful failure
	Memory manager
		+ wrappers for calloc
					 malloc
					 realloc
		+ memory counter
		+ performance_counter integration
		+ allocate fails to log and fall through to quit.
		- Reroute all memory calls to memory.c (most mapped, not all)

30-10-24

	== COMMENT == 
	new memory manager will eliminate a lot of code bloat if executed correctly, although I am worried about access times traversing the memoryBlock_t array.
	Other benefits:
		- Exact knowledge of memory footprint for game.
		- Possible breakdown of what's usign the most memory.
		- The end(hopefully) of memory leaks.
		- Takes my program's seriousness to the next level.
		- Lays very convenient groundwork for debug on low resource systems to see the true requirements of the game.

31-10-24
	+ E1M0 is E1M1 of Wolfenstein 3D.
	+ Debug function trigger mem_Realloc problems.
	- mapState = 0 sees no graphics at intro loop
	+ Update message strings
	+ Menu still off-center when coming out of Story so Far from menu screen
	+ Tweak enemy damage at lower difficulties
	* Heap corruption in 320x200 - Limit detail change below medium. 
		* 426x240 - corruption in ultra low.
		* 640x360 - Corruption in low.
	+ Kill counter off in low resolutions, horizontally - check spacing
	+ If player makes high score table at the bottom, 11th slot also shows.
	+ Sound effects for invinci/invisibility
	* Basic spritesheets for the 2 bosses - Police inspector and Yakov Yurovsky

	=== COMMENT ===
	I think we are playing with fire allowing buffers so low that they exceed TEXTURE_WIDTH.
	Set limits to buffer sizes to ensure basic arithmetic doesn't go out of bounds.
	Buffer drawing (at least) for these decals, in low resolutions, are accessing out of bounds.
	Almost definitely an out of bounds access for the decal buffer construction.
	Confirmed - in the decal creation functions. Out of bounds because of RENDER_PIXEL_COUNT
	 being lower than TEXTURE_WIDTH * TEXTURE_HEIGHT
	Bug fixed.


1-11-24
	+ You can shoot the dog through a wall in E1M1 (Shoot inside the cage).
	* floorf -> (int)

	== COMMENT ==
	Function trigger memory bug is due to this line:	
		    functionTriggerCount--;
   			functionTriggers = mem_Realloc(functionTriggers, sizeof(functionTrigger_t) * functionTriggerCount, MEMORY_FUNCTION_TRIGGERS, 1);
	When the last trigger is finished, the trigger array is actualyl freed due to realloc quirk (realloc(ptr, 0) = free)
	Bugfixed.

2-11-24
	* Police spritesheet
	- intelligent sprite search. (minor performance increase, only on very dense levels)
	* Grin during invuln should take it back to invuln after.
	- Police sound effects and attack pattern (not enough time.)

4-11-24

	* Sprites sorted by type.
	* Indexes into type allows skipping irrelevant sprites.

	== COMMENT ==
	After mulling the whole sprites issue around in my head, I've decided to do this:
		- All sprites get sorted at level Load.
			- Blocking decorations first, then non blocking second, then enemies, finally items.
		- We store the count positions of the last of these sprites.
		- Now we can skip amny static elements of the array when checking for different parts of the array.
		For example, enemise never need to check items as that will never effect them.
		Movement code needn't check anything other than blocking sprites and enemies.

		Next enhancement will be the sprites found in cells.

5-11-24
	+ Doors open during a save and load don't close.
	* Intermission screen now draws correctly.
	Level end screen:
		- + Flags don't draw properly according to the end of level.
		- + Red square not aligned properly to level.
		- + Ticker ticks too slow.
		- + Ticker needs a replacement sound.
		- + Tickers stop before the correct percentage is reached.
		- + Royal treasure doesn't light up properly.
		- + Interrupts with keyboard should be more responsive.
		- + New sound effect for new level load.
		- + Menu should pause level end screen's tickers.
		- + Menu destroys the level end.
		- * play holy relic sound effect when relic revealed.
		- + Spazzy sound effect for some reason.
		- * Treasure count bugged on save loads. - Due to freeSprites() resetting from load.
	Segfault in load screen: (guard against null introduced in function).
			Thread 1 received signal SIGSEGV, Segmentation fault.
		0x00007ff73977170b in drawFromBuffer (srcBuffer=srcBuffer@entry=0x0, destBuffer=destBuffer@entry=BUFFER_UI,
			originalWidth=originalWidth@entry=160, originalHeight=originalHeight@entry=90, newWidth=newWidth@entry=160, newHeight=90,
			startX=startX@entry=139, startY=startY@entry=272) at src/graphics_draw.c:473
		473                 const color_t texel = srcBuffer[srcYByOriginalWidth + srcX];

6-11-24
	+ The level end code is a bit of an unworkable mess - We'll refactor today:
	+ Loaded games don't load doors (all doors appear open).
	+ Load game doesn't reset end level screen.

7-11-24
	* Weird graphics glitch in loaded games (see E1M1 save)
	* Problem picking up treasure in a saved game (Item number not found) PROVISIONAL FIX - fallback to whole sprite array search.
	* Memory manager for secrets
	- Catch unroutable allocation in MEMORY_MISC as a number
	* End level timer counter can still bug out.
	Fix dead dog sprite
	+ end level 3 acts like level 2? (Red box flash at wrong positon etc?)
	+ 0% treasure tallies refuses to progress to next screen (Not sure here..)
	* Starting a new game doesn't wipe your royal treasure status.

	== COMMENT == 
	So when a saved game is loaded, we are simply replacing the arrays (which have been assigned by the memory manager and placed into reference blocks) with the arrays from the saved file.
	THe problem is, the memory manager is not informed of this reallocation, as it is simply:
		doors = savedDoors.
	So, the question is, what should we do in cases where we need to replace the entire memory block? Loop through the 
	Fixed bug.

