What do you guys think? I just cooked this one up while reading the thread. It probably needs a few tweaks.
It could work; the measurements would need to be extremely accurate, though.
Simply pulling on it to ensure it is taut should provide the required accuracy.
I disagree; gravitation will cause it to sag.
Yes it is in a 2 dimensional plane but there is no need to over complicate with differential equations that are not required as a simple bit of logic will suffice.
First you tell me I have no idea what I am talking about in simple maths, and then you say there's no need to overcomplicate things. To be honest, I didn't think there was much need to bring calculus into it either, but apparently your lack of understanding of what curvature is mandates it.
The total internal reflection with a curved light would cause scattering and probably end up with very distorted signal as the light disperses and signal attenuation occurs.
Why would it cause scattering, and why would there be signal attenuation?
I used 1um as an example due to the wavelengths of light signals in fibers. Fiber optics usually use IR (from 850 - 1600 nm wavelength) thus if you distort the signal by even this small amount over 1 meter (well within the curvature of the earth or the 'bend' on your theory) it would have the effect of steering the wave into the wall of the cable gradually increasing the angle of reflection and by doing so increasing scattering and attenuation. This would therefore not work over long distances and the fiber would be useless.
Yes, it would steer the wave into the wall of the cable. But so does bending the optic fibre itself, and fibres seem to work just fine in all sorts of shapes. Why would the angle of reflection gradually increase?
Well...a micrometre deviation in a fibre optic over a metre of distance is a pretty big one, especially considering the wavelengths at work. TIR would be distorted - to the point where attenuation would increase considerably due to magnificently larger occurences of reflection, and over a long distance of a few KM for example - there would be serious signal degradation to the point where typical error correction methods probably couldn't cope. A parity bit for example would be not nearly enough to deal with the issues.
How can there be larger occurences of reflection when all of the light is already being reflected?
Theory. You always talk about theory and as I have said earlier anyone can devise some equation which draws some graphs with parabola. But actual data, measurements, experiments? You don't have any way to detect or measure your EA and you never will.
How am I supposed to come up with a way to measure something that hasn't been comprehensively described yet?