On Sun, 13 Jun 2004, Tim Wegner wrote: (...)
Another group project I have been thinking about for a long time is to make an animated really really deep zoom.
Pardon me, but on a VCD or SVCD, that might amount to ten minutes of video, and seems to be a whole other disk. The batch files wouldn't take much space, though, so it seems that the software for making an AVI or MPEG or FLC or quicktime (multi-image GIF) or something might fit the project. (...)
4. Rendering the images. This would be the "group" part of the project. List members could render the images and send them to a central place.
On the classic set with phone lines, this would take more time to arrange and upload than it would take for one person to render with tesseral or solid guessing (where the latter has results as good as tesseral on the classic set). (...)
There's an animation like this on the Image Lab 2nd Edfition CD, but it only goes a bit past the limit of double precvision. I rendered the images with a hacked up version of Fractint I have that uses 80 bit long doubles instead of doubles.
I was wondering if arbitrary precision took advantage of extended precision when the numbers matched. I guess this answers my question.