	Problem: When a saved game is loaded, a truncated version of the sprites array is loaded (including the last part, which is pickup items.)
	Bug - Sprites are not being calculated correctly.

8-11-24
	- Flag positions should be calculated via the relative position of the pixel in the map .png (if the dimensions of the window are kept to my specs, this isn't necessary)
	+ Intermission text function moved to graphics_draw so it can be used by many cutscenes.

10-11-24
	- Doors should only block from enemies (not badly placed sprites) (Just place sprites well, check for doors that don't close, inBlockingSprites check start at enemies)

	* segfault looking into automat secret in police offices: (Never could replicate this, marking fixed)
		Verified - engine bug
	Thread 1 received signal SIGSEGV, Segmentation fault.
	0x00007ff793feb8dc in windowProjectionDrawRoutine (windowId=60) at src/wall_window.c:80
	80                  if (windowId != ray->windowHits[i].windowId)
	(gdb) bt
	#0  0x00007ff793feb8dc in windowProjectionDrawRoutine (windowId=60) at src/wall_window.c:80
	#1  renderWindowProjection () at src/wall_window.c:43
	#2  0x0000000000000000 in ?? ()
	(gdb)

	+ Police walk speed does not increase on aggro.
	+ Add police boss into game.

11-11-24
	* E1M5 rough sketch complete.
	* Implement all police boss logic.
		* Item drop logic.
		* Movement logic.
		* animation cycle logic.
		* attack logic.
	* Implement intermission 1 & 2.
	+ E1M6 start with knife and 50% hp (nightmare knife + 25%hp), no ammo.
	+ some mids still reload before looping properly? (E1M5 music, investigate)
	* Players with zero ammo picking up avtomat should auto switch to avtomat & same for lewis
	+ After new life if in mid fire, fire animation still goes off
	* Map placements E1M5:
		+Medkit in courtyard.
		+Health in the toilets.
		+more ammo in the stables.

12-11-24
	+ E1M6 draft.	
	* A collected royal treasure on a level reset removes that treasure.
	Sprites behind pushwalls still visible on minimap (getVisibleCells??)
	Sprites being culled incorrectly in some angle combinations

13-11-24
	+ Invuln redraw BUFFER_BG still bugged.
	+ Message when picking up a avtomat from 0 ammo still implies you have acquired it.
	+ picking up weapons shouldn't grin after initial pickup.
	* Percentage counters in end level screen should be shifted to the right unless at 100%.
	* Police will continue to fire even with no LoS

14-11-24
	+ E1M7 start
	+ MIDI creation start
	+ Packaging textures, levels, sfx, midi into .b3d files (when all texture spots and sfx etc are finalised)
	- Draggable windows
	+ Sprite projection is not accurate - see E1M7 intro

15-11-24
	* Fix the new bug introduced for sprite culling along X Axis
	+ When firing z fighting happens with lights placed over other objects
	- 1 pixel culling of top of sprite at medium range (not serious/visible enough to bother investigating)
	* Reduce health of guards.

	== COMMENT == 
	Issues were introduced using tanf to project the sprite.
	changing to  linear mapping has fixed the bug.

	Composing begins today.
		Some interesting instruments for Bolsheviktion	
			33 - Fingered bass
			45 - Pizzicato
			14 - Tubular Bells

16-11-24
	* E1M7 draft
	+ Tighten up door drawing for smoother animation
	+ Menu sound effect replacement

17-11-24
	* E1M1 music draft
	+ E1M8 draft
	+ Implement basic boss mechanics
		+ 15 seconds temporary invisibility
		+ Sound recordings for boss

	+ Improve invisibility effect
	+ handle lower resolution buffer read access
	+ Toilet flushes randomly sometimes, nowhere near a toilet. (under investigation)
	* Police should have only a 50% chance to stagger.
	+ Don't put an extra life in plain sight on train level.
	+ Take invisibility cloak away from boss room / only put it on easier difficulties.
	* Royal Treasure offers much less points.
	+ Half the available medkits on higher difficulties in train gangways.
	+ Police pain state should be holding a gun
	* Decals don't show when shooting enemy behind doors.
	+ SHooting accuracy tolerance should be improved.
	* Update and enhance random number generator.
	* Update Lewis gun pickup sound.
	- Can't load midi file handle error
	- What happens when no sprites in level? (make sure there's at least 1 sprite)

	== COMMENT == 
	See comment for 27/11 for phantom toilet flushing

18-11-24

	* Sound effects for Yurov
	* More boss mechanics - at certain levels of hp he whines in pain and drops items
	+ Final cutscene sequence intialisation after end level screen for E1M8 and "The End"

19-11-24

	* Final cutscene graphics first pass
	* Custom function for transparent boss so as not to impact regular sprite draw performance.
	+ stop spamming end level switch

20-11-24
	
	+ Finish final cutscene logic
	* include pbitem7 for relic aquisition
	* Triggers for INTERMISSION_3

21-11-24

	* Complete game loop established!
	* Boss should drop gold and silver keys (replacing orlov sceptre)
	* End level screen blocks can't be spammed.
	+ Finish rough edges final intermission
	+ Start work menu fine-tuning
	- Fade out music at the change to going back to main demo loop
	- Functionality added - Fade out (Disabled)
	+ Preload Hardcoded engine midis into memory for easy loading.

22-11-24

	* Finish draft E1M8
	+ Draft E1M9 (secret level)
	+ Lower render resolutions don't sample correctly for invisibility
	+ Timer counter in end level can still break
	* Sound positioning directions fixed
	* Bullets can travel beyond doors and pushwalls.
	* Story so far click throughs don't act in the same way as keyboard space presses
	* Mouse movements in Menu shouldn't spam option select noise
	- SFX when played and player moves from original source, sfx need updating directionally
	* Hitting escape during end level tickers opens map, then takes user back to level
	* Escape permanently pauses intermission

	== COMMENT ==
	Preloaded game midis sees a marginal increase in game cohesion.

