Yes! I'm so glad I wasn't the first person so say this - a complete misunderstanding of the scale of the solar system is one of my greatest pet peeves.

Imagine a perfect scale model of the universe where the sun is 4.3 feet in diameter (a foot longer than a meter stick). How big and how far away do you imagine earth to be?

Nobody can intuitively come close to the right answer because there's absolutely no attempt at any level of general education for conveying the scale of things in this realm. The correct answer is precisely 0.476 inches, a little smaller than a dime, and 465 feet away, about two football fields.

This simulation - while cool - at times depicts Jupiter as being twice as big as the sun, but at this scale it would be 5.332 inches in diameter (about the size of a cantaloupe).

At the very least, I'd sleep much better if his "scale" toggle was off by default.

The app creator made the right decision when he defaulted to a false scale. Scale is not the only important thing to convey, one also wants to show relationships and have a model of the whole that fits in a monkey brain.

Also, in this interface, the planets serve as big "buttons" to navigate to a celestial body of interest. So they have to be larger.

The thing would be unusable, and boring, if it showed specks of dust orbiting a mustard seed.

Here's another take on that, a scale model of the solar system that stretches across much of Sweden:

http://en.wikipedia.org/wiki/Sweden_Solar_System

The sun is 110 m in diameter ("including corona" they say). A 65 cm Earth is 7.6 km away.

There is a switch to scale to realistic proportions. Actually, there are three separate switches for sun, planets, and moons.

It also features a gravity grid.

So everything you want is already in there?

First impressions can make a huge difference. Especially in the case of the spatial organization of things, they set the frame we use orient ourselves in a broader context.

I should have mentioned that I'd noticed the inclusion of scale-change settings the first time around. But I have strong sense (based in part on years of research in the area) that allowing parts of the environment to change scale arbitrarily (at least early in the user's interaction) will hurt their capacity to form a veridical and long-lasting representation of the organization of space.

Of course, I've never done anything close to this in awesomeness; no doubt there are trade-offs of which I'm unaware.

Have you seen this thing from NASA?

(http://apod.nasa.gov/apod/ap120312.html)

It was posted to HN a while ago here - (http://news.ycombinator.com/item?id=3756044)

(Only got 7 up votes, which is baffling.)

There's also this - (http://blogs.discovermagazine.com/badastronomy/2012/03/23/an...)

Which was posted to HN here (http://news.ycombinator.com/item?id=3758047)