8 direction diagonal movement difficulties

Questions about the LÖVE API, installing LÖVE and other support related questions go here.
Forum rules
Before you make a thread asking for help, read this.
User avatar
bdjnk
Citizen
Posts: 81
Joined: Wed Jul 03, 2013 11:44 pm

8 direction diagonal movement difficulties

Post by bdjnk » Fri Jul 05, 2013 10:37 pm

So I was doing some of the tutorials from the wiki, and when I got to the hamster ball and efficient tile-based scrolling, I noticed the diagonal movement was too fast. Here's what I added to the hamster ball tutorial as an attempted fix:

Code: Select all

function love.update(dt)
	dx, dy = 0, 0

   if love.keyboard.isDown("right") then
      dx = dx + speed * dt
   end
   if love.keyboard.isDown("left") then
      dx = dx - speed * dt
   end
   if love.keyboard.isDown("down") then
      dy = dy + speed * dt
   end
   if love.keyboard.isDown("up") then
      dy = dy - speed * dt
   end
   
   if dx ~= 0 and dy ~= 0 then
		dx = dx < 0 and -((dx^2)/2)^0.5 or ((dx^2)/2)^0.5
		dy = dy < 0 and -((dy^2)/2)^0.5 or ((dy^2)/2)^0.5
	end
   x = x + dx
   y = y + dy
Several aspects of this are unexpectedly broken. For example, if I hold up and down then press right I move right, as expected; but if I press left I don't move left. The only difference I can see is that left subtracts while right adds.

Here's the .love file. If you have any idea what might be going on, please help me out.
Last edited by bdjnk on Fri Jul 05, 2013 11:19 pm, edited 1 time in total.

User avatar
raidho36
Party member
Posts: 1927
Joined: Mon Jun 17, 2013 12:00 pm

Re: 8 direction diagonal movement speed difficulties

Post by raidho36 » Fri Jul 05, 2013 10:55 pm

It's a lot simplier than that. Just use math:

Code: Select all

if dx ~= 0 and dy ~= 0 then
    dx = dx * 1.41421 -- (roughly) sqare root of 2
    dy = dy * 1.41421
end
The trick behind this is trigonometry. Namely, you need to find specific ratio to apply to diagonal movement, that is you need to find how much longer diagonal path than straight (in terms of speed). That's finding hypotenuse of a triangle. Pythagoras' theorem which states that hypotenuse squared is a sum of both catheti squared. Therefore square root of sum of catheti squared is your hypotenuse. Since we only need ratio, assume that catheti are both unit length. 1² = 1, so the ratio is √1+1 (√2). And that's about it.
Last edited by raidho36 on Fri Jul 05, 2013 11:07 pm, edited 1 time in total.

User avatar
micha
Inner party member
Posts: 1083
Joined: Wed Sep 26, 2012 5:13 pm

Re: 8 direction diagonal movement speed difficulties

Post by micha » Fri Jul 05, 2013 11:03 pm

raidho36 wrote:It's a lot simplier than that. Just use math:

Code: Select all

if dx ~= 0 and dy ~= 0 then
    dx = dx * 1.41421 -- (roughly) sqare root of 2
    dy = dy * 1.41421
end
Small typo. You'd want to divide not multiply.

Code: Select all

if dx ~= 0 and dy ~= 0 then
    dx = dx / 1.41421 -- (roughly) sqare root of 2
    dy = dy / 1.41421
end
or multiply by the square root of 0.5:

Code: Select all

if dx ~= 0 and dy ~= 0 then
    dx = dx * 0.707106781
    dy = dy * 0.707106781
end

User avatar
raidho36
Party member
Posts: 1927
Joined: Mon Jun 17, 2013 12:00 pm

Re: 8 direction diagonal movement speed difficulties

Post by raidho36 » Fri Jul 05, 2013 11:07 pm

Ah, yes. Well, it was easy to figure anyway.

User avatar
bdjnk
Citizen
Posts: 81
Joined: Wed Jul 03, 2013 11:44 pm

Re: 8 direction diagonal movement speed difficulties

Post by bdjnk » Fri Jul 05, 2013 11:15 pm

Hm, very clever. Thanks :D

However, it doesn't solve the issue, which is that somehow, under certain conditions, dx and dy are not being updated appropriately.