	Menu problem -> GAMESTATE_END_LEVEL -> GAMESTATE_MENU -> GAMESTATE_RUN

23-11-24
	
	* Menu inaccessible during dying.
	* pbclev change to level 6 gives gun and 7 bullets.
	* Black dots where there are transparent window textures
	- Editor: Background copy the lower color of the level
	* Editor: Holding down Shift Selects blocking automatically
								 Vertical Alignment of window

	== COMMENT ==
	Black dots are at the end of the buffer.
	LOL black dots were actually on the texture

24-11-24
	
	* Weapon ammo check fixed.
	* boss dmg and drops tweak 

25-11-24
	
	* Weapon feel and animation improved.
	+ Avtomat and lewis sound effect needs beefing up
	Bolsheviktor modifiers:
		* 50% setup time for fire
		E1M1
			+ another helath pack in wood room
	+ New game interrupting old game confirm dialog
	+ Fade from E1M9 -> new game
	+ segfault in game over
	+ Continuing to fire after an enemy falls continues to produce blood decal for a time.
	more than sav_09 doesn't show
	+ Save file on game load shows medium difficulty, but loading it and checking it shows it as BOLSHEVIKTOR
	+ Map fixing and general polishing
	+ Y and N for exit in other contexts

26-11-24

	* Dying enemies don't trigger ENEMYSTATE_PAIN
	End level screen disallows next level continuing (sometimes - what are conditions??)
	* End level up-down keys in menu still make sounds
	* Flags and squares for map screen
	+ end level menu music doesn't loop correctly.
	+ Menu sounds - military?
	* Decals shorter duration
	+ Decals being generated through windows? -- Check fix for decal visibility implemented earlier
	* Entering E1M9 triggers end intermission
	+ Intermission enter key not working.
	+ Recheck knife attack range (see E1M6 intro)

	== COMMENT ==
	Z fighting seems to happen on more than merely firing - certain other conditions trigger it too - we need a more robust system for cases where sprites are stacking.
	Bug fixed - Reordered sprite draw orders and as a rule forced lower spriteIds to always be drawn first.

27-11-24

	* End level switch can't be pressed 2nd time
	+ Intermission music doesn't play
	* Refactored all menu input commands into input_fns_menu.c
	* E1M9 draft complete.
	+ 2 more Wolf3D textures to replace.
	Some ambush modes don't work?
	+ BUFFER_3D drawing in all black on sometimes menu screen
	* Screenshot ratios in load and save game fixed.
	+ Loading new Bolsheviktor game from already started game does not fade out/fade in properly.

	== COMMENT == 
	Phantom toilet flush seems to be an unclipped ray cast check
	They were due to a double bug - one I was double calculating buffer fo rthe function, extending reach way further
	I was also using simple loop calculations to clamp the range of activate, which changes depending on the angle.
	So now we get the abs of the x and y deltas between the calculation check and player, and the first one to hit half tile size stops the check (32 map units)

28-11-24

	* Saves are now ordered newest first.
	* Disable menu access during RENDERSTATE transitions
	- Segfault at demo end propagate sound
	F10 quit - instant quit on demo record.
	+ Program arguments reworking
		$ gdb --args raycast-debug.exe –demo_E1M4.dmo p

29-11-24

	+ Menus don't work in demo playback (--playback parameter)
	F10 doesn't work in demo playback (--playback parameter)	
	Magic header for demo files
	- It should actually be frame 3 of firing weapon that enemies get hit, as that's when the bullet technically travels
	+ Demo recording is NOT capturing all events faithfully.
	* New setting - position for starting window
	+ Add primitive logging into the game.
	- Investigate TinyPNG feasibility - Tiny strips some necessary data for uPNG out.
	+ Directory cleanup in anticipation of 0.9 BETA
	- Directory separation of source files (future projects perhaps)

	== COMMENT ==
	Program arguments are now much more robust and clearer. 
	File existence and type checks before loading any command line file into program.
	Very unhappy with spaghetti code for demo playback.
	keyStateFire seems to be starting at the correc ttime 
	Bug - Weapon fires are not deterministic from record to playback
		KeyStateFire is NOT on in recording, but ON in playback.
		So, what determines how keyStateFire is on?

30-11-24

	- Enemies exactly on the other side of the door may make it hard to open (see E1M4 room with invis)
	+ Determinate demos
	* Can't escape out of a main menu demo running (try going in and out of menus etc.)
	* Demos loaded from normal main menu don't disable mouse cursor.
	+ Game over music won't play.
	* HUD visible during game over.
	* Pressing escape during game over resets DYING_2 
	+ Hi-Scores continually writes to file

	== COMMENT == 
	After many more hours of debugging, I have somewhat improved the demo accuracy.
	After some more hours, I found a potential non-deterministic bug, and I also coupled shots to levelframes.
	Now we seem to have demo consistency!
	Not all the time it seems however... very disappointing

1-12-24
	BETA 1 day
	+ At final end intermission screen, stretched image isn't scaled correctly.
	+ Can't escape out of main menu on BOLSHEVIKTION 3D splash
	+ Set Par Times for all levels
	* Wolf3D level 0
	* Map default in settings.cfg
	* Entry parameters for --level and --record extended to provide difficulty
	
2-12-24
	* X button doesn't kill game
	* Activate is very wonky and unreliable.
	- Precompute player angles once per tick using rayAngle_t struct
	* if mapStart == 2, black screenwipe on first startup
 	+ Hi scores shouldn't write unless actually cracked top10.
	clev changes introduce black buffer again
	+ New game from end level does not reset the end level exit switch spam
	- E1M4 timer should stop at intermission
	* Intermission music E1M4 doesn't go off.
	+ Disable input during the gameover
	+ Increase ammo of Elite guard
	+ Police sometimes only drop one clip
	* Police should have a 50% chance of dropping dogfood
	+ Invisibility should evade aggro, however if aggro, should simply reduce accuracy
	* Final boss does no damage?

