On Saturday 14 October 2006 12:52, Jonathan Osuch wrote:

Hal,

...

Does anyone know if arbitrary precision math works with formula type fractals in FractInt?

The arbitrary precision math has been implemented for only four fractal types. mandel, julia, manzpower, and julzpower.

what does arbitrary precision mean ? sammi

On Saturday 14 October 2006 18:17, Paul N. Lee wrote:

I believe Tim Wegner was the main individual responsible for bringing together arbitrary precision into FractInt.

thank you paul, most helpful :) sammi

Paul wrote:

I believe Tim Wegner was the main individual responsible for bringing together arbitrary precision into FractInt.

We Loewer wrote the library and I used it to add deep zooming support for the fractals Jonathan mentioned. Sammi asked:

so ,just to make sure i understand the text it is a type of magnification (as well) ?

"Arbitrary precision" is poorly named. Generally a programmer uses a fixed data type built in to the computing environment such as the 128 bit double precision type. Any floating point type has a maximum space for representing both the exponent and mantissa of a floating point number. This creates a practical limit as to how much magnification one can use to zoom into a fractal - at some point the difference between the numbers associated with adjacent pixels becomes too small to represent. What you see on the screen is the image degenerating into blocky rectangles. An arbitrary precision library gives the programmer control over how much memory space is used to represent the exponent and mantissa of the numbers used in the fractal calculation, so as you zoom deeper, more memory can be allocated for numbers, more digits can be represented, and in particular, there are enough digits to distinguish the numbers associated with adjacent pixels. Computer hardware supports double precision calculations directly, so calculations are very rapid. Arbitrary precision calculations, on the other hand, are handled in software and are slow to begin with, and increasingly slow as the precision increases. What's "arbitrary" about arbitrary precision is that the programmer can "arbitrarily" choose how much memory is allocated for a number, and hence the digits available for calculations. So while arbitrary precision is not a type of magnification, as you asked, it is a programming technique that lets the programmer support higher levels of magnification, so the two are related. I would love to see a programming environment that had arbitrary precision built-in, but the truth of the matter is that ordinary IEEE double precision numbers can handle a very large class of mathematical calculations. Fractal calculations are one of the exceptions. Tim