This is what I found during building my Advanced WIFI Fax


The most effective delayer is a tube filled with water. But it will decrease the frequency of spark significantly, bad for continuous spark but best for single spark input.

For continuous sparks, we have to use wire delayer instead. I've found the design that have 2 pixels with one space is the most effective. I know neither code nor how it functions, but it works.



Normally the speed of spark is 2 pixels/frame on average. But the one on the third design was 1.5 pixels/frame, which means you gain extra 25% delay time with a same length of wire. Slightly more effective, uh?

Felix


Test
Oh, thanks for that. I didnt continue my experiments on 4-1 or higher than 3-1, I thought they are the same.

plead-for-destruction
Despite what they wrote in the code, what I've found:

1.Normally spark move at a speed of 2pixels/frame: sometimes 2 ppf(pixels per frame), sometimes stand still for 1 frame, then move at 4 ppf the next frame, that is to say, it's speed is something like this: 2 2 2 2 4 0 2 2 0 4 2 2 2 0 4 0 4 2 4 0 2 2...

2.Spark can get throught a one-pixel gap, but that always takes 1 frame to stop at the edge of the gap, and 1 frame to get through it.

With 2-1 design, spark start from the first pixel of the 2 MTEL, then have to stop at the 2nd pixel of METL, the spark through the gap. It's speed is 1, 2 per 2-1 design, on average 1.5 ppf. Something like 'the spark lost its speed at the edge', but not for 1-1 3-1 design, when they stop at the edge, the don't 'lost speed'

So we can say that: any design like x-1, when x is a even number, it have a delay effect. It's average speed is (2x+2)/(x+2)
When x=2, ie. a 2-1 design, the speed is 1.5
x=4, ie. 4-1 design, the speed is 1.666...~=1.67
x=6, 6-1 design, the speed is 1.75
x=inf, a continuous wire, the speed is 2.

2-1 is the best effective design for delayer.