3-12-24

	+ Overlapping trees causing a segfault? - Confirmed negative index access bug

4-12-24
	* beefier Lewis sound effect
	- New sound effect for ENEMY_ELITE shots

5-12-24
	+ E1M9 completion does not trigger tricolor flag on finish
	Invisibility:
		* 20 seconds only
		* Evade aggro, reduce accuracy if already aggroed
		+ Make effect more prominent
		+ Fix buffer access in low render detail
			
	+ Holding cells red square should be defined, should player go into it from E1M0
	+ Some enemies don't switch from ambush after aggro (See E1M4)

	== COMMENT ==
	The backup buffer is used because for some reason the 3D Buffer is full TRANSPARENT_COLOR.
	Bug fixed - we were using a default calculation to get the correct values, but at different render qualities the number needed to be updated.

6-12-24
	* Hook up secret level boolean for proper flag drawing. (this will also need to be saved.)

7-12-24
	Loading corrupt saves shouldnt' crash program.
	* Bonus for beating par time working

8-12-14
	LINUX:
		* Buffer overflow on level change.
		* Buffer overflows on level complete.
		* Buffer overflows in save game menu
		Level End treasure portraits not aligned correctly.
		Buffer writing problems in audio module - What triggers? -- Race conditions with pulseaudio
		+ Music on 2nd repeat drops certain notes
		Malloc crash in renderColorBuffers?
			#19 0x000055555556d95d in renderColorBuffers () at src/graphics.c:406
			#20 0x000055555557930d in main_render () at src/main_render.c:138
			#21 0x000055555555d592 in main (argc=<optimised out>, argv=<optimised out>)
				at src/main.c:149
				malloc(): mismatching next->prev_size (unsorted)
			Thread 1 "bn3d-debug" received signal SIGABRT, Aborted.
			__pthread_kill_implementation (no_tid=0, signo=6, threadid=140737327904832) at ./nptl/pthread_kill.c:44
			44	./nptl/pthread_kill.c: No such file or directory.
			(gdb) bt
			#0  __pthread_kill_implementation (no_tid=0, signo=6, threadid=140737327904832)
				at ./nptl/pthread_kill.c:44
			#1  __pthread_kill_internal (signo=6, threadid=140737327904832)
				at ./nptl/pthread_kill.c:78
			#2  __GI___pthread_kill (threadid=140737327904832, signo=signo@entry=6)
				at ./nptl/pthread_kill.c:89
			#3  0x00007ffff79c7476 in __GI_raise (sig=sig@entry=6)
				at ../sysdeps/posix/raise.c:26
			#4  0x00007ffff79ad7f3 in __GI_abort () at ./stdlib/abort.c:79
			#5  0x00007ffff7a0e676 in __libc_message
				(action=action@entry=do_abort, fmt=fmt@entry=0x7ffff7b60b77 "%s\n")
				at ../sysdeps/posix/libc_fatal.c:155
			#6  0x00007ffff7a25cfc in malloc_printerr
				(str=str@entry=0x7ffff7b63c10 "malloc(): mismatching next->prev_size (unsorted)") at ./malloc/malloc.c:5664
			#7  0x00007ffff7a2937c in _int_malloc
				(av=av@entry=0x7ffff7b9fc80 <main_arena>, bytes=bytes@entry=632)
				at ./malloc/malloc.c:4007
			#8  0x00007ffff7a2b5f9 in __libc_calloc
	
	+ Extra life at 100,000 points
	+ Increase points death penalty (to 50,000)
	* Advance program number
	- Black pop in when loading levels
	* Mountain Outpost flag says complete when entering the level
	* Border crossing square needs to be aligned a bit better
	* Ipatiev house should be a bit thicker
	* All existing cheats disable upon new Bolsheviktor game
	* Menu SFX
	* E1M9 dog pen has a window bug
	* E1M9 can't pick up royal treasure
	+ E1M2 extend silver key a bit, change routing for level - too linear
	* Clear up pointer not found errors upon exit.
	+ Quicksave/quickload not working
	* Player can still shoot when render is black

	== COMMENT ==
	Happily, after 5 months away from Linux development, coming back I got the game working within 10 minutes.

9-12-24
	+ Demo reset must reset all invis/invuln flags.
	+ Move thread memory out of manager to avoid race conditions and mutex issues.
	+ seenMap not being correctly freed
	+ Initial opening of menu in GAMESTATE_INTRO doesn't produce sound effect
	+ Time in end of level E1M4 illustrates bug.
	Movement:
		+ Clipping on walls behind player when movement is forward -- To investigate more..
		* Wallrunning through closed doors (E1M1 silver key)

	== COMMENT ==
	Black fade is not looking great because we are stopping the static hud drawing while transitioning.
	Movement:
		problem: In strafe, player can be facing the NEXT tile, and glide through a door.
		Adding new door checks fixes this bug.

10-12-24
	* Train station level tweak
	* Logging
	* Add verbose logging setting
	+ End level timer -- I can't reproduce this properly!
	+ Black screen in new game from normal startup -- Deeper issue at play here.
	* Invuln face timer tick not paused on level pause.
	+ Route all printf errors to the logger (Ongoing)
	* Hi scores should only save when needed.
	+ Warning for new game if starting new game from already running game (Implemented, but check to see if works in all situations)
	* Difficulty not being added into demos.
	* Add difficulty to the map screen and prettify
	- Hi scores, if an identical high score is achieved, should continue to write to the table.
	* Previously held weapon is not loaded from save.
	+ In some conditions, an enemy can walk through a wall. (See E1M1 save).

	== COMMENT ==
	This function trigger bug was potentially a very hard-to-detect problem - That the timer would increase even if the levelFrames were not increasing.