As I said before. "For example, if I hold up and down then press right I move right, as expected; but if I press left I don't move left."

User avatar
raidho36
Party member
Posts: 1927
Joined: Mon Jun 17, 2013 12:00 pm

Re: 8 direction diagonal movement difficulties

Post by raidho36 » Fri Jul 05, 2013 11:26 pm

First off, why would you hold up and down, those are mutually exclusive controls.

Second, is your keyboard is ps/2, or is it USB? Because it might simply have buffer overflow when you press too many buttons simultaneously (how many exactly depends on particular key combo). Because I didn't had any such issues, and the code couldn't have been a cause for that anyway.

User avatar
micha
Inner party member
Posts: 1083
Joined: Wed Sep 26, 2012 5:13 pm

Re: 8 direction diagonal movement difficulties

Post by micha » Fri Jul 05, 2013 11:39 pm

I can confirm that. If I press up+down+left, it does not move. As raidho said, it probably a hardware problem. At least the code is correct. You can check if this correct by letting the progam output which buttons are pressed:

Code: Select all

if love.keyboard.isDown('up') then
  love.graphics.print('Up-key is pressed',10,10)
end
and so on for all four direction buttons. That way you can determine if the input is detected in the program or not.

User avatar
bdjnk
Citizen
Posts: 81
Joined: Wed Jul 03, 2013 11:44 pm

Re: 8 direction diagonal movement difficulties

Post by bdjnk » Fri Jul 05, 2013 11:41 pm

raidho36 wrote:First off, why would you hold up and down, those are mutually exclusive controls.

Second, is your keyboard is ps/2, or is it USB? Because it might simply have buffer overflow when you press too many characters simultaneously (how many exactly depends on particular key combo). Because I didn't had any such issues, and the code couldn't have been a cause for that anyway.
It isn't so much that I'd want to do it, as that people might. You've played the type of game where you're a fighter jet / space ship flying against a hoard of enemy craft and installations, right? Well the way I play those games is to hold both the right and left arrow keys and release the one I want to move in the opposite direction of. Why do I do this? I don't know, but I can't be the only one. Anyway, often in those games you can also move forward and backward as well (this becomes particularly important against bosses). But really, why does it matter? It's not working, and that's enough of a reason for me.

I have a PS/2 keyboard, but I'm really having a hard time understanding why that might cause problems. (Google is not yielding helpful data...)

User avatar
raidho36
Party member
Posts: 1927
Joined: Mon Jun 17, 2013 12:00 pm

Re: 8 direction diagonal movement difficulties

Post by raidho36 » Fri Jul 05, 2013 11:52 pm

PS/2 port has legacy-related issues with bandwitdh or something, I never went into details too deep, but the sympthoms of that is that if you can't have more than specific amount of keys pressed simultaneously. Hell knows what logic picks first few that pass, the rest are ignored (and buffer overflow signal is sent, this normally results in PC speaker short beep).

User avatar
bdjnk
Citizen
Posts: 81
Joined: Wed Jul 03, 2013 11:44 pm

Re: 8 direction diagonal movement difficulties

Post by bdjnk » Sat Jul 06, 2013 12:03 am

micha wrote:I can confirm that. If I press up+down+left, it does not move. As raidho said, it probably a hardware problem. At least the code is correct. You can check if this correct by letting the progam output which buttons are pressed:

Code: Select all

if love.keyboard.isDown('up') then
  love.graphics.print('Up-key is pressed',10,10)
end
and so on for all four direction buttons. That way you can determine if the input is detected in the program or not.
raidho36 wrote:PS/2 port has legacy-related issues with bandwitdh or something, I never went into details too deep, but the sympthoms of that is that if you can't have more than specific amount of keys pressed simultaneously. Hell knows what logic picks first few that pass, the rest are ignored (and buffer overflow signal is sent, this normally results in PC speaker short beep).
Okay, so I checked in the program (the original link is updated to include output for the arrow keys currently being held) and sure enough the key presses aren't being seen. To verify I checked with xev then showkey (p.s. I'm running Linux). Yeah, the arrow keys aren't entering the system at all.

Thanks for the assistance. Guess it's just broken :(

Post Reply

Who is online

Users browsing this forum: No registered users and 13 guests