RippleGen is a lot less efficient than it should be

[stevenroose]

I don't know exactly what the 50% estimate means, but on my system, it takes RippleGen usually 20x the estimated time to find a matching address.

I've been searching for a 4-letter pattern (excluding the leading r) for hours on end without success while the 50% estimate is 39 minutes.

When I started experimenting a little with other patterns, all of them were around 20x slower than estimated.

I run the latest Ubuntu Linux, 64-bit. Any idea how this comes, or is it supposed to be like this?