Matthew Ernest comments on the other day's wind direction graph:
A possibility would be to put the current time in the center, not the starting time. The idea is that the current conditions are of more immediate interest, so they should be accessible faster. The trouble is that at the very center all directions look the same. Perhaps the "center" should not be the center point but the edge of a circle around the center point.
That sounds good (by which I mean I was thinking putting now on the edge when I was reading the middle of the paragraph!). That was an unsettling thing, that the farther along the week you got, the more erratic the curve had to be to show the same change. If you coordinate the thing of most immediate interest (the current wind direction) with the most visible thing in the graph, well, there you go.
You can also think of graphing the wind direction on the interior of a cylinder, then looking inside, so that now is the closer end of the cylinder and the opposite end is far away in both space and time. The further back in time the conditions, the farther away in "space" they are, so the smaller they are in the two-dimensional projection.
(Matthew also linked to the Halting Poem, which is so wrong, in so right a way.)
