I uploaded a cut down version of something I had that does an offset relative to bullet angle/speed rather than absolute co-ordinates.

It takes the angle [ or angle of velocity] and checks for colliding at offset relative to that, which might help where you've got extremely fast moving stuff and you need to check ahead a bit to see where it might collide, especially if it's small.

So I have an arrow that spawns and points in the direction of another sprite, linked to angle towards it based on UID/IID.

The only problem I'm having though, is not spawning duplicate arrows.

Is there an easy way to check if a sprite with a particular instance variable exists already before spawning?

<img src="http://i.imgur.com/IaqC7I1.png" border="0" />

Love the music - inspired my Aphex twin? :D

Got any gameplay footage yet?

Okay, so I'll start with a link to the game so you can see what I'm trying to do ( Click image to play ) :

So I have a tethered chain of physics objects tethered to an immovable object, and presently they are set not to collide with each other and have 360 degree freedom of movement.

What I'd like to do is set them as limited revolute joints, but have them spawning relatively to the initial object.

Like so - what I've got now kinda works, but as you can see in the second shot [ or playtesting the game ] things go kinda borked.

<img src="http://i.imgur.com/NoAMhqs.jpg" border="0">

Any suggestions welcome. <img src="smileys/smiley1.gif" border="0" align="middle">

Wow, I ended up doing something really convoluted the last time I ran into this issue by making all the objects immovable until collided with, yielding mixed results.

I thought the pixel rounding was purely a visual thing.

I always wonder how games like Angry Birds solve it, whether they use a totally different engine, or a heavily modded version of Box2d when it comes to their rather complex layouts.

Late to the party, my PC is limping along a little and throwing up blue screens. Gonna try do something useful if I can.

The last one was really fun. <img src="smileys/smiley1.gif" border="0" align="middle" />

It skips steps below 12 pixels, you can adjust it so the gap is less.

I may have missed the point though, do you want it to stop when it hits a 1px wall, or move over it?

I removed the odd time when it stopped. <img src="smileys/smiley36.gif" border="0" align="middle" />

I think this sorta works, as a workaround - it would be cool to specify a 'passable step' dimension for the platform object though.

Example

Download

Sorry - forgot about spritefont!

Did test on firefox and got a friend to test on iphone, seems it's chrome's fault, which is disappointing, as it's included as the default browser for the latest android installs.

Remembering back to google's response on the sticky key bug thing that you guys were chasing them on, I guess I'll refer people to firefox.

<img src="smileys/smiley12.gif" border="0" align="middle" />

If you think of it like the difference between force and impulse in the physics behaviour, instead of applying a vector adjustment once, which would straight away be affected by deceleration, instead, set a variable, and while true , set vector.x to a certain amount.

I did this for wall jumps in a game, I set a variable to 80, then each tick, subtracted 1 from variable, and set vector-x to a fixed number, so the character would move in the air/repel from walls.

Once it hit 0, the vector.x wasn't affected any more.

DoublePost!

Link to .capx file (required!):

PhoneTest.capx

Steps to reproduce:

1. Run game on a phone

2. Touch more than three times

3. Observe touch count.

Observed result:

Touch count goes up, but doesn't go down.

Expected result:

Touch count returns to zero with no touches applied to screen.

Browsers affected:

Chrome: yes

Firefox: Not tested

Internet Explorer: Is there a mobile version?

Operating system & service pack:

Android 4.2

Construct 2 version:

131 I think, the latest one as of this post.

VIDEO :

[tube]YwgpoAYjkl4[/tube]