11-12-24

	+ Midi creation E1M1.mid + E1M2.mid start
	+ Invuln in E1M9 is blocking
	+ Final pass 1 of maps - 
	+ bundle maps together and read maps from file
	* Mouse stops working on menus during demo playback
	+ Tiny space between HUD and bottom of screen. (calculate)
	* Release hides console (but enabled via settings.cfg)
	+ Blue changing parameters disabled if not compatible in cmd
	* Key unlock sound effect for doors
	* Picked up a an russian... message
	Failed loads and saves show properly, but should not be choosable to open
	* Random death sound if yes is pressed for quitting game

12-12-24

	* Choosing Yes for exit with keys should switch the modal button to yes as well.
	* Intro midi done.
	* Arrows in save/load disproproptionate in scale.

13-12-24

	* Expand out printError so as to take many arguments so we don't need to print to string buffers.
	+ Demo fire frames are all single shot even when auto is enabled.
	* Intro loop should align with music.
	+ New game from started new game needs to be tweaked and perfected.
	* First shot after new game bug persists for demos.
	* Add version information to intro card and registration information.
		Be sure to validate all upper case for registration information.
	* Cheats only for registered
	* Custom level loaders, demo record/playback only for registered
	+ Registered log info
	* settings.cfg level of verbosity for metrics
	* custom name in HISCORES for registered players

	Map Debug for registered players - + Entity info (Shift tab enables these)
									   + pathing
									   + snd propagation
									   + map boxes
	+ Cheat Unlimited ammo (99 ammo and ammo doesn't decrease)
	+ Hidden rooms for no clipping
	* Framerate info enabled/disabled via f1 (granularity included)

	== COMMENT == 
	Registered mode provides the following benefits
		Custom Demo recording and playback
		Custom level loading through parameters
		Cheats enabled
		Custom name in HISCORES.b3d
		"Thanks for playing, "name!" in wedding screen
		Custom soundfont loading
	
14-12-24
	
	* Metrics in demos can be enabled
	+ Mysterious desync on demos, try loading three demos in and watch after a few reloads

15-12-24
	Website Pre-planning - flask + deployment (find cheapo)
	* Title on Window
	+ Before release - 9 nice demos
	+ some E1M5 lights are marked as blocking
	* Put registration info bottom right of title card
	- disable demo record for final release. (Not needed as underlying issues fixed)

	== COMMENT ==
	we don't want to effect the main game's gameplay.
	I just can't justify any more time on these demo syncs, it's been months of banging head against the wall.
	I have to just let that bug go into the final release. Demos won't be reliable.
	I'll create a few different demos, see how they play back in the engine after the fact, choose some OK ones.
	In that case, disable demo record in command line for final release.


16-12-24
	* Sndpropagation doesn't make aggro sounds to enemies.
	* Player move refactor to make easier to debug.
	* Player score ticks up in an animated fashion
	+ Split and enhance audio features of main midi.


	== COMMENT == 
	Looks like a slide is occurring into sprites as well
	Bug fixed - in the slide, there would a double buffer add that was propelling player through some obstacles.

17-12-24

	- Double yell on one of the death sound
	F10 exits immediately even on non-demo
	* Recalculation of score upon level change. (should reset)
	Guard death 16 + 18 are identical, 17+19 too
	- Right now, no way to access Bolsheviktor! E1M0.txt without commandline
	* Trees near the Secret clubhouse in e1m9 are not blocking
	pbclev cheat always enable for all modes.
	* Map finalising png

18-12-24
	+ Game reset after death shouldn't recaclulate score climb
	* E1M1 final mid done
	* E1M2 map finalise	
		* Lamp in wood room needs blcoking
		* Sacks need blocking
		* Maze needs ambush
		* one bunker needs doortracks
		* Update par
	* Fix up any logic errors in point assignment and bonus assignment
	+ enemy->seesPlayer true even if LoS.
	* A lot of enemy code has been refactored, modularised.
	- Windows are drawn after enemy map info.

19-12-24

	- Player pain grunt
	Level 3
		* Silver key on level 3
		* Flag in next room after key needs blocking
		* Ammo in normal mode is too plentiful
		* Put something in the secret pushwall to the main hallway?
	+ Guard aggro should be tweaked
	+ Pistol sprite

	E1M4 - * green barrels do something (only cleared on hard/bolsheviktor)
		* secret locked door show proper texture
		*Fixed cursed hallway sprites

	+ E1M5 - Sacks in jail
	* Knife doesn't guarantee stagger
	+ E1M6 - Connecting area from freeze carraige texture
	* Above 9 lives affected by bonus from 100000 doesn't draw num to screen.
	+ E1M7 - +Well in courtyard needs blocking
			+ Add some more cruficied/burned people at start
	+ E1M8 - make sure statues don't get blocked by chandeliers
		Normal mode - Yurov needs to hit doublely as hard
	* High score Writes ALL the same name to the high score file.
	+ MEMORY_WINDOWS taking some time to free?
	- E1M1 Dogs still aggro at some points (Sndprogagate?)
	+ Refactor file to prepare for the .b3d archive read.

20-12-24

	Refactor mostly complete -
		* Pushwalls not working
		* Next level queue up
		* reading from binary
	
21-12-24
	+ All dev folders moved from directory
	+ Custom packager .py
	+ loading All assets from .b3d

22-12-24
	+ Disable saving from --level parameter - not having custom code to load up potential files that are not there.
	+ Disable custom demo creation from custom maps - only can load up official maps.
	* Modify custom level loading parameters - can now take numbers (official levels) and .txt files (custom)
	+ Clear memory for the b3dFile_t
	* Restructure midis, read from binary
	* Textures, SFX, and Music #'s replaced with %
	* Too much ammo on Bolsheviktor! E1M1
	+ A door that has been jammed open should be uninteractable
	* E1M1 - Move ammo away from barrel to make it easier to collect in Bolsheviktor

