That was the only vaguely sensible thing spamboy said.  It's a
completely meaningless comparison.  The eye doesn't work like a camera
at all.

And you are just spamming what is almost certainly a complete-waste-of-
time-website.

So, FOAD, would you?

Nice post. Thanks for that and the interesting link.

In the earliest days of CCD photography people tried to make such non-
rectangular pattern image chips.  The results were very problematic.
One of the key methods of IC layout these days is the use of repeating
cells.  That is not usable with a variable pitch photomosaic.  Also
the electrical characteristics of the detector cell depends on size-
if we have a variable size detector/well, it is very hard to
normalize.

The main reason for these efforts were helmet mounted displays and
helmet sights. I can see no great reason to try overly hard to develop
such focal planes for digital photography, since most people want to
scan their eyes over a print.  Do you really want prints with sharp
centers and fuzzy edges? Actually, I have seen some lenses that do a
pretty good job of simulating that anyway.  :-)

Your logic and your arithmetic need a little improvement.

Around 2000 digicams had 2 Megapixels.  Now in 2007 a lot have 10 Mp.

The rate of increase seems to be around 1 Mp per year, so assuming that rate
of increase is consant your "soon" looks a bit more like100 years.

Roy G