Originally Posted by svtmike

Or thought through and dismissed because it's not that complex.

Originally Posted by svtmike

You can't get away from the fact that the same rear cogs would be available to both (or "all" since we're being imaginative) front chainrings. You can't get away from the fact that the gears will be more spread out at the edges and more crowded where the ranges intersect. It's simple linear math.

Originally Posted by svtmike

The current Di2 is a simple, elegant way to consistently transform rider input into bicycle reaction. It takes away the simple but annoying task of trimming the front derailleur. I can't think of any time (especially in a fast-moving paceline) that I'd want anything but an immediate, specific reaction from the bike to my shifting demand, or that I'd want an autonomous shifting action from the bicycle. I can compute my gearing needs much more quickly and accurately than any external computer can estimate them. And I can predict them too, something no computer can do.

When it comes to taking action on the bicycle based on what the rider is doing, the best computer possible is already on the bike -- and it's the rider.

Originally Posted by svtmike

I've read it before and am unimpressed. All it does is skip gears where the ranges intersect. Some of you guys seem to think that shifting is some kind of black art or rocket science. Of course, no one ever went broke underestimating....