You will have to do this mathematically.

It should look something like this:

Set X to: choose(100, 200, 300, 400) + background.X % 100

Set Y to: choose(100, 200) + background.Y % 100

You may have to experiment there a bit, unless that works perfectly in the first go.

Of course you're confused, your avatar pic says it all.

But, what seems to be wrong? The video preview shows correct changes in readings and you tested them all well. And it worked well, didn't it?

So now you know what each of them say, how to reset them (in the beginning if you have to), and it can detect where the phone is tilting exactly.

Except, two problems: the Beta -179 goes to plus 180, suddenly which can cause problems if you're performing actions relative to Beta values... The sudden jump. But it can easily be resolved by some math. Search on this forum you'll find many similar problems and solutions.

Second, if you're worried about the sudden wrong readings due to spikes and jerks, you can easily eliminate that by 'easing' out the values. i.e. Instead of showing "exactly" what it the sensor reads, let it gradually go towards the value, which helps in filtering wrong values in between (i.e. jerks).

You can do that by using the 'lerp' function... Look it up.

Example: alpha_value = lerp(alpha_value, Touch.alpha, 0.4)

Thanks oosyrag ! That clears it up! I'll try this soon, if worked, I'll change the title of this post as [Solved].

Cheers!

Hi, I am getting confused on syncing Touch.X and Touch.Y only of the Peer values to Host, not vice versa.

All the examples / tutorials sync objects and then syncing their variables, however, in my case all I want is the Host knows where the Peer is touching.

My questions are:

1. Do I 'have' to sync an object? can I simply sync Variables (or instance variables) ?
2. How do I avoid syncing the host object to the peer? Let's say I only want the Peer object (or instance variables) to be sending data to the Host, not vice versa.

I would appreciate if someone can list the minimum actions that I will require to do this.

THanks!

Okay so I am about to make a large semi-circle boundary in which objects can't enter.

My game is already too heavy (lots of sprites, particles and events), so I was wondering which of these will consume lesser CPU/GPU usage:

1. Having a semi-circle sprite (1600x800 px) with around 25 point collision polygon
--OR--
2. Use Compass-Arc method and only draw a dotted-boundary using 50 small sprites ?

(Meaning, using one thin radius-long sprite as a temporary guide, rotate around the Origin, while spawning total 50 small dots [10x10px] equally along its edge using Repeat and Rotate event-action)

I hope it makes sense? Simpler question is: 1 Large Sprite vs. N Small Objects

Thank you

I tried. Works fine for me, no windows cursor no matter what I do. (Chrome, i7, no graphics card).

So let's debug it: It only happens on Clicks only? Confirm this by disabling On Click conditions. Then (assuming it works then) Enable those again, and now try replacing its actions (path following?) with some other action, maybe use Tween to reach there directly. See if the problem still exist?

Oh and also, try changing the event of Set Mouse Cursor, maybe set it only on Mouse click, or changing the position of that event elsewhere. See what happens.