workergnome / exquisite_donut Goto Github PK

What happens if two clients try to both have the same ID?

There should be a ID 0 example server written in OpenFrameworks.

Should we have some sort of offset, that allows particles to go offscreen? This would allow bezels, and would also handle larger particles that might not want to vanish when their center point hits the edge...

Since you can only have one application listening on a given port, you can only have one instance of this running per-computer.

The issue is what happens when you send a particle off your left side—the message will be sent to yourself, but you won't be able to tell if it should be on your left or your right, because both your left and your right sides are yourself. (Man, this is confusing to try and write.)

So. I'm thinking that if a particle message comes in from an ID that is equal to your own ID and the velocity is positive, assign it to the left side, otherwise the right side.

Any issues?

It would be nice to have a web-based reference implementation of this.

Since we can determine entry side from the direction of velocity, so we still need sender ID?

Accelerations and velocities are defined in terms of the width, even for yacc and yvel, so that they mean the same thing in pixels/centimeters for both. However, ypos is inconsistently defined in terms of the height. This means that code is going to have to look like:

xpos += xvel;
ypos += (yvel / ASPECT_RATIO);
xvel += xacc;
yvel += yacc; // I almost divided by ASPECT_RATIO here too!

I think it would be simpler to let ypos also be defined in terms of width, so that it would range from 0-1.778 for portrait mode 1920x1080 monitors.

Even Rule 5.1 gets this wrong by saying:

YPOS should be modified by adding YVEL to it.

(I'm working with Pras, btw)

There should be an example sketch for Processing.

From @jmarsico: should this be 30? if i remember, i had trouble hitting 60fps in processing when i was a beginner.