Don’t use a fixed increment, use current precision

[LeaVerou]

Currently, incrementing a number like 0.123456 will produce 1.123456 which is rarely useful. There are modifiers for getting 1/10th of the increment, but that is little help with these cases.

Ideally, you'd want to use the actual precision of the number, meaning 0.5678 would be incremented like (assuming we just keep hitting the up arrow):

0.5678, 0.5679,
0.568, 0.569,
0.57, 0.58, 0.59,
0.6, 0.7, 0.8, 0.9,
1
2
3
...

The step would still be capped to 1, i.e. incrementing 100000000 would still produce 100000001.