[video output=day047 member=cmuratori stream_platform=twitch stream_username=handmade_hero project=code title="Vector Lengths" vod_platform=youtube id=5KzJ0TDeLxQ annotator=csnover annotator=schme annotator=theinternetftw]
[0:00][Intro]
[1:33][Recap]
[2:50][Diagonal move recap]
[4:40][New analog stick problem]
[10:33][Changing vector length without changing direction]
[14:00][Doing it aggressively]
[15:48][An important note about ratios and squares]
[18:27][Why use length=1?]
[20:10][Back to the wall sticking problem!]
[21:36][Explanation of search in T implementation]
[26:07][Equation for search in T]
[29:37][Simplest possible case]
[32:20][Division by zero]
[35:57][The equation to solve for infinite vertical and horizontal walls]
[37:15][Fixing equation for finite wall sizes]
[38:30][Implementing the new collision code]
[40:40][How to find the tiles to check]
[45:50][What to do about multiple wall collisions]
[54:13][Q&A][:speech]
[55:25][Why did you change from search in p to search in t?]
[59:32][How does it handle the state of two separate players on different floors?]
[1:00:18][Why is the code complexity for p-search so much bigger than t-search?]
[1:03:19][Why not treat the map as a matrix and validate objects and collisions through coordinates?]
[1:04:23][Can you elaborate more on how the coordinate system works with the collision detection?]
[1:06:51][Can this be extended to 3D collision?]
[1:07:04][Could you check the player position and check if it's in the inner side of the wall vectors?]
[1:10:42][What's your point on zero-length vector normalization?]
[1:12:54][How come sometimes players can glitch through walls or corners? Is it art or programming?]
[1:17:04][Do you need to do the math with solid walls and objects or is it basic program arithmetics?]
[1:18:42][Chat adds: I think he means graphically. Walls can be hollow inside.]
[/video]