23-12-24
	+ Segfault in mouse motion while menu is quitting
	+ Loading custom levels + midi file.
	DLL Cutdown
	* E1M2 midi
	- Toilet bug STILL there... (too rare to bother with fixing.)
	* Switching guns mid fire shows the fire frame for the new weapon
	Start readme.txt
	* Reset levels kills midi

24-12-24
	* state.c refactor
	* Music from custom levels playing .mid
	+ Record some demos
	Watertight loading and saving
	LICENCES.txt
	+ registration generation and checking

25-12-24

	* Demo playback option quits the game after demo has run.
	+ Demo, if dying on invuln, continues invuln in next demo.
	WINDOWS 7
		* Console set to true still closes window
		* Cannot find saves and demo directory

	+ Bug in enemy movement code can allow them to clip into walls.
	* End level performance is slow. Color buffer to blame?
		-	* New texture - empty portrait of tsar
	+ Pillar in secret pushwall not blocking E1M2
		+ Invuln tanks FPS on slower cpus? (inlined function) (using a LUT)
	* Full screen doesn't actually full screen on lower resolutions
	* Write logs to state when a program has exited successfully

	== COMMENT ==
	Trying to debug the Windows 7 build of the game - Finally got access to a debuger with w64devkit
	Fixed, cut down too many dlls 

26-12-24

	* Full screen stretched doesn't draw flash properly 
	* Loading save changes input state
	* Loading saves doesn't load new game music.
	+ Segfault in memory freeing.
	+ Quickload doesn't recognise new loads after the first one is made.
	+ Secret area at the beginning of E1M3 has non blocking things tha tshould be blocking.
	* Invuln not only sticks on after demos, but also after death sometimes. Seems if the player picked it up at all during the previous run.
	Remove fluidsynth debug log in main folder.
	+ Remove Federov on all difficulties other than medium in E1M4 courtyard.
	* ENemies still believe they can see player when they cannot.
	* If in custom level load, midifile cannot be found, fallback
	* Moderate improvement in the rendering of invuln.
	* Choosing to quit game with mouse hangs game.
	* New texture pass and new company logo
	
	== COMMENT ==
	Windows XP support looks unlikely at this point. Newer versions of SDL2 don't provide support for it.
	Perhaps a custom build with an XP toolchain - Not high priority.

27-12-24

	* Message on why saving is unavailable
	* Move quit and save exclusion code to input_fns.c
	* #include <windows.h> should be conditional with #ifdef
	* CoreAudio added to macos fluidsynth initialisation
	* MacOS build
	+ newline eof gcc/clang requirement
	+ Mouse offset in menu? (fullscreen?)
	Go through each module, think of ways to break.
	+ settings.cfg solid fallback
	+ hiscore.b3d solid fallback

	== COMMENT ==
	After some annoying faffing with permissions and having to build binaries for hours, I have a working version of BN3D on Mac.
	Runs amazing, speed is great! All I have to etst it on is a MacBook Air

28-12-24
	* Only loading from param and .b3d for demo files.
	* Gamestate run in record demo -> pause -> return to game as GAMESTATE_PAUSE
	* Do all .b3d files exist? File check upon startup
	* Demo desync - First load and subsequent load.
	* Smaller font for text in bolsheviktion splash
	* Map zooming for demo Playback
	if the palyer is ever in the way of a patrol and forces a turn around, that should aggro the guard.

29-12-24
	* Error messages come up if logger unavailable.
	+ E1M4 music
	* settings.cfg validation and integrity
	* Save file validation
	* if saving while score increasing, make sure to finalise the score and not save as incrementing.
	* quicksaves should load properly now.
	+ registration

30-12-24
	* Free username
	* Advance program to 0.95 Preview
	+ New game fade out from menu / no fade out from a game already started.
	+ Load game from main menu no fade out
	start going through some unusual button combos and load pathways.
		1) LoadLevel -> gameover -> normal base game
		2) loadLevel -> finish level -> normal base
		3) new Game from demo record/demo playback disabled.
	+ Spelling and grammar/wording 
	+ Faberge egg misplaced pixel
	+ Occasional Crash on non-gdb debug at demo clearing - To me, it looks like Fluidsynth doing something
	+ bn3d logo should fade to black more obviously
	+ Take over to Valgrind

	== COMMENT ==
	All features are now finalised!
	Now for some cosmetic improvements, starting with fade in/fade out

31-12-24
	+ Crash confirmed All systems outside of debugger.
	* Malloc issue in b3d on linux.
	* Unused results for fread in file_savegame.C
	* file_settings.c please eitehr fix or delete string seems to over run
	* graphics_map.c 326 fix sprintf call
	* Linux error box for error messages.

	== COMMENT ==
	After hours fiddling with clang to no avail, going to do the old fashioned way - debug via printf's.
	The problem with this bug is it seems more and more unlikely to happen as I keep the game running - perhaps it was just memory corruption?

1-1-25
	+ No intro card for Mac, but it shows on repeat loopings.
	* Few image enhancements.
	* EnemyQuadrant bug fixed.
	+ High scores still writes the same name many times. (Non debugger only)
	+ Toilet in e1M5 should be blocking.
	Encounter with the tsar's last word in his text goes off-screen on 2560x1440
	* Corrupted main Logo
	+ Hype vid

	== COMMENT ==
	After successfully gaining access to a crashdump, we are seeing a crash inside fluidsynth.
	I just cannot get demo synchronisation, I'm giving up. Humiliating defeat. But I did find a desync cause inside the quadrant decision function which I fixed.
	Some thoughts going forward - Early planning I could've integrated the entire engine tightly around determinism - the idea to have a playback for the demo loop was
	a decision that came later, and looking back I should never have considered it, given the amount of time wasted on trying to make it work.
	Additionally, I believe C++ would be better suited to functionality like this, keeping the code tightly together.

