Leaderboard

I find it is often a good idea to learn about things by studying, rather than just making stuff up based on total ignorance.

So you might kill a mouse and leave it on the doorstep? Cruise people's backyards at night? Walk across the kitchen counter? Very strange question you have.

I like that.

Interestingly the first measuring devices for time were (not very) linear devices, not cyclic devices as modern clocks are. But the first devices for measuring distance were not linear devices like rulers or tapes, they were cyclic devices - measuring wheels.

That's really the kicker we can only measure the rate of change and compare different rates of change. One can literally throw away the word time which is just a convenient label and simply describe the rate of change in any process. Another common misconception is thinking time controls rate of change. Time isn't a force or substance that can directly influence anything.

You’d move the table until your indicator read the desired value, yes.

If you really think this is unique to time, how about defining length, without any circular references. Perhaps it’s what a ruler measures? And if your concern is about physics having a mathematical construct that doesn’t physically exist, I fear you haven’t been exposed to much physics.

Thanks. That's clear enough. We know the Earth's not expanding so we know we can discount one of the theories. I suggest you hold a vigil for the lost soul of your idea and then lay it peacefully to rest. It's not coming back.

No it isn't. C is the offspring of, you guessed it, B. Which in turn derived from BCPL. Which was influenced by algol. And there are a great many other languages that owe nothing to C (lisp, Cobol, ML...)

If you want supercomputer/computer/microcomputer/smartphone to process the more data at the same time you should do: 1) Optimize your algorithm. 2) Review compiler assembler output to check what has been generated. e.g. if you won't turn on Floating Model: Fast in Visual Studio, floating point operations can be utterly tragic slow. Default compiler option is Precise. Not all programmers (especially people who don't know assembler) are aware of what this option is doing (therefor they don't switch it every time they're making new Visual Studio C/C++ project). This shows how important is programmer's intelligence, knowledge and competence. 3) Replace Java code by C/C++ (the most time consuming tasks), if it's not enough replace it by assembler. 4) Use gfx cards (OpenCL and CUDA) instead of CPU. Now they have 1024+ cores. And you can buy motherboards which allow couple such cards to be inserted in them at the same time. etc. etc.

First up, partial fractions to make this vaguely approachable. [math]y=\frac{k x}{1+x}=k-\frac{k}{1-x}=k\left( 1-\frac{1}{1-x}\right)[/math] Then some substitution: take [imath]u:=\ln(y)[/imath] and [imath]v:=\ln(x)[/imath]. [math]e^{u} = k\left( 1-\frac{1}{1-e^v}\right)[/math] Take logs. [math]u = \ln(k) + \ln(1 - \tfrac{1}{1-e^v} )[/math] Finally, differentiate. [math]\frac{du}{dv} = \frac{ \tfrac{d}{dv} \left( 1-\tfrac{1}{1-e^v} \right)}{1-\tfrac{1}{1-e^v}}=- \frac{ \tfrac{e^v}{(1-e^v)^2} }{1-\tfrac{1}{1-e^v}}[/math] Throw [imath]x[/imath] back in and simplify as much as you feel like... [math]\frac{du}{dv} = \frac{-x}{(1-x)^2 (1-\tfrac{1}{1-x})}= \frac{-x}{x^2 - x}= \frac{-1}{x-1}[/math] And voilà! [math]\frac{d\ln(y)}{d\ln(x)}= \frac{-1}{x-1}[/math] And no, I have no idea why you would want to do that, or why it would be relevant to anything.

So, conclusions about ether would be off-topic in this thread. I don't think that you're joking, either.