2-1-25
	* Consolidated sin/cos angle calculations.
	+ Fix the high scores - Not happening in a debugger.
	* Loading save from end level deosn't reset state bool
	Download module for Website
	* Bonus PAR score not 10000 but 15000

	Buy module for Website
		Crypto
		Paypal
		Stripe? -- Copy code from Private Language Teachers
	* Save game name corruption in non-debugger - This is same bug from October
	Hype video:
		* - Music cut better
		* - End burn out transition
		* - BEtter font in JANUARY 2025
	+ end of level, automatically finish countup.

	== COMMENT == 
	Save game bug was quite thorny, but ultimately came down to not reassigning a freed save game pool when creating a new one.

3-1-25
	* Post hype to socials
	* Countup should cancel and all points set in stone at level end, demo end, death, saving
	* Warning when user chooses unsupported screen resolutions.
	* Updating high score table happens multiple times.
	* Refactor high score drawing code in cutscene.c
	+ Code check - Null termination
	* String drawing functions optimised using c99 LUT tricks
	* ANSI terminal compatibility checks to disable ugly terminal command strings where it's not supported.

	== COMMENT ==
	Code to check if a file has already been written was using the longer of a registered user, without truncation.

4-1-25
	- Dog pathfinding broken
	+ Crash @ release build after new game from game over. (spam enter)
	* Game over using old texture / Negant 1895 update
	* segfault when selecting new gmae with mouse
		* check for null	
		* fix null call
	* No response during no input and alt-tab out
	* All menu items after game over should be reset??

5-1-25
	* E1M4.mid finish
	* Ptr error from demo freeMemory - Active demo
	* demo_t records direction nums as unsigned, causing movement issues in playback

	== COMMENT == 
	Like a fool, I just simply cannot accept defeat on deterministic demos.
	So I made a special debug version of the program and, alongside the main demo file, output a version of the enemies array at each frame in the original playthrough.
	Then on replay, compare those frames, enemy by enemy, frame by frame.
	This is brilliant, and literally would've saved me a dozen hours had I just done this from teh start.

		== Restructure ==
		The key to accurate reproducability is all about taking out and injecting the states of the variables at the same places in which
		they were originally taken out. So the new refactor follows this pattern, and uncouples the activation of keyState fire to the function of the firing weapon/.


6-1-25
	* Move action polled events.
	* inject player movement data in proper position.
	* Update midi.b3d with new E1M4.mid.
	- recouple firing with a keystatefire to fix the issue with the firing animation in demos.
	+ Rerecord all demos and repackage.
	- Sentinel value on demo, -1 not end, 0 end death, 1 end no death
	+ Better hide the hidden pikuma room

	== COMMENT == 
	VICTORY! Determinism has been achieved.

7-1-25
	+ Demo record mode disallows all new games, save games, load games
	- Demos not needing to be loaded when mode set to demo record.
	- Last ditch effort to synch fire visuals - Store all relevant timings into a struct - new struct for attack.c

8-1-25
	* Enemy->awake not initialised.
	* Loading level from parameters doesn't account for difficulty
	CHecking:
		+ movement ok
		- ammo mismatch??? (Acceptable)

	* Able to fire knife on next frame after out of ammo

	== COMMENT ==
	Spoke too soon, had to go through my loop, document extensively, and rethink the capture order of the gamestates.


9-1-25
	* Weapons switching at 0 ammo in playback should sync.
	* Demo will not overrun demo frames.
	* Weapon frames now consistent among demo and playback.
	+ Stress test all levels with demo record.
	Outstanding issues: + Enemies clipping into walls.
					     + Decal appears when shooting from the side of walls.
	f10 still brings up dialog during record.
	* new, save and load disabled for demo record.
	* Determinism for demoplayback parameter, but not for normal in game demos.
		* Frame 0 not rendering from the demo. (Loop function triggers must be called at start of function)
		* Uninitialised rotateDraw value from file
	* Determinism Achieved for both --p and 1st demo of demo loop
	+ Missing determinism for 2nd, 3rd demo loops

10-1-25
	* Update all functions to atan2f
	* Wipe the bullet hit struct in player on demo end reset
	* HP change bug with police for determinism
	* drop sprite logic updated and refactored.
	* timeSinceLosCheck needs an update on getLevelFramesRand() to account for 0 frames.
	* Table in first room E1M7 not blocking.
	* Enemies occasionally clipping into walls
	* Debug array is not getting reset from demo to demo.
	+ Remember at release to deactivate all debug arrays. Keep code in game. (of #ifdef DEBUG it out);
	+ String referencing bodies is too big.
	+ End of Demo red fade out like he's dead?


	== COMMENT ==
	Walking through walls debug - this one is gonna need a lot of printf's to pin the culprit.
	We know that there should be checks prohibiting walking into walls, mainly. getMapAt().
		Plan:
			1) get enemy who dies' enemy number.
			2) print out the mapAt location of this enemy each frame.
			3) pinpoint frame he walks through the wall.
			4) 
	Bug confirmed - buffer units to wall in edge cases was being applied when it shouldn't.
	We're going to rewrite the way buffers are applied.
	Bug (fixed) - keeping a sentinel error reporter if I detect enemies going into walls.

13-01-25

	+ Demo fade seamless
	* Decals in wrong place
	Eliminate all Fluidsynth unrecognised meta events
	+ E1M6 music
	* Fix death reset loop
	+ E1M1 0x20 & Loop ugly in game
	* "It will quit this game" even after death -> reset to intro loop

	== COMMENT ==
	We might be able to just filter it out on wall slice passes within the renderer.	
	I've decided to implement a final LoS check for the decal at its souce - the creation in shootRay.

14-01-25

	* Advance game to 0.99
	+ map open in demo should close on death
	- Investigate potential memory leak on demo loop (no memory leak!)
	- Critical error bug in SDL Custom Free (occasional)
	+ Map E1M1 no clip run to the end, shows 00:07 when the output shows 00:10
	loop vars += put into the for statement

15-01-25
	+ All music finalised.
	MacOS icon
	* New game transition from menu
		* graphics_transition.c has issues

16-01-25
	+ E1M9 Music remix
	- CLicking saves and loads

17-01-25
	Paypal
	Asha logos video
	Final stress test + packing installer
	* saves folder missing, ptr wiping goes bananas?
	Disable console by default
	+ Fade from demos?
	+ Half bonus on easy mode looks silly 
	+ Train midi stutter loops

18-01-25
	* Starting new game from end level screen does not clear the end level secreen and reset it
	- Segfault when loading a save game AS a demo is being loaded in? - Can't replicate
	* Demo starting from the cutscene doesn't fade in
	* Replace E1M0 music for copyright
	* Thanks for playing message in console
	* start from level -> die to end of game -> demo loop -> after first demo, game quits

19-01-25
	- Black buffer still resides for BUFFER_3D
	* custom tick callback for fluidsynth not being called for midis loaded from memory
	* Tsar image not fixed properly in end screen
	- Stutter on Stardust loop
	- Desync on demo: Finish game with cheats, back to main loop (No idea what happened here...)
	Crash on linux after a long time with libFluidSynth
	+ Desync on linux with shotFired (float precision issue)

	== COMMENT ==
	Long standing issue with regards to the end level timer I think has been fixed - issue was the way counting up was applied and FPS of the game.

	RC1 - No crashes for all platforms stress tested over 60 minutes!!
	Will independently go through a full playthrough of the game on all systems tonight, but I think the game is ready.


20-01-25
	Valgrind:
		* get_terminal_width() fallbacks
		* calloc to initialise color buffers to 0.
		* pollDemoFile shouldn't continue on null ptr
		* destroyAudio check for valid pointers before free
		* unload soundfont
		* Increase usleep time for linux to avoid segfault?

	=== COMMENT ===
	Tried to build a custom version of fluidsynth so I could debug the issue, but it was pretty much a dead end.

21-01-25
	Build website pathways for creating registration file for players who pay

22-01-25
	* Sped up angle conversion function
	Pathway and email generation complete
	Valgrind:
		* reading invalid data in readLevel()
		* reading invalid data in pollDemoFile()
	* noPrintF and stripping unnecessary printf's.
	* Centering menu
	* Advance code to 1.0 release!
	+ Save code corrupts the sprites array in some cases
	+ Extend dog attack range
	enemies can still move while in attack mode?

23-01-25
	* Tweak animation
	* Make all enemies have an end fire frame before going back to chase mode (see enemy guard)
	Better fade for loading games
	- Grain sack and bone pile in TRAIN STATION should be blocking
	Watertight saving and loading
	* Rerecord demos
	* grunts for being attacked
	* process guard death 15
	* Demo ending at end switch doesn't record properly

24-01-25
	* Y/N doesn't work for the are you sure prompts
	Mac Malloc error
	* E1M1 plant in first big hallway needs BLOCKING
	Quick load loads the new save before fade out/in
	* Change tab shift to something else
	die fade out and in doesn't always work
	* Ammo in hut on E1M9 is blocking 
	* If scrolling to the bottom of load game, the option gun goes off the screen
	* Treasure value and enemy kill value needs to be readjusted, as does end of level bonuses.
		I'm on level 4 and I already have 725,000 score in BOLSHEVIKTOR!'
		Level 7 960k
		* Make death penalty harsher as you increase in mode
	* Memory leak when loading load games menu
	* Items in E1M5 silver key room need BLOCKING
	* Edge sprite not seen case (Very old engine bug)
	* Last carriage in GULAG EXPRESS has a table not blocking
	* TRAIN midi can be extended near the end to be a bit more epic.
	* Congratulations! Level # Complete
	PTR not found from the following memory sources:
		SDL_Custom_free x 2,
		Sprite decal buffer clear
		Render sprite precalculation
	* Quicksaves going above 9 corrupts
	* ScoreNum continues to increase, even though it doesn't show on screen. Has ramifications for dying etc.
	E1M5 music 0x21 fluidsynth meta event
	Box Art?
	* Burnt at stake E1M7 in mid island needs BLOCKING
	* False reported corruption of seenMapNum when cheat has revealed the whole map.
	* Update object names for treasure
	SeenMap ptr hanging
	* If a new game is started from a loaded game, program doesn't know that game has already started.
	* HUD alignment
	Upgrade hud drawing routine


	== COMMENT ==
	Did a full, uninterrupted playthrough of Bolsheviktor! mode on Mac - Very close for full release.

25-01-25
	Email sent to christienm7@gmail.com / Set up on phone to receive messages too
	Finalise 1.0 and set up installer for PC and Mac

26-01-25
	Other crypto methods included in payment option

27-01-25
	* Rerecord E1M1
	- Demo desync protection?
	* If added to the high score list, flash? (Hiightlight)
	- Decal buffer not being flushed from level to level??
	- Can our hero look angrier as he gets more and more hurt?
	* Hide hidden room better in E1M9
	* Treasure in last screen needs to be reset
	- Wood door texture get rid of black border?
	* Seen map double free it seems
	* In Save load, names of difficulties should be BOLSHEVIKTOR!
	* Segfault when loading a game when a demo is loading.
	* Reduce the elite guard's volume

	== COMMENT ==
	Did 3 full playthroughs in different difficulties successfully and picked up a few bugs along the way.
	Provisional fix - making sure to update the memoryType when inside my mem_realloc function.
	It's a very rare crash the segfault when laoding game, but the right timing can cause it to happen.

28-01-25
	Release date.

1-2-25
	* MacOS official app pathways added for dynamic directory usage.
	Memory address concerns:
		* Patrol.c not checking for ptr before freeing.
		* Seenmap being freed by free() at the wrong place, so memory manager cannot free it

2-2-25
	Add icon to .app
	change some description of where files we be checked
	sign app and notarise
	Build for MacOS and Linux APPImage

-- CURRENT KNOWN BUGS --
Finishing a level with god mode enabled keeps it on until the game logic can turn it off again (